==================================================================================================== #loader.py ### canonically ### torchrun --standalone --nproc_per_node=8 loader.py ### but for us, probably ### set USE_LIBUV=0 ### set RANK ### set TORCH_CUDNN_SDPA_ENABLED=1 ### torchrun --standalone --nproc_per_node=1 loader.py import os import sys with open(sys.argv[0]) as f: code = f.read() # read the code of this file ASAP, for logging import uuid import glob import time from dataclasses import dataclass import numpy as np import torch import bitsandbytes as bnb from torch import nn import torch.nn.functional as F import torch.distributed as dist import torch._inductor.config as config from torch.nn.parallel import DistributedDataParallel as DDP import pgptlformer ### modded-nanogpt distributed dataset loader # ----------------------------------------------------------------------------- # their simple Distributed Data Loader def _peek_data_shard(filename): # only reads the header, returns header data with open(filename, "rb") as f: # first read the header, which is 256 int32 integers (4 bytes each) header = np.frombuffer(f.read(256*4), dtype=np.int32) if header[0] != 20240520: print("ERROR: magic number mismatch in the data .bin file!") print("---> HINT: Are you passing in a correct file with --input_bin?") print("---> HINT: Dataset encoding changed recently, re-run data prepro or refer again to README") print("---> HINT: For example re-run: `python dev/data/tinyshakespeare.py`, then re-try") exit(1) assert header[1] == 1, "unsupported version" ntok = header[2] # number of tokens (claimed) return ntok # for now just return the number of tokens def _load_data_shard(filename): with open(filename, "rb") as f: # first read the header, which is 256 int32 integers (4 bytes each) header = np.frombuffer(f.read(256*4), dtype=np.int32) assert header[0] == 20240520, "magic number mismatch in the data .bin file" assert header[1] == 1, "unsupported version" ntok = header[2] # number of tokens (claimed) # the rest of it are tokens, stored as uint16 tokens = np.frombuffer(f.read(), dtype=np.uint16) assert len(tokens) == ntok, "number of tokens read does not match header?" return tokens class DistributedDataLoader: def __init__(self, filename_pattern, B, T, process_rank, num_processes): self.process_rank = process_rank self.num_processes = num_processes self.B = B self.T = T # glob files that match the pattern self.files = sorted(glob.glob(filename_pattern)) assert len(self.files) > 0, f"did not find any files that match the pattern {filename_pattern}" # load and validate all data shards, count number of tokens in total ntok_total = 0 for fname in self.files: shard_ntok = _peek_data_shard(fname) assert shard_ntok >= num_processes * B * T + 1 ntok_total += int(shard_ntok) self.ntok_total = ntok_total # kick things off self.reset() def reset(self): self.current_shard = 0 self.current_position = self.process_rank * self.B * self.T self.tokens = _load_data_shard(self.files[self.current_shard]) def advance(self): # advance to next data shard self.current_shard = (self.current_shard + 1) % len(self.files) self.current_position = self.process_rank * self.B * self.T self.tokens = _load_data_shard(self.files[self.current_shard]) def next_batch(self): B = self.B T = self.T buf = self.tokens[self.current_position : self.current_position+B*T+1] buf = torch.tensor(buf.astype(np.int32), dtype=torch.long) x = (buf[:-1]).view(B, T) # inputs y = (buf[1:]).view(B, T) # targets # advance current position and load next shard if necessary self.current_position += B * T * self.num_processes if self.current_position + (B * T * self.num_processes + 1) > len(self.tokens): self.advance() return x.cuda(), y.cuda() # ----------------------------------------------------------------------------- # downgrade to poor man's data loader: # maybe superfluous bc distributed data loader started working # delete? [ ] def get_batch(split): # We recreate np.memmap every batch to avoid a memory leak, as per # https://stackoverflow.com/questions/45132940/numpy-memmap-memory-usage-want-to-iterate-once/61472122#61472122 block_size = args.sequence_length batch_size = args.batch_size if split == 'train': data = np.memmap(os.path.join(data_dir, 'train.bin'), dtype=np.uint16, mode='r') else: data = np.memmap(os.path.join(data_dir, 'val.bin'), dtype=np.uint16, mode='r') ix = torch.randint(len(data) - block_size, (batch_size,)) x = torch.stack([torch.from_numpy((data[i:i+block_size]).astype(np.int64)) for i in ix]) y = torch.stack([torch.from_numpy((data[i+1:i+1+block_size]).astype(np.int64)) for i in ix]) if device_type == 'cuda': # pin arrays x,y, which allows us to move them to GPU asynchronously (non_blocking=True) x, y = x.pin_memory().to(device, non_blocking=True), y.pin_memory().to(device, non_blocking=True) else: x, y = x.to(device), y.to(device) return x, y ### modded-nanogpt ### either 24/16*20=30 batches per 4090 or 24/32*20=15 batches per 4090, ### depending on what kind of v100 tinystories used. @dataclass class Hyperparameters: # data hyperparams input_bin : str = 'data/tinystories-pqt/tinystories-pqt_train_*.bin' # input .bin to train on input_val_bin : str = 'data/tinystories-pqt/tinystories-pqt_val_*.bin' # input .bin to eval validation loss on # optimization hyperparams batch_size : int = 4*32 # macrobatch size, in sequences, across all devices device_batch_size : int = 32 # batch size, in sequences, per device. try to increase/decrease by powers of 2 sequence_length : int = 512 # sequence length, in tokens num_iterations : int = 6250 # number of iterations to run attack : int = 40 # 2*(1-betas)^-1 release : int = 256 # number of iterations of linear warmup/warmdown for triangular or trapezoidal schedule weight_decay : float = 0 # evaluation and logging hyperparams val_loss_every : int = 2000 # every how many steps to evaluate val loss? 0 for only at the end val_tokens : int = 5242880 # how many tokens of validation data? it's important to keep this fixed for consistent comparisons save_every : int = 0 # every how many steps to save the checkpoint? 0 for only at the end run_name : str = "re-pqt-rmsXrmsx2x2-ATTNII" # supercompute boilerplate ddp_run : bool = False #this stuff is so nyannoying device = "cuda" # examples: 'cpu', 'cuda', 'cuda:0', 'cuda:1' etc., or try 'mps' on macbooks torch_compile = False #hahahaha use_z_loss = True z_loss_coefficient = 1e-4 args = Hyperparameters() # convenience variables B, T = args.device_batch_size, args.sequence_length # set up DDP (distributed data parallel). torchrun sets this env variable assert torch.cuda.is_available() if args.ddp_run == True: dist.init_process_group(backend='nccl') ddp_rank = int(os.environ['RANK']) ddp_local_rank = int(os.environ['LOCAL_RANK']) ddp_world_size = int(os.environ['WORLD_SIZE']) device = f'cuda:{ddp_local_rank}' torch.cuda.set_device(device) print(f"using device: {device}") master_process = (ddp_rank == 0) # this process will do logging, checkpointing etc. else: # if not ddp, we are running on a single gpu, and one process master_process = True seed_offset = 0 ddp_world_size = 1 ddp_rank = 0 device = args.device #tokens_per_iter = train_accumulation_steps * ddp_world_size * batch_size * block_size #print(f"tokens per iteration will be: {tokens_per_iter:,}") # calculate the number of steps to take in the val loop. assert args.val_tokens % (B * T * ddp_world_size) == 0 val_steps = args.val_tokens // (B * T * ddp_world_size) # calculate the steps of gradient accumulation required to attain the desired global batch size. assert args.batch_size % (B * ddp_world_size) == 0 train_accumulation_steps = args.batch_size // (B * ddp_world_size) # load tokens train_loader = DistributedDataLoader(args.input_bin, B, T, ddp_rank, ddp_world_size) val_loader = DistributedDataLoader(args.input_val_bin, B, T, ddp_rank, ddp_world_size) if master_process: print(f"Training DataLoader: total number of tokens: {train_loader.ntok_total} across {len(train_loader.files)} files") print(f"Validation DataLoader: total number of tokens: {val_loader.ntok_total} across {len(val_loader.files)} files") x, y = train_loader.next_batch() if master_process: print("Building model...") #tinystories #num_vocab=50304 for non-tinystories models #qknorm="identitynorm" for nonqknorm models layer_prefab = {"dim":512,"dim_head":64,"headcount":8,"ff_mult":4, "lambda":True,"layerwisenorm":"rmsnorm","qknorm":"dynamic_shape_rmsnorm", "attention_deux":True, "training_seqlen":args.sequence_length} #global_prefab = {"vocab_size":8192, "num_layers":4} #weird errors global_prefab = {"vocab_size":50304, "num_layers":8} config = {} config.update(layer_prefab) config.update(global_prefab) model = pgptlformer.PGPT_Lformer(config) if hasattr(config, "coordinate_descent_tuning"): config.coordinate_descent_tuning = True # suggested by @Chillee model = model.to(device) if args.torch_compile: model = torch.compile(model) # here we wrap model into DDP container if args.ddp_run: model = DDP(model, device_ids=[ddp_local_rank]) #raw_model = model.modules() # always contains the "raw" unwrapped model ctx = torch.amp.autocast(device_type='cuda', dtype=torch.bfloat16) if master_process: print("Model built.") # CUDNN attention is ~4ms faster than Flash, but doesn't get selected by default in PyTorch 2.5.1 from torch.backends.cuda import enable_cudnn_sdp, enable_flash_sdp, enable_math_sdp, enable_mem_efficient_sdp enable_cudnn_sdp(True) enable_flash_sdp(True) enable_mem_efficient_sdp(True) enable_math_sdp(False) # modded-nanogpt optimizer inits adam1 = torch.optim.Adam([model.lambdaformer.what_the_embedder_doin.weight], lr=0.3, betas=(0.9, 0.95) ) adam2 = torch.optim.Adam([model.tokenpicker_head.weight], lr=0.002, betas=(0.9, 0.95) ) params = list(model.lambdaformer.blocks.parameters()) matrix_params = [p for p in params if p.ndim == 2] scalar_params = [p for p in params if p.ndim < 2] adam3 = bnb.optim.Adam8bit(matrix_params, lr=0.02, betas=(0.9, 0.95) ) #tune this, sensitive adam4 = bnb.optim.Adam8bit(scalar_params, lr=0.02, betas=(0.9, 0.95) ) #???, less sensitive optim_ensemble = [adam1, adam2, adam3, adam4] # lr scheduler def get_ASR_env(it): assert it <= args.num_iterations # A) the famous linear warmup back at it again if it < args.attack: return (it+1) / args.attack # S) constant sustain elif it < args.num_iterations - args.release: return 1.0 # R) release else: release_ratio = (args.num_iterations - it) / args.release return release_ratio schedulers = [torch.optim.lr_scheduler.LambdaLR(opt, get_ASR_env) for opt in optim_ensemble] # begin logging if master_process: run_id = str(uuid.uuid4()) if args.run_name is not None: sep="-" run_id = sep.join([args.run_name, run_id]) logdir = 'logs/%s/' % run_id os.makedirs(logdir, exist_ok=True) logfile = 'logs/%s.txt' % run_id # create the log file with open(logfile, "w") as f: # begin the log by printing this file (the Python code) f.write('='*100 + '\n') f.write(code) f.write('='*100 + '\n') # log information about the hardware/software environment this is running on # and print the full `nvidia-smi` to file f.write(f"Running pytorch {torch.version.__version__} compiled for CUDA {torch.version.cuda}\nnvidia-smi:\n") import subprocess result = subprocess.run(['nvidia-smi'], stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True) f.write(f'{result.stdout}\n') f.write('='*100 + '\n') training_time_ms = 0 # start the clock torch.cuda.synchronize() t0 = time.time() # begin training train_loader.reset() for step in range(args.num_iterations + 1): last_step = (step == args.num_iterations) # This effectively ignores timing first 10 steps, which are slower for weird reasons. # Alternately, and slightly more correctly in terms of benchmarking, we could do 10 # steps with dummy data first, and then re-initialize the model and reset the loader. if step == 10: training_time_ms = 0 t0 = time.time() timed_steps = float('nan') if step <= 11 else (step - 10) + 1 # <= 11 to avoid bug in val # once in a while evaluate the validation dataset if (last_step or (args.val_loss_every > 0 and step % args.val_loss_every == 0)): # stop the clock torch.cuda.synchronize() training_time_ms += 1000 * (time.time() - t0) # run validation batches model.eval() val_loader.reset() val_loss = 0.0 val_aux_loss = 0.0 for _ in range(val_steps): x_val, y_val = val_loader.next_batch() with ctx: # of course, we'd like to use no_grad() here too, but that creates a torch.compile error for some reason _, loss, z_loss = model(x_val, y_val, return_logits=False, return_zloss=args.use_z_loss) val_loss += loss.detach() if z_loss is not None: val_aux_loss += z_loss.detach()*args.z_loss_coefficient del loss, z_loss if args.ddp_run: dist.all_reduce(val_loss, op=dist.ReduceOp.AVG) dist.all_reduce(val_aux_loss, op=dist.ReduceOp.AVG) val_loss /= val_steps val_aux_loss /= val_steps # log val loss to console and to logfile if master_process: print(f'step:{step}/{args.num_iterations} val_loss:{val_loss:.4f} val_aux_loss:{val_aux_loss:.4f} train_time:{training_time_ms:.0f}ms step_avg:{training_time_ms/(timed_steps-1):.2f}ms') with open(logfile, "a") as f: f.write(f'step:{step}/{args.num_iterations} val_loss:{val_loss:.4f} val_aux_loss:{val_aux_loss:.4f} train_time:{training_time_ms:.0f}ms step_avg:{training_time_ms/(timed_steps-1):.2f}ms\n') # start the clock again torch.cuda.synchronize() t0 = time.time() if master_process and (last_step or (args.save_every > 0 and step % args.save_every == 0)): # stop the clock torch.cuda.synchronize() training_time_ms += 1000 * (time.time() - t0) # save the state of the training process log = dict(step=step, code=code, model=model.state_dict(), model_args=config, optim_ensemble=[opt.state_dict() for opt in optim_ensemble]) torch.save(log, 'logs/%s/state_step%06d.pt' % (run_id, step)) # start the clock again torch.cuda.synchronize() t0 = time.time() # bit confusing: we want to make sure to eval on 0th iteration # but also after the very last iteration. so we loop for step <= num_iterations # instead of just < num_iterations (one extra due to <=), only to do # the validation/sampling one last time, and then we break right here as we're done. if last_step: break # --- train time --- model.train() for i in range(1, train_accumulation_steps+1): # forward pass with ctx: _, loss, z_loss = model(x, y, return_logits=False, return_zloss=args.use_z_loss) train_loss = loss.detach() if z_loss is not None: train_aux_loss = z_loss.detach()*args.z_loss_coefficient loss = loss+z_loss*args.z_loss_coefficient else: train_aux_loss = 0 # advance the dataset for the next batch x, y = train_loader.next_batch() # backward pass if args.ddp_run: if i < train_accumulation_steps: with model.no_sync(): # there's no need to sync gradients every accumulation step loss.backward() else: loss.backward() # just sync on the last step for p in model.parameters(): #grad accum normalization? p.grad /= train_accumulation_steps # skip muon momentum warmup since we're adaming it #... # step the optimizers and schedulers for opt, sched in zip(optim_ensemble, schedulers): opt.step() sched.step() # null the gradients model.zero_grad(set_to_none=True) # --- train time is already over --- #dist.all_reduce(train_loss, op=dist.ReduceOp.AVG) # all-reducing the training loss would be more correct in terms of logging, but slower if master_process: approx_time = training_time_ms + 1000 * (time.time() - t0) print(f"step:{step+1}/{args.num_iterations} train_loss:{train_loss.item():.4f} aux_loss:{train_aux_loss.item():.4f} train_time:{approx_time:.0f}ms step_avg:{approx_time/timed_steps:.2f}ms") with open(logfile, "a") as f: f.write(f"step:{step+1}/{args.num_iterations} train_loss:{train_loss.item():.4f} aux_loss:{train_aux_loss.item():.4f} train_time:{approx_time:.0f}ms step_avg:{approx_time/timed_steps:.2f}ms\n") if master_process: print(f"peak memory consumption: {torch.cuda.max_memory_allocated() // 1024 // 1024} MiB") # clean up nice if args.ddp_run: dist.destroy_process_group()==================================================================================================== Running pytorch 2.4.0+cu124 compiled for CUDA 12.4 nvidia-smi: Fri Jan 31 20:54:29 2025 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 551.61 Driver Version: 551.61 CUDA Version: 12.4 | |-----------------------------------------+------------------------+----------------------+ | GPU Name TCC/WDDM | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| | 0 NVIDIA GeForce RTX 4090 WDDM | 00000000:01:00.0 On | Off | | 36% 51C P2 61W / 350W | 2547MiB / 24564MiB | 1% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ +-----------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=========================================================================================| | 0 N/A N/A 4924 C+G ...64__8wekyb3d8bbwe\CalculatorApp.exe N/A | | 0 N/A N/A 15616 C+G C:\Program Files\VSCodium\VSCodium.exe N/A | | 0 N/A N/A 15884 C+G ...aming\Telegram Desktop\Telegram.exe N/A | | 0 N/A N/A 16300 C+G ...B\system_tray\lghub_system_tray.exe N/A | | 0 N/A N/A 17516 C+G ...64__8wekyb3d8bbwe\CalculatorApp.exe N/A | | 0 N/A N/A 27144 C+G ...64__8wekyb3d8bbwe\CalculatorApp.exe N/A | | 0 N/A N/A 32740 C+G ....Search_cw5n1h2txyewy\SearchApp.exe N/A | | 0 N/A N/A 41308 C+G C:\Windows\explorer.exe N/A | | 0 N/A N/A 42544 C+G ...IP STUDIO PAINT\CLIPStudioPaint.exe N/A | | 0 N/A N/A 44208 C+G ...ekyb3d8bbwe\PhoneExperienceHost.exe N/A | | 0 N/A N/A 49400 C+G ...al\Discord\app-1.0.9179\Discord.exe N/A | | 0 N/A N/A 52028 C ...rograms\Python\Python310\python.exe N/A | | 0 N/A N/A 62540 C+G ...1.0_x64__8wekyb3d8bbwe\Video.UI.exe N/A | | 0 N/A N/A 73872 C+G ...5n1h2txyewy\ShellExperienceHost.exe N/A | | 0 N/A N/A 76236 C+G ....Search_cw5n1h2txyewy\SearchApp.exe N/A | | 0 N/A N/A 80180 C+G ...a\Local\Mozilla Firefox\firefox.exe N/A | | 0 N/A N/A 80476 C+G C:\Hydrus\client.exe N/A | | 0 N/A N/A 81528 C+G ...les (x86)\foobar2000\foobar2000.exe N/A | | 0 N/A N/A 84480 C+G ...siveControlPanel\SystemSettings.exe N/A | | 0 N/A N/A 84668 C+G ...64__8wekyb3d8bbwe\CalculatorApp.exe N/A | | 0 N/A N/A 85528 C+G ...CBS_cw5n1h2txyewy\TextInputHost.exe N/A | | 0 N/A N/A 94588 C+G ...\cef\cef.win7x64\steamwebhelper.exe N/A | | 0 N/A N/A 100784 C+G ...cal\Microsoft\OneDrive\OneDrive.exe N/A | | 0 N/A N/A 117496 C+G ...t.LockApp_cw5n1h2txyewy\LockApp.exe N/A | | 0 N/A N/A 117932 C+G ...64__8wekyb3d8bbwe\CalculatorApp.exe N/A | +-----------------------------------------------------------------------------------------+ ==================================================================================================== step:0/6250 val_loss:10.8259 val_aux_loss:0.0421 train_time:80ms step_avg:nanms step:1/6250 train_loss:10.8258 aux_loss:0.0421 train_time:2169ms step_avg:nanms step:2/6250 train_loss:10.8231 aux_loss:0.0421 train_time:3652ms step_avg:nanms step:3/6250 train_loss:10.7940 aux_loss:0.0421 train_time:4945ms step_avg:nanms step:4/6250 train_loss:10.7497 aux_loss:0.0420 train_time:6368ms step_avg:nanms step:5/6250 train_loss:10.6270 aux_loss:0.0417 train_time:10204ms step_avg:nanms step:6/6250 train_loss:10.5167 aux_loss:0.0415 train_time:12139ms step_avg:nanms step:7/6250 train_loss:10.3312 aux_loss:0.0412 train_time:14687ms step_avg:nanms step:8/6250 train_loss:10.1390 aux_loss:0.0409 train_time:16266ms step_avg:nanms step:9/6250 train_loss:10.0221 aux_loss:0.0410 train_time:20518ms step_avg:nanms step:10/6250 train_loss:9.5647 aux_loss:0.0402 train_time:23811ms step_avg:nanms step:11/6250 train_loss:9.2276 aux_loss:0.0399 train_time:1556ms step_avg:nanms step:12/6250 train_loss:8.9287 aux_loss:0.0401 train_time:3714ms step_avg:nanms step:13/6250 train_loss:8.5375 aux_loss:0.0400 train_time:5456ms step_avg:1818.64ms step:14/6250 train_loss:8.1577 aux_loss:0.0403 train_time:7003ms step_avg:1750.83ms step:15/6250 train_loss:7.8432 aux_loss:0.0410 train_time:16094ms step_avg:3218.81ms step:16/6250 train_loss:7.4508 aux_loss:0.0412 train_time:17596ms step_avg:2932.68ms step:17/6250 train_loss:7.1908 aux_loss:0.0420 train_time:18932ms step_avg:2704.54ms step:18/6250 train_loss:6.9528 aux_loss:0.0430 train_time:20333ms step_avg:2541.63ms step:19/6250 train_loss:6.7638 aux_loss:0.0439 train_time:21825ms step_avg:2425.03ms step:20/6250 train_loss:6.5677 aux_loss:0.0445 train_time:23196ms step_avg:2319.60ms step:21/6250 train_loss:6.5161 aux_loss:0.0459 train_time:24691ms step_avg:2244.67ms step:22/6250 train_loss:6.3443 aux_loss:0.0467 train_time:26186ms step_avg:2182.20ms step:23/6250 train_loss:6.4845 aux_loss:0.0484 train_time:27653ms step_avg:2127.14ms step:24/6250 train_loss:6.1531 aux_loss:0.0465 train_time:29105ms step_avg:2078.95ms step:25/6250 train_loss:6.2610 aux_loss:0.0474 train_time:30552ms step_avg:2036.82ms step:26/6250 train_loss:6.1556 aux_loss:0.0444 train_time:32089ms step_avg:2005.57ms step:27/6250 train_loss:9.3239 aux_loss:0.0402 train_time:33713ms step_avg:1983.14ms step:28/6250 train_loss:6.1337 aux_loss:0.0425 train_time:35564ms step_avg:1975.76ms step:29/6250 train_loss:6.1234 aux_loss:0.0442 train_time:37346ms step_avg:1965.60ms step:30/6250 train_loss:6.1876 aux_loss:0.0446 train_time:38850ms step_avg:1942.52ms step:31/6250 train_loss:6.0766 aux_loss:0.0438 train_time:40240ms step_avg:1916.20ms step:32/6250 train_loss:6.0857 aux_loss:0.0432 train_time:41670ms step_avg:1894.11ms step:33/6250 train_loss:6.0752 aux_loss:0.0434 train_time:43425ms step_avg:1888.06ms step:34/6250 train_loss:5.9583 aux_loss:0.0426 train_time:45256ms step_avg:1885.68ms step:35/6250 train_loss:6.0333 aux_loss:0.0444 train_time:47164ms step_avg:1886.56ms step:36/6250 train_loss:5.9911 aux_loss:0.0441 train_time:48763ms step_avg:1875.52ms step:37/6250 train_loss:5.9690 aux_loss:0.0438 train_time:50090ms step_avg:1855.18ms step:38/6250 train_loss:5.9515 aux_loss:0.0427 train_time:51441ms step_avg:1837.16ms step:39/6250 train_loss:5.9433 aux_loss:0.0431 train_time:52834ms step_avg:1821.86ms step:40/6250 train_loss:5.9615 aux_loss:0.0431 train_time:54362ms step_avg:1812.08ms step:41/6250 train_loss:6.3068 aux_loss:0.0478 train_time:57887ms step_avg:1867.33ms step:42/6250 train_loss:6.2232 aux_loss:0.0449 train_time:59506ms step_avg:1859.57ms step:43/6250 train_loss:5.9971 aux_loss:0.0408 train_time:60948ms step_avg:1846.91ms step:44/6250 train_loss:5.9864 aux_loss:0.0381 train_time:62753ms step_avg:1845.69ms step:45/6250 train_loss:6.0427 aux_loss:0.0464 train_time:64346ms step_avg:1838.47ms step:46/6250 train_loss:5.9152 aux_loss:0.0385 train_time:65935ms step_avg:1831.53ms step:47/6250 train_loss:5.9338 aux_loss:0.0362 train_time:67332ms step_avg:1819.78ms step:48/6250 train_loss:6.0376 aux_loss:0.0405 train_time:69783ms step_avg:1836.39ms step:49/6250 train_loss:5.9729 aux_loss:0.0398 train_time:71380ms step_avg:1830.25ms step:50/6250 train_loss:5.9900 aux_loss:0.0375 train_time:73969ms step_avg:1849.23ms step:51/6250 train_loss:6.0085 aux_loss:0.0372 train_time:76702ms step_avg:1870.77ms step:52/6250 train_loss:5.8804 aux_loss:0.0387 train_time:78088ms step_avg:1859.25ms step:53/6250 train_loss:5.9332 aux_loss:0.0400 train_time:79439ms step_avg:1847.42ms step:54/6250 train_loss:5.9099 aux_loss:0.0403 train_time:80896ms step_avg:1838.55ms step:55/6250 train_loss:5.8947 aux_loss:0.0397 train_time:82468ms step_avg:1832.63ms step:56/6250 train_loss:5.9217 aux_loss:0.0389 train_time:83892ms step_avg:1823.73ms step:57/6250 train_loss:5.9223 aux_loss:0.0384 train_time:85381ms step_avg:1816.62ms step:58/6250 train_loss:5.9116 aux_loss:0.0386 train_time:86745ms step_avg:1807.19ms step:59/6250 train_loss:5.8840 aux_loss:0.0391 train_time:88084ms step_avg:1797.64ms step:60/6250 train_loss:5.8557 aux_loss:0.0396 train_time:89471ms step_avg:1789.42ms step:61/6250 train_loss:5.8707 aux_loss:0.0397 train_time:90795ms step_avg:1780.29ms step:62/6250 train_loss:5.9502 aux_loss:0.0394 train_time:92554ms step_avg:1779.89ms step:63/6250 train_loss:5.8758 aux_loss:0.0388 train_time:94209ms step_avg:1777.53ms step:64/6250 train_loss:5.7920 aux_loss:0.0387 train_time:96220ms step_avg:1781.86ms step:65/6250 train_loss:5.8986 aux_loss:0.0387 train_time:98649ms step_avg:1793.62ms step:66/6250 train_loss:5.8595 aux_loss:0.0390 train_time:101499ms step_avg:1812.49ms step:67/6250 train_loss:5.8711 aux_loss:0.0394 train_time:102867ms step_avg:1804.69ms step:68/6250 train_loss:5.8593 aux_loss:0.0397 train_time:105184ms step_avg:1813.51ms step:69/6250 train_loss:5.8711 aux_loss:0.0392 train_time:106678ms step_avg:1808.10ms step:70/6250 train_loss:5.8486 aux_loss:0.0388 train_time:108027ms step_avg:1800.45ms step:71/6250 train_loss:5.8501 aux_loss:0.0388 train_time:109703ms step_avg:1798.42ms step:72/6250 train_loss:5.8372 aux_loss:0.0389 train_time:111827ms step_avg:1803.67ms step:73/6250 train_loss:5.7948 aux_loss:0.0391 train_time:113994ms step_avg:1809.43ms step:74/6250 train_loss:5.8339 aux_loss:0.0391 train_time:115714ms step_avg:1808.04ms step:75/6250 train_loss:5.7900 aux_loss:0.0389 train_time:117525ms step_avg:1808.07ms step:76/6250 train_loss:5.8801 aux_loss:0.0402 train_time:119167ms step_avg:1805.56ms step:77/6250 train_loss:5.6615 aux_loss:0.0385 train_time:121295ms step_avg:1810.38ms step:78/6250 train_loss:5.6764 aux_loss:0.0387 train_time:122977ms step_avg:1808.49ms step:79/6250 train_loss:5.6245 aux_loss:0.0390 train_time:124484ms step_avg:1804.12ms step:80/6250 train_loss:5.5443 aux_loss:0.0388 train_time:126033ms step_avg:1800.48ms step:81/6250 train_loss:5.5350 aux_loss:0.0394 train_time:127442ms step_avg:1794.95ms step:82/6250 train_loss:5.4985 aux_loss:0.0402 train_time:129398ms step_avg:1797.20ms step:83/6250 train_loss:5.4439 aux_loss:0.0395 train_time:130941ms step_avg:1793.71ms step:84/6250 train_loss:5.4428 aux_loss:0.0411 train_time:132758ms step_avg:1794.02ms step:85/6250 train_loss:5.4924 aux_loss:0.0404 train_time:134519ms step_avg:1793.59ms step:86/6250 train_loss:5.3827 aux_loss:0.0398 train_time:135960ms step_avg:1788.95ms step:87/6250 train_loss:5.4211 aux_loss:0.0420 train_time:137338ms step_avg:1783.61ms step:88/6250 train_loss:5.3330 aux_loss:0.0417 train_time:138778ms step_avg:1779.20ms step:89/6250 train_loss:5.4365 aux_loss:0.0407 train_time:140294ms step_avg:1775.88ms step:90/6250 train_loss:5.2397 aux_loss:0.0413 train_time:142418ms step_avg:1780.23ms step:91/6250 train_loss:5.2378 aux_loss:0.0419 train_time:143979ms step_avg:1777.51ms step:92/6250 train_loss:5.1862 aux_loss:0.0425 train_time:145525ms step_avg:1774.70ms step:93/6250 train_loss:5.2286 aux_loss:0.0424 train_time:146933ms step_avg:1770.28ms step:94/6250 train_loss:5.1639 aux_loss:0.0421 train_time:148372ms step_avg:1766.34ms step:95/6250 train_loss:5.0016 aux_loss:0.0428 train_time:150462ms step_avg:1770.15ms step:96/6250 train_loss:5.0912 aux_loss:0.0437 train_time:152308ms step_avg:1771.02ms step:97/6250 train_loss:5.1650 aux_loss:0.0434 train_time:154316ms step_avg:1773.75ms step:98/6250 train_loss:5.0497 aux_loss:0.0427 train_time:155870ms step_avg:1771.25ms step:99/6250 train_loss:5.0271 aux_loss:0.0427 train_time:157846ms step_avg:1773.55ms step:100/6250 train_loss:5.0161 aux_loss:0.0438 train_time:159408ms step_avg:1771.20ms step:101/6250 train_loss:4.9641 aux_loss:0.0443 train_time:160815ms step_avg:1767.20ms step:102/6250 train_loss:4.9538 aux_loss:0.0433 train_time:162207ms step_avg:1763.12ms step:103/6250 train_loss:4.9213 aux_loss:0.0431 train_time:163655ms step_avg:1759.74ms step:104/6250 train_loss:4.9341 aux_loss:0.0443 train_time:165210ms step_avg:1757.56ms step:105/6250 train_loss:4.8399 aux_loss:0.0448 train_time:166933ms step_avg:1757.19ms step:106/6250 train_loss:4.9238 aux_loss:0.0442 train_time:168423ms step_avg:1754.41ms step:107/6250 train_loss:4.8393 aux_loss:0.0448 train_time:169858ms step_avg:1751.11ms step:108/6250 train_loss:4.7484 aux_loss:0.0455 train_time:171485ms step_avg:1749.85ms step:109/6250 train_loss:4.7526 aux_loss:0.0457 train_time:172883ms step_avg:1746.29ms step:110/6250 train_loss:4.6624 aux_loss:0.0463 train_time:174589ms step_avg:1745.89ms step:111/6250 train_loss:4.7095 aux_loss:0.0461 train_time:176331ms step_avg:1745.85ms step:112/6250 train_loss:4.7246 aux_loss:0.0471 train_time:177988ms step_avg:1744.98ms step:113/6250 train_loss:4.6445 aux_loss:0.0470 train_time:179395ms step_avg:1741.70ms step:114/6250 train_loss:4.6070 aux_loss:0.0477 train_time:181061ms step_avg:1740.97ms step:115/6250 train_loss:4.6619 aux_loss:0.0471 train_time:182684ms step_avg:1739.85ms step:116/6250 train_loss:4.6360 aux_loss:0.0499 train_time:184123ms step_avg:1737.01ms step:117/6250 train_loss:4.6473 aux_loss:0.0476 train_time:185957ms step_avg:1737.91ms step:118/6250 train_loss:4.6048 aux_loss:0.0505 train_time:187466ms step_avg:1735.80ms step:119/6250 train_loss:4.6529 aux_loss:0.0485 train_time:188962ms step_avg:1733.60ms step:120/6250 train_loss:4.6086 aux_loss:0.0479 train_time:190401ms step_avg:1730.92ms step:121/6250 train_loss:4.5512 aux_loss:0.0497 train_time:191809ms step_avg:1728.01ms step:122/6250 train_loss:4.5686 aux_loss:0.0510 train_time:193442ms step_avg:1727.16ms step:123/6250 train_loss:4.5125 aux_loss:0.0493 train_time:194994ms step_avg:1725.61ms step:124/6250 train_loss:4.4987 aux_loss:0.0489 train_time:196571ms step_avg:1724.31ms step:125/6250 train_loss:4.4716 aux_loss:0.0501 train_time:198120ms step_avg:1722.78ms step:126/6250 train_loss:4.6177 aux_loss:0.0506 train_time:199728ms step_avg:1721.80ms step:127/6250 train_loss:4.4598 aux_loss:0.0493 train_time:201343ms step_avg:1720.88ms step:128/6250 train_loss:4.4710 aux_loss:0.0494 train_time:202864ms step_avg:1719.18ms step:129/6250 train_loss:4.4659 aux_loss:0.0521 train_time:204595ms step_avg:1719.29ms step:130/6250 train_loss:4.5183 aux_loss:0.0506 train_time:206101ms step_avg:1717.51ms step:131/6250 train_loss:4.5882 aux_loss:0.0491 train_time:207614ms step_avg:1715.82ms step:132/6250 train_loss:4.6276 aux_loss:0.0498 train_time:209091ms step_avg:1713.86ms step:133/6250 train_loss:4.4737 aux_loss:0.0498 train_time:210502ms step_avg:1711.40ms step:134/6250 train_loss:4.5656 aux_loss:0.0487 train_time:212039ms step_avg:1709.99ms step:135/6250 train_loss:4.3949 aux_loss:0.0491 train_time:213577ms step_avg:1708.61ms step:136/6250 train_loss:4.3870 aux_loss:0.0507 train_time:215101ms step_avg:1707.15ms step:137/6250 train_loss:4.4220 aux_loss:0.0518 train_time:216724ms step_avg:1706.49ms step:138/6250 train_loss:4.3368 aux_loss:0.0510 train_time:218138ms step_avg:1704.21ms step:139/6250 train_loss:4.3697 aux_loss:0.0495 train_time:219622ms step_avg:1702.49ms step:140/6250 train_loss:4.3691 aux_loss:0.0493 train_time:221079ms step_avg:1700.61ms step:141/6250 train_loss:4.2743 aux_loss:0.0489 train_time:222477ms step_avg:1698.30ms step:142/6250 train_loss:4.2852 aux_loss:0.0500 train_time:223983ms step_avg:1696.84ms step:143/6250 train_loss:4.2738 aux_loss:0.0510 train_time:225447ms step_avg:1695.09ms step:144/6250 train_loss:4.2509 aux_loss:0.0503 train_time:226879ms step_avg:1693.13ms step:145/6250 train_loss:4.3215 aux_loss:0.0512 train_time:228373ms step_avg:1691.65ms step:146/6250 train_loss:4.2593 aux_loss:0.0518 train_time:229790ms step_avg:1689.64ms step:147/6250 train_loss:4.2211 aux_loss:0.0506 train_time:231241ms step_avg:1687.89ms step:148/6250 train_loss:4.2278 aux_loss:0.0502 train_time:232620ms step_avg:1685.65ms step:149/6250 train_loss:4.1848 aux_loss:0.0501 train_time:234665ms step_avg:1688.24ms step:150/6250 train_loss:4.1603 aux_loss:0.0514 train_time:236323ms step_avg:1688.02ms step:151/6250 train_loss:4.2316 aux_loss:0.0499 train_time:237853ms step_avg:1686.90ms step:152/6250 train_loss:4.1528 aux_loss:0.0505 train_time:242789ms step_avg:1709.78ms step:153/6250 train_loss:4.1450 aux_loss:0.0509 train_time:244182ms step_avg:1707.56ms step:154/6250 train_loss:4.1387 aux_loss:0.0527 train_time:245613ms step_avg:1705.65ms step:155/6250 train_loss:4.1284 aux_loss:0.0492 train_time:246996ms step_avg:1703.42ms step:156/6250 train_loss:4.0900 aux_loss:0.0526 train_time:250850ms step_avg:1718.15ms step:157/6250 train_loss:4.0965 aux_loss:0.0505 train_time:253320ms step_avg:1723.26ms step:158/6250 train_loss:4.1405 aux_loss:0.0493 train_time:255084ms step_avg:1723.54ms step:159/6250 train_loss:4.0942 aux_loss:0.0516 train_time:256531ms step_avg:1721.68ms step:160/6250 train_loss:4.1167 aux_loss:0.0514 train_time:258031ms step_avg:1720.21ms step:161/6250 train_loss:4.1872 aux_loss:0.0499 train_time:260080ms step_avg:1722.38ms step:162/6250 train_loss:4.0243 aux_loss:0.0509 train_time:262109ms step_avg:1724.40ms step:163/6250 train_loss:4.0292 aux_loss:0.0535 train_time:264556ms step_avg:1729.13ms step:164/6250 train_loss:4.0876 aux_loss:0.0513 train_time:266909ms step_avg:1733.18ms step:165/6250 train_loss:3.9792 aux_loss:0.0506 train_time:270806ms step_avg:1747.14ms step:166/6250 train_loss:4.0059 aux_loss:0.0522 train_time:272882ms step_avg:1749.24ms step:167/6250 train_loss:4.0675 aux_loss:0.0521 train_time:276002ms step_avg:1757.97ms step:168/6250 train_loss:4.1965 aux_loss:0.0508 train_time:278970ms step_avg:1765.63ms step:169/6250 train_loss:4.0478 aux_loss:0.0523 train_time:282322ms step_avg:1775.61ms step:170/6250 train_loss:4.0392 aux_loss:0.0517 train_time:284748ms step_avg:1779.67ms step:171/6250 train_loss:3.9833 aux_loss:0.0510 train_time:287223ms step_avg:1784.00ms step:172/6250 train_loss:4.0204 aux_loss:0.0527 train_time:289450ms step_avg:1786.73ms step:173/6250 train_loss:4.0033 aux_loss:0.0512 train_time:291937ms step_avg:1791.03ms step:174/6250 train_loss:4.0539 aux_loss:0.0519 train_time:294070ms step_avg:1793.11ms step:175/6250 train_loss:3.9701 aux_loss:0.0510 train_time:296129ms step_avg:1794.72ms step:176/6250 train_loss:4.0079 aux_loss:0.0523 train_time:297928ms step_avg:1794.74ms step:177/6250 train_loss:3.9562 aux_loss:0.0508 train_time:299994ms step_avg:1796.37ms step:178/6250 train_loss:3.9809 aux_loss:0.0523 train_time:302499ms step_avg:1800.59ms step:179/6250 train_loss:3.9152 aux_loss:0.0515 train_time:305046ms step_avg:1805.01ms step:180/6250 train_loss:3.9037 aux_loss:0.0525 train_time:307774ms step_avg:1810.44ms step:181/6250 train_loss:3.9353 aux_loss:0.0521 train_time:309909ms step_avg:1812.33ms step:182/6250 train_loss:3.8985 aux_loss:0.0527 train_time:312807ms step_avg:1818.64ms step:183/6250 train_loss:3.9259 aux_loss:0.0511 train_time:315051ms step_avg:1821.10ms step:184/6250 train_loss:3.9972 aux_loss:0.0545 train_time:317525ms step_avg:1824.86ms step:185/6250 train_loss:3.8747 aux_loss:0.0510 train_time:320162ms step_avg:1829.50ms step:186/6250 train_loss:3.9007 aux_loss:0.0517 train_time:322280ms step_avg:1831.14ms step:187/6250 train_loss:3.8463 aux_loss:0.0532 train_time:324388ms step_avg:1832.70ms step:188/6250 train_loss:3.9380 aux_loss:0.0520 train_time:326790ms step_avg:1835.90ms step:189/6250 train_loss:3.8638 aux_loss:0.0518 train_time:328882ms step_avg:1837.33ms step:190/6250 train_loss:3.8576 aux_loss:0.0518 train_time:330797ms step_avg:1837.76ms step:191/6250 train_loss:3.9116 aux_loss:0.0525 train_time:332588ms step_avg:1837.50ms step:192/6250 train_loss:3.8395 aux_loss:0.0521 train_time:336240ms step_avg:1847.47ms step:193/6250 train_loss:3.9304 aux_loss:0.0517 train_time:340009ms step_avg:1857.97ms step:194/6250 train_loss:3.8217 aux_loss:0.0532 train_time:345347ms step_avg:1876.89ms step:195/6250 train_loss:3.8615 aux_loss:0.0522 train_time:349518ms step_avg:1889.29ms step:196/6250 train_loss:3.8922 aux_loss:0.0513 train_time:352381ms step_avg:1894.52ms step:197/6250 train_loss:3.7595 aux_loss:0.0522 train_time:355752ms step_avg:1902.42ms step:198/6250 train_loss:3.8158 aux_loss:0.0516 train_time:358069ms step_avg:1904.62ms step:199/6250 train_loss:3.8083 aux_loss:0.0518 train_time:360649ms step_avg:1908.19ms step:200/6250 train_loss:3.8418 aux_loss:0.0515 train_time:362926ms step_avg:1910.14ms step:201/6250 train_loss:3.8815 aux_loss:0.0539 train_time:365206ms step_avg:1912.07ms step:202/6250 train_loss:3.8272 aux_loss:0.0499 train_time:367534ms step_avg:1914.24ms step:203/6250 train_loss:3.8171 aux_loss:0.0548 train_time:368950ms step_avg:1911.66ms step:204/6250 train_loss:3.9183 aux_loss:0.0533 train_time:370330ms step_avg:1908.92ms step:205/6250 train_loss:3.8906 aux_loss:0.0510 train_time:371776ms step_avg:1906.55ms step:206/6250 train_loss:3.9199 aux_loss:0.0535 train_time:373310ms step_avg:1904.64ms step:207/6250 train_loss:3.7570 aux_loss:0.0525 train_time:375136ms step_avg:1904.25ms step:208/6250 train_loss:3.8078 aux_loss:0.0516 train_time:376761ms step_avg:1902.83ms step:209/6250 train_loss:3.8305 aux_loss:0.0517 train_time:378279ms step_avg:1900.90ms step:210/6250 train_loss:3.8007 aux_loss:0.0521 train_time:379909ms step_avg:1899.55ms step:211/6250 train_loss:3.7665 aux_loss:0.0517 train_time:381544ms step_avg:1898.23ms step:212/6250 train_loss:3.7819 aux_loss:0.0518 train_time:383220ms step_avg:1897.13ms step:213/6250 train_loss:3.7774 aux_loss:0.0524 train_time:386922ms step_avg:1906.02ms step:214/6250 train_loss:3.8458 aux_loss:0.0522 train_time:389159ms step_avg:1907.64ms step:215/6250 train_loss:3.7429 aux_loss:0.0516 train_time:390543ms step_avg:1905.09ms step:216/6250 train_loss:3.6851 aux_loss:0.0515 train_time:392031ms step_avg:1903.06ms step:217/6250 train_loss:3.7920 aux_loss:0.0521 train_time:393530ms step_avg:1901.11ms step:218/6250 train_loss:3.8035 aux_loss:0.0529 train_time:395037ms step_avg:1899.22ms step:219/6250 train_loss:3.7329 aux_loss:0.0524 train_time:396466ms step_avg:1896.96ms step:220/6250 train_loss:3.7478 aux_loss:0.0519 train_time:397926ms step_avg:1894.89ms step:221/6250 train_loss:3.7708 aux_loss:0.0527 train_time:400217ms step_avg:1896.76ms step:222/6250 train_loss:3.7716 aux_loss:0.0521 train_time:402873ms step_avg:1900.34ms step:223/6250 train_loss:3.7188 aux_loss:0.0525 train_time:404780ms step_avg:1900.38ms step:224/6250 train_loss:3.7600 aux_loss:0.0514 train_time:406847ms step_avg:1901.16ms step:225/6250 train_loss:3.7685 aux_loss:0.0511 train_time:408677ms step_avg:1900.83ms step:226/6250 train_loss:3.7441 aux_loss:0.0528 train_time:410535ms step_avg:1900.62ms step:227/6250 train_loss:3.7466 aux_loss:0.0502 train_time:412520ms step_avg:1901.01ms step:228/6250 train_loss:3.7478 aux_loss:0.0535 train_time:414044ms step_avg:1899.29ms step:229/6250 train_loss:3.7457 aux_loss:0.0499 train_time:415576ms step_avg:1897.61ms step:230/6250 train_loss:3.7105 aux_loss:0.0566 train_time:417481ms step_avg:1897.64ms step:231/6250 train_loss:3.7753 aux_loss:0.0527 train_time:419443ms step_avg:1897.93ms step:232/6250 train_loss:3.6768 aux_loss:0.0508 train_time:421675ms step_avg:1899.44ms step:233/6250 train_loss:3.7583 aux_loss:0.0521 train_time:423495ms step_avg:1899.08ms step:234/6250 train_loss:3.7565 aux_loss:0.0530 train_time:425451ms step_avg:1899.33ms step:235/6250 train_loss:3.5812 aux_loss:0.0518 train_time:427651ms step_avg:1900.67ms step:236/6250 train_loss:3.6688 aux_loss:0.0517 train_time:431004ms step_avg:1907.10ms step:237/6250 train_loss:3.7249 aux_loss:0.0514 train_time:432478ms step_avg:1905.19ms step:238/6250 train_loss:3.6308 aux_loss:0.0512 train_time:434512ms step_avg:1905.76ms step:239/6250 train_loss:3.5873 aux_loss:0.0525 train_time:436494ms step_avg:1906.09ms step:240/6250 train_loss:3.7836 aux_loss:0.0520 train_time:469888ms step_avg:2042.99ms step:241/6250 train_loss:3.7553 aux_loss:0.0505 train_time:518009ms step_avg:2242.46ms step:242/6250 train_loss:3.6754 aux_loss:0.0508 train_time:574744ms step_avg:2477.35ms step:243/6250 train_loss:3.6333 aux_loss:0.0523 train_time:629498ms step_avg:2701.71ms step:244/6250 train_loss:3.6930 aux_loss:0.0526 train_time:680734ms step_avg:2909.12ms step:245/6250 train_loss:3.6480 aux_loss:0.0508 train_time:719576ms step_avg:3062.03ms step:246/6250 train_loss:3.6631 aux_loss:0.0506 train_time:771676ms step_avg:3269.81ms step:247/6250 train_loss:3.5274 aux_loss:0.0525 train_time:815109ms step_avg:3439.28ms step:248/6250 train_loss:3.6009 aux_loss:0.0513 train_time:870349ms step_avg:3656.93ms step:249/6250 train_loss:3.6087 aux_loss:0.0517 train_time:914517ms step_avg:3826.43ms step:250/6250 train_loss:3.5473 aux_loss:0.0531 train_time:954431ms step_avg:3976.80ms step:251/6250 train_loss:3.6270 aux_loss:0.0519 train_time:957228ms step_avg:3971.90ms step:252/6250 train_loss:3.6138 aux_loss:0.0505 train_time:963030ms step_avg:3979.46ms step:253/6250 train_loss:3.6320 aux_loss:0.0522 train_time:965219ms step_avg:3972.10ms step:254/6250 train_loss:3.6491 aux_loss:0.0519 train_time:966781ms step_avg:3962.22ms step:255/6250 train_loss:3.6284 aux_loss:0.0520 train_time:968206ms step_avg:3951.86ms step:256/6250 train_loss:3.5601 aux_loss:0.0506 train_time:969849ms step_avg:3942.48ms step:257/6250 train_loss:3.6123 aux_loss:0.0519 train_time:972406ms step_avg:3936.87ms step:258/6250 train_loss:3.6303 aux_loss:0.0521 train_time:975389ms step_avg:3933.02ms step:259/6250 train_loss:3.6460 aux_loss:0.0507 train_time:986605ms step_avg:3962.27ms step:260/6250 train_loss:3.6258 aux_loss:0.0515 train_time:998476ms step_avg:3993.90ms step:261/6250 train_loss:3.6320 aux_loss:0.0527 train_time:1000579ms step_avg:3986.37ms step:262/6250 train_loss:3.5962 aux_loss:0.0519 train_time:1002298ms step_avg:3977.37ms step:263/6250 train_loss:3.4494 aux_loss:0.0503 train_time:1005172ms step_avg:3973.01ms step:264/6250 train_loss:3.5767 aux_loss:0.0532 train_time:1007317ms step_avg:3965.82ms step:265/6250 train_loss:3.5390 aux_loss:0.0518 train_time:1009717ms step_avg:3959.68ms step:266/6250 train_loss:3.5916 aux_loss:0.0518 train_time:1011926ms step_avg:3952.84ms step:267/6250 train_loss:3.4899 aux_loss:0.0520 train_time:1023714ms step_avg:3983.32ms step:268/6250 train_loss:3.5705 aux_loss:0.0508 train_time:1032144ms step_avg:4000.56ms step:269/6250 train_loss:3.4770 aux_loss:0.0511 train_time:1036789ms step_avg:4003.05ms step:270/6250 train_loss:3.4937 aux_loss:0.0516 train_time:1039168ms step_avg:3996.80ms step:271/6250 train_loss:3.5110 aux_loss:0.0512 train_time:1041514ms step_avg:3990.47ms step:272/6250 train_loss:3.5357 aux_loss:0.0513 train_time:1043904ms step_avg:3984.37ms step:273/6250 train_loss:3.4845 aux_loss:0.0508 train_time:1046142ms step_avg:3977.73ms step:274/6250 train_loss:3.5302 aux_loss:0.0513 train_time:1048275ms step_avg:3970.74ms step:275/6250 train_loss:3.4566 aux_loss:0.0508 train_time:1050472ms step_avg:3964.05ms step:276/6250 train_loss:3.5183 aux_loss:0.0508 train_time:1052687ms step_avg:3957.47ms step:277/6250 train_loss:3.5650 aux_loss:0.0522 train_time:1055213ms step_avg:3952.11ms step:278/6250 train_loss:3.5235 aux_loss:0.0515 train_time:1067795ms step_avg:3984.31ms step:279/6250 train_loss:3.4232 aux_loss:0.0517 train_time:1077112ms step_avg:4004.13ms step:280/6250 train_loss:3.3560 aux_loss:0.0518 train_time:1078541ms step_avg:3994.60ms step:281/6250 train_loss:3.5171 aux_loss:0.0515 train_time:1079980ms step_avg:3985.17ms step:282/6250 train_loss:3.3773 aux_loss:0.0508 train_time:1081375ms step_avg:3975.64ms step:283/6250 train_loss:3.4196 aux_loss:0.0515 train_time:1083390ms step_avg:3968.46ms step:284/6250 train_loss:3.4991 aux_loss:0.0504 train_time:1085260ms step_avg:3960.80ms step:285/6250 train_loss:3.5146 aux_loss:0.0520 train_time:1086906ms step_avg:3952.38ms step:286/6250 train_loss:3.4356 aux_loss:0.0505 train_time:1088319ms step_avg:3943.18ms step:287/6250 train_loss:3.4948 aux_loss:0.0508 train_time:1090174ms step_avg:3935.65ms step:288/6250 train_loss:3.5075 aux_loss:0.0515 train_time:1095926ms step_avg:3942.18ms step:289/6250 train_loss:3.4301 aux_loss:0.0508 train_time:1097651ms step_avg:3934.23ms step:290/6250 train_loss:3.4565 aux_loss:0.0506 train_time:1098827ms step_avg:3924.38ms step:291/6250 train_loss:3.5094 aux_loss:0.0517 train_time:1100529ms step_avg:3916.47ms step:292/6250 train_loss:3.3949 aux_loss:0.0502 train_time:1103606ms step_avg:3913.50ms step:293/6250 train_loss:3.4112 aux_loss:0.0512 train_time:1105208ms step_avg:3905.33ms step:294/6250 train_loss:3.3810 aux_loss:0.0520 train_time:1106074ms step_avg:3894.63ms step:295/6250 train_loss:3.4444 aux_loss:0.0512 train_time:1107433ms step_avg:3885.73ms step:296/6250 train_loss:3.4961 aux_loss:0.0507 train_time:1121542ms step_avg:3921.47ms step:297/6250 train_loss:3.3923 aux_loss:0.0517 train_time:1122640ms step_avg:3911.64ms step:298/6250 train_loss:3.4715 aux_loss:0.0524 train_time:1123577ms step_avg:3901.31ms step:299/6250 train_loss:3.4627 aux_loss:0.0495 train_time:1124624ms step_avg:3891.43ms step:300/6250 train_loss:3.3386 aux_loss:0.0517 train_time:1125750ms step_avg:3881.90ms step:301/6250 train_loss:3.4552 aux_loss:0.0515 train_time:1126701ms step_avg:3871.83ms step:302/6250 train_loss:3.4597 aux_loss:0.0503 train_time:1127680ms step_avg:3861.92ms step:303/6250 train_loss:3.4786 aux_loss:0.0510 train_time:1128883ms step_avg:3852.84ms step:304/6250 train_loss:3.3740 aux_loss:0.0503 train_time:1129926ms step_avg:3843.28ms step:305/6250 train_loss:3.2869 aux_loss:0.0505 train_time:1130813ms step_avg:3833.26ms step:306/6250 train_loss:3.3938 aux_loss:0.0522 train_time:1131750ms step_avg:3823.48ms step:307/6250 train_loss:3.3974 aux_loss:0.0513 train_time:1132737ms step_avg:3813.93ms step:308/6250 train_loss:3.5051 aux_loss:0.0502 train_time:1133829ms step_avg:3804.79ms step:309/6250 train_loss:3.2689 aux_loss:0.0510 train_time:1135029ms step_avg:3796.08ms step:310/6250 train_loss:3.3691 aux_loss:0.0513 train_time:1146692ms step_avg:3822.31ms step:311/6250 train_loss:3.3348 aux_loss:0.0503 train_time:1147836ms step_avg:3813.41ms step:312/6250 train_loss:3.3992 aux_loss:0.0509 train_time:1149394ms step_avg:3805.94ms step:313/6250 train_loss:3.5519 aux_loss:0.0514 train_time:1156735ms step_avg:3817.61ms step:314/6250 train_loss:3.3504 aux_loss:0.0502 train_time:1157563ms step_avg:3807.77ms step:315/6250 train_loss:3.4849 aux_loss:0.0508 train_time:1158417ms step_avg:3798.09ms step:316/6250 train_loss:3.3125 aux_loss:0.0507 train_time:1159264ms step_avg:3788.45ms step:317/6250 train_loss:3.4298 aux_loss:0.0512 train_time:1160094ms step_avg:3778.81ms step:318/6250 train_loss:3.3369 aux_loss:0.0500 train_time:1160946ms step_avg:3769.31ms step:319/6250 train_loss:3.3376 aux_loss:0.0508 train_time:1161791ms step_avg:3759.84ms step:320/6250 train_loss:3.3840 aux_loss:0.0513 train_time:1162642ms step_avg:3750.46ms step:321/6250 train_loss:3.2925 aux_loss:0.0508 train_time:1163470ms step_avg:3741.06ms step:322/6250 train_loss:3.3678 aux_loss:0.0504 train_time:1164300ms step_avg:3731.73ms step:323/6250 train_loss:3.3092 aux_loss:0.0510 train_time:1165159ms step_avg:3722.55ms step:324/6250 train_loss:3.4200 aux_loss:0.0503 train_time:1165999ms step_avg:3713.37ms step:325/6250 train_loss:3.3778 aux_loss:0.0499 train_time:1166846ms step_avg:3704.27ms step:326/6250 train_loss:3.4117 aux_loss:0.0502 train_time:1169061ms step_avg:3699.56ms step:327/6250 train_loss:3.3071 aux_loss:0.0502 train_time:1170781ms step_avg:3693.32ms step:328/6250 train_loss:3.3998 aux_loss:0.0500 train_time:1183232ms step_avg:3720.86ms step:329/6250 train_loss:3.3176 aux_loss:0.0503 train_time:1184832ms step_avg:3714.21ms step:330/6250 train_loss:3.3591 aux_loss:0.0505 train_time:1190082ms step_avg:3719.01ms step:331/6250 train_loss:3.2376 aux_loss:0.0504 train_time:1191238ms step_avg:3711.02ms step:332/6250 train_loss:3.3870 aux_loss:0.0507 train_time:1192905ms step_avg:3704.68ms step:333/6250 train_loss:3.3560 aux_loss:0.0504 train_time:1194678ms step_avg:3698.69ms step:334/6250 train_loss:3.3404 aux_loss:0.0495 train_time:1196252ms step_avg:3692.14ms step:335/6250 train_loss:3.2373 aux_loss:0.0509 train_time:1201966ms step_avg:3698.36ms step:336/6250 train_loss:3.3125 aux_loss:0.0500 train_time:1203508ms step_avg:3691.74ms step:337/6250 train_loss:3.4087 aux_loss:0.0508 train_time:1204905ms step_avg:3684.73ms step:338/6250 train_loss:3.2524 aux_loss:0.0501 train_time:1207097ms step_avg:3680.17ms step:339/6250 train_loss:3.4597 aux_loss:0.0507 train_time:1208489ms step_avg:3673.22ms step:340/6250 train_loss:3.2823 aux_loss:0.0503 train_time:1209981ms step_avg:3666.61ms step:341/6250 train_loss:3.3403 aux_loss:0.0500 train_time:1211561ms step_avg:3660.30ms step:342/6250 train_loss:3.2238 aux_loss:0.0516 train_time:1213051ms step_avg:3653.77ms step:343/6250 train_loss:3.3059 aux_loss:0.0514 train_time:1214522ms step_avg:3647.21ms step:344/6250 train_loss:3.4193 aux_loss:0.0510 train_time:1216448ms step_avg:3642.06ms step:345/6250 train_loss:3.3194 aux_loss:0.0504 train_time:1217765ms step_avg:3635.12ms step:346/6250 train_loss:3.2138 aux_loss:0.0501 train_time:1219321ms step_avg:3628.93ms step:347/6250 train_loss:3.2711 aux_loss:0.0498 train_time:1220729ms step_avg:3622.34ms step:348/6250 train_loss:3.3009 aux_loss:0.0502 train_time:1223904ms step_avg:3621.02ms step:349/6250 train_loss:3.1748 aux_loss:0.0503 train_time:1225414ms step_avg:3614.79ms step:350/6250 train_loss:3.3501 aux_loss:0.0497 train_time:1232887ms step_avg:3626.14ms step:351/6250 train_loss:3.2170 aux_loss:0.0509 train_time:1234258ms step_avg:3619.52ms step:352/6250 train_loss:3.2530 aux_loss:0.0509 train_time:1235832ms step_avg:3613.54ms step:353/6250 train_loss:3.2734 aux_loss:0.0497 train_time:1237278ms step_avg:3607.22ms step:354/6250 train_loss:3.1405 aux_loss:0.0502 train_time:1238658ms step_avg:3600.75ms step:355/6250 train_loss:3.3553 aux_loss:0.0502 train_time:1240303ms step_avg:3595.08ms step:356/6250 train_loss:3.2231 aux_loss:0.0501 train_time:1241586ms step_avg:3588.40ms step:357/6250 train_loss:3.3100 aux_loss:0.0501 train_time:1243157ms step_avg:3582.58ms step:358/6250 train_loss:3.2793 aux_loss:0.0506 train_time:1250683ms step_avg:3593.92ms step:359/6250 train_loss:3.1787 aux_loss:0.0498 train_time:1252281ms step_avg:3588.20ms step:360/6250 train_loss:3.3081 aux_loss:0.0502 train_time:1253506ms step_avg:3581.45ms step:361/6250 train_loss:3.1641 aux_loss:0.0504 train_time:1254752ms step_avg:3574.79ms step:362/6250 train_loss:3.3008 aux_loss:0.0497 train_time:1256004ms step_avg:3568.19ms step:363/6250 train_loss:3.2472 aux_loss:0.0496 train_time:1257186ms step_avg:3561.43ms step:364/6250 train_loss:3.1824 aux_loss:0.0503 train_time:1258398ms step_avg:3554.80ms step:365/6250 train_loss:3.1201 aux_loss:0.0506 train_time:1259612ms step_avg:3548.20ms step:366/6250 train_loss:3.1301 aux_loss:0.0505 train_time:1260829ms step_avg:3541.65ms step:367/6250 train_loss:3.2720 aux_loss:0.0500 train_time:1262006ms step_avg:3535.03ms step:368/6250 train_loss:3.2253 aux_loss:0.0507 train_time:1263194ms step_avg:3528.48ms step:369/6250 train_loss:3.2245 aux_loss:0.0502 train_time:1264443ms step_avg:3522.12ms step:370/6250 train_loss:3.2134 aux_loss:0.0514 train_time:1265750ms step_avg:3515.97ms step:371/6250 train_loss:3.1930 aux_loss:0.0497 train_time:1267008ms step_avg:3509.72ms step:372/6250 train_loss:3.2492 aux_loss:0.0508 train_time:1268253ms step_avg:3503.46ms step:373/6250 train_loss:3.2162 aux_loss:0.0496 train_time:1269463ms step_avg:3497.14ms step:374/6250 train_loss:3.1190 aux_loss:0.0497 train_time:1270697ms step_avg:3490.93ms step:375/6250 train_loss:3.1948 aux_loss:0.0505 train_time:1271889ms step_avg:3484.63ms step:376/6250 train_loss:3.1620 aux_loss:0.0500 train_time:1273326ms step_avg:3479.03ms step:377/6250 train_loss:3.1655 aux_loss:0.0491 train_time:1274605ms step_avg:3473.04ms step:378/6250 train_loss:3.2662 aux_loss:0.0492 train_time:1275954ms step_avg:3467.27ms step:379/6250 train_loss:3.2340 aux_loss:0.0502 train_time:1277813ms step_avg:3462.91ms step:380/6250 train_loss:3.2682 aux_loss:0.0501 train_time:1279381ms step_avg:3457.79ms step:381/6250 train_loss:3.1775 aux_loss:0.0497 train_time:1280911ms step_avg:3452.59ms step:382/6250 train_loss:3.2768 aux_loss:0.0497 train_time:1282488ms step_avg:3447.55ms step:383/6250 train_loss:3.1648 aux_loss:0.0501 train_time:1284026ms step_avg:3442.43ms step:384/6250 train_loss:3.2128 aux_loss:0.0495 train_time:1285299ms step_avg:3436.63ms step:385/6250 train_loss:3.2819 aux_loss:0.0492 train_time:1286488ms step_avg:3430.64ms step:386/6250 train_loss:3.1653 aux_loss:0.0495 train_time:1287723ms step_avg:3424.79ms step:387/6250 train_loss:3.0554 aux_loss:0.0496 train_time:1296436ms step_avg:3438.82ms step:388/6250 train_loss:3.1640 aux_loss:0.0499 train_time:1297885ms step_avg:3433.56ms step:389/6250 train_loss:3.1420 aux_loss:0.0496 train_time:1303180ms step_avg:3438.47ms step:390/6250 train_loss:3.1320 aux_loss:0.0507 train_time:1307936ms step_avg:3441.94ms step:391/6250 train_loss:3.1684 aux_loss:0.0499 train_time:1309304ms step_avg:3436.49ms step:392/6250 train_loss:3.1374 aux_loss:0.0496 train_time:1310773ms step_avg:3431.34ms step:393/6250 train_loss:3.1703 aux_loss:0.0501 train_time:1312266ms step_avg:3426.28ms step:394/6250 train_loss:3.1370 aux_loss:0.0499 train_time:1313718ms step_avg:3421.14ms step:395/6250 train_loss:3.1424 aux_loss:0.0497 train_time:1315110ms step_avg:3415.87ms step:396/6250 train_loss:3.0410 aux_loss:0.0498 train_time:1316412ms step_avg:3410.39ms step:397/6250 train_loss:3.2253 aux_loss:0.0497 train_time:1318018ms step_avg:3405.73ms step:398/6250 train_loss:3.0648 aux_loss:0.0494 train_time:1319702ms step_avg:3401.29ms step:399/6250 train_loss:3.1196 aux_loss:0.0496 train_time:1333647ms step_avg:3428.40ms step:400/6250 train_loss:3.0897 aux_loss:0.0496 train_time:1339916ms step_avg:3435.68ms step:401/6250 train_loss:3.1308 aux_loss:0.0493 train_time:1341761ms step_avg:3431.61ms step:402/6250 train_loss:3.1462 aux_loss:0.0489 train_time:1351386ms step_avg:3447.41ms step:403/6250 train_loss:3.0847 aux_loss:0.0506 train_time:1359919ms step_avg:3460.35ms step:404/6250 train_loss:3.2342 aux_loss:0.0488 train_time:1360988ms step_avg:3454.29ms step:405/6250 train_loss:3.0780 aux_loss:0.0489 train_time:1361964ms step_avg:3448.01ms step:406/6250 train_loss:3.0915 aux_loss:0.0501 train_time:1366852ms step_avg:3451.65ms step:407/6250 train_loss:3.1201 aux_loss:0.0498 train_time:1367783ms step_avg:3445.30ms step:408/6250 train_loss:3.0519 aux_loss:0.0499 train_time:1368762ms step_avg:3439.10ms step:409/6250 train_loss:3.0870 aux_loss:0.0497 train_time:1369732ms step_avg:3432.91ms step:410/6250 train_loss:3.0701 aux_loss:0.0498 train_time:1370679ms step_avg:3426.70ms step:411/6250 train_loss:3.1656 aux_loss:0.0497 train_time:1371600ms step_avg:3420.45ms step:412/6250 train_loss:3.1872 aux_loss:0.0499 train_time:1372515ms step_avg:3414.22ms step:413/6250 train_loss:3.1751 aux_loss:0.0487 train_time:1373442ms step_avg:3408.05ms step:414/6250 train_loss:3.0827 aux_loss:0.0497 train_time:1374341ms step_avg:3401.83ms step:415/6250 train_loss:3.1557 aux_loss:0.0493 train_time:1375257ms step_avg:3395.70ms step:416/6250 train_loss:3.0391 aux_loss:0.0491 train_time:1376135ms step_avg:3389.50ms step:417/6250 train_loss:3.1017 aux_loss:0.0490 train_time:1377118ms step_avg:3383.58ms step:418/6250 train_loss:3.1414 aux_loss:0.0491 train_time:1378050ms step_avg:3377.57ms step:419/6250 train_loss:3.1502 aux_loss:0.0489 train_time:1378900ms step_avg:3371.39ms step:420/6250 train_loss:3.0928 aux_loss:0.0491 train_time:1379853ms step_avg:3365.49ms step:421/6250 train_loss:3.0117 aux_loss:0.0492 train_time:1380783ms step_avg:3359.57ms step:422/6250 train_loss:2.9904 aux_loss:0.0496 train_time:1381826ms step_avg:3353.95ms step:423/6250 train_loss:3.1461 aux_loss:0.0491 train_time:1384347ms step_avg:3351.93ms step:424/6250 train_loss:3.0961 aux_loss:0.0492 train_time:1390658ms step_avg:3359.08ms step:425/6250 train_loss:3.0950 aux_loss:0.0486 train_time:1392676ms step_avg:3355.85ms step:426/6250 train_loss:3.1386 aux_loss:0.0488 train_time:1393534ms step_avg:3349.84ms step:427/6250 train_loss:3.1074 aux_loss:0.0488 train_time:1394407ms step_avg:3343.90ms step:428/6250 train_loss:3.1083 aux_loss:0.0492 train_time:1395293ms step_avg:3338.02ms step:429/6250 train_loss:3.0101 aux_loss:0.0493 train_time:1396170ms step_avg:3332.15ms step:430/6250 train_loss:3.0261 aux_loss:0.0496 train_time:1397057ms step_avg:3326.33ms step:431/6250 train_loss:3.0115 aux_loss:0.0493 train_time:1397979ms step_avg:3320.62ms step:432/6250 train_loss:2.9807 aux_loss:0.0491 train_time:1399062ms step_avg:3315.31ms step:433/6250 train_loss:3.0679 aux_loss:0.0492 train_time:1399994ms step_avg:3309.68ms step:434/6250 train_loss:3.0749 aux_loss:0.0492 train_time:1400905ms step_avg:3304.02ms step:435/6250 train_loss:3.0854 aux_loss:0.0489 train_time:1401830ms step_avg:3298.42ms step:436/6250 train_loss:2.9969 aux_loss:0.0492 train_time:1406685ms step_avg:3302.08ms step:437/6250 train_loss:3.1384 aux_loss:0.0490 train_time:1407612ms step_avg:3296.52ms step:438/6250 train_loss:3.0586 aux_loss:0.0494 train_time:1408561ms step_avg:3291.03ms step:439/6250 train_loss:3.1054 aux_loss:0.0481 train_time:1409476ms step_avg:3285.49ms step:440/6250 train_loss:3.0277 aux_loss:0.0498 train_time:1410413ms step_avg:3280.03ms step:441/6250 train_loss:3.0366 aux_loss:0.0494 train_time:1411376ms step_avg:3274.65ms step:442/6250 train_loss:3.1881 aux_loss:0.0480 train_time:1412322ms step_avg:3269.26ms step:443/6250 train_loss:3.0881 aux_loss:0.0487 train_time:1413224ms step_avg:3263.80ms step:444/6250 train_loss:3.1104 aux_loss:0.0488 train_time:1414157ms step_avg:3258.43ms step:445/6250 train_loss:2.9419 aux_loss:0.0492 train_time:1415190ms step_avg:3253.31ms step:446/6250 train_loss:3.0860 aux_loss:0.0486 train_time:1416051ms step_avg:3247.82ms step:447/6250 train_loss:3.0582 aux_loss:0.0487 train_time:1416906ms step_avg:3242.35ms step:448/6250 train_loss:3.0249 aux_loss:0.0489 train_time:1417797ms step_avg:3236.98ms step:449/6250 train_loss:3.0823 aux_loss:0.0483 train_time:1418657ms step_avg:3231.57ms step:450/6250 train_loss:3.0322 aux_loss:0.0488 train_time:1419546ms step_avg:3226.24ms step:451/6250 train_loss:2.9660 aux_loss:0.0477 train_time:1420447ms step_avg:3220.97ms step:452/6250 train_loss:2.9778 aux_loss:0.0493 train_time:1421344ms step_avg:3215.71ms step:453/6250 train_loss:3.0312 aux_loss:0.0490 train_time:1422323ms step_avg:3210.66ms step:454/6250 train_loss:2.9606 aux_loss:0.0485 train_time:1423304ms step_avg:3205.64ms step:455/6250 train_loss:3.0750 aux_loss:0.0483 train_time:1424240ms step_avg:3200.54ms step:456/6250 train_loss:3.0860 aux_loss:0.0481 train_time:1425128ms step_avg:3195.35ms step:457/6250 train_loss:2.9244 aux_loss:0.0491 train_time:1426065ms step_avg:3190.30ms step:458/6250 train_loss:3.0265 aux_loss:0.0487 train_time:1427016ms step_avg:3185.30ms step:459/6250 train_loss:2.9723 aux_loss:0.0483 train_time:1428012ms step_avg:3180.43ms step:460/6250 train_loss:3.0593 aux_loss:0.0486 train_time:1428956ms step_avg:3175.46ms step:461/6250 train_loss:2.9907 aux_loss:0.0484 train_time:1429866ms step_avg:3170.44ms step:462/6250 train_loss:3.0048 aux_loss:0.0483 train_time:1430755ms step_avg:3165.39ms step:463/6250 train_loss:3.0376 aux_loss:0.0486 train_time:1431708ms step_avg:3160.50ms step:464/6250 train_loss:3.0044 aux_loss:0.0482 train_time:1432690ms step_avg:3155.70ms step:465/6250 train_loss:2.9987 aux_loss:0.0481 train_time:1433679ms step_avg:3150.94ms step:466/6250 train_loss:2.9122 aux_loss:0.0484 train_time:1434572ms step_avg:3145.99ms step:467/6250 train_loss:2.8957 aux_loss:0.0486 train_time:1435505ms step_avg:3141.15ms step:468/6250 train_loss:2.9803 aux_loss:0.0482 train_time:1436413ms step_avg:3136.27ms step:469/6250 train_loss:2.9685 aux_loss:0.0481 train_time:1437362ms step_avg:3131.51ms step:470/6250 train_loss:3.0313 aux_loss:0.0485 train_time:1438257ms step_avg:3126.65ms step:471/6250 train_loss:2.9834 aux_loss:0.0483 train_time:1439207ms step_avg:3121.92ms step:472/6250 train_loss:2.8936 aux_loss:0.0480 train_time:1440173ms step_avg:3117.26ms step:473/6250 train_loss:2.9922 aux_loss:0.0488 train_time:1441103ms step_avg:3112.53ms step:474/6250 train_loss:2.9851 aux_loss:0.0484 train_time:1442058ms step_avg:3107.88ms step:475/6250 train_loss:2.9790 aux_loss:0.0485 train_time:1442998ms step_avg:3103.22ms step:476/6250 train_loss:2.9637 aux_loss:0.0480 train_time:1443972ms step_avg:3098.65ms step:477/6250 train_loss:2.9588 aux_loss:0.0478 train_time:1444910ms step_avg:3094.03ms step:478/6250 train_loss:2.9156 aux_loss:0.0476 train_time:1445894ms step_avg:3089.52ms step:479/6250 train_loss:2.9222 aux_loss:0.0478 train_time:1446847ms step_avg:3084.96ms step:480/6250 train_loss:2.9798 aux_loss:0.0482 train_time:1447743ms step_avg:3080.30ms step:481/6250 train_loss:2.9393 aux_loss:0.0486 train_time:1448610ms step_avg:3075.61ms step:482/6250 train_loss:2.9713 aux_loss:0.0479 train_time:1449533ms step_avg:3071.04ms step:483/6250 train_loss:3.0382 aux_loss:0.0481 train_time:1450399ms step_avg:3066.38ms step:484/6250 train_loss:3.0352 aux_loss:0.0481 train_time:1451269ms step_avg:3061.75ms step:485/6250 train_loss:2.8951 aux_loss:0.0485 train_time:1452165ms step_avg:3057.19ms step:486/6250 train_loss:3.0568 aux_loss:0.0475 train_time:1453026ms step_avg:3052.57ms step:487/6250 train_loss:2.9483 aux_loss:0.0483 train_time:1453885ms step_avg:3047.98ms step:488/6250 train_loss:2.9646 aux_loss:0.0479 train_time:1454785ms step_avg:3043.48ms step:489/6250 train_loss:2.9283 aux_loss:0.0479 train_time:1455738ms step_avg:3039.12ms step:490/6250 train_loss:2.9237 aux_loss:0.0481 train_time:1457156ms step_avg:3035.74ms step:491/6250 train_loss:2.8942 aux_loss:0.0479 train_time:1458907ms step_avg:3033.07ms step:492/6250 train_loss:2.9621 aux_loss:0.0471 train_time:1460711ms step_avg:3030.52ms step:493/6250 train_loss:2.9590 aux_loss:0.0476 train_time:1461943ms step_avg:3026.80ms step:494/6250 train_loss:2.9098 aux_loss:0.0478 train_time:1463426ms step_avg:3023.61ms step:495/6250 train_loss:2.9224 aux_loss:0.0476 train_time:1464540ms step_avg:3019.67ms step:496/6250 train_loss:2.8687 aux_loss:0.0473 train_time:1465759ms step_avg:3015.97ms step:497/6250 train_loss:2.9409 aux_loss:0.0485 train_time:1466741ms step_avg:3011.79ms step:498/6250 train_loss:2.8587 aux_loss:0.0477 train_time:1467695ms step_avg:3007.57ms step:499/6250 train_loss:2.9637 aux_loss:0.0479 train_time:1468740ms step_avg:3003.56ms step:500/6250 train_loss:2.9355 aux_loss:0.0474 train_time:1470095ms step_avg:3000.19ms step:501/6250 train_loss:2.8624 aux_loss:0.0472 train_time:1471250ms step_avg:2996.44ms step:502/6250 train_loss:3.0373 aux_loss:0.0471 train_time:1472271ms step_avg:2992.42ms step:503/6250 train_loss:2.9247 aux_loss:0.0475 train_time:1473169ms step_avg:2988.17ms step:504/6250 train_loss:2.9826 aux_loss:0.0472 train_time:1474025ms step_avg:2983.86ms step:505/6250 train_loss:2.9452 aux_loss:0.0476 train_time:1474921ms step_avg:2979.64ms step:506/6250 train_loss:2.9551 aux_loss:0.0479 train_time:1475825ms step_avg:2975.45ms step:507/6250 train_loss:2.9185 aux_loss:0.0474 train_time:1476697ms step_avg:2971.22ms step:508/6250 train_loss:2.9441 aux_loss:0.0477 train_time:1477545ms step_avg:2966.96ms step:509/6250 train_loss:3.0217 aux_loss:0.0476 train_time:1478504ms step_avg:2962.93ms step:510/6250 train_loss:3.0070 aux_loss:0.0469 train_time:1480012ms step_avg:2960.02ms step:511/6250 train_loss:2.8994 aux_loss:0.0470 train_time:1480898ms step_avg:2955.88ms step:512/6250 train_loss:3.0133 aux_loss:0.0462 train_time:1481997ms step_avg:2952.18ms step:513/6250 train_loss:2.9029 aux_loss:0.0476 train_time:1483043ms step_avg:2948.40ms step:514/6250 train_loss:2.9467 aux_loss:0.0472 train_time:1485553ms step_avg:2947.53ms step:515/6250 train_loss:2.9678 aux_loss:0.0468 train_time:1486635ms step_avg:2943.83ms step:516/6250 train_loss:2.9434 aux_loss:0.0471 train_time:1487486ms step_avg:2939.70ms step:517/6250 train_loss:2.8504 aux_loss:0.0473 train_time:1488345ms step_avg:2935.59ms step:518/6250 train_loss:2.9463 aux_loss:0.0473 train_time:1489187ms step_avg:2931.47ms step:519/6250 train_loss:2.9265 aux_loss:0.0474 train_time:1490045ms step_avg:2927.40ms step:520/6250 train_loss:2.9381 aux_loss:0.0470 train_time:1490891ms step_avg:2923.32ms step:521/6250 train_loss:2.9267 aux_loss:0.0474 train_time:1491734ms step_avg:2919.25ms step:522/6250 train_loss:2.8690 aux_loss:0.0467 train_time:1492598ms step_avg:2915.23ms step:523/6250 train_loss:2.9032 aux_loss:0.0468 train_time:1493443ms step_avg:2911.20ms step:524/6250 train_loss:2.9159 aux_loss:0.0470 train_time:1494290ms step_avg:2907.18ms step:525/6250 train_loss:2.8525 aux_loss:0.0466 train_time:1495152ms step_avg:2903.21ms step:526/6250 train_loss:2.8731 aux_loss:0.0467 train_time:1496002ms step_avg:2899.23ms step:527/6250 train_loss:2.8873 aux_loss:0.0471 train_time:1496874ms step_avg:2895.31ms step:528/6250 train_loss:2.9176 aux_loss:0.0467 train_time:1497764ms step_avg:2891.44ms step:529/6250 train_loss:2.9267 aux_loss:0.0470 train_time:1498953ms step_avg:2888.16ms step:530/6250 train_loss:2.8844 aux_loss:0.0472 train_time:1499847ms step_avg:2884.32ms step:531/6250 train_loss:2.8815 aux_loss:0.0469 train_time:1500767ms step_avg:2880.55ms step:532/6250 train_loss:2.7962 aux_loss:0.0471 train_time:1505024ms step_avg:2883.19ms step:533/6250 train_loss:2.8269 aux_loss:0.0467 train_time:1506584ms step_avg:2880.66ms step:534/6250 train_loss:2.8414 aux_loss:0.0466 train_time:1507465ms step_avg:2876.84ms step:535/6250 train_loss:2.8953 aux_loss:0.0467 train_time:1510137ms step_avg:2876.45ms step:536/6250 train_loss:2.8131 aux_loss:0.0464 train_time:1511375ms step_avg:2873.34ms step:537/6250 train_loss:2.8160 aux_loss:0.0469 train_time:1513176ms step_avg:2871.30ms step:538/6250 train_loss:2.8556 aux_loss:0.0471 train_time:1514075ms step_avg:2867.57ms step:539/6250 train_loss:2.8816 aux_loss:0.0466 train_time:1514981ms step_avg:2863.86ms step:540/6250 train_loss:2.8876 aux_loss:0.0466 train_time:1515880ms step_avg:2860.15ms step:541/6250 train_loss:2.8358 aux_loss:0.0465 train_time:1516774ms step_avg:2856.45ms step:542/6250 train_loss:2.8252 aux_loss:0.0466 train_time:1517888ms step_avg:2853.17ms step:543/6250 train_loss:2.9481 aux_loss:0.0470 train_time:1518816ms step_avg:2849.56ms step:544/6250 train_loss:2.8186 aux_loss:0.0465 train_time:1519700ms step_avg:2845.88ms step:545/6250 train_loss:2.9179 aux_loss:0.0462 train_time:1520537ms step_avg:2842.13ms step:546/6250 train_loss:2.7116 aux_loss:0.0463 train_time:1521430ms step_avg:2838.49ms step:547/6250 train_loss:2.8072 aux_loss:0.0462 train_time:1522283ms step_avg:2834.79ms step:548/6250 train_loss:2.8951 aux_loss:0.0468 train_time:1523118ms step_avg:2831.08ms step:549/6250 train_loss:2.9218 aux_loss:0.0464 train_time:1523978ms step_avg:2827.42ms step:550/6250 train_loss:2.8687 aux_loss:0.0462 train_time:1524827ms step_avg:2823.75ms step:551/6250 train_loss:2.9104 aux_loss:0.0463 train_time:1525657ms step_avg:2820.07ms step:552/6250 train_loss:2.8019 aux_loss:0.0461 train_time:1526494ms step_avg:2816.41ms step:553/6250 train_loss:2.8722 aux_loss:0.0461 train_time:1527360ms step_avg:2812.82ms step:554/6250 train_loss:2.9270 aux_loss:0.0461 train_time:1528222ms step_avg:2809.23ms step:555/6250 train_loss:2.8239 aux_loss:0.0464 train_time:1529081ms step_avg:2805.65ms step:556/6250 train_loss:2.8192 aux_loss:0.0463 train_time:1529918ms step_avg:2802.05ms step:557/6250 train_loss:2.9051 aux_loss:0.0457 train_time:1530766ms step_avg:2798.48ms step:558/6250 train_loss:2.7622 aux_loss:0.0458 train_time:1531609ms step_avg:2794.91ms step:559/6250 train_loss:2.8232 aux_loss:0.0464 train_time:1532459ms step_avg:2791.36ms step:560/6250 train_loss:2.8694 aux_loss:0.0457 train_time:1534173ms step_avg:2789.41ms step:561/6250 train_loss:2.8904 aux_loss:0.0457 train_time:1535031ms step_avg:2785.90ms step:562/6250 train_loss:2.9081 aux_loss:0.0454 train_time:1535875ms step_avg:2782.38ms step:563/6250 train_loss:2.7986 aux_loss:0.0460 train_time:1536731ms step_avg:2778.90ms step:564/6250 train_loss:2.8054 aux_loss:0.0460 train_time:1537757ms step_avg:2775.74ms step:565/6250 train_loss:2.8214 aux_loss:0.0463 train_time:1538697ms step_avg:2772.43ms step:566/6250 train_loss:2.7382 aux_loss:0.0462 train_time:1539711ms step_avg:2769.26ms step:567/6250 train_loss:2.8519 aux_loss:0.0458 train_time:1540546ms step_avg:2765.79ms step:568/6250 train_loss:2.8276 aux_loss:0.0462 train_time:1541407ms step_avg:2762.38ms step:569/6250 train_loss:2.8315 aux_loss:0.0458 train_time:1542266ms step_avg:2758.97ms step:570/6250 train_loss:2.8400 aux_loss:0.0462 train_time:1543112ms step_avg:2755.56ms step:571/6250 train_loss:2.8556 aux_loss:0.0455 train_time:1543968ms step_avg:2752.17ms step:572/6250 train_loss:2.8367 aux_loss:0.0462 train_time:1544809ms step_avg:2748.77ms step:573/6250 train_loss:2.7432 aux_loss:0.0462 train_time:1545658ms step_avg:2745.40ms step:574/6250 train_loss:2.7787 aux_loss:0.0457 train_time:1546501ms step_avg:2742.02ms step:575/6250 train_loss:2.8057 aux_loss:0.0459 train_time:1547359ms step_avg:2738.69ms step:576/6250 train_loss:2.7557 aux_loss:0.0460 train_time:1548213ms step_avg:2735.36ms step:577/6250 train_loss:2.7797 aux_loss:0.0455 train_time:1549061ms step_avg:2732.03ms step:578/6250 train_loss:2.7410 aux_loss:0.0460 train_time:1549909ms step_avg:2728.71ms step:579/6250 train_loss:2.7258 aux_loss:0.0464 train_time:1550767ms step_avg:2725.43ms step:580/6250 train_loss:2.7519 aux_loss:0.0459 train_time:1551603ms step_avg:2722.11ms step:581/6250 train_loss:2.7665 aux_loss:0.0458 train_time:1552471ms step_avg:2718.86ms step:582/6250 train_loss:2.8485 aux_loss:0.0458 train_time:1553314ms step_avg:2715.58ms step:583/6250 train_loss:2.8634 aux_loss:0.0456 train_time:1554171ms step_avg:2712.34ms step:584/6250 train_loss:2.8693 aux_loss:0.0454 train_time:1555014ms step_avg:2709.08ms step:585/6250 train_loss:2.7143 aux_loss:0.0454 train_time:1555865ms step_avg:2705.85ms step:586/6250 train_loss:2.8392 aux_loss:0.0454 train_time:1556725ms step_avg:2702.65ms step:587/6250 train_loss:2.8761 aux_loss:0.0457 train_time:1557584ms step_avg:2699.45ms step:588/6250 train_loss:2.8508 aux_loss:0.0453 train_time:1558433ms step_avg:2696.25ms step:589/6250 train_loss:2.7959 aux_loss:0.0455 train_time:1559270ms step_avg:2693.04ms step:590/6250 train_loss:2.8007 aux_loss:0.0454 train_time:1560120ms step_avg:2689.86ms step:591/6250 train_loss:2.7776 aux_loss:0.0457 train_time:1560978ms step_avg:2686.71ms step:592/6250 train_loss:2.8533 aux_loss:0.0454 train_time:1561838ms step_avg:2683.57ms step:593/6250 train_loss:2.8095 aux_loss:0.0454 train_time:1562692ms step_avg:2680.43ms step:594/6250 train_loss:2.7319 aux_loss:0.0452 train_time:1563530ms step_avg:2677.28ms step:595/6250 train_loss:2.7753 aux_loss:0.0451 train_time:1564382ms step_avg:2674.16ms step:596/6250 train_loss:2.8722 aux_loss:0.0448 train_time:1565225ms step_avg:2671.03ms step:597/6250 train_loss:2.8881 aux_loss:0.0448 train_time:1566070ms step_avg:2667.92ms step:598/6250 train_loss:2.7749 aux_loss:0.0449 train_time:1566981ms step_avg:2664.93ms step:599/6250 train_loss:2.6990 aux_loss:0.0452 train_time:1567823ms step_avg:2661.84ms step:600/6250 train_loss:2.6521 aux_loss:0.0455 train_time:1568699ms step_avg:2658.81ms step:601/6250 train_loss:2.7595 aux_loss:0.0452 train_time:1569554ms step_avg:2655.76ms step:602/6250 train_loss:2.7805 aux_loss:0.0454 train_time:1570427ms step_avg:2652.75ms step:603/6250 train_loss:2.7799 aux_loss:0.0448 train_time:1571284ms step_avg:2649.72ms step:604/6250 train_loss:2.8004 aux_loss:0.0451 train_time:1572124ms step_avg:2646.67ms step:605/6250 train_loss:2.6938 aux_loss:0.0455 train_time:1572984ms step_avg:2643.67ms step:606/6250 train_loss:2.8118 aux_loss:0.0449 train_time:1573835ms step_avg:2640.66ms step:607/6250 train_loss:2.7710 aux_loss:0.0446 train_time:1576201ms step_avg:2640.20ms step:608/6250 train_loss:2.6884 aux_loss:0.0448 train_time:1577293ms step_avg:2637.61ms step:609/6250 train_loss:2.7879 aux_loss:0.0442 train_time:1579247ms step_avg:2636.47ms step:610/6250 train_loss:2.6641 aux_loss:0.0448 train_time:1580607ms step_avg:2634.35ms step:611/6250 train_loss:2.7543 aux_loss:0.0450 train_time:1581618ms step_avg:2631.64ms step:612/6250 train_loss:2.7907 aux_loss:0.0450 train_time:1583239ms step_avg:2629.96ms step:613/6250 train_loss:2.8265 aux_loss:0.0446 train_time:1585581ms step_avg:2629.49ms step:614/6250 train_loss:2.7364 aux_loss:0.0444 train_time:1587132ms step_avg:2627.70ms step:615/6250 train_loss:2.7539 aux_loss:0.0449 train_time:1588635ms step_avg:2625.84ms step:616/6250 train_loss:2.7025 aux_loss:0.0449 train_time:1589499ms step_avg:2622.94ms step:617/6250 train_loss:2.8004 aux_loss:0.0445 train_time:1590431ms step_avg:2620.15ms step:618/6250 train_loss:2.6636 aux_loss:0.0444 train_time:1591371ms step_avg:2617.39ms step:619/6250 train_loss:2.6868 aux_loss:0.0448 train_time:1592330ms step_avg:2614.66ms step:620/6250 train_loss:2.7629 aux_loss:0.0448 train_time:1593277ms step_avg:2611.93ms step:621/6250 train_loss:2.8405 aux_loss:0.0446 train_time:1594194ms step_avg:2609.16ms step:622/6250 train_loss:2.6576 aux_loss:0.0446 train_time:1595128ms step_avg:2606.42ms step:623/6250 train_loss:2.7784 aux_loss:0.0443 train_time:1595979ms step_avg:2603.55ms step:624/6250 train_loss:2.7637 aux_loss:0.0445 train_time:1596838ms step_avg:2600.71ms step:625/6250 train_loss:2.7188 aux_loss:0.0450 train_time:1597715ms step_avg:2597.91ms step:626/6250 train_loss:2.7660 aux_loss:0.0443 train_time:1598638ms step_avg:2595.19ms step:627/6250 train_loss:2.7566 aux_loss:0.0444 train_time:1599594ms step_avg:2592.54ms step:628/6250 train_loss:2.7211 aux_loss:0.0444 train_time:1600555ms step_avg:2589.89ms step:629/6250 train_loss:2.7733 aux_loss:0.0443 train_time:1601482ms step_avg:2587.21ms step:630/6250 train_loss:2.7866 aux_loss:0.0442 train_time:1602456ms step_avg:2584.61ms step:631/6250 train_loss:2.7609 aux_loss:0.0441 train_time:1603363ms step_avg:2581.90ms step:632/6250 train_loss:2.7635 aux_loss:0.0441 train_time:1604327ms step_avg:2579.30ms step:633/6250 train_loss:2.8527 aux_loss:0.0436 train_time:1605226ms step_avg:2576.61ms step:634/6250 train_loss:2.7396 aux_loss:0.0442 train_time:1606132ms step_avg:2573.93ms step:635/6250 train_loss:2.7097 aux_loss:0.0444 train_time:1607036ms step_avg:2571.26ms step:636/6250 train_loss:2.7720 aux_loss:0.0445 train_time:1608034ms step_avg:2568.74ms step:637/6250 train_loss:2.5813 aux_loss:0.0441 train_time:1608893ms step_avg:2566.02ms step:638/6250 train_loss:2.7517 aux_loss:0.0441 train_time:1609905ms step_avg:2563.54ms step:639/6250 train_loss:2.6988 aux_loss:0.0447 train_time:1610811ms step_avg:2560.91ms step:640/6250 train_loss:2.5682 aux_loss:0.0445 train_time:1611679ms step_avg:2558.22ms step:641/6250 train_loss:2.7850 aux_loss:0.0443 train_time:1612596ms step_avg:2555.62ms step:642/6250 train_loss:2.7264 aux_loss:0.0441 train_time:1613511ms step_avg:2553.02ms step:643/6250 train_loss:2.7266 aux_loss:0.0439 train_time:1614496ms step_avg:2550.55ms step:644/6250 train_loss:2.6561 aux_loss:0.0436 train_time:1615403ms step_avg:2547.95ms step:645/6250 train_loss:2.6801 aux_loss:0.0438 train_time:1616306ms step_avg:2545.36ms step:646/6250 train_loss:2.7596 aux_loss:0.0438 train_time:1617197ms step_avg:2542.76ms step:647/6250 train_loss:2.5952 aux_loss:0.0442 train_time:1618096ms step_avg:2540.18ms step:648/6250 train_loss:2.6932 aux_loss:0.0440 train_time:1618978ms step_avg:2537.58ms step:649/6250 train_loss:2.6248 aux_loss:0.0438 train_time:1619840ms step_avg:2534.96ms step:650/6250 train_loss:2.6202 aux_loss:0.0435 train_time:1620703ms step_avg:2532.35ms step:651/6250 train_loss:2.7450 aux_loss:0.0433 train_time:1621581ms step_avg:2529.77ms step:652/6250 train_loss:2.7724 aux_loss:0.0436 train_time:1622523ms step_avg:2527.30ms step:653/6250 train_loss:2.6568 aux_loss:0.0436 train_time:1623432ms step_avg:2524.78ms step:654/6250 train_loss:2.6132 aux_loss:0.0439 train_time:1624307ms step_avg:2522.22ms step:655/6250 train_loss:2.6913 aux_loss:0.0436 train_time:1625175ms step_avg:2519.65ms step:656/6250 train_loss:2.7416 aux_loss:0.0435 train_time:1626099ms step_avg:2517.18ms step:657/6250 train_loss:2.6323 aux_loss:0.0437 train_time:1627032ms step_avg:2514.73ms step:658/6250 train_loss:2.7092 aux_loss:0.0435 train_time:1627920ms step_avg:2512.22ms step:659/6250 train_loss:2.7043 aux_loss:0.0434 train_time:1628796ms step_avg:2509.70ms step:660/6250 train_loss:2.6164 aux_loss:0.0434 train_time:1629715ms step_avg:2507.25ms step:661/6250 train_loss:2.7643 aux_loss:0.0439 train_time:1630605ms step_avg:2504.77ms step:662/6250 train_loss:2.6340 aux_loss:0.0434 train_time:1631518ms step_avg:2502.33ms step:663/6250 train_loss:2.6757 aux_loss:0.0436 train_time:1632502ms step_avg:2500.00ms step:664/6250 train_loss:2.6612 aux_loss:0.0435 train_time:1633452ms step_avg:2497.63ms step:665/6250 train_loss:2.7739 aux_loss:0.0434 train_time:1634370ms step_avg:2495.22ms step:666/6250 train_loss:2.7278 aux_loss:0.0432 train_time:1635347ms step_avg:2492.91ms step:667/6250 train_loss:2.7703 aux_loss:0.0432 train_time:1636295ms step_avg:2490.56ms step:668/6250 train_loss:2.7325 aux_loss:0.0432 train_time:1637226ms step_avg:2488.19ms step:669/6250 train_loss:2.6587 aux_loss:0.0432 train_time:1638178ms step_avg:2485.85ms step:670/6250 train_loss:2.6944 aux_loss:0.0434 train_time:1639105ms step_avg:2483.49ms step:671/6250 train_loss:2.5643 aux_loss:0.0439 train_time:1640017ms step_avg:2481.11ms step:672/6250 train_loss:2.6365 aux_loss:0.0438 train_time:1640889ms step_avg:2478.68ms step:673/6250 train_loss:2.6278 aux_loss:0.0440 train_time:1641759ms step_avg:2476.26ms step:674/6250 train_loss:2.7162 aux_loss:0.0434 train_time:1642638ms step_avg:2473.85ms step:675/6250 train_loss:2.6324 aux_loss:0.0434 train_time:1643527ms step_avg:2471.47ms step:676/6250 train_loss:2.6868 aux_loss:0.0430 train_time:1644405ms step_avg:2469.08ms step:677/6250 train_loss:2.6633 aux_loss:0.0434 train_time:1645340ms step_avg:2466.78ms step:678/6250 train_loss:2.6787 aux_loss:0.0432 train_time:1646302ms step_avg:2464.52ms step:679/6250 train_loss:2.6833 aux_loss:0.0430 train_time:1647244ms step_avg:2462.25ms step:680/6250 train_loss:2.7495 aux_loss:0.0430 train_time:1648171ms step_avg:2459.96ms step:681/6250 train_loss:2.6455 aux_loss:0.0428 train_time:1649111ms step_avg:2457.69ms step:682/6250 train_loss:2.6617 aux_loss:0.0430 train_time:1650027ms step_avg:2455.40ms step:683/6250 train_loss:2.7547 aux_loss:0.0430 train_time:1650966ms step_avg:2453.14ms step:684/6250 train_loss:2.7230 aux_loss:0.0432 train_time:1651918ms step_avg:2450.92ms step:685/6250 train_loss:2.6513 aux_loss:0.0432 train_time:1652795ms step_avg:2448.58ms step:686/6250 train_loss:2.6964 aux_loss:0.0431 train_time:1653747ms step_avg:2446.37ms step:687/6250 train_loss:2.6913 aux_loss:0.0428 train_time:1654596ms step_avg:2444.01ms step:688/6250 train_loss:2.7355 aux_loss:0.0429 train_time:1655463ms step_avg:2441.69ms step:689/6250 train_loss:2.7745 aux_loss:0.0431 train_time:1656334ms step_avg:2439.37ms step:690/6250 train_loss:2.6974 aux_loss:0.0426 train_time:1657179ms step_avg:2437.03ms step:691/6250 train_loss:2.6713 aux_loss:0.0428 train_time:1658022ms step_avg:2434.69ms step:692/6250 train_loss:2.7365 aux_loss:0.0428 train_time:1658883ms step_avg:2432.38ms step:693/6250 train_loss:2.7284 aux_loss:0.0428 train_time:1659743ms step_avg:2430.08ms step:694/6250 train_loss:2.6549 aux_loss:0.0430 train_time:1660610ms step_avg:2427.79ms step:695/6250 train_loss:2.6183 aux_loss:0.0429 train_time:1661488ms step_avg:2425.53ms step:696/6250 train_loss:2.7114 aux_loss:0.0428 train_time:1662437ms step_avg:2423.38ms step:697/6250 train_loss:2.7344 aux_loss:0.0426 train_time:1663337ms step_avg:2421.16ms step:698/6250 train_loss:2.7378 aux_loss:0.0423 train_time:1664273ms step_avg:2419.00ms step:699/6250 train_loss:2.6540 aux_loss:0.0426 train_time:1665176ms step_avg:2416.80ms step:700/6250 train_loss:2.7265 aux_loss:0.0426 train_time:1666102ms step_avg:2414.64ms step:701/6250 train_loss:2.7400 aux_loss:0.0422 train_time:1667031ms step_avg:2412.49ms step:702/6250 train_loss:2.6311 aux_loss:0.0425 train_time:1668009ms step_avg:2410.42ms step:703/6250 train_loss:2.6130 aux_loss:0.0423 train_time:1668912ms step_avg:2408.24ms step:704/6250 train_loss:2.7101 aux_loss:0.0422 train_time:1669802ms step_avg:2406.05ms step:705/6250 train_loss:2.6063 aux_loss:0.0425 train_time:1670756ms step_avg:2403.97ms step:706/6250 train_loss:2.5749 aux_loss:0.0426 train_time:1671710ms step_avg:2401.88ms step:707/6250 train_loss:2.6333 aux_loss:0.0427 train_time:1672605ms step_avg:2399.72ms step:708/6250 train_loss:2.6381 aux_loss:0.0427 train_time:1673605ms step_avg:2397.72ms step:709/6250 train_loss:2.6307 aux_loss:0.0426 train_time:1674475ms step_avg:2395.53ms step:710/6250 train_loss:2.5904 aux_loss:0.0424 train_time:1675376ms step_avg:2393.39ms step:711/6250 train_loss:2.6043 aux_loss:0.0425 train_time:1676325ms step_avg:2391.33ms step:712/6250 train_loss:2.5995 aux_loss:0.0423 train_time:1677203ms step_avg:2389.18ms step:713/6250 train_loss:2.5843 aux_loss:0.0424 train_time:1678063ms step_avg:2387.00ms step:714/6250 train_loss:2.6713 aux_loss:0.0423 train_time:1679026ms step_avg:2384.98ms step:715/6250 train_loss:2.6615 aux_loss:0.0422 train_time:1679904ms step_avg:2382.84ms step:716/6250 train_loss:2.6066 aux_loss:0.0424 train_time:1680845ms step_avg:2380.80ms step:717/6250 train_loss:2.5579 aux_loss:0.0422 train_time:1681709ms step_avg:2378.65ms step:718/6250 train_loss:2.6938 aux_loss:0.0419 train_time:1682587ms step_avg:2376.53ms step:719/6250 train_loss:2.6335 aux_loss:0.0417 train_time:1683456ms step_avg:2374.41ms step:720/6250 train_loss:2.6803 aux_loss:0.0418 train_time:1684325ms step_avg:2372.29ms step:721/6250 train_loss:2.5582 aux_loss:0.0421 train_time:1685188ms step_avg:2370.17ms step:722/6250 train_loss:2.6177 aux_loss:0.0424 train_time:1686051ms step_avg:2368.05ms step:723/6250 train_loss:2.5912 aux_loss:0.0420 train_time:1686895ms step_avg:2365.91ms step:724/6250 train_loss:2.6180 aux_loss:0.0420 train_time:1687806ms step_avg:2363.87ms step:725/6250 train_loss:2.6425 aux_loss:0.0420 train_time:1688640ms step_avg:2361.73ms step:726/6250 train_loss:2.5047 aux_loss:0.0424 train_time:1689709ms step_avg:2359.93ms step:727/6250 train_loss:2.6687 aux_loss:0.0423 train_time:1690597ms step_avg:2357.88ms step:728/6250 train_loss:2.4619 aux_loss:0.0424 train_time:1691484ms step_avg:2355.83ms step:729/6250 train_loss:2.5895 aux_loss:0.0419 train_time:1692384ms step_avg:2353.80ms step:730/6250 train_loss:2.6579 aux_loss:0.0421 train_time:1693225ms step_avg:2351.70ms step:731/6250 train_loss:2.7051 aux_loss:0.0420 train_time:1694102ms step_avg:2349.66ms step:732/6250 train_loss:2.6362 aux_loss:0.0416 train_time:1694972ms step_avg:2347.61ms step:733/6250 train_loss:2.6131 aux_loss:0.0418 train_time:1695843ms step_avg:2345.56ms step:734/6250 train_loss:2.5412 aux_loss:0.0421 train_time:1696716ms step_avg:2343.53ms step:735/6250 train_loss:2.6059 aux_loss:0.0421 train_time:1697570ms step_avg:2341.48ms step:736/6250 train_loss:2.6337 aux_loss:0.0418 train_time:1698418ms step_avg:2339.42ms step:737/6250 train_loss:2.6413 aux_loss:0.0423 train_time:1699272ms step_avg:2337.38ms step:738/6250 train_loss:2.6096 aux_loss:0.0420 train_time:1700135ms step_avg:2335.35ms step:739/6250 train_loss:2.5980 aux_loss:0.0419 train_time:1700999ms step_avg:2333.33ms step:740/6250 train_loss:2.6612 aux_loss:0.0418 train_time:1701864ms step_avg:2331.32ms step:741/6250 train_loss:2.5488 aux_loss:0.0420 train_time:1702721ms step_avg:2329.30ms step:742/6250 train_loss:2.6217 aux_loss:0.0415 train_time:1703581ms step_avg:2327.30ms step:743/6250 train_loss:2.5340 aux_loss:0.0417 train_time:1704436ms step_avg:2325.29ms step:744/6250 train_loss:2.5578 aux_loss:0.0416 train_time:1705325ms step_avg:2323.33ms step:745/6250 train_loss:2.6140 aux_loss:0.0416 train_time:1706206ms step_avg:2321.37ms step:746/6250 train_loss:2.5343 aux_loss:0.0419 train_time:1707066ms step_avg:2319.38ms step:747/6250 train_loss:2.6670 aux_loss:0.0418 train_time:1707943ms step_avg:2317.43ms step:748/6250 train_loss:2.7048 aux_loss:0.0418 train_time:1708810ms step_avg:2315.46ms step:749/6250 train_loss:2.4809 aux_loss:0.0416 train_time:1709656ms step_avg:2313.47ms step:750/6250 train_loss:2.5260 aux_loss:0.0415 train_time:1710718ms step_avg:2311.78ms step:751/6250 train_loss:2.5326 aux_loss:0.0414 train_time:1711624ms step_avg:2309.88ms step:752/6250 train_loss:2.5810 aux_loss:0.0413 train_time:1712638ms step_avg:2308.14ms step:753/6250 train_loss:2.6977 aux_loss:0.0414 train_time:1713494ms step_avg:2306.18ms step:754/6250 train_loss:2.5845 aux_loss:0.0410 train_time:1714544ms step_avg:2304.49ms step:755/6250 train_loss:2.5890 aux_loss:0.0414 train_time:1715418ms step_avg:2302.57ms step:756/6250 train_loss:2.5603 aux_loss:0.0414 train_time:1716384ms step_avg:2300.78ms step:757/6250 train_loss:2.6091 aux_loss:0.0413 train_time:1717279ms step_avg:2298.90ms step:758/6250 train_loss:2.4869 aux_loss:0.0414 train_time:1718158ms step_avg:2297.00ms step:759/6250 train_loss:2.5611 aux_loss:0.0414 train_time:1719023ms step_avg:2295.09ms step:760/6250 train_loss:2.5881 aux_loss:0.0412 train_time:1719882ms step_avg:2293.18ms step:761/6250 train_loss:2.6074 aux_loss:0.0411 train_time:1720829ms step_avg:2291.38ms step:762/6250 train_loss:2.5989 aux_loss:0.0412 train_time:1721703ms step_avg:2289.50ms step:763/6250 train_loss:2.6916 aux_loss:0.0411 train_time:1722633ms step_avg:2287.69ms step:764/6250 train_loss:2.4674 aux_loss:0.0411 train_time:1723577ms step_avg:2285.91ms step:765/6250 train_loss:2.6463 aux_loss:0.0408 train_time:1724535ms step_avg:2284.15ms step:766/6250 train_loss:2.6167 aux_loss:0.0408 train_time:1725455ms step_avg:2282.35ms step:767/6250 train_loss:2.5281 aux_loss:0.0408 train_time:1726306ms step_avg:2280.46ms step:768/6250 train_loss:2.4802 aux_loss:0.0410 train_time:1727173ms step_avg:2278.59ms step:769/6250 train_loss:2.5600 aux_loss:0.0408 train_time:1728128ms step_avg:2276.85ms step:770/6250 train_loss:2.5056 aux_loss:0.0411 train_time:1729058ms step_avg:2275.08ms step:771/6250 train_loss:2.6503 aux_loss:0.0410 train_time:1729936ms step_avg:2273.24ms step:772/6250 train_loss:2.6383 aux_loss:0.0411 train_time:1730789ms step_avg:2271.38ms step:773/6250 train_loss:2.6364 aux_loss:0.0408 train_time:1731673ms step_avg:2269.56ms step:774/6250 train_loss:2.5091 aux_loss:0.0410 train_time:1732629ms step_avg:2267.84ms step:775/6250 train_loss:2.6345 aux_loss:0.0409 train_time:1733531ms step_avg:2266.05ms step:776/6250 train_loss:2.5121 aux_loss:0.0409 train_time:1734441ms step_avg:2264.28ms step:777/6250 train_loss:2.5993 aux_loss:0.0407 train_time:1735355ms step_avg:2262.52ms step:778/6250 train_loss:2.5644 aux_loss:0.0408 train_time:1736328ms step_avg:2260.84ms step:779/6250 train_loss:2.5628 aux_loss:0.0409 train_time:1737197ms step_avg:2259.03ms step:780/6250 train_loss:2.5926 aux_loss:0.0406 train_time:1738140ms step_avg:2257.32ms step:781/6250 train_loss:2.6010 aux_loss:0.0408 train_time:1739068ms step_avg:2255.60ms step:782/6250 train_loss:2.5927 aux_loss:0.0408 train_time:1739964ms step_avg:2253.84ms step:783/6250 train_loss:2.5682 aux_loss:0.0406 train_time:1740827ms step_avg:2252.04ms step:784/6250 train_loss:2.5319 aux_loss:0.0406 train_time:1741746ms step_avg:2250.32ms step:785/6250 train_loss:2.5969 aux_loss:0.0405 train_time:1743842ms step_avg:2250.12ms step:786/6250 train_loss:2.5824 aux_loss:0.0404 train_time:1744780ms step_avg:2248.43ms step:787/6250 train_loss:2.6109 aux_loss:0.0406 train_time:1746179ms step_avg:2247.33ms step:788/6250 train_loss:2.5398 aux_loss:0.0404 train_time:1747049ms step_avg:2245.56ms step:789/6250 train_loss:2.4462 aux_loss:0.0407 train_time:1747955ms step_avg:2243.85ms step:790/6250 train_loss:2.5478 aux_loss:0.0410 train_time:1749299ms step_avg:2242.69ms step:791/6250 train_loss:2.5580 aux_loss:0.0407 train_time:1752039ms step_avg:2243.33ms step:792/6250 train_loss:2.5620 aux_loss:0.0404 train_time:1754372ms step_avg:2243.44ms step:793/6250 train_loss:2.5359 aux_loss:0.0405 train_time:1756997ms step_avg:2243.93ms step:794/6250 train_loss:2.5415 aux_loss:0.0405 train_time:1758544ms step_avg:2243.04ms step:795/6250 train_loss:2.5648 aux_loss:0.0409 train_time:1760337ms step_avg:2242.47ms step:796/6250 train_loss:2.5024 aux_loss:0.0405 train_time:1761408ms step_avg:2240.98ms step:797/6250 train_loss:2.5175 aux_loss:0.0402 train_time:1762269ms step_avg:2239.22ms step:798/6250 train_loss:2.5966 aux_loss:0.0405 train_time:1763127ms step_avg:2237.47ms step:799/6250 train_loss:2.4864 aux_loss:0.0405 train_time:1763976ms step_avg:2235.71ms step:800/6250 train_loss:2.5649 aux_loss:0.0404 train_time:1764835ms step_avg:2233.97ms step:801/6250 train_loss:2.5783 aux_loss:0.0406 train_time:1766767ms step_avg:2233.59ms step:802/6250 train_loss:2.5255 aux_loss:0.0403 train_time:1769408ms step_avg:2234.10ms step:803/6250 train_loss:2.5932 aux_loss:0.0399 train_time:1770860ms step_avg:2233.11ms step:804/6250 train_loss:2.5660 aux_loss:0.0402 train_time:1771771ms step_avg:2231.45ms step:805/6250 train_loss:2.5414 aux_loss:0.0399 train_time:1772619ms step_avg:2229.71ms step:806/6250 train_loss:2.5005 aux_loss:0.0399 train_time:1774027ms step_avg:2228.68ms step:807/6250 train_loss:2.6114 aux_loss:0.0400 train_time:1775210ms step_avg:2227.36ms step:808/6250 train_loss:2.5366 aux_loss:0.0404 train_time:1789945ms step_avg:2243.04ms step:809/6250 train_loss:2.5457 aux_loss:0.0404 train_time:1826586ms step_avg:2286.09ms step:810/6250 train_loss:2.5996 aux_loss:0.0400 train_time:1832023ms step_avg:2290.03ms step:811/6250 train_loss:2.5331 aux_loss:0.0400 train_time:1834723ms step_avg:2290.54ms step:812/6250 train_loss:2.6284 aux_loss:0.0402 train_time:1836313ms step_avg:2289.67ms step:813/6250 train_loss:2.6141 aux_loss:0.0400 train_time:1838539ms step_avg:2289.59ms step:814/6250 train_loss:2.5082 aux_loss:0.0400 train_time:1839880ms step_avg:2288.41ms step:815/6250 train_loss:2.5941 aux_loss:0.0400 train_time:1840900ms step_avg:2286.83ms step:816/6250 train_loss:2.5645 aux_loss:0.0397 train_time:1841853ms step_avg:2285.18ms step:817/6250 train_loss:2.4191 aux_loss:0.0400 train_time:1844287ms step_avg:2285.36ms step:818/6250 train_loss:2.5195 aux_loss:0.0398 train_time:1846140ms step_avg:2284.83ms step:819/6250 train_loss:2.5418 aux_loss:0.0397 train_time:1847692ms step_avg:2283.92ms step:820/6250 train_loss:2.4076 aux_loss:0.0401 train_time:1850113ms step_avg:2284.09ms step:821/6250 train_loss:2.6208 aux_loss:0.0396 train_time:1851678ms step_avg:2283.20ms step:822/6250 train_loss:2.4827 aux_loss:0.0396 train_time:1852595ms step_avg:2281.52ms step:823/6250 train_loss:2.5088 aux_loss:0.0397 train_time:1854380ms step_avg:2280.91ms step:824/6250 train_loss:2.4631 aux_loss:0.0398 train_time:1856226ms step_avg:2280.38ms step:825/6250 train_loss:2.4750 aux_loss:0.0398 train_time:1857900ms step_avg:2279.63ms step:826/6250 train_loss:2.5310 aux_loss:0.0397 train_time:1860816ms step_avg:2280.41ms step:827/6250 train_loss:2.5149 aux_loss:0.0400 train_time:1861799ms step_avg:2278.82ms step:828/6250 train_loss:2.5256 aux_loss:0.0401 train_time:1862755ms step_avg:2277.21ms step:829/6250 train_loss:2.4644 aux_loss:0.0397 train_time:1866042ms step_avg:2278.44ms step:830/6250 train_loss:2.4689 aux_loss:0.0402 train_time:1868170ms step_avg:2278.26ms step:831/6250 train_loss:2.5653 aux_loss:0.0399 train_time:1869589ms step_avg:2277.21ms step:832/6250 train_loss:2.4664 aux_loss:0.0397 train_time:1871136ms step_avg:2276.32ms step:833/6250 train_loss:2.5050 aux_loss:0.0394 train_time:1873361ms step_avg:2276.26ms step:834/6250 train_loss:2.5503 aux_loss:0.0396 train_time:1874498ms step_avg:2274.88ms step:835/6250 train_loss:2.6117 aux_loss:0.0393 train_time:1875419ms step_avg:2273.23ms step:836/6250 train_loss:2.5482 aux_loss:0.0393 train_time:1876676ms step_avg:2272.01ms step:837/6250 train_loss:2.3906 aux_loss:0.0397 train_time:1878396ms step_avg:2271.34ms step:838/6250 train_loss:2.4945 aux_loss:0.0396 train_time:1880142ms step_avg:2270.70ms step:839/6250 train_loss:2.4391 aux_loss:0.0396 train_time:1882763ms step_avg:2271.13ms step:840/6250 train_loss:2.5024 aux_loss:0.0397 train_time:1884429ms step_avg:2270.40ms step:841/6250 train_loss:2.5680 aux_loss:0.0398 train_time:1885850ms step_avg:2269.37ms step:842/6250 train_loss:2.5047 aux_loss:0.0396 train_time:1887604ms step_avg:2268.76ms step:843/6250 train_loss:2.5604 aux_loss:0.0394 train_time:1889712ms step_avg:2268.56ms step:844/6250 train_loss:2.4480 aux_loss:0.0393 train_time:1893518ms step_avg:2270.41ms step:845/6250 train_loss:2.4496 aux_loss:0.0390 train_time:1894572ms step_avg:2268.95ms step:846/6250 train_loss:2.4772 aux_loss:0.0390 train_time:1895513ms step_avg:2267.36ms step:847/6250 train_loss:2.4631 aux_loss:0.0393 train_time:1896642ms step_avg:2266.00ms step:848/6250 train_loss:2.5691 aux_loss:0.0389 train_time:1897540ms step_avg:2264.37ms step:849/6250 train_loss:2.5655 aux_loss:0.0387 train_time:1898474ms step_avg:2262.78ms step:850/6250 train_loss:2.5729 aux_loss:0.0389 train_time:1899395ms step_avg:2261.18ms step:851/6250 train_loss:2.5035 aux_loss:0.0396 train_time:1900519ms step_avg:2259.83ms step:852/6250 train_loss:2.5234 aux_loss:0.0389 train_time:1905229ms step_avg:2262.74ms step:853/6250 train_loss:2.5184 aux_loss:0.0388 train_time:1906237ms step_avg:2261.25ms step:854/6250 train_loss:2.6116 aux_loss:0.0392 train_time:1907196ms step_avg:2259.71ms step:855/6250 train_loss:2.5523 aux_loss:0.0393 train_time:1908120ms step_avg:2258.13ms step:856/6250 train_loss:2.4391 aux_loss:0.0390 train_time:1909050ms step_avg:2256.56ms step:857/6250 train_loss:2.4274 aux_loss:0.0393 train_time:1909988ms step_avg:2255.00ms step:858/6250 train_loss:2.4790 aux_loss:0.0389 train_time:1911008ms step_avg:2253.55ms step:859/6250 train_loss:2.5417 aux_loss:0.0392 train_time:1912018ms step_avg:2252.08ms step:860/6250 train_loss:2.5981 aux_loss:0.0392 train_time:1912974ms step_avg:2250.56ms step:861/6250 train_loss:2.5260 aux_loss:0.0388 train_time:1914235ms step_avg:2249.40ms step:862/6250 train_loss:2.5448 aux_loss:0.0388 train_time:1915391ms step_avg:2248.11ms step:863/6250 train_loss:2.5104 aux_loss:0.0391 train_time:1917055ms step_avg:2247.43ms step:864/6250 train_loss:2.5210 aux_loss:0.0391 train_time:1918781ms step_avg:2246.82ms step:865/6250 train_loss:2.5511 aux_loss:0.0389 train_time:1920286ms step_avg:2245.95ms step:866/6250 train_loss:2.4983 aux_loss:0.0391 train_time:1921573ms step_avg:2244.83ms step:867/6250 train_loss:2.4657 aux_loss:0.0389 train_time:1923097ms step_avg:2243.99ms step:868/6250 train_loss:2.4681 aux_loss:0.0391 train_time:1924979ms step_avg:2243.57ms step:869/6250 train_loss:2.5115 aux_loss:0.0389 train_time:1926399ms step_avg:2242.61ms step:870/6250 train_loss:2.5225 aux_loss:0.0388 train_time:1927793ms step_avg:2241.62ms step:871/6250 train_loss:2.4602 aux_loss:0.0387 train_time:1929509ms step_avg:2241.01ms step:872/6250 train_loss:2.5303 aux_loss:0.0388 train_time:1931210ms step_avg:2240.38ms step:873/6250 train_loss:2.4614 aux_loss:0.0391 train_time:1932638ms step_avg:2239.44ms step:874/6250 train_loss:2.4446 aux_loss:0.0391 train_time:1933918ms step_avg:2238.33ms step:875/6250 train_loss:2.4108 aux_loss:0.0386 train_time:1934978ms step_avg:2236.97ms step:876/6250 train_loss:2.5716 aux_loss:0.0384 train_time:1935865ms step_avg:2235.41ms step:877/6250 train_loss:2.5137 aux_loss:0.0383 train_time:1938041ms step_avg:2235.34ms step:878/6250 train_loss:2.4454 aux_loss:0.0385 train_time:1939987ms step_avg:2235.01ms step:879/6250 train_loss:2.5686 aux_loss:0.0386 train_time:1941843ms step_avg:2234.57ms step:880/6250 train_loss:2.5277 aux_loss:0.0384 train_time:1943650ms step_avg:2234.08ms step:881/6250 train_loss:2.4534 aux_loss:0.0384 train_time:1956095ms step_avg:2245.80ms step:882/6250 train_loss:2.4528 aux_loss:0.0388 train_time:1957526ms step_avg:2244.87ms step:883/6250 train_loss:2.5147 aux_loss:0.0385 train_time:1962050ms step_avg:2247.48ms step:884/6250 train_loss:2.5061 aux_loss:0.0384 train_time:1963600ms step_avg:2246.68ms step:885/6250 train_loss:2.4027 aux_loss:0.0387 train_time:1965327ms step_avg:2246.09ms step:886/6250 train_loss:2.4393 aux_loss:0.0385 train_time:1966865ms step_avg:2245.28ms step:887/6250 train_loss:2.3811 aux_loss:0.0382 train_time:1968357ms step_avg:2244.42ms step:888/6250 train_loss:2.3805 aux_loss:0.0382 train_time:1969857ms step_avg:2243.57ms step:889/6250 train_loss:2.4509 aux_loss:0.0382 train_time:1971221ms step_avg:2242.57ms step:890/6250 train_loss:2.5305 aux_loss:0.0378 train_time:1972670ms step_avg:2241.67ms step:891/6250 train_loss:2.4316 aux_loss:0.0380 train_time:1973977ms step_avg:2240.61ms step:892/6250 train_loss:2.3627 aux_loss:0.0383 train_time:1980613ms step_avg:2245.59ms step:893/6250 train_loss:2.5342 aux_loss:0.0381 train_time:1982148ms step_avg:2244.79ms step:894/6250 train_loss:2.4594 aux_loss:0.0383 train_time:1984079ms step_avg:2244.43ms step:895/6250 train_loss:2.5441 aux_loss:0.0380 train_time:1985820ms step_avg:2243.86ms step:896/6250 train_loss:2.5055 aux_loss:0.0381 train_time:1987691ms step_avg:2243.44ms step:897/6250 train_loss:2.4573 aux_loss:0.0382 train_time:1989401ms step_avg:2242.84ms step:898/6250 train_loss:2.4148 aux_loss:0.0381 train_time:1990841ms step_avg:2241.94ms step:899/6250 train_loss:2.3717 aux_loss:0.0384 train_time:1992437ms step_avg:2241.21ms step:900/6250 train_loss:2.4117 aux_loss:0.0381 train_time:1994320ms step_avg:2240.81ms step:901/6250 train_loss:2.4926 aux_loss:0.0380 train_time:1996019ms step_avg:2240.20ms step:902/6250 train_loss:2.5256 aux_loss:0.0380 train_time:1998076ms step_avg:2240.00ms step:903/6250 train_loss:2.4227 aux_loss:0.0380 train_time:1999771ms step_avg:2239.39ms step:904/6250 train_loss:2.4208 aux_loss:0.0380 train_time:2001130ms step_avg:2238.40ms step:905/6250 train_loss:2.4518 aux_loss:0.0382 train_time:2002940ms step_avg:2237.92ms step:906/6250 train_loss:2.4445 aux_loss:0.0381 train_time:2004557ms step_avg:2237.23ms step:907/6250 train_loss:2.5674 aux_loss:0.0377 train_time:2005927ms step_avg:2236.26ms step:908/6250 train_loss:2.3971 aux_loss:0.0381 train_time:2007327ms step_avg:2235.33ms step:909/6250 train_loss:2.4789 aux_loss:0.0382 train_time:2008843ms step_avg:2234.53ms step:910/6250 train_loss:2.4698 aux_loss:0.0380 train_time:2010418ms step_avg:2233.80ms step:911/6250 train_loss:2.4330 aux_loss:0.0378 train_time:2011991ms step_avg:2233.06ms step:912/6250 train_loss:2.4339 aux_loss:0.0376 train_time:2013523ms step_avg:2232.29ms step:913/6250 train_loss:2.4867 aux_loss:0.0376 train_time:2014937ms step_avg:2231.38ms step:914/6250 train_loss:2.5082 aux_loss:0.0375 train_time:2016299ms step_avg:2230.42ms step:915/6250 train_loss:2.4386 aux_loss:0.0377 train_time:2017601ms step_avg:2229.39ms step:916/6250 train_loss:2.4018 aux_loss:0.0379 train_time:2019009ms step_avg:2228.49ms step:917/6250 train_loss:2.3962 aux_loss:0.0375 train_time:2020709ms step_avg:2227.90ms step:918/6250 train_loss:2.5126 aux_loss:0.0376 train_time:2022319ms step_avg:2227.22ms step:919/6250 train_loss:2.3829 aux_loss:0.0377 train_time:2023670ms step_avg:2226.26ms step:920/6250 train_loss:2.4204 aux_loss:0.0380 train_time:2025028ms step_avg:2225.31ms step:921/6250 train_loss:2.4016 aux_loss:0.0374 train_time:2026399ms step_avg:2224.37ms step:922/6250 train_loss:2.4630 aux_loss:0.0376 train_time:2027756ms step_avg:2223.42ms step:923/6250 train_loss:2.4347 aux_loss:0.0377 train_time:2029231ms step_avg:2222.60ms step:924/6250 train_loss:2.4599 aux_loss:0.0374 train_time:2030798ms step_avg:2221.88ms step:925/6250 train_loss:2.5124 aux_loss:0.0373 train_time:2032333ms step_avg:2221.13ms step:926/6250 train_loss:2.3688 aux_loss:0.0375 train_time:2033613ms step_avg:2220.10ms step:927/6250 train_loss:2.3434 aux_loss:0.0373 train_time:2034976ms step_avg:2219.17ms step:928/6250 train_loss:2.5488 aux_loss:0.0374 train_time:2036334ms step_avg:2218.23ms step:929/6250 train_loss:2.4305 aux_loss:0.0373 train_time:2037722ms step_avg:2217.33ms step:930/6250 train_loss:2.4620 aux_loss:0.0372 train_time:2039042ms step_avg:2216.35ms step:931/6250 train_loss:2.5115 aux_loss:0.0372 train_time:2040363ms step_avg:2215.38ms step:932/6250 train_loss:2.5144 aux_loss:0.0375 train_time:2041794ms step_avg:2214.53ms step:933/6250 train_loss:2.4205 aux_loss:0.0375 train_time:2043187ms step_avg:2213.64ms step:934/6250 train_loss:2.3895 aux_loss:0.0373 train_time:2044594ms step_avg:2212.76ms step:935/6250 train_loss:2.5177 aux_loss:0.0373 train_time:2045866ms step_avg:2211.75ms step:936/6250 train_loss:2.5247 aux_loss:0.0374 train_time:2047269ms step_avg:2210.87ms step:937/6250 train_loss:2.3917 aux_loss:0.0375 train_time:2048544ms step_avg:2209.86ms step:938/6250 train_loss:2.4747 aux_loss:0.0373 train_time:2049869ms step_avg:2208.91ms step:939/6250 train_loss:2.4531 aux_loss:0.0374 train_time:2051243ms step_avg:2208.01ms step:940/6250 train_loss:2.4518 aux_loss:0.0375 train_time:2052642ms step_avg:2207.14ms step:941/6250 train_loss:2.4242 aux_loss:0.0377 train_time:2053948ms step_avg:2206.17ms step:942/6250 train_loss:2.4050 aux_loss:0.0374 train_time:2055240ms step_avg:2205.19ms step:943/6250 train_loss:2.4061 aux_loss:0.0374 train_time:2056621ms step_avg:2204.31ms step:944/6250 train_loss:2.4284 aux_loss:0.0373 train_time:2057957ms step_avg:2203.38ms step:945/6250 train_loss:2.4365 aux_loss:0.0376 train_time:2059203ms step_avg:2202.36ms step:946/6250 train_loss:2.3290 aux_loss:0.0371 train_time:2060468ms step_avg:2201.35ms step:947/6250 train_loss:2.3706 aux_loss:0.0371 train_time:2061832ms step_avg:2200.46ms step:948/6250 train_loss:2.3567 aux_loss:0.0370 train_time:2063125ms step_avg:2199.49ms step:949/6250 train_loss:2.5944 aux_loss:0.0370 train_time:2064500ms step_avg:2198.62ms step:950/6250 train_loss:2.4396 aux_loss:0.0373 train_time:2065851ms step_avg:2197.71ms step:951/6250 train_loss:2.4436 aux_loss:0.0371 train_time:2067163ms step_avg:2196.77ms step:952/6250 train_loss:2.4071 aux_loss:0.0370 train_time:2068542ms step_avg:2195.90ms step:953/6250 train_loss:2.4017 aux_loss:0.0370 train_time:2069822ms step_avg:2194.93ms step:954/6250 train_loss:2.4293 aux_loss:0.0370 train_time:2071149ms step_avg:2194.01ms step:955/6250 train_loss:2.3896 aux_loss:0.0370 train_time:2072448ms step_avg:2193.07ms step:956/6250 train_loss:2.5731 aux_loss:0.0366 train_time:2073799ms step_avg:2192.18ms step:957/6250 train_loss:2.4002 aux_loss:0.0368 train_time:2075110ms step_avg:2191.25ms step:958/6250 train_loss:2.5243 aux_loss:0.0369 train_time:2076475ms step_avg:2190.37ms step:959/6250 train_loss:2.4108 aux_loss:0.0368 train_time:2077817ms step_avg:2189.48ms step:960/6250 train_loss:2.4958 aux_loss:0.0366 train_time:2079221ms step_avg:2188.65ms step:961/6250 train_loss:2.3900 aux_loss:0.0370 train_time:2080723ms step_avg:2187.93ms step:962/6250 train_loss:2.4351 aux_loss:0.0369 train_time:2082049ms step_avg:2187.03ms step:963/6250 train_loss:2.4323 aux_loss:0.0368 train_time:2083312ms step_avg:2186.06ms step:964/6250 train_loss:2.4816 aux_loss:0.0369 train_time:2084685ms step_avg:2185.20ms step:965/6250 train_loss:2.3200 aux_loss:0.0372 train_time:2086061ms step_avg:2184.36ms step:966/6250 train_loss:2.3560 aux_loss:0.0370 train_time:2087363ms step_avg:2183.43ms step:967/6250 train_loss:2.3928 aux_loss:0.0367 train_time:2088653ms step_avg:2182.50ms step:968/6250 train_loss:2.4513 aux_loss:0.0368 train_time:2089939ms step_avg:2181.56ms step:969/6250 train_loss:2.4238 aux_loss:0.0368 train_time:2091354ms step_avg:2180.77ms step:970/6250 train_loss:2.4286 aux_loss:0.0366 train_time:2092723ms step_avg:2179.92ms step:971/6250 train_loss:2.3720 aux_loss:0.0367 train_time:2094120ms step_avg:2179.10ms step:972/6250 train_loss:2.4497 aux_loss:0.0366 train_time:2095490ms step_avg:2178.26ms step:973/6250 train_loss:2.4698 aux_loss:0.0366 train_time:2096806ms step_avg:2177.37ms step:974/6250 train_loss:2.3511 aux_loss:0.0364 train_time:2098153ms step_avg:2176.51ms step:975/6250 train_loss:2.4524 aux_loss:0.0361 train_time:2099405ms step_avg:2175.55ms step:976/6250 train_loss:2.5109 aux_loss:0.0360 train_time:2100708ms step_avg:2174.65ms step:977/6250 train_loss:2.4539 aux_loss:0.0365 train_time:2102060ms step_avg:2173.80ms step:978/6250 train_loss:2.3567 aux_loss:0.0366 train_time:2103426ms step_avg:2172.96ms step:979/6250 train_loss:2.3966 aux_loss:0.0363 train_time:2104808ms step_avg:2172.14ms step:980/6250 train_loss:2.4210 aux_loss:0.0364 train_time:2106241ms step_avg:2171.38ms step:981/6250 train_loss:2.4241 aux_loss:0.0365 train_time:2107583ms step_avg:2170.53ms step:982/6250 train_loss:2.4354 aux_loss:0.0365 train_time:2109044ms step_avg:2169.80ms step:983/6250 train_loss:2.3512 aux_loss:0.0368 train_time:2110450ms step_avg:2169.01ms step:984/6250 train_loss:2.4885 aux_loss:0.0367 train_time:2111792ms step_avg:2168.16ms step:985/6250 train_loss:2.4852 aux_loss:0.0364 train_time:2113146ms step_avg:2167.33ms step:986/6250 train_loss:2.4045 aux_loss:0.0366 train_time:2114475ms step_avg:2166.47ms step:987/6250 train_loss:2.5804 aux_loss:0.0365 train_time:2115851ms step_avg:2165.66ms step:988/6250 train_loss:2.3402 aux_loss:0.0366 train_time:2117263ms step_avg:2164.89ms step:989/6250 train_loss:2.4876 aux_loss:0.0360 train_time:2118687ms step_avg:2164.13ms step:990/6250 train_loss:2.4926 aux_loss:0.0361 train_time:2120077ms step_avg:2163.34ms step:991/6250 train_loss:2.4035 aux_loss:0.0361 train_time:2121335ms step_avg:2162.42ms step:992/6250 train_loss:2.4182 aux_loss:0.0361 train_time:2122735ms step_avg:2161.64ms step:993/6250 train_loss:2.4140 aux_loss:0.0363 train_time:2124196ms step_avg:2160.93ms step:994/6250 train_loss:2.3530 aux_loss:0.0363 train_time:2125565ms step_avg:2160.13ms step:995/6250 train_loss:2.3170 aux_loss:0.0360 train_time:2126948ms step_avg:2159.34ms step:996/6250 train_loss:2.4775 aux_loss:0.0359 train_time:2128299ms step_avg:2158.52ms step:997/6250 train_loss:2.2956 aux_loss:0.0363 train_time:2129714ms step_avg:2157.76ms step:998/6250 train_loss:2.4628 aux_loss:0.0363 train_time:2131026ms step_avg:2156.91ms step:999/6250 train_loss:2.3647 aux_loss:0.0363 train_time:2132405ms step_avg:2156.12ms step:1000/6250 train_loss:2.3553 aux_loss:0.0358 train_time:2133731ms step_avg:2155.28ms step:1001/6250 train_loss:2.3736 aux_loss:0.0359 train_time:2135031ms step_avg:2154.42ms step:1002/6250 train_loss:2.3917 aux_loss:0.0361 train_time:2136436ms step_avg:2153.67ms step:1003/6250 train_loss:2.4500 aux_loss:0.0359 train_time:2137782ms step_avg:2152.85ms step:1004/6250 train_loss:2.4123 aux_loss:0.0362 train_time:2139077ms step_avg:2151.99ms step:1005/6250 train_loss:2.4503 aux_loss:0.0358 train_time:2140461ms step_avg:2151.22ms step:1006/6250 train_loss:2.3495 aux_loss:0.0359 train_time:2141771ms step_avg:2150.37ms step:1007/6250 train_loss:2.3650 aux_loss:0.0358 train_time:2143156ms step_avg:2149.61ms step:1008/6250 train_loss:2.3891 aux_loss:0.0361 train_time:2144641ms step_avg:2148.94ms step:1009/6250 train_loss:2.2739 aux_loss:0.0358 train_time:2145952ms step_avg:2148.10ms step:1010/6250 train_loss:2.4290 aux_loss:0.0356 train_time:2147373ms step_avg:2147.37ms step:1011/6250 train_loss:2.3222 aux_loss:0.0361 train_time:2148704ms step_avg:2146.56ms step:1012/6250 train_loss:2.4532 aux_loss:0.0362 train_time:2150105ms step_avg:2145.81ms step:1013/6250 train_loss:2.3045 aux_loss:0.0357 train_time:2151485ms step_avg:2145.05ms step:1014/6250 train_loss:2.4138 aux_loss:0.0356 train_time:2152840ms step_avg:2144.26ms step:1015/6250 train_loss:2.3264 aux_loss:0.0358 train_time:2154123ms step_avg:2143.41ms step:1016/6250 train_loss:2.3640 aux_loss:0.0357 train_time:2155401ms step_avg:2142.55ms step:1017/6250 train_loss:2.4225 aux_loss:0.0357 train_time:2156706ms step_avg:2141.71ms step:1018/6250 train_loss:2.4800 aux_loss:0.0356 train_time:2158088ms step_avg:2140.96ms step:1019/6250 train_loss:2.2985 aux_loss:0.0359 train_time:2159483ms step_avg:2140.22ms step:1020/6250 train_loss:2.3413 aux_loss:0.0360 train_time:2160782ms step_avg:2139.39ms step:1021/6250 train_loss:2.4565 aux_loss:0.0358 train_time:2162194ms step_avg:2138.67ms step:1022/6250 train_loss:2.5030 aux_loss:0.0359 train_time:2163578ms step_avg:2137.92ms step:1023/6250 train_loss:2.3980 aux_loss:0.0358 train_time:2164929ms step_avg:2137.15ms step:1024/6250 train_loss:2.3448 aux_loss:0.0354 train_time:2166302ms step_avg:2136.39ms step:1025/6250 train_loss:2.3416 aux_loss:0.0354 train_time:2167673ms step_avg:2135.64ms step:1026/6250 train_loss:2.3431 aux_loss:0.0353 train_time:2168976ms step_avg:2134.82ms step:1027/6250 train_loss:2.3250 aux_loss:0.0357 train_time:2170368ms step_avg:2134.09ms step:1028/6250 train_loss:2.4092 aux_loss:0.0356 train_time:2171683ms step_avg:2133.28ms step:1029/6250 train_loss:2.4052 aux_loss:0.0357 train_time:2173018ms step_avg:2132.50ms step:1030/6250 train_loss:2.4460 aux_loss:0.0356 train_time:2174343ms step_avg:2131.71ms step:1031/6250 train_loss:2.3036 aux_loss:0.0354 train_time:2175662ms step_avg:2130.91ms step:1032/6250 train_loss:2.3180 aux_loss:0.0357 train_time:2176986ms step_avg:2130.12ms step:1033/6250 train_loss:2.2927 aux_loss:0.0356 train_time:2178255ms step_avg:2129.28ms step:1034/6250 train_loss:2.4020 aux_loss:0.0355 train_time:2179591ms step_avg:2128.51ms step:1035/6250 train_loss:2.3735 aux_loss:0.0355 train_time:2180940ms step_avg:2127.75ms step:1036/6250 train_loss:2.3611 aux_loss:0.0356 train_time:2182254ms step_avg:2126.95ms step:1037/6250 train_loss:2.3430 aux_loss:0.0356 train_time:2183545ms step_avg:2126.14ms step:1038/6250 train_loss:2.3747 aux_loss:0.0355 train_time:2184908ms step_avg:2125.40ms step:1039/6250 train_loss:2.3723 aux_loss:0.0353 train_time:2186244ms step_avg:2124.63ms step:1040/6250 train_loss:2.3268 aux_loss:0.0354 train_time:2187559ms step_avg:2123.84ms step:1041/6250 train_loss:2.4492 aux_loss:0.0356 train_time:2188839ms step_avg:2123.03ms step:1042/6250 train_loss:2.4087 aux_loss:0.0358 train_time:2190162ms step_avg:2122.25ms step:1043/6250 train_loss:2.3390 aux_loss:0.0356 train_time:2191448ms step_avg:2121.44ms step:1044/6250 train_loss:2.4489 aux_loss:0.0352 train_time:2192862ms step_avg:2120.76ms step:1045/6250 train_loss:2.3417 aux_loss:0.0350 train_time:2194267ms step_avg:2120.07ms step:1046/6250 train_loss:2.3468 aux_loss:0.0362 train_time:2195508ms step_avg:2119.22ms step:1047/6250 train_loss:2.3482 aux_loss:0.0357 train_time:2196846ms step_avg:2118.46ms step:1048/6250 train_loss:2.3552 aux_loss:0.0351 train_time:2198258ms step_avg:2117.78ms step:1049/6250 train_loss:2.3843 aux_loss:0.0352 train_time:2199561ms step_avg:2117.00ms step:1050/6250 train_loss:2.2624 aux_loss:0.0356 train_time:2200925ms step_avg:2116.27ms step:1051/6250 train_loss:2.4017 aux_loss:0.0356 train_time:2202266ms step_avg:2115.53ms step:1052/6250 train_loss:2.3628 aux_loss:0.0352 train_time:2203614ms step_avg:2114.79ms step:1053/6250 train_loss:2.3085 aux_loss:0.0351 train_time:2204902ms step_avg:2114.00ms step:1054/6250 train_loss:2.4269 aux_loss:0.0351 train_time:2206222ms step_avg:2113.24ms step:1055/6250 train_loss:2.3898 aux_loss:0.0350 train_time:2207610ms step_avg:2112.55ms step:1056/6250 train_loss:2.3530 aux_loss:0.0353 train_time:2208928ms step_avg:2111.79ms step:1057/6250 train_loss:2.4346 aux_loss:0.0350 train_time:2210368ms step_avg:2111.14ms step:1058/6250 train_loss:2.3522 aux_loss:0.0349 train_time:2211762ms step_avg:2110.46ms step:1059/6250 train_loss:2.4174 aux_loss:0.0349 train_time:2213096ms step_avg:2109.72ms step:1060/6250 train_loss:2.3736 aux_loss:0.0350 train_time:2214354ms step_avg:2108.91ms step:1061/6250 train_loss:2.4170 aux_loss:0.0351 train_time:2215687ms step_avg:2108.17ms step:1062/6250 train_loss:2.3034 aux_loss:0.0354 train_time:2217799ms step_avg:2108.17ms step:1063/6250 train_loss:2.3922 aux_loss:0.0350 train_time:2218703ms step_avg:2107.03ms step:1064/6250 train_loss:2.3756 aux_loss:0.0349 train_time:2219615ms step_avg:2105.90ms step:1065/6250 train_loss:2.3368 aux_loss:0.0348 train_time:2220512ms step_avg:2104.75ms step:1066/6250 train_loss:2.4468 aux_loss:0.0346 train_time:2221443ms step_avg:2103.64ms step:1067/6250 train_loss:2.3406 aux_loss:0.0348 train_time:2222366ms step_avg:2102.52ms step:1068/6250 train_loss:2.3677 aux_loss:0.0351 train_time:2223273ms step_avg:2101.39ms step:1069/6250 train_loss:2.4508 aux_loss:0.0351 train_time:2224193ms step_avg:2100.28ms step:1070/6250 train_loss:2.3478 aux_loss:0.0347 train_time:2225121ms step_avg:2099.17ms step:1071/6250 train_loss:2.3533 aux_loss:0.0347 train_time:2226050ms step_avg:2098.07ms step:1072/6250 train_loss:2.3569 aux_loss:0.0349 train_time:2226967ms step_avg:2096.96ms step:1073/6250 train_loss:2.2824 aux_loss:0.0351 train_time:2227887ms step_avg:2095.85ms step:1074/6250 train_loss:2.2717 aux_loss:0.0346 train_time:2228819ms step_avg:2094.75ms step:1075/6250 train_loss:2.3572 aux_loss:0.0347 train_time:2229694ms step_avg:2093.61ms step:1076/6250 train_loss:2.3867 aux_loss:0.0349 train_time:2230598ms step_avg:2092.49ms step:1077/6250 train_loss:2.3392 aux_loss:0.0350 train_time:2231898ms step_avg:2091.75ms step:1078/6250 train_loss:2.3289 aux_loss:0.0347 train_time:2233228ms step_avg:2091.04ms step:1079/6250 train_loss:2.3844 aux_loss:0.0346 train_time:2234509ms step_avg:2090.28ms step:1080/6250 train_loss:2.3517 aux_loss:0.0347 train_time:2236481ms step_avg:2090.17ms step:1081/6250 train_loss:2.2925 aux_loss:0.0350 train_time:2237666ms step_avg:2089.32ms step:1082/6250 train_loss:2.4318 aux_loss:0.0351 train_time:2238570ms step_avg:2088.22ms step:1083/6250 train_loss:2.3442 aux_loss:0.0351 train_time:2239536ms step_avg:2087.17ms step:1084/6250 train_loss:2.3322 aux_loss:0.0345 train_time:2240938ms step_avg:2086.53ms step:1085/6250 train_loss:2.3818 aux_loss:0.0344 train_time:2242186ms step_avg:2085.75ms step:1086/6250 train_loss:2.3286 aux_loss:0.0344 train_time:2243156ms step_avg:2084.72ms step:1087/6250 train_loss:2.3398 aux_loss:0.0347 train_time:2244149ms step_avg:2083.70ms step:1088/6250 train_loss:2.4241 aux_loss:0.0343 train_time:2245399ms step_avg:2082.93ms step:1089/6250 train_loss:2.4332 aux_loss:0.0344 train_time:2246662ms step_avg:2082.17ms step:1090/6250 train_loss:2.2624 aux_loss:0.0344 train_time:2247793ms step_avg:2081.29ms step:1091/6250 train_loss:2.3496 aux_loss:0.0343 train_time:2248826ms step_avg:2080.32ms step:1092/6250 train_loss:2.2869 aux_loss:0.0342 train_time:2249857ms step_avg:2079.35ms step:1093/6250 train_loss:2.3637 aux_loss:0.0342 train_time:2250915ms step_avg:2078.41ms step:1094/6250 train_loss:2.3268 aux_loss:0.0346 train_time:2251895ms step_avg:2077.39ms step:1095/6250 train_loss:2.3806 aux_loss:0.0343 train_time:2252843ms step_avg:2076.35ms step:1096/6250 train_loss:2.3556 aux_loss:0.0343 train_time:2253792ms step_avg:2075.32ms step:1097/6250 train_loss:2.3879 aux_loss:0.0344 train_time:2254929ms step_avg:2074.45ms step:1098/6250 train_loss:2.3649 aux_loss:0.0344 train_time:2255885ms step_avg:2073.42ms step:1099/6250 train_loss:2.4019 aux_loss:0.0343 train_time:2256842ms step_avg:2072.40ms step:1100/6250 train_loss:2.3467 aux_loss:0.0339 train_time:2258418ms step_avg:2071.94ms step:1101/6250 train_loss:2.2873 aux_loss:0.0341 train_time:2259690ms step_avg:2071.21ms step:1102/6250 train_loss:2.3053 aux_loss:0.0344 train_time:2260698ms step_avg:2070.24ms step:1103/6250 train_loss:2.3251 aux_loss:0.0341 train_time:2261643ms step_avg:2069.21ms step:1104/6250 train_loss:2.4471 aux_loss:0.0342 train_time:2262601ms step_avg:2068.19ms step:1105/6250 train_loss:2.4393 aux_loss:0.0341 train_time:2263548ms step_avg:2067.17ms step:1106/6250 train_loss:2.3253 aux_loss:0.0343 train_time:2264783ms step_avg:2066.41ms step:1107/6250 train_loss:2.3917 aux_loss:0.0346 train_time:2266074ms step_avg:2065.70ms step:1108/6250 train_loss:2.2979 aux_loss:0.0348 train_time:2267468ms step_avg:2065.09ms step:1109/6250 train_loss:2.3604 aux_loss:0.0342 train_time:2268602ms step_avg:2064.24ms step:1110/6250 train_loss:2.4742 aux_loss:0.0339 train_time:2269668ms step_avg:2063.33ms step:1111/6250 train_loss:2.3191 aux_loss:0.0339 train_time:2270593ms step_avg:2062.30ms step:1112/6250 train_loss:2.3451 aux_loss:0.0339 train_time:2271567ms step_avg:2061.31ms step:1113/6250 train_loss:2.3187 aux_loss:0.0339 train_time:2272522ms step_avg:2060.31ms step:1114/6250 train_loss:2.1898 aux_loss:0.0340 train_time:2273798ms step_avg:2059.60ms step:1115/6250 train_loss:2.3453 aux_loss:0.0339 train_time:2275085ms step_avg:2058.90ms step:1116/6250 train_loss:2.2776 aux_loss:0.0338 train_time:2276286ms step_avg:2058.12ms step:1117/6250 train_loss:2.2778 aux_loss:0.0338 train_time:2277281ms step_avg:2057.16ms step:1118/6250 train_loss:2.2850 aux_loss:0.0338 train_time:2278229ms step_avg:2056.16ms step:1119/6250 train_loss:2.3398 aux_loss:0.0340 train_time:2279130ms step_avg:2055.12ms step:1120/6250 train_loss:2.2960 aux_loss:0.0340 train_time:2280508ms step_avg:2054.51ms step:1121/6250 train_loss:2.1764 aux_loss:0.0341 train_time:2281827ms step_avg:2053.85ms step:1122/6250 train_loss:2.3185 aux_loss:0.0343 train_time:2283252ms step_avg:2053.28ms step:1123/6250 train_loss:2.3845 aux_loss:0.0336 train_time:2284282ms step_avg:2052.36ms step:1124/6250 train_loss:2.2916 aux_loss:0.0335 train_time:2285285ms step_avg:2051.42ms step:1125/6250 train_loss:2.3550 aux_loss:0.0336 train_time:2286262ms step_avg:2050.46ms step:1126/6250 train_loss:2.2619 aux_loss:0.0335 train_time:2287218ms step_avg:2049.48ms step:1127/6250 train_loss:2.4187 aux_loss:0.0333 train_time:2288198ms step_avg:2048.52ms step:1128/6250 train_loss:2.3013 aux_loss:0.0336 train_time:2289131ms step_avg:2047.52ms step:1129/6250 train_loss:2.3187 aux_loss:0.0338 train_time:2290086ms step_avg:2046.55ms step:1130/6250 train_loss:2.3960 aux_loss:0.0338 train_time:2291082ms step_avg:2045.61ms step:1131/6250 train_loss:2.3207 aux_loss:0.0337 train_time:2292822ms step_avg:2045.34ms step:1132/6250 train_loss:2.3095 aux_loss:0.0336 train_time:2294675ms step_avg:2045.16ms step:1133/6250 train_loss:2.3956 aux_loss:0.0334 train_time:2297176ms step_avg:2045.57ms step:1134/6250 train_loss:2.3160 aux_loss:0.0340 train_time:2299543ms step_avg:2045.86ms step:1135/6250 train_loss:2.3514 aux_loss:0.0335 train_time:2301480ms step_avg:2045.76ms step:1136/6250 train_loss:2.3503 aux_loss:0.0333 train_time:2303477ms step_avg:2045.72ms step:1137/6250 train_loss:2.3723 aux_loss:0.0337 train_time:2304764ms step_avg:2045.04ms step:1138/6250 train_loss:2.3475 aux_loss:0.0335 train_time:2308085ms step_avg:2046.17ms step:1139/6250 train_loss:2.3192 aux_loss:0.0335 train_time:2309044ms step_avg:2045.21ms step:1140/6250 train_loss:2.2639 aux_loss:0.0337 train_time:2309997ms step_avg:2044.25ms step:1141/6250 train_loss:2.3699 aux_loss:0.0337 train_time:2310941ms step_avg:2043.27ms step:1142/6250 train_loss:2.3407 aux_loss:0.0334 train_time:2311909ms step_avg:2042.32ms step:1143/6250 train_loss:2.2160 aux_loss:0.0338 train_time:2312908ms step_avg:2041.40ms step:1144/6250 train_loss:2.3920 aux_loss:0.0337 train_time:2313848ms step_avg:2040.43ms step:1145/6250 train_loss:2.3072 aux_loss:0.0338 train_time:2314812ms step_avg:2039.48ms step:1146/6250 train_loss:2.3471 aux_loss:0.0335 train_time:2315741ms step_avg:2038.50ms step:1147/6250 train_loss:2.2710 aux_loss:0.0331 train_time:2316648ms step_avg:2037.51ms step:1148/6250 train_loss:2.3825 aux_loss:0.0335 train_time:2318516ms step_avg:2037.36ms step:1149/6250 train_loss:2.2627 aux_loss:0.0334 train_time:2319926ms step_avg:2036.81ms step:1150/6250 train_loss:2.2668 aux_loss:0.0333 train_time:2321711ms step_avg:2036.59ms step:1151/6250 train_loss:2.3716 aux_loss:0.0333 train_time:2322657ms step_avg:2035.63ms step:1152/6250 train_loss:2.3520 aux_loss:0.0337 train_time:2324096ms step_avg:2035.11ms step:1153/6250 train_loss:2.2946 aux_loss:0.0336 train_time:2325247ms step_avg:2034.34ms step:1154/6250 train_loss:2.2763 aux_loss:0.0333 train_time:2327293ms step_avg:2034.35ms step:1155/6250 train_loss:2.4017 aux_loss:0.0331 train_time:2329429ms step_avg:2034.44ms step:1156/6250 train_loss:2.2867 aux_loss:0.0333 train_time:2330279ms step_avg:2033.40ms step:1157/6250 train_loss:2.3493 aux_loss:0.0331 train_time:2331145ms step_avg:2032.38ms step:1158/6250 train_loss:2.4012 aux_loss:0.0332 train_time:2332032ms step_avg:2031.39ms step:1159/6250 train_loss:2.3240 aux_loss:0.0332 train_time:2333540ms step_avg:2030.93ms step:1160/6250 train_loss:2.2446 aux_loss:0.0333 train_time:2334913ms step_avg:2030.36ms step:1161/6250 train_loss:2.2963 aux_loss:0.0331 train_time:2336465ms step_avg:2029.94ms step:1162/6250 train_loss:2.1749 aux_loss:0.0330 train_time:2338000ms step_avg:2029.51ms step:1163/6250 train_loss:2.3208 aux_loss:0.0330 train_time:2339134ms step_avg:2028.74ms step:1164/6250 train_loss:2.2929 aux_loss:0.0329 train_time:2341353ms step_avg:2028.90ms step:1165/6250 train_loss:2.3173 aux_loss:0.0334 train_time:2342546ms step_avg:2028.18ms step:1166/6250 train_loss:2.3007 aux_loss:0.0337 train_time:2343407ms step_avg:2027.17ms step:1167/6250 train_loss:2.3652 aux_loss:0.0330 train_time:2344266ms step_avg:2026.16ms step:1168/6250 train_loss:2.2490 aux_loss:0.0329 train_time:2345141ms step_avg:2025.16ms step:1169/6250 train_loss:2.2916 aux_loss:0.0332 train_time:2346011ms step_avg:2024.17ms step:1170/6250 train_loss:2.2783 aux_loss:0.0331 train_time:2346871ms step_avg:2023.16ms step:1171/6250 train_loss:2.2695 aux_loss:0.0331 train_time:2347758ms step_avg:2022.19ms step:1172/6250 train_loss:2.3085 aux_loss:0.0328 train_time:2348814ms step_avg:2021.35ms step:1173/6250 train_loss:2.2970 aux_loss:0.0329 train_time:2349672ms step_avg:2020.35ms step:1174/6250 train_loss:2.2916 aux_loss:0.0332 train_time:2350662ms step_avg:2019.47ms step:1175/6250 train_loss:2.4051 aux_loss:0.0333 train_time:2352281ms step_avg:2019.13ms step:1176/6250 train_loss:2.2619 aux_loss:0.0332 train_time:2353140ms step_avg:2018.13ms step:1177/6250 train_loss:2.2714 aux_loss:0.0332 train_time:2353989ms step_avg:2017.13ms step:1178/6250 train_loss:2.2764 aux_loss:0.0330 train_time:2354851ms step_avg:2016.14ms step:1179/6250 train_loss:2.3198 aux_loss:0.0330 train_time:2355697ms step_avg:2015.14ms step:1180/6250 train_loss:2.3282 aux_loss:0.0327 train_time:2356532ms step_avg:2014.13ms step:1181/6250 train_loss:2.3007 aux_loss:0.0330 train_time:2357393ms step_avg:2013.15ms step:1182/6250 train_loss:2.2592 aux_loss:0.0329 train_time:2358260ms step_avg:2012.17ms step:1183/6250 train_loss:2.3596 aux_loss:0.0329 train_time:2359690ms step_avg:2011.67ms step:1184/6250 train_loss:2.3719 aux_loss:0.0328 train_time:2361242ms step_avg:2011.28ms step:1185/6250 train_loss:2.2767 aux_loss:0.0329 train_time:2362427ms step_avg:2010.58ms step:1186/6250 train_loss:2.1625 aux_loss:0.0329 train_time:2363650ms step_avg:2009.91ms step:1187/6250 train_loss:2.2651 aux_loss:0.0326 train_time:2365673ms step_avg:2009.92ms step:1188/6250 train_loss:2.3548 aux_loss:0.0325 train_time:2367286ms step_avg:2009.58ms step:1189/6250 train_loss:2.3109 aux_loss:0.0326 train_time:2368852ms step_avg:2009.20ms step:1190/6250 train_loss:2.3477 aux_loss:0.0327 train_time:2370606ms step_avg:2008.99ms step:1191/6250 train_loss:2.3311 aux_loss:0.0326 train_time:2371453ms step_avg:2008.00ms step:1192/6250 train_loss:2.2410 aux_loss:0.0331 train_time:2377075ms step_avg:2011.06ms step:1193/6250 train_loss:2.3924 aux_loss:0.0330 train_time:2378711ms step_avg:2010.75ms step:1194/6250 train_loss:2.3868 aux_loss:0.0324 train_time:2382728ms step_avg:2012.44ms step:1195/6250 train_loss:2.2495 aux_loss:0.0326 train_time:2383762ms step_avg:2011.61ms step:1196/6250 train_loss:2.3372 aux_loss:0.0328 train_time:2385221ms step_avg:2011.15ms step:1197/6250 train_loss:2.3031 aux_loss:0.0326 train_time:2390331ms step_avg:2013.76ms step:1198/6250 train_loss:2.2371 aux_loss:0.0326 train_time:2396537ms step_avg:2017.29ms step:1199/6250 train_loss:2.3277 aux_loss:0.0327 train_time:2398361ms step_avg:2017.12ms step:1200/6250 train_loss:2.2139 aux_loss:0.0330 train_time:2399211ms step_avg:2016.14ms step:1201/6250 train_loss:2.4012 aux_loss:0.0327 train_time:2400069ms step_avg:2015.17ms step:1202/6250 train_loss:2.2377 aux_loss:0.0323 train_time:2400935ms step_avg:2014.21ms step:1203/6250 train_loss:2.2870 aux_loss:0.0324 train_time:2401805ms step_avg:2013.25ms step:1204/6250 train_loss:2.2664 aux_loss:0.0328 train_time:2402680ms step_avg:2012.30ms step:1205/6250 train_loss:2.2664 aux_loss:0.0324 train_time:2403509ms step_avg:2011.30ms step:1206/6250 train_loss:2.3178 aux_loss:0.0324 train_time:2404373ms step_avg:2010.35ms step:1207/6250 train_loss:2.2645 aux_loss:0.0322 train_time:2405222ms step_avg:2009.38ms step:1208/6250 train_loss:2.2786 aux_loss:0.0322 train_time:2406079ms step_avg:2008.41ms step:1209/6250 train_loss:2.2231 aux_loss:0.0323 train_time:2407219ms step_avg:2007.69ms step:1210/6250 train_loss:2.2550 aux_loss:0.0325 train_time:2408257ms step_avg:2006.88ms step:1211/6250 train_loss:2.2829 aux_loss:0.0329 train_time:2409117ms step_avg:2005.93ms step:1212/6250 train_loss:2.3106 aux_loss:0.0328 train_time:2410224ms step_avg:2005.18ms step:1213/6250 train_loss:2.1926 aux_loss:0.0324 train_time:2411259ms step_avg:2004.37ms step:1214/6250 train_loss:2.3559 aux_loss:0.0321 train_time:2412137ms step_avg:2003.44ms step:1215/6250 train_loss:2.2759 aux_loss:0.0324 train_time:2413015ms step_avg:2002.50ms step:1216/6250 train_loss:2.3004 aux_loss:0.0323 train_time:2413894ms step_avg:2001.57ms step:1217/6250 train_loss:2.3337 aux_loss:0.0326 train_time:2414757ms step_avg:2000.63ms step:1218/6250 train_loss:2.3034 aux_loss:0.0323 train_time:2415611ms step_avg:1999.68ms step:1219/6250 train_loss:2.2967 aux_loss:0.0326 train_time:2416457ms step_avg:1998.72ms step:1220/6250 train_loss:2.2950 aux_loss:0.0323 train_time:2417464ms step_avg:1997.90ms step:1221/6250 train_loss:2.3019 aux_loss:0.0323 train_time:2418467ms step_avg:1997.08ms step:1222/6250 train_loss:2.2624 aux_loss:0.0325 train_time:2419309ms step_avg:1996.13ms step:1223/6250 train_loss:2.2110 aux_loss:0.0325 train_time:2420168ms step_avg:1995.19ms step:1224/6250 train_loss:2.2442 aux_loss:0.0324 train_time:2421027ms step_avg:1994.26ms step:1225/6250 train_loss:2.3106 aux_loss:0.0320 train_time:2421873ms step_avg:1993.31ms step:1226/6250 train_loss:2.3172 aux_loss:0.0323 train_time:2422733ms step_avg:1992.38ms step:1227/6250 train_loss:2.3198 aux_loss:0.0322 train_time:2423574ms step_avg:1991.43ms step:1228/6250 train_loss:2.2590 aux_loss:0.0324 train_time:2424453ms step_avg:1990.52ms step:1229/6250 train_loss:2.2837 aux_loss:0.0323 train_time:2425321ms step_avg:1989.60ms step:1230/6250 train_loss:2.2976 aux_loss:0.0321 train_time:2426183ms step_avg:1988.67ms step:1231/6250 train_loss:2.2404 aux_loss:0.0321 train_time:2427049ms step_avg:1987.76ms step:1232/6250 train_loss:2.2487 aux_loss:0.0321 train_time:2427929ms step_avg:1986.85ms step:1233/6250 train_loss:2.3938 aux_loss:0.0322 train_time:2428782ms step_avg:1985.92ms step:1234/6250 train_loss:2.3355 aux_loss:0.0321 train_time:2429637ms step_avg:1985.00ms step:1235/6250 train_loss:2.2934 aux_loss:0.0322 train_time:2430486ms step_avg:1984.07ms step:1236/6250 train_loss:2.2337 aux_loss:0.0320 train_time:2431345ms step_avg:1983.15ms step:1237/6250 train_loss:2.2816 aux_loss:0.0322 train_time:2432219ms step_avg:1982.25ms step:1238/6250 train_loss:2.2549 aux_loss:0.0317 train_time:2433066ms step_avg:1981.32ms step:1239/6250 train_loss:2.3337 aux_loss:0.0321 train_time:2433924ms step_avg:1980.41ms step:1240/6250 train_loss:2.3189 aux_loss:0.0322 train_time:2434773ms step_avg:1979.49ms step:1241/6250 train_loss:2.3399 aux_loss:0.0321 train_time:2435616ms step_avg:1978.57ms step:1242/6250 train_loss:2.2790 aux_loss:0.0322 train_time:2436451ms step_avg:1977.64ms step:1243/6250 train_loss:2.3159 aux_loss:0.0315 train_time:2437311ms step_avg:1976.73ms step:1244/6250 train_loss:2.3147 aux_loss:0.0319 train_time:2438347ms step_avg:1975.97ms step:1245/6250 train_loss:2.2705 aux_loss:0.0317 train_time:2439207ms step_avg:1975.07ms step:1246/6250 train_loss:2.2704 aux_loss:0.0316 train_time:2440078ms step_avg:1974.17ms step:1247/6250 train_loss:2.2459 aux_loss:0.0315 train_time:2440939ms step_avg:1973.27ms step:1248/6250 train_loss:2.3149 aux_loss:0.0317 train_time:2441802ms step_avg:1972.38ms step:1249/6250 train_loss:2.2247 aux_loss:0.0320 train_time:2442660ms step_avg:1971.48ms step:1250/6250 train_loss:2.2032 aux_loss:0.0319 train_time:2443507ms step_avg:1970.57ms step:1251/6250 train_loss:2.2468 aux_loss:0.0319 train_time:2444371ms step_avg:1969.68ms step:1252/6250 train_loss:2.3044 aux_loss:0.0318 train_time:2445215ms step_avg:1968.77ms step:1253/6250 train_loss:2.3363 aux_loss:0.0316 train_time:2446076ms step_avg:1967.88ms step:1254/6250 train_loss:2.2552 aux_loss:0.0318 train_time:2446937ms step_avg:1966.99ms step:1255/6250 train_loss:2.2342 aux_loss:0.0315 train_time:2447810ms step_avg:1966.11ms step:1256/6250 train_loss:2.2707 aux_loss:0.0319 train_time:2448659ms step_avg:1965.22ms step:1257/6250 train_loss:2.2161 aux_loss:0.0317 train_time:2449489ms step_avg:1964.31ms step:1258/6250 train_loss:2.3062 aux_loss:0.0313 train_time:2450531ms step_avg:1963.57ms step:1259/6250 train_loss:2.2599 aux_loss:0.0312 train_time:2451402ms step_avg:1962.69ms step:1260/6250 train_loss:2.2886 aux_loss:0.0317 train_time:2452250ms step_avg:1961.80ms step:1261/6250 train_loss:2.2538 aux_loss:0.0318 train_time:2453127ms step_avg:1960.93ms step:1262/6250 train_loss:2.3127 aux_loss:0.0318 train_time:2453991ms step_avg:1960.06ms step:1263/6250 train_loss:2.2868 aux_loss:0.0313 train_time:2454840ms step_avg:1959.17ms step:1264/6250 train_loss:2.1986 aux_loss:0.0313 train_time:2455703ms step_avg:1958.30ms step:1265/6250 train_loss:2.2975 aux_loss:0.0315 train_time:2456569ms step_avg:1957.43ms step:1266/6250 train_loss:2.2967 aux_loss:0.0317 train_time:2457432ms step_avg:1956.55ms step:1267/6250 train_loss:2.2875 aux_loss:0.0317 train_time:2458307ms step_avg:1955.69ms step:1268/6250 train_loss:2.1780 aux_loss:0.0316 train_time:2459168ms step_avg:1954.82ms step:1269/6250 train_loss:2.2634 aux_loss:0.0316 train_time:2460028ms step_avg:1953.95ms step:1270/6250 train_loss:2.3731 aux_loss:0.0321 train_time:2460886ms step_avg:1953.08ms step:1271/6250 train_loss:2.2606 aux_loss:0.0317 train_time:2461745ms step_avg:1952.22ms step:1272/6250 train_loss:2.2530 aux_loss:0.0310 train_time:2462594ms step_avg:1951.34ms step:1273/6250 train_loss:2.2156 aux_loss:0.0314 train_time:2463470ms step_avg:1950.49ms step:1274/6250 train_loss:2.3412 aux_loss:0.0317 train_time:2464338ms step_avg:1949.63ms step:1275/6250 train_loss:2.2742 aux_loss:0.0316 train_time:2465198ms step_avg:1948.77ms step:1276/6250 train_loss:2.3003 aux_loss:0.0312 train_time:2466059ms step_avg:1947.91ms step:1277/6250 train_loss:2.2895 aux_loss:0.0313 train_time:2466932ms step_avg:1947.07ms step:1278/6250 train_loss:2.2743 aux_loss:0.0318 train_time:2467846ms step_avg:1946.25ms step:1279/6250 train_loss:2.3554 aux_loss:0.0314 train_time:2468990ms step_avg:1945.62ms step:1280/6250 train_loss:2.2312 aux_loss:0.0313 train_time:2474059ms step_avg:1948.08ms step:1281/6250 train_loss:2.2253 aux_loss:0.0315 train_time:2475087ms step_avg:1947.35ms step:1282/6250 train_loss:2.2277 aux_loss:0.0314 train_time:2475953ms step_avg:1946.50ms step:1283/6250 train_loss:2.3156 aux_loss:0.0314 train_time:2480980ms step_avg:1948.92ms step:1284/6250 train_loss:2.2802 aux_loss:0.0311 train_time:2481826ms step_avg:1948.06ms step:1285/6250 train_loss:2.2049 aux_loss:0.0313 train_time:2482687ms step_avg:1947.21ms step:1286/6250 train_loss:2.2885 aux_loss:0.0312 train_time:2483529ms step_avg:1946.34ms step:1287/6250 train_loss:2.2729 aux_loss:0.0313 train_time:2484379ms step_avg:1945.48ms step:1288/6250 train_loss:2.1944 aux_loss:0.0310 train_time:2485251ms step_avg:1944.64ms step:1289/6250 train_loss:2.2476 aux_loss:0.0309 train_time:2486096ms step_avg:1943.78ms step:1290/6250 train_loss:2.2677 aux_loss:0.0313 train_time:2486997ms step_avg:1942.97ms step:1291/6250 train_loss:2.1649 aux_loss:0.0312 train_time:2488237ms step_avg:1942.42ms step:1292/6250 train_loss:2.1894 aux_loss:0.0313 train_time:2490386ms step_avg:1942.58ms step:1293/6250 train_loss:2.2716 aux_loss:0.0310 train_time:2491783ms step_avg:1942.15ms step:1294/6250 train_loss:2.2081 aux_loss:0.0312 train_time:2492644ms step_avg:1941.31ms step:1295/6250 train_loss:2.3204 aux_loss:0.0314 train_time:2493484ms step_avg:1940.45ms step:1296/6250 train_loss:2.2796 aux_loss:0.0311 train_time:2494340ms step_avg:1939.61ms step:1297/6250 train_loss:2.2525 aux_loss:0.0310 train_time:2495726ms step_avg:1939.18ms step:1298/6250 train_loss:2.3140 aux_loss:0.0310 train_time:2497000ms step_avg:1938.66ms step:1299/6250 train_loss:2.2275 aux_loss:0.0309 train_time:2497858ms step_avg:1937.83ms step:1300/6250 train_loss:2.2291 aux_loss:0.0312 train_time:2498879ms step_avg:1937.12ms step:1301/6250 train_loss:2.2308 aux_loss:0.0310 train_time:2499735ms step_avg:1936.28ms step:1302/6250 train_loss:2.1217 aux_loss:0.0304 train_time:2500606ms step_avg:1935.45ms step:1303/6250 train_loss:2.2441 aux_loss:0.0305 train_time:2506280ms step_avg:1938.34ms step:1304/6250 train_loss:2.2295 aux_loss:0.0308 train_time:2507375ms step_avg:1937.69ms step:1305/6250 train_loss:2.2249 aux_loss:0.0309 train_time:2508278ms step_avg:1936.89ms step:1306/6250 train_loss:2.2620 aux_loss:0.0307 train_time:2509582ms step_avg:1936.41ms step:1307/6250 train_loss:2.2636 aux_loss:0.0307 train_time:2511438ms step_avg:1936.34ms step:1308/6250 train_loss:2.2795 aux_loss:0.0304 train_time:2512298ms step_avg:1935.51ms step:1309/6250 train_loss:2.1995 aux_loss:0.0307 train_time:2513407ms step_avg:1934.88ms step:1310/6250 train_loss:2.3265 aux_loss:0.0313 train_time:2514484ms step_avg:1934.22ms step:1311/6250 train_loss:2.3253 aux_loss:0.0309 train_time:2515345ms step_avg:1933.39ms step:1312/6250 train_loss:2.3439 aux_loss:0.0308 train_time:2516221ms step_avg:1932.58ms step:1313/6250 train_loss:2.2568 aux_loss:0.0310 train_time:2517088ms step_avg:1931.76ms step:1314/6250 train_loss:2.2554 aux_loss:0.0310 train_time:2518231ms step_avg:1931.16ms step:1315/6250 train_loss:2.3377 aux_loss:0.0308 train_time:2519506ms step_avg:1930.66ms step:1316/6250 train_loss:2.2725 aux_loss:0.0306 train_time:2520363ms step_avg:1929.83ms step:1317/6250 train_loss:2.3023 aux_loss:0.0306 train_time:2525089ms step_avg:1931.97ms step:1318/6250 train_loss:2.3230 aux_loss:0.0308 train_time:2525938ms step_avg:1931.15ms step:1319/6250 train_loss:2.3033 aux_loss:0.0304 train_time:2526789ms step_avg:1930.32ms step:1320/6250 train_loss:2.1392 aux_loss:0.0305 train_time:2527622ms step_avg:1929.48ms step:1321/6250 train_loss:2.3588 aux_loss:0.0309 train_time:2528471ms step_avg:1928.66ms step:1322/6250 train_loss:2.2288 aux_loss:0.0303 train_time:2529315ms step_avg:1927.83ms step:1323/6250 train_loss:2.2260 aux_loss:0.0304 train_time:2530232ms step_avg:1927.06ms step:1324/6250 train_loss:2.2838 aux_loss:0.0305 train_time:2531319ms step_avg:1926.42ms step:1325/6250 train_loss:2.2944 aux_loss:0.0304 train_time:2532179ms step_avg:1925.61ms step:1326/6250 train_loss:2.4040 aux_loss:0.0304 train_time:2533220ms step_avg:1924.94ms step:1327/6250 train_loss:2.2078 aux_loss:0.0308 train_time:2534275ms step_avg:1924.28ms step:1328/6250 train_loss:2.2807 aux_loss:0.0306 train_time:2535113ms step_avg:1923.45ms step:1329/6250 train_loss:2.3499 aux_loss:0.0302 train_time:2535974ms step_avg:1922.65ms step:1330/6250 train_loss:2.2431 aux_loss:0.0306 train_time:2536980ms step_avg:1921.95ms step:1331/6250 train_loss:2.2769 aux_loss:0.0303 train_time:2537842ms step_avg:1921.15ms step:1332/6250 train_loss:2.3010 aux_loss:0.0304 train_time:2539178ms step_avg:1920.71ms step:1333/6250 train_loss:2.1617 aux_loss:0.0304 train_time:2543914ms step_avg:1922.84ms step:1334/6250 train_loss:2.2971 aux_loss:0.0304 train_time:2544775ms step_avg:1922.04ms step:1335/6250 train_loss:2.2173 aux_loss:0.0305 train_time:2545618ms step_avg:1921.22ms step:1336/6250 train_loss:2.2293 aux_loss:0.0301 train_time:2546467ms step_avg:1920.41ms step:1337/6250 train_loss:2.2202 aux_loss:0.0303 train_time:2547316ms step_avg:1919.61ms step:1338/6250 train_loss:2.2129 aux_loss:0.0305 train_time:2548667ms step_avg:1919.18ms step:1339/6250 train_loss:2.3462 aux_loss:0.0305 train_time:2549503ms step_avg:1918.36ms step:1340/6250 train_loss:2.3048 aux_loss:0.0300 train_time:2551345ms step_avg:1918.30ms step:1341/6250 train_loss:2.2728 aux_loss:0.0301 train_time:2552219ms step_avg:1917.52ms step:1342/6250 train_loss:2.2731 aux_loss:0.0306 train_time:2553246ms step_avg:1916.85ms step:1343/6250 train_loss:2.3048 aux_loss:0.0301 train_time:2558078ms step_avg:1919.04ms step:1344/6250 train_loss:2.1705 aux_loss:0.0299 train_time:2559161ms step_avg:1918.41ms step:1345/6250 train_loss:2.2840 aux_loss:0.0303 train_time:2560042ms step_avg:1917.63ms step:1346/6250 train_loss:2.1904 aux_loss:0.0303 train_time:2561071ms step_avg:1916.97ms step:1347/6250 train_loss:2.1853 aux_loss:0.0303 train_time:2562211ms step_avg:1916.39ms step:1348/6250 train_loss:2.2262 aux_loss:0.0305 train_time:2563080ms step_avg:1915.61ms step:1349/6250 train_loss:2.2972 aux_loss:0.0304 train_time:2564630ms step_avg:1915.33ms step:1350/6250 train_loss:2.1754 aux_loss:0.0305 train_time:2570409ms step_avg:1918.22ms step:1351/6250 train_loss:2.2909 aux_loss:0.0307 train_time:2571257ms step_avg:1917.42ms step:1352/6250 train_loss:2.2961 aux_loss:0.0302 train_time:2572109ms step_avg:1916.62ms step:1353/6250 train_loss:2.2915 aux_loss:0.0302 train_time:2572961ms step_avg:1915.83ms step:1354/6250 train_loss:2.1978 aux_loss:0.0303 train_time:2573846ms step_avg:1915.06ms step:1355/6250 train_loss:2.2760 aux_loss:0.0305 train_time:2574931ms step_avg:1914.45ms step:1356/6250 train_loss:2.2408 aux_loss:0.0308 train_time:2576308ms step_avg:1914.05ms step:1357/6250 train_loss:2.2305 aux_loss:0.0301 train_time:2577402ms step_avg:1913.44ms step:1358/6250 train_loss:2.1515 aux_loss:0.0300 train_time:2578257ms step_avg:1912.65ms step:1359/6250 train_loss:2.1905 aux_loss:0.0301 train_time:2579784ms step_avg:1912.37ms step:1360/6250 train_loss:2.1549 aux_loss:0.0301 train_time:2580640ms step_avg:1911.59ms step:1361/6250 train_loss:2.2547 aux_loss:0.0302 train_time:2581484ms step_avg:1910.80ms step:1362/6250 train_loss:2.1654 aux_loss:0.0302 train_time:2582323ms step_avg:1910.00ms step:1363/6250 train_loss:2.2463 aux_loss:0.0300 train_time:2583724ms step_avg:1909.63ms step:1364/6250 train_loss:2.3373 aux_loss:0.0301 train_time:2584673ms step_avg:1908.92ms step:1365/6250 train_loss:2.2147 aux_loss:0.0299 train_time:2585528ms step_avg:1908.14ms step:1366/6250 train_loss:2.3076 aux_loss:0.0297 train_time:2594970ms step_avg:1913.69ms step:1367/6250 train_loss:2.2690 aux_loss:0.0297 train_time:2596395ms step_avg:1913.33ms step:1368/6250 train_loss:2.2776 aux_loss:0.0303 train_time:2597701ms step_avg:1912.89ms step:1369/6250 train_loss:2.1726 aux_loss:0.0300 train_time:2600138ms step_avg:1913.27ms step:1370/6250 train_loss:2.1834 aux_loss:0.0298 train_time:2601604ms step_avg:1912.94ms step:1371/6250 train_loss:2.2559 aux_loss:0.0299 train_time:2603135ms step_avg:1912.66ms step:1372/6250 train_loss:2.2449 aux_loss:0.0302 train_time:2604953ms step_avg:1912.59ms step:1373/6250 train_loss:2.2424 aux_loss:0.0301 train_time:2606164ms step_avg:1912.08ms step:1374/6250 train_loss:2.2063 aux_loss:0.0299 train_time:2607049ms step_avg:1911.33ms step:1375/6250 train_loss:2.2014 aux_loss:0.0296 train_time:2608053ms step_avg:1910.66ms step:1376/6250 train_loss:2.2573 aux_loss:0.0297 train_time:2608932ms step_avg:1909.91ms step:1377/6250 train_loss:2.2525 aux_loss:0.0299 train_time:2609790ms step_avg:1909.14ms step:1378/6250 train_loss:2.2017 aux_loss:0.0301 train_time:2610644ms step_avg:1908.37ms step:1379/6250 train_loss:2.1978 aux_loss:0.0295 train_time:2611491ms step_avg:1907.59ms step:1380/6250 train_loss:2.0922 aux_loss:0.0295 train_time:2612350ms step_avg:1906.82ms step:1381/6250 train_loss:2.1769 aux_loss:0.0296 train_time:2613224ms step_avg:1906.07ms step:1382/6250 train_loss:2.2273 aux_loss:0.0300 train_time:2614078ms step_avg:1905.30ms step:1383/6250 train_loss:2.2360 aux_loss:0.0301 train_time:2614940ms step_avg:1904.55ms step:1384/6250 train_loss:2.2879 aux_loss:0.0302 train_time:2615810ms step_avg:1903.79ms step:1385/6250 train_loss:2.2192 aux_loss:0.0294 train_time:2617051ms step_avg:1903.31ms step:1386/6250 train_loss:2.2575 aux_loss:0.0293 train_time:2617906ms step_avg:1902.55ms step:1387/6250 train_loss:2.1775 aux_loss:0.0295 train_time:2618771ms step_avg:1901.79ms step:1388/6250 train_loss:2.2582 aux_loss:0.0298 train_time:2620116ms step_avg:1901.39ms step:1389/6250 train_loss:2.1968 aux_loss:0.0295 train_time:2620980ms step_avg:1900.64ms step:1390/6250 train_loss:2.2088 aux_loss:0.0296 train_time:2622531ms step_avg:1900.39ms step:1391/6250 train_loss:2.2512 aux_loss:0.0297 train_time:2623397ms step_avg:1899.64ms step:1392/6250 train_loss:2.2335 aux_loss:0.0297 train_time:2624382ms step_avg:1898.97ms step:1393/6250 train_loss:2.2356 aux_loss:0.0298 train_time:2625773ms step_avg:1898.61ms step:1394/6250 train_loss:2.2051 aux_loss:0.0297 train_time:2626621ms step_avg:1897.85ms step:1395/6250 train_loss:2.2209 aux_loss:0.0296 train_time:2627465ms step_avg:1897.09ms step:1396/6250 train_loss:2.2652 aux_loss:0.0298 train_time:2628336ms step_avg:1896.35ms step:1397/6250 train_loss:2.0929 aux_loss:0.0299 train_time:2629530ms step_avg:1895.84ms step:1398/6250 train_loss:2.2420 aux_loss:0.0293 train_time:2630616ms step_avg:1895.26ms step:1399/6250 train_loss:2.2352 aux_loss:0.0293 train_time:2631471ms step_avg:1894.51ms step:1400/6250 train_loss:2.1992 aux_loss:0.0298 train_time:2632345ms step_avg:1893.77ms step:1401/6250 train_loss:2.1761 aux_loss:0.0301 train_time:2633209ms step_avg:1893.03ms step:1402/6250 train_loss:2.2838 aux_loss:0.0295 train_time:2635173ms step_avg:1893.08ms step:1403/6250 train_loss:2.2156 aux_loss:0.0294 train_time:2636033ms step_avg:1892.34ms step:1404/6250 train_loss:2.2001 aux_loss:0.0296 train_time:2637103ms step_avg:1891.75ms step:1405/6250 train_loss:2.2012 aux_loss:0.0300 train_time:2639139ms step_avg:1891.86ms step:1406/6250 train_loss:2.1552 aux_loss:0.0296 train_time:2640345ms step_avg:1891.36ms step:1407/6250 train_loss:2.1900 aux_loss:0.0294 train_time:2641707ms step_avg:1890.99ms step:1408/6250 train_loss:2.2106 aux_loss:0.0297 train_time:2644207ms step_avg:1891.42ms step:1409/6250 train_loss:2.2876 aux_loss:0.0297 train_time:2645532ms step_avg:1891.02ms step:1410/6250 train_loss:2.2086 aux_loss:0.0293 train_time:2646782ms step_avg:1890.56ms step:1411/6250 train_loss:2.2495 aux_loss:0.0292 train_time:2652936ms step_avg:1893.60ms step:1412/6250 train_loss:2.1195 aux_loss:0.0294 train_time:2655064ms step_avg:1893.77ms step:1413/6250 train_loss:2.2075 aux_loss:0.0294 train_time:2656275ms step_avg:1893.28ms step:1414/6250 train_loss:2.2124 aux_loss:0.0294 train_time:2657670ms step_avg:1892.93ms step:1415/6250 train_loss:2.1806 aux_loss:0.0294 train_time:2659141ms step_avg:1892.63ms step:1416/6250 train_loss:2.3079 aux_loss:0.0292 train_time:2660464ms step_avg:1892.22ms step:1417/6250 train_loss:2.2595 aux_loss:0.0292 train_time:2661759ms step_avg:1891.80ms step:1418/6250 train_loss:2.2552 aux_loss:0.0295 train_time:2663315ms step_avg:1891.56ms step:1419/6250 train_loss:2.1876 aux_loss:0.0294 train_time:2664766ms step_avg:1891.25ms step:1420/6250 train_loss:2.2577 aux_loss:0.0298 train_time:2666272ms step_avg:1890.97ms step:1421/6250 train_loss:2.3146 aux_loss:0.0295 train_time:2667564ms step_avg:1890.55ms step:1422/6250 train_loss:2.2209 aux_loss:0.0296 train_time:2668922ms step_avg:1890.17ms step:1423/6250 train_loss:2.2064 aux_loss:0.0295 train_time:2670515ms step_avg:1889.96ms step:1424/6250 train_loss:2.1873 aux_loss:0.0292 train_time:2671996ms step_avg:1889.67ms step:1425/6250 train_loss:2.2787 aux_loss:0.0292 train_time:2673344ms step_avg:1889.29ms step:1426/6250 train_loss:2.1219 aux_loss:0.0294 train_time:2674756ms step_avg:1888.95ms step:1427/6250 train_loss:2.1529 aux_loss:0.0294 train_time:2676303ms step_avg:1888.71ms step:1428/6250 train_loss:2.2618 aux_loss:0.0291 train_time:2677829ms step_avg:1888.46ms step:1429/6250 train_loss:2.1593 aux_loss:0.0292 train_time:2679315ms step_avg:1888.17ms step:1430/6250 train_loss:2.2848 aux_loss:0.0295 train_time:2680757ms step_avg:1887.86ms step:1431/6250 train_loss:2.2624 aux_loss:0.0298 train_time:2682438ms step_avg:1887.71ms step:1432/6250 train_loss:2.1824 aux_loss:0.0296 train_time:2684220ms step_avg:1887.64ms step:1433/6250 train_loss:2.2014 aux_loss:0.0293 train_time:2685505ms step_avg:1887.21ms step:1434/6250 train_loss:2.1930 aux_loss:0.0290 train_time:2686977ms step_avg:1886.92ms step:1435/6250 train_loss:2.1976 aux_loss:0.0294 train_time:2688531ms step_avg:1886.69ms step:1436/6250 train_loss:2.1810 aux_loss:0.0287 train_time:2689750ms step_avg:1886.22ms step:1437/6250 train_loss:2.2099 aux_loss:0.0285 train_time:2690957ms step_avg:1885.74ms step:1438/6250 train_loss:2.1763 aux_loss:0.0288 train_time:2692211ms step_avg:1885.30ms step:1439/6250 train_loss:2.1289 aux_loss:0.0287 train_time:2700012ms step_avg:1889.44ms step:1440/6250 train_loss:2.1816 aux_loss:0.0286 train_time:2701877ms step_avg:1889.42ms step:1441/6250 train_loss:2.2954 aux_loss:0.0291 train_time:2703444ms step_avg:1889.20ms step:1442/6250 train_loss:2.2818 aux_loss:0.0294 train_time:2704603ms step_avg:1888.69ms step:1443/6250 train_loss:2.1404 aux_loss:0.0292 train_time:2706241ms step_avg:1888.51ms step:1444/6250 train_loss:2.1785 aux_loss:0.0289 train_time:2708504ms step_avg:1888.78ms step:1445/6250 train_loss:2.2636 aux_loss:0.0290 train_time:2713464ms step_avg:1890.92ms step:1446/6250 train_loss:2.2483 aux_loss:0.0289 train_time:2715867ms step_avg:1891.27ms step:1447/6250 train_loss:2.1446 aux_loss:0.0294 train_time:2717456ms step_avg:1891.06ms step:1448/6250 train_loss:2.2698 aux_loss:0.0292 train_time:2718339ms step_avg:1890.36ms step:1449/6250 train_loss:2.2308 aux_loss:0.0292 train_time:2719216ms step_avg:1889.66ms step:1450/6250 train_loss:2.2807 aux_loss:0.0292 train_time:2720640ms step_avg:1889.33ms step:1451/6250 train_loss:2.2150 aux_loss:0.0293 train_time:2721482ms step_avg:1888.61ms step:1452/6250 train_loss:2.2082 aux_loss:0.0290 train_time:2722329ms step_avg:1887.88ms step:1453/6250 train_loss:2.2039 aux_loss:0.0287 train_time:2723233ms step_avg:1887.20ms step:1454/6250 train_loss:2.1915 aux_loss:0.0290 train_time:2724107ms step_avg:1886.50ms step:1455/6250 train_loss:2.2582 aux_loss:0.0289 train_time:2725111ms step_avg:1885.89ms step:1456/6250 train_loss:2.2187 aux_loss:0.0290 train_time:2726420ms step_avg:1885.49ms step:1457/6250 train_loss:2.2218 aux_loss:0.0291 train_time:2728032ms step_avg:1885.30ms step:1458/6250 train_loss:2.3199 aux_loss:0.0292 train_time:2729753ms step_avg:1885.19ms step:1459/6250 train_loss:2.1697 aux_loss:0.0288 train_time:2731349ms step_avg:1884.99ms step:1460/6250 train_loss:2.2259 aux_loss:0.0287 train_time:2735013ms step_avg:1886.22ms step:1461/6250 train_loss:2.2187 aux_loss:0.0287 train_time:2736010ms step_avg:1885.60ms step:1462/6250 train_loss:2.2183 aux_loss:0.0286 train_time:2736870ms step_avg:1884.90ms step:1463/6250 train_loss:2.1618 aux_loss:0.0290 train_time:2737729ms step_avg:1884.19ms step:1464/6250 train_loss:2.2469 aux_loss:0.0289 train_time:2738573ms step_avg:1883.48ms step:1465/6250 train_loss:2.1827 aux_loss:0.0288 train_time:2739426ms step_avg:1882.77ms step:1466/6250 train_loss:2.1684 aux_loss:0.0291 train_time:2740296ms step_avg:1882.07ms step:1467/6250 train_loss:2.1451 aux_loss:0.0289 train_time:2741182ms step_avg:1881.39ms step:1468/6250 train_loss:2.1820 aux_loss:0.0284 train_time:2742059ms step_avg:1880.70ms step:1469/6250 train_loss:2.2520 aux_loss:0.0286 train_time:2742923ms step_avg:1880.00ms step:1470/6250 train_loss:2.1917 aux_loss:0.0290 train_time:2743794ms step_avg:1879.31ms step:1471/6250 train_loss:2.1823 aux_loss:0.0284 train_time:2744662ms step_avg:1878.62ms step:1472/6250 train_loss:2.2365 aux_loss:0.0285 train_time:2745512ms step_avg:1877.92ms step:1473/6250 train_loss:2.1807 aux_loss:0.0286 train_time:2747341ms step_avg:1877.88ms step:1474/6250 train_loss:2.1345 aux_loss:0.0291 train_time:2748479ms step_avg:1877.38ms step:1475/6250 train_loss:2.1215 aux_loss:0.0291 train_time:2749429ms step_avg:1876.74ms step:1476/6250 train_loss:2.1990 aux_loss:0.0285 train_time:2750862ms step_avg:1876.44ms step:1477/6250 train_loss:2.1905 aux_loss:0.0284 train_time:2751999ms step_avg:1875.94ms step:1478/6250 train_loss:2.1881 aux_loss:0.0283 train_time:2753982ms step_avg:1876.01ms step:1479/6250 train_loss:2.1341 aux_loss:0.0284 train_time:2755812ms step_avg:1875.98ms step:1480/6250 train_loss:2.0812 aux_loss:0.0282 train_time:2757173ms step_avg:1875.63ms step:1481/6250 train_loss:2.3037 aux_loss:0.0283 train_time:2759142ms step_avg:1875.69ms step:1482/6250 train_loss:2.1160 aux_loss:0.0283 train_time:2761065ms step_avg:1875.72ms step:1483/6250 train_loss:2.2874 aux_loss:0.0283 train_time:2762349ms step_avg:1875.32ms step:1484/6250 train_loss:2.1086 aux_loss:0.0283 train_time:2763893ms step_avg:1875.10ms step:1485/6250 train_loss:2.2163 aux_loss:0.0284 train_time:2765416ms step_avg:1874.86ms step:1486/6250 train_loss:2.2900 aux_loss:0.0284 train_time:2768227ms step_avg:1875.49ms step:1487/6250 train_loss:2.1281 aux_loss:0.0286 train_time:2769823ms step_avg:1875.30ms step:1488/6250 train_loss:2.1783 aux_loss:0.0282 train_time:2770995ms step_avg:1874.83ms step:1489/6250 train_loss:2.1584 aux_loss:0.0283 train_time:2772655ms step_avg:1874.68ms step:1490/6250 train_loss:2.2158 aux_loss:0.0290 train_time:2774194ms step_avg:1874.46ms step:1491/6250 train_loss:2.1184 aux_loss:0.0285 train_time:2775445ms step_avg:1874.03ms step:1492/6250 train_loss:2.2293 aux_loss:0.0283 train_time:2776458ms step_avg:1873.45ms step:1493/6250 train_loss:2.2246 aux_loss:0.0287 train_time:2778099ms step_avg:1873.30ms step:1494/6250 train_loss:2.1391 aux_loss:0.0283 train_time:2779571ms step_avg:1873.03ms step:1495/6250 train_loss:2.2123 aux_loss:0.0282 train_time:2780447ms step_avg:1872.36ms step:1496/6250 train_loss:2.1596 aux_loss:0.0281 train_time:2781315ms step_avg:1871.68ms step:1497/6250 train_loss:2.2315 aux_loss:0.0283 train_time:2782192ms step_avg:1871.01ms step:1498/6250 train_loss:2.1679 aux_loss:0.0284 train_time:2783070ms step_avg:1870.34ms step:1499/6250 train_loss:2.1715 aux_loss:0.0283 train_time:2783942ms step_avg:1869.67ms step:1500/6250 train_loss:2.2707 aux_loss:0.0283 train_time:2784817ms step_avg:1869.00ms step:1501/6250 train_loss:2.1557 aux_loss:0.0281 train_time:2785694ms step_avg:1868.34ms step:1502/6250 train_loss:2.1408 aux_loss:0.0286 train_time:2786562ms step_avg:1867.67ms step:1503/6250 train_loss:2.1869 aux_loss:0.0284 train_time:2787457ms step_avg:1867.02ms step:1504/6250 train_loss:2.1566 aux_loss:0.0282 train_time:2788379ms step_avg:1866.39ms step:1505/6250 train_loss:2.2020 aux_loss:0.0283 train_time:2789568ms step_avg:1865.93ms step:1506/6250 train_loss:2.2593 aux_loss:0.0283 train_time:2790630ms step_avg:1865.39ms step:1507/6250 train_loss:2.2409 aux_loss:0.0281 train_time:2791785ms step_avg:1864.92ms step:1508/6250 train_loss:2.2118 aux_loss:0.0279 train_time:2792678ms step_avg:1864.27ms step:1509/6250 train_loss:2.2304 aux_loss:0.0285 train_time:2793528ms step_avg:1863.59ms step:1510/6250 train_loss:2.2683 aux_loss:0.0282 train_time:2794510ms step_avg:1863.01ms step:1511/6250 train_loss:2.2268 aux_loss:0.0281 train_time:2795660ms step_avg:1862.53ms step:1512/6250 train_loss:2.2411 aux_loss:0.0283 train_time:2796764ms step_avg:1862.03ms step:1513/6250 train_loss:2.2055 aux_loss:0.0280 train_time:2797665ms step_avg:1861.39ms step:1514/6250 train_loss:2.2651 aux_loss:0.0283 train_time:2798815ms step_avg:1860.91ms step:1515/6250 train_loss:2.1212 aux_loss:0.0279 train_time:2800270ms step_avg:1860.64ms step:1516/6250 train_loss:2.2995 aux_loss:0.0281 train_time:2801194ms step_avg:1860.02ms step:1517/6250 train_loss:2.2295 aux_loss:0.0282 train_time:2802396ms step_avg:1859.59ms step:1518/6250 train_loss:2.2040 aux_loss:0.0281 train_time:2803575ms step_avg:1859.13ms step:1519/6250 train_loss:2.0924 aux_loss:0.0279 train_time:2804724ms step_avg:1858.66ms step:1520/6250 train_loss:2.1724 aux_loss:0.0277 train_time:2805562ms step_avg:1857.99ms step:1521/6250 train_loss:2.1235 aux_loss:0.0280 train_time:2806425ms step_avg:1857.33ms step:1522/6250 train_loss:2.1507 aux_loss:0.0281 train_time:2807487ms step_avg:1856.80ms step:1523/6250 train_loss:2.1261 aux_loss:0.0279 train_time:2808588ms step_avg:1856.30ms step:1524/6250 train_loss:2.1021 aux_loss:0.0279 train_time:2809453ms step_avg:1855.65ms step:1525/6250 train_loss:2.2583 aux_loss:0.0281 train_time:2810319ms step_avg:1855.00ms step:1526/6250 train_loss:2.1997 aux_loss:0.0280 train_time:2811352ms step_avg:1854.45ms step:1527/6250 train_loss:2.1210 aux_loss:0.0276 train_time:2812215ms step_avg:1853.80ms step:1528/6250 train_loss:2.1509 aux_loss:0.0278 train_time:2813091ms step_avg:1853.16ms step:1529/6250 train_loss:2.0498 aux_loss:0.0283 train_time:2814083ms step_avg:1852.59ms step:1530/6250 train_loss:2.1626 aux_loss:0.0282 train_time:2815075ms step_avg:1852.02ms step:1531/6250 train_loss:2.1839 aux_loss:0.0280 train_time:2816093ms step_avg:1851.47ms step:1532/6250 train_loss:2.1564 aux_loss:0.0273 train_time:2817559ms step_avg:1851.22ms step:1533/6250 train_loss:2.1642 aux_loss:0.0279 train_time:2818432ms step_avg:1850.58ms step:1534/6250 train_loss:2.2043 aux_loss:0.0279 train_time:2819282ms step_avg:1849.92ms step:1535/6250 train_loss:2.1385 aux_loss:0.0276 train_time:2820147ms step_avg:1849.28ms step:1536/6250 train_loss:2.1395 aux_loss:0.0278 train_time:2821121ms step_avg:1848.70ms step:1537/6250 train_loss:2.2153 aux_loss:0.0281 train_time:2821967ms step_avg:1848.05ms step:1538/6250 train_loss:2.1014 aux_loss:0.0278 train_time:2822851ms step_avg:1847.42ms step:1539/6250 train_loss:2.3043 aux_loss:0.0276 train_time:2823725ms step_avg:1846.78ms step:1540/6250 train_loss:2.1796 aux_loss:0.0277 train_time:2824575ms step_avg:1846.13ms step:1541/6250 train_loss:2.2054 aux_loss:0.0278 train_time:2825439ms step_avg:1845.49ms step:1542/6250 train_loss:2.2224 aux_loss:0.0276 train_time:2826378ms step_avg:1844.89ms step:1543/6250 train_loss:2.1744 aux_loss:0.0276 train_time:2827247ms step_avg:1844.26ms step:1544/6250 train_loss:2.1349 aux_loss:0.0276 train_time:2828098ms step_avg:1843.61ms step:1545/6250 train_loss:2.1990 aux_loss:0.0274 train_time:2828969ms step_avg:1842.98ms step:1546/6250 train_loss:2.1848 aux_loss:0.0275 train_time:2829839ms step_avg:1842.34ms step:1547/6250 train_loss:2.1567 aux_loss:0.0278 train_time:2830702ms step_avg:1841.71ms step:1548/6250 train_loss:2.1747 aux_loss:0.0278 train_time:2831564ms step_avg:1841.07ms step:1549/6250 train_loss:2.2542 aux_loss:0.0276 train_time:2832440ms step_avg:1840.44ms step:1550/6250 train_loss:2.2115 aux_loss:0.0276 train_time:2833288ms step_avg:1839.80ms step:1551/6250 train_loss:2.1505 aux_loss:0.0276 train_time:2834178ms step_avg:1839.18ms step:1552/6250 train_loss:2.3169 aux_loss:0.0274 train_time:2835054ms step_avg:1838.56ms step:1553/6250 train_loss:2.1954 aux_loss:0.0272 train_time:2835908ms step_avg:1837.92ms step:1554/6250 train_loss:2.1495 aux_loss:0.0272 train_time:2836787ms step_avg:1837.30ms step:1555/6250 train_loss:2.2040 aux_loss:0.0278 train_time:2837652ms step_avg:1836.67ms step:1556/6250 train_loss:2.2724 aux_loss:0.0275 train_time:2838508ms step_avg:1836.03ms step:1557/6250 train_loss:2.1369 aux_loss:0.0278 train_time:2839377ms step_avg:1835.41ms step:1558/6250 train_loss:2.1732 aux_loss:0.0275 train_time:2840255ms step_avg:1834.79ms step:1559/6250 train_loss:2.0994 aux_loss:0.0273 train_time:2841139ms step_avg:1834.18ms step:1560/6250 train_loss:2.2694 aux_loss:0.0279 train_time:2842046ms step_avg:1833.58ms step:1561/6250 train_loss:2.0997 aux_loss:0.0275 train_time:2843227ms step_avg:1833.16ms step:1562/6250 train_loss:2.1356 aux_loss:0.0273 train_time:2844254ms step_avg:1832.64ms step:1563/6250 train_loss:2.1106 aux_loss:0.0276 train_time:2845132ms step_avg:1832.02ms step:1564/6250 train_loss:2.1276 aux_loss:0.0275 train_time:2846010ms step_avg:1831.41ms step:1565/6250 train_loss:2.1824 aux_loss:0.0273 train_time:2846955ms step_avg:1830.84ms step:1566/6250 train_loss:2.1309 aux_loss:0.0273 train_time:2847838ms step_avg:1830.23ms step:1567/6250 train_loss:2.1335 aux_loss:0.0275 train_time:2848705ms step_avg:1829.61ms step:1568/6250 train_loss:2.1279 aux_loss:0.0273 train_time:2849570ms step_avg:1828.99ms step:1569/6250 train_loss:2.1436 aux_loss:0.0272 train_time:2850455ms step_avg:1828.39ms step:1570/6250 train_loss:2.0534 aux_loss:0.0274 train_time:2851468ms step_avg:1827.86ms step:1571/6250 train_loss:2.1609 aux_loss:0.0271 train_time:2852322ms step_avg:1827.24ms step:1572/6250 train_loss:2.1679 aux_loss:0.0279 train_time:2853328ms step_avg:1826.71ms step:1573/6250 train_loss:2.1679 aux_loss:0.0278 train_time:2855061ms step_avg:1826.65ms step:1574/6250 train_loss:2.1708 aux_loss:0.0274 train_time:2856202ms step_avg:1826.22ms step:1575/6250 train_loss:2.0969 aux_loss:0.0272 train_time:2857060ms step_avg:1825.60ms step:1576/6250 train_loss:2.0640 aux_loss:0.0270 train_time:2858000ms step_avg:1825.03ms step:1577/6250 train_loss:2.3208 aux_loss:0.0272 train_time:2859070ms step_avg:1824.55ms step:1578/6250 train_loss:2.0428 aux_loss:0.0274 train_time:2859945ms step_avg:1823.94ms step:1579/6250 train_loss:2.1362 aux_loss:0.0274 train_time:2860999ms step_avg:1823.45ms step:1580/6250 train_loss:2.2109 aux_loss:0.0273 train_time:2862160ms step_avg:1823.03ms step:1581/6250 train_loss:2.1656 aux_loss:0.0274 train_time:2863295ms step_avg:1822.59ms step:1582/6250 train_loss:2.2398 aux_loss:0.0274 train_time:2864669ms step_avg:1822.31ms step:1583/6250 train_loss:2.1978 aux_loss:0.0270 train_time:2869659ms step_avg:1824.32ms step:1584/6250 train_loss:2.1623 aux_loss:0.0271 train_time:2871522ms step_avg:1824.35ms step:1585/6250 train_loss:2.0531 aux_loss:0.0272 train_time:2873446ms step_avg:1824.41ms step:1586/6250 train_loss:2.3091 aux_loss:0.0269 train_time:2875364ms step_avg:1824.47ms step:1587/6250 train_loss:2.1800 aux_loss:0.0269 train_time:2878524ms step_avg:1825.32ms step:1588/6250 train_loss:2.1132 aux_loss:0.0273 train_time:2879875ms step_avg:1825.02ms step:1589/6250 train_loss:2.0804 aux_loss:0.0271 train_time:2881442ms step_avg:1824.85ms step:1590/6250 train_loss:2.2231 aux_loss:0.0269 train_time:2882816ms step_avg:1824.57ms step:1591/6250 train_loss:2.1945 aux_loss:0.0273 train_time:2884177ms step_avg:1824.27ms step:1592/6250 train_loss:2.1577 aux_loss:0.0273 train_time:2885490ms step_avg:1823.95ms step:1593/6250 train_loss:2.1366 aux_loss:0.0271 train_time:2886981ms step_avg:1823.74ms step:1594/6250 train_loss:2.1737 aux_loss:0.0272 train_time:2888406ms step_avg:1823.49ms step:1595/6250 train_loss:2.1799 aux_loss:0.0271 train_time:2890096ms step_avg:1823.40ms step:1596/6250 train_loss:2.2107 aux_loss:0.0275 train_time:2891653ms step_avg:1823.24ms step:1597/6250 train_loss:2.2620 aux_loss:0.0273 train_time:2892508ms step_avg:1822.63ms step:1598/6250 train_loss:2.0688 aux_loss:0.0270 train_time:2893893ms step_avg:1822.35ms step:1599/6250 train_loss:2.1203 aux_loss:0.0271 train_time:2894778ms step_avg:1821.76ms step:1600/6250 train_loss:2.1178 aux_loss:0.0270 train_time:2896074ms step_avg:1821.43ms step:1601/6250 train_loss:2.2152 aux_loss:0.0274 train_time:2896958ms step_avg:1820.84ms step:1602/6250 train_loss:2.2050 aux_loss:0.0271 train_time:2897841ms step_avg:1820.25ms step:1603/6250 train_loss:2.2154 aux_loss:0.0274 train_time:2898727ms step_avg:1819.67ms step:1604/6250 train_loss:2.1414 aux_loss:0.0272 train_time:2899578ms step_avg:1819.06ms step:1605/6250 train_loss:2.2136 aux_loss:0.0273 train_time:2900568ms step_avg:1818.54ms step:1606/6250 train_loss:2.1125 aux_loss:0.0275 train_time:2902033ms step_avg:1818.32ms step:1607/6250 train_loss:2.1949 aux_loss:0.0271 train_time:2902913ms step_avg:1817.73ms step:1608/6250 train_loss:2.1400 aux_loss:0.0268 train_time:2903786ms step_avg:1817.14ms step:1609/6250 train_loss:2.2017 aux_loss:0.0270 train_time:2904654ms step_avg:1816.54ms step:1610/6250 train_loss:2.1982 aux_loss:0.0271 train_time:2905505ms step_avg:1815.94ms step:1611/6250 train_loss:2.1702 aux_loss:0.0267 train_time:2906409ms step_avg:1815.37ms step:1612/6250 train_loss:2.1745 aux_loss:0.0269 train_time:2907276ms step_avg:1814.78ms step:1613/6250 train_loss:2.1703 aux_loss:0.0272 train_time:2908159ms step_avg:1814.20ms step:1614/6250 train_loss:2.1545 aux_loss:0.0272 train_time:2909041ms step_avg:1813.62ms step:1615/6250 train_loss:2.0793 aux_loss:0.0269 train_time:2909915ms step_avg:1813.03ms step:1616/6250 train_loss:2.0780 aux_loss:0.0269 train_time:2910773ms step_avg:1812.44ms step:1617/6250 train_loss:2.1355 aux_loss:0.0269 train_time:2911643ms step_avg:1811.85ms step:1618/6250 train_loss:2.1523 aux_loss:0.0267 train_time:2912484ms step_avg:1811.25ms step:1619/6250 train_loss:2.1501 aux_loss:0.0264 train_time:2913352ms step_avg:1810.66ms step:1620/6250 train_loss:2.1530 aux_loss:0.0264 train_time:2914210ms step_avg:1810.07ms step:1621/6250 train_loss:2.0223 aux_loss:0.0267 train_time:2915086ms step_avg:1809.49ms step:1622/6250 train_loss:2.2654 aux_loss:0.0269 train_time:2915947ms step_avg:1808.90ms step:1623/6250 train_loss:2.1985 aux_loss:0.0267 train_time:2916835ms step_avg:1808.33ms step:1624/6250 train_loss:2.1793 aux_loss:0.0266 train_time:2917984ms step_avg:1807.92ms step:1625/6250 train_loss:2.2245 aux_loss:0.0270 train_time:2918885ms step_avg:1807.36ms step:1626/6250 train_loss:2.1774 aux_loss:0.0271 train_time:2919745ms step_avg:1806.77ms step:1627/6250 train_loss:2.1819 aux_loss:0.0269 train_time:2920595ms step_avg:1806.18ms step:1628/6250 train_loss:2.2714 aux_loss:0.0270 train_time:2921473ms step_avg:1805.61ms step:1629/6250 train_loss:2.1443 aux_loss:0.0270 train_time:2922348ms step_avg:1805.03ms step:1630/6250 train_loss:2.1264 aux_loss:0.0268 train_time:2923405ms step_avg:1804.57ms step:1631/6250 train_loss:2.1226 aux_loss:0.0266 train_time:2924266ms step_avg:1803.99ms step:1632/6250 train_loss:2.2072 aux_loss:0.0269 train_time:2925172ms step_avg:1803.44ms step:1633/6250 train_loss:2.1033 aux_loss:0.0267 train_time:2926149ms step_avg:1802.93ms step:1634/6250 train_loss:2.2053 aux_loss:0.0266 train_time:2927015ms step_avg:1802.35ms step:1635/6250 train_loss:2.0858 aux_loss:0.0267 train_time:2927866ms step_avg:1801.76ms step:1636/6250 train_loss:2.1303 aux_loss:0.0268 train_time:2928804ms step_avg:1801.23ms step:1637/6250 train_loss:2.2310 aux_loss:0.0271 train_time:2929968ms step_avg:1800.84ms step:1638/6250 train_loss:2.2162 aux_loss:0.0268 train_time:2931023ms step_avg:1800.38ms step:1639/6250 train_loss:2.1072 aux_loss:0.0270 train_time:2931907ms step_avg:1799.82ms step:1640/6250 train_loss:2.1159 aux_loss:0.0271 train_time:2933252ms step_avg:1799.54ms step:1641/6250 train_loss:2.1490 aux_loss:0.0266 train_time:2934104ms step_avg:1798.96ms step:1642/6250 train_loss:2.1921 aux_loss:0.0270 train_time:2934966ms step_avg:1798.39ms step:1643/6250 train_loss:2.1670 aux_loss:0.0266 train_time:2936123ms step_avg:1797.99ms step:1644/6250 train_loss:2.2052 aux_loss:0.0264 train_time:2937008ms step_avg:1797.43ms step:1645/6250 train_loss:2.1597 aux_loss:0.0268 train_time:2938185ms step_avg:1797.06ms step:1646/6250 train_loss:2.2134 aux_loss:0.0268 train_time:2939053ms step_avg:1796.49ms step:1647/6250 train_loss:2.0754 aux_loss:0.0265 train_time:2940116ms step_avg:1796.04ms step:1648/6250 train_loss:2.1937 aux_loss:0.0265 train_time:2940987ms step_avg:1795.47ms step:1649/6250 train_loss:2.1862 aux_loss:0.0265 train_time:2941874ms step_avg:1794.92ms step:1650/6250 train_loss:2.1850 aux_loss:0.0270 train_time:2942743ms step_avg:1794.36ms step:1651/6250 train_loss:2.1696 aux_loss:0.0265 train_time:2943611ms step_avg:1793.79ms step:1652/6250 train_loss:2.1248 aux_loss:0.0264 train_time:2944487ms step_avg:1793.23ms step:1653/6250 train_loss:2.1039 aux_loss:0.0264 train_time:2945366ms step_avg:1792.68ms step:1654/6250 train_loss:2.0362 aux_loss:0.0265 train_time:2946357ms step_avg:1792.19ms step:1655/6250 train_loss:2.1911 aux_loss:0.0263 train_time:2947246ms step_avg:1791.64ms step:1656/6250 train_loss:2.2646 aux_loss:0.0260 train_time:2948164ms step_avg:1791.11ms step:1657/6250 train_loss:2.1640 aux_loss:0.0264 train_time:2949764ms step_avg:1790.99ms step:1658/6250 train_loss:2.0414 aux_loss:0.0263 train_time:2950875ms step_avg:1790.58ms step:1659/6250 train_loss:2.2410 aux_loss:0.0261 train_time:2951958ms step_avg:1790.15ms step:1660/6250 train_loss:2.0918 aux_loss:0.0261 train_time:2952833ms step_avg:1789.60ms step:1661/6250 train_loss:2.1692 aux_loss:0.0261 train_time:2953695ms step_avg:1789.03ms step:1662/6250 train_loss:2.0806 aux_loss:0.0265 train_time:2955108ms step_avg:1788.81ms step:1663/6250 train_loss:2.0515 aux_loss:0.0263 train_time:2957017ms step_avg:1788.88ms step:1664/6250 train_loss:2.2394 aux_loss:0.0262 train_time:2958186ms step_avg:1788.50ms step:1665/6250 train_loss:2.2462 aux_loss:0.0263 train_time:2959045ms step_avg:1787.94ms step:1666/6250 train_loss:2.1635 aux_loss:0.0265 train_time:2959883ms step_avg:1787.37ms step:1667/6250 train_loss:2.0799 aux_loss:0.0263 train_time:2960740ms step_avg:1786.81ms step:1668/6250 train_loss:2.1922 aux_loss:0.0262 train_time:2961592ms step_avg:1786.24ms step:1669/6250 train_loss:2.2378 aux_loss:0.0264 train_time:2962766ms step_avg:1785.87ms step:1670/6250 train_loss:2.1266 aux_loss:0.0265 train_time:2963738ms step_avg:1785.38ms step:1671/6250 train_loss:2.1325 aux_loss:0.0264 train_time:2965651ms step_avg:1785.46ms step:1672/6250 train_loss:2.1793 aux_loss:0.0263 train_time:2967195ms step_avg:1785.32ms step:1673/6250 train_loss:2.1115 aux_loss:0.0261 train_time:2968687ms step_avg:1785.14ms step:1674/6250 train_loss:2.0916 aux_loss:0.0262 train_time:2971009ms step_avg:1785.46ms step:1675/6250 train_loss:2.0952 aux_loss:0.0260 train_time:2971894ms step_avg:1784.92ms step:1676/6250 train_loss:2.2127 aux_loss:0.0263 train_time:2972793ms step_avg:1784.39ms step:1677/6250 train_loss:2.1659 aux_loss:0.0266 train_time:2973667ms step_avg:1783.84ms step:1678/6250 train_loss:2.1751 aux_loss:0.0260 train_time:2974558ms step_avg:1783.31ms step:1679/6250 train_loss:1.9890 aux_loss:0.0263 train_time:2975430ms step_avg:1782.76ms step:1680/6250 train_loss:2.1906 aux_loss:0.0270 train_time:2976312ms step_avg:1782.22ms step:1681/6250 train_loss:2.1130 aux_loss:0.0261 train_time:2977277ms step_avg:1781.73ms step:1682/6250 train_loss:2.0975 aux_loss:0.0261 train_time:2978224ms step_avg:1781.23ms step:1683/6250 train_loss:2.1391 aux_loss:0.0262 train_time:2979109ms step_avg:1780.70ms step:1684/6250 train_loss:2.1033 aux_loss:0.0263 train_time:2980069ms step_avg:1780.21ms step:1685/6250 train_loss:2.2140 aux_loss:0.0264 train_time:2981031ms step_avg:1779.72ms step:1686/6250 train_loss:2.0854 aux_loss:0.0256 train_time:2981992ms step_avg:1779.23ms step:1687/6250 train_loss:2.1450 aux_loss:0.0262 train_time:2982876ms step_avg:1778.70ms step:1688/6250 train_loss:2.0783 aux_loss:0.0259 train_time:2983805ms step_avg:1778.19ms step:1689/6250 train_loss:2.1149 aux_loss:0.0257 train_time:2984707ms step_avg:1777.67ms step:1690/6250 train_loss:2.2020 aux_loss:0.0259 train_time:2985580ms step_avg:1777.13ms step:1691/6250 train_loss:2.1824 aux_loss:0.0260 train_time:2986467ms step_avg:1776.60ms step:1692/6250 train_loss:2.1848 aux_loss:0.0256 train_time:2987361ms step_avg:1776.08ms step:1693/6250 train_loss:2.1289 aux_loss:0.0255 train_time:2988259ms step_avg:1775.55ms step:1694/6250 train_loss:2.2064 aux_loss:0.0259 train_time:2989168ms step_avg:1775.04ms step:1695/6250 train_loss:2.1021 aux_loss:0.0263 train_time:2990048ms step_avg:1774.51ms step:1696/6250 train_loss:2.1410 aux_loss:0.0260 train_time:2990919ms step_avg:1773.97ms step:1697/6250 train_loss:2.0770 aux_loss:0.0260 train_time:2991796ms step_avg:1773.44ms step:1698/6250 train_loss:2.2299 aux_loss:0.0260 train_time:2992738ms step_avg:1772.95ms step:1699/6250 train_loss:2.1060 aux_loss:0.0256 train_time:2993770ms step_avg:1772.51ms step:1700/6250 train_loss:2.1077 aux_loss:0.0253 train_time:2994668ms step_avg:1771.99ms step:1701/6250 train_loss:2.1203 aux_loss:0.0256 train_time:2995557ms step_avg:1771.47ms step:1702/6250 train_loss:2.0806 aux_loss:0.0258 train_time:2997264ms step_avg:1771.43ms step:1703/6250 train_loss:2.0911 aux_loss:0.0255 train_time:2998887ms step_avg:1771.35ms step:1704/6250 train_loss:2.1241 aux_loss:0.0252 train_time:2999757ms step_avg:1770.81ms step:1705/6250 train_loss:2.0534 aux_loss:0.0258 train_time:3000607ms step_avg:1770.27ms step:1706/6250 train_loss:2.0548 aux_loss:0.0263 train_time:3002058ms step_avg:1770.08ms step:1707/6250 train_loss:2.1284 aux_loss:0.0261 train_time:3003136ms step_avg:1769.67ms step:1708/6250 train_loss:2.1073 aux_loss:0.0259 train_time:3004260ms step_avg:1769.29ms step:1709/6250 train_loss:2.1620 aux_loss:0.0261 train_time:3005587ms step_avg:1769.03ms step:1710/6250 train_loss:2.2049 aux_loss:0.0263 train_time:3006643ms step_avg:1768.61ms step:1711/6250 train_loss:2.1168 aux_loss:0.0260 train_time:3007537ms step_avg:1768.10ms step:1712/6250 train_loss:2.0937 aux_loss:0.0257 train_time:3008511ms step_avg:1767.63ms step:1713/6250 train_loss:2.0722 aux_loss:0.0261 train_time:3009462ms step_avg:1767.15ms step:1714/6250 train_loss:2.1358 aux_loss:0.0262 train_time:3010557ms step_avg:1766.76ms step:1715/6250 train_loss:2.1603 aux_loss:0.0259 train_time:3011437ms step_avg:1766.24ms step:1716/6250 train_loss:2.1749 aux_loss:0.0257 train_time:3012311ms step_avg:1765.72ms step:1717/6250 train_loss:2.1466 aux_loss:0.0259 train_time:3013368ms step_avg:1765.30ms step:1718/6250 train_loss:2.1991 aux_loss:0.0258 train_time:3014315ms step_avg:1764.82ms step:1719/6250 train_loss:2.1316 aux_loss:0.0258 train_time:3015539ms step_avg:1764.51ms step:1720/6250 train_loss:2.0987 aux_loss:0.0257 train_time:3016413ms step_avg:1763.98ms step:1721/6250 train_loss:2.1519 aux_loss:0.0258 train_time:3017289ms step_avg:1763.47ms step:1722/6250 train_loss:2.0733 aux_loss:0.0252 train_time:3018162ms step_avg:1762.95ms step:1723/6250 train_loss:2.1358 aux_loss:0.0255 train_time:3019246ms step_avg:1762.55ms step:1724/6250 train_loss:2.0747 aux_loss:0.0255 train_time:3020113ms step_avg:1762.03ms step:1725/6250 train_loss:2.1189 aux_loss:0.0259 train_time:3020983ms step_avg:1761.51ms step:1726/6250 train_loss:2.1575 aux_loss:0.0257 train_time:3021941ms step_avg:1761.04ms step:1727/6250 train_loss:2.1741 aux_loss:0.0257 train_time:3023309ms step_avg:1760.81ms step:1728/6250 train_loss:2.1685 aux_loss:0.0255 train_time:3024446ms step_avg:1760.45ms step:1729/6250 train_loss:2.1934 aux_loss:0.0257 train_time:3025618ms step_avg:1760.10ms step:1730/6250 train_loss:2.0865 aux_loss:0.0254 train_time:3027571ms step_avg:1760.22ms step:1731/6250 train_loss:2.2173 aux_loss:0.0257 train_time:3028962ms step_avg:1760.00ms step:1732/6250 train_loss:2.0872 aux_loss:0.0254 train_time:3030446ms step_avg:1759.84ms step:1733/6250 train_loss:2.1217 aux_loss:0.0259 train_time:3031609ms step_avg:1759.49ms step:1734/6250 train_loss:2.0640 aux_loss:0.0258 train_time:3032799ms step_avg:1759.16ms step:1735/6250 train_loss:2.0349 aux_loss:0.0257 train_time:3033898ms step_avg:1758.78ms step:1736/6250 train_loss:2.0844 aux_loss:0.0256 train_time:3035019ms step_avg:1758.41ms step:1737/6250 train_loss:2.1406 aux_loss:0.0260 train_time:3036065ms step_avg:1758.00ms step:1738/6250 train_loss:2.1789 aux_loss:0.0253 train_time:3036942ms step_avg:1757.49ms step:1739/6250 train_loss:2.1813 aux_loss:0.0256 train_time:3037906ms step_avg:1757.03ms step:1740/6250 train_loss:2.0638 aux_loss:0.0254 train_time:3039266ms step_avg:1756.80ms step:1741/6250 train_loss:2.0860 aux_loss:0.0256 train_time:3040281ms step_avg:1756.37ms step:1742/6250 train_loss:2.1518 aux_loss:0.0258 train_time:3041257ms step_avg:1755.92ms step:1743/6250 train_loss:2.1381 aux_loss:0.0256 train_time:3043156ms step_avg:1756.00ms step:1744/6250 train_loss:2.1318 aux_loss:0.0255 train_time:3044306ms step_avg:1755.66ms step:1745/6250 train_loss:2.0744 aux_loss:0.0257 train_time:3046086ms step_avg:1755.67ms step:1746/6250 train_loss:2.1603 aux_loss:0.0252 train_time:3047205ms step_avg:1755.30ms step:1747/6250 train_loss:2.1254 aux_loss:0.0255 train_time:3048695ms step_avg:1755.15ms step:1748/6250 train_loss:2.0917 aux_loss:0.0257 train_time:3050692ms step_avg:1755.29ms step:1749/6250 train_loss:2.0653 aux_loss:0.0252 train_time:3051533ms step_avg:1754.76ms step:1750/6250 train_loss:2.1542 aux_loss:0.0255 train_time:3052396ms step_avg:1754.25ms step:1751/6250 train_loss:2.1171 aux_loss:0.0258 train_time:3053258ms step_avg:1753.74ms step:1752/6250 train_loss:2.0693 aux_loss:0.0255 train_time:3054123ms step_avg:1753.23ms step:1753/6250 train_loss:2.0866 aux_loss:0.0255 train_time:3055443ms step_avg:1752.98ms step:1754/6250 train_loss:2.1312 aux_loss:0.0255 train_time:3056623ms step_avg:1752.65ms step:1755/6250 train_loss:2.1490 aux_loss:0.0254 train_time:3057499ms step_avg:1752.15ms step:1756/6250 train_loss:2.0433 aux_loss:0.0253 train_time:3058368ms step_avg:1751.64ms step:1757/6250 train_loss:2.1702 aux_loss:0.0252 train_time:3059240ms step_avg:1751.14ms step:1758/6250 train_loss:2.1748 aux_loss:0.0252 train_time:3060110ms step_avg:1750.64ms step:1759/6250 train_loss:2.1178 aux_loss:0.0248 train_time:3060972ms step_avg:1750.13ms step:1760/6250 train_loss:2.1318 aux_loss:0.0252 train_time:3061839ms step_avg:1749.62ms step:1761/6250 train_loss:2.1082 aux_loss:0.0252 train_time:3062695ms step_avg:1749.11ms step:1762/6250 train_loss:2.0453 aux_loss:0.0253 train_time:3063703ms step_avg:1748.69ms step:1763/6250 train_loss:2.0289 aux_loss:0.0256 train_time:3064558ms step_avg:1748.18ms step:1764/6250 train_loss:2.2105 aux_loss:0.0255 train_time:3065906ms step_avg:1747.95ms step:1765/6250 train_loss:2.0632 aux_loss:0.0254 train_time:3066767ms step_avg:1747.45ms step:1766/6250 train_loss:2.1292 aux_loss:0.0255 train_time:3067617ms step_avg:1746.93ms step:1767/6250 train_loss:2.1688 aux_loss:0.0253 train_time:3068479ms step_avg:1746.43ms step:1768/6250 train_loss:2.1335 aux_loss:0.0253 train_time:3069418ms step_avg:1745.97ms step:1769/6250 train_loss:2.1735 aux_loss:0.0252 train_time:3070354ms step_avg:1745.51ms step:1770/6250 train_loss:2.1144 aux_loss:0.0252 train_time:3071550ms step_avg:1745.20ms step:1771/6250 train_loss:2.1171 aux_loss:0.0256 train_time:3072579ms step_avg:1744.79ms step:1772/6250 train_loss:2.1224 aux_loss:0.0255 train_time:3073660ms step_avg:1744.42ms step:1773/6250 train_loss:2.1293 aux_loss:0.0252 train_time:3074512ms step_avg:1743.91ms step:1774/6250 train_loss:2.1098 aux_loss:0.0255 train_time:3075373ms step_avg:1743.41ms step:1775/6250 train_loss:2.1236 aux_loss:0.0254 train_time:3076246ms step_avg:1742.92ms step:1776/6250 train_loss:2.1698 aux_loss:0.0252 train_time:3077104ms step_avg:1742.41ms step:1777/6250 train_loss:2.1116 aux_loss:0.0252 train_time:3079045ms step_avg:1742.53ms step:1778/6250 train_loss:2.0972 aux_loss:0.0256 train_time:3080261ms step_avg:1742.23ms step:1779/6250 train_loss:2.0293 aux_loss:0.0255 train_time:3081149ms step_avg:1741.75ms step:1780/6250 train_loss:2.1592 aux_loss:0.0250 train_time:3082018ms step_avg:1741.25ms step:1781/6250 train_loss:2.1524 aux_loss:0.0248 train_time:3082894ms step_avg:1740.76ms step:1782/6250 train_loss:2.1162 aux_loss:0.0251 train_time:3083756ms step_avg:1740.27ms step:1783/6250 train_loss:2.0995 aux_loss:0.0248 train_time:3084596ms step_avg:1739.76ms step:1784/6250 train_loss:2.0708 aux_loss:0.0249 train_time:3085641ms step_avg:1739.37ms step:1785/6250 train_loss:2.0859 aux_loss:0.0251 train_time:3086544ms step_avg:1738.90ms step:1786/6250 train_loss:2.0981 aux_loss:0.0250 train_time:3087587ms step_avg:1738.51ms step:1787/6250 train_loss:2.1495 aux_loss:0.0249 train_time:3088909ms step_avg:1738.27ms step:1788/6250 train_loss:2.1536 aux_loss:0.0255 train_time:3089808ms step_avg:1737.80ms step:1789/6250 train_loss:1.9784 aux_loss:0.0254 train_time:3090692ms step_avg:1737.32ms step:1790/6250 train_loss:2.0698 aux_loss:0.0251 train_time:3091553ms step_avg:1736.83ms step:1791/6250 train_loss:2.1822 aux_loss:0.0249 train_time:3092497ms step_avg:1736.38ms step:1792/6250 train_loss:2.1178 aux_loss:0.0249 train_time:3093685ms step_avg:1736.07ms step:1793/6250 train_loss:2.0817 aux_loss:0.0250 train_time:3094536ms step_avg:1735.58ms step:1794/6250 train_loss:2.1432 aux_loss:0.0250 train_time:3095423ms step_avg:1735.10ms step:1795/6250 train_loss:2.0608 aux_loss:0.0249 train_time:3096686ms step_avg:1734.84ms step:1796/6250 train_loss:2.0831 aux_loss:0.0252 train_time:3097778ms step_avg:1734.48ms step:1797/6250 train_loss:2.0933 aux_loss:0.0252 train_time:3098646ms step_avg:1733.99ms step:1798/6250 train_loss:2.0654 aux_loss:0.0249 train_time:3099487ms step_avg:1733.49ms step:1799/6250 train_loss:2.1780 aux_loss:0.0251 train_time:3100368ms step_avg:1733.02ms step:1800/6250 train_loss:2.1328 aux_loss:0.0251 train_time:3101237ms step_avg:1732.53ms step:1801/6250 train_loss:2.1389 aux_loss:0.0248 train_time:3102099ms step_avg:1732.05ms step:1802/6250 train_loss:2.0981 aux_loss:0.0250 train_time:3102965ms step_avg:1731.57ms step:1803/6250 train_loss:2.1562 aux_loss:0.0249 train_time:3103845ms step_avg:1731.09ms step:1804/6250 train_loss:2.2126 aux_loss:0.0247 train_time:3104777ms step_avg:1730.64ms step:1805/6250 train_loss:2.1398 aux_loss:0.0248 train_time:3105981ms step_avg:1730.35ms step:1806/6250 train_loss:2.0791 aux_loss:0.0248 train_time:3106983ms step_avg:1729.95ms step:1807/6250 train_loss:2.1418 aux_loss:0.0254 train_time:3108237ms step_avg:1729.68ms step:1808/6250 train_loss:2.0394 aux_loss:0.0249 train_time:3109591ms step_avg:1729.47ms step:1809/6250 train_loss:2.0772 aux_loss:0.0245 train_time:3110698ms step_avg:1729.13ms step:1810/6250 train_loss:2.1155 aux_loss:0.0248 train_time:3112033ms step_avg:1728.91ms step:1811/6250 train_loss:2.1864 aux_loss:0.0247 train_time:3113503ms step_avg:1728.76ms step:1812/6250 train_loss:2.0334 aux_loss:0.0246 train_time:3115256ms step_avg:1728.78ms step:1813/6250 train_loss:2.1441 aux_loss:0.0244 train_time:3116156ms step_avg:1728.32ms step:1814/6250 train_loss:2.0880 aux_loss:0.0248 train_time:3117022ms step_avg:1727.84ms step:1815/6250 train_loss:2.0376 aux_loss:0.0246 train_time:3119815ms step_avg:1728.43ms step:1816/6250 train_loss:2.0616 aux_loss:0.0249 train_time:3122119ms step_avg:1728.75ms step:1817/6250 train_loss:2.1322 aux_loss:0.0244 train_time:3124989ms step_avg:1729.38ms step:1818/6250 train_loss:2.0672 aux_loss:0.0244 train_time:3126489ms step_avg:1729.25ms step:1819/6250 train_loss:2.0218 aux_loss:0.0245 train_time:3127502ms step_avg:1728.86ms step:1820/6250 train_loss:2.0658 aux_loss:0.0245 train_time:3128848ms step_avg:1728.65ms step:1821/6250 train_loss:2.0972 aux_loss:0.0250 train_time:3129893ms step_avg:1728.27ms step:1822/6250 train_loss:2.1328 aux_loss:0.0253 train_time:3130999ms step_avg:1727.92ms step:1823/6250 train_loss:2.0891 aux_loss:0.0245 train_time:3132414ms step_avg:1727.75ms step:1824/6250 train_loss:2.1674 aux_loss:0.0251 train_time:3133810ms step_avg:1727.57ms step:1825/6250 train_loss:2.0898 aux_loss:0.0249 train_time:3134929ms step_avg:1727.23ms step:1826/6250 train_loss:2.1594 aux_loss:0.0244 train_time:3135807ms step_avg:1726.77ms step:1827/6250 train_loss:2.0915 aux_loss:0.0246 train_time:3136663ms step_avg:1726.29ms step:1828/6250 train_loss:2.0520 aux_loss:0.0248 train_time:3137513ms step_avg:1725.80ms step:1829/6250 train_loss:2.2947 aux_loss:0.0249 train_time:3138377ms step_avg:1725.33ms step:1830/6250 train_loss:2.0555 aux_loss:0.0248 train_time:3139327ms step_avg:1724.91ms step:1831/6250 train_loss:2.1787 aux_loss:0.0245 train_time:3140388ms step_avg:1724.54ms step:1832/6250 train_loss:2.0336 aux_loss:0.0245 train_time:3141233ms step_avg:1724.06ms step:1833/6250 train_loss:2.2013 aux_loss:0.0251 train_time:3142093ms step_avg:1723.58ms step:1834/6250 train_loss:2.1321 aux_loss:0.0246 train_time:3143384ms step_avg:1723.35ms step:1835/6250 train_loss:2.1079 aux_loss:0.0242 train_time:3144276ms step_avg:1722.89ms step:1836/6250 train_loss:2.2174 aux_loss:0.0246 train_time:3145148ms step_avg:1722.43ms step:1837/6250 train_loss:2.0887 aux_loss:0.0248 train_time:3146036ms step_avg:1721.97ms step:1838/6250 train_loss:2.1275 aux_loss:0.0245 train_time:3146902ms step_avg:1721.50ms step:1839/6250 train_loss:2.0757 aux_loss:0.0243 train_time:3147782ms step_avg:1721.04ms step:1840/6250 train_loss:2.0798 aux_loss:0.0245 train_time:3148651ms step_avg:1720.57ms step:1841/6250 train_loss:2.1323 aux_loss:0.0249 train_time:3149963ms step_avg:1720.35ms step:1842/6250 train_loss:2.0801 aux_loss:0.0245 train_time:3150842ms step_avg:1719.89ms step:1843/6250 train_loss:2.1659 aux_loss:0.0244 train_time:3151713ms step_avg:1719.43ms step:1844/6250 train_loss:2.1710 aux_loss:0.0247 train_time:3152559ms step_avg:1718.95ms step:1845/6250 train_loss:2.0877 aux_loss:0.0247 train_time:3153423ms step_avg:1718.49ms step:1846/6250 train_loss:2.1396 aux_loss:0.0248 train_time:3154553ms step_avg:1718.17ms step:1847/6250 train_loss:2.1482 aux_loss:0.0244 train_time:3157288ms step_avg:1718.72ms step:1848/6250 train_loss:1.9876 aux_loss:0.0244 train_time:3158638ms step_avg:1718.52ms step:1849/6250 train_loss:2.1567 aux_loss:0.0247 train_time:3159511ms step_avg:1718.06ms step:1850/6250 train_loss:2.1658 aux_loss:0.0246 train_time:3160637ms step_avg:1717.74ms step:1851/6250 train_loss:2.0394 aux_loss:0.0246 train_time:3161501ms step_avg:1717.27ms step:1852/6250 train_loss:2.0387 aux_loss:0.0244 train_time:3162431ms step_avg:1716.85ms step:1853/6250 train_loss:2.1285 aux_loss:0.0247 train_time:3163331ms step_avg:1716.40ms step:1854/6250 train_loss:2.0750 aux_loss:0.0247 train_time:3164287ms step_avg:1715.99ms step:1855/6250 train_loss:2.0536 aux_loss:0.0249 train_time:3165253ms step_avg:1715.58ms step:1856/6250 train_loss:2.1694 aux_loss:0.0247 train_time:3166331ms step_avg:1715.24ms step:1857/6250 train_loss:2.1605 aux_loss:0.0249 train_time:3167874ms step_avg:1715.15ms step:1858/6250 train_loss:2.0721 aux_loss:0.0243 train_time:3169106ms step_avg:1714.88ms step:1859/6250 train_loss:2.2088 aux_loss:0.0247 train_time:3170129ms step_avg:1714.51ms step:1860/6250 train_loss:2.0007 aux_loss:0.0246 train_time:3171091ms step_avg:1714.10ms step:1861/6250 train_loss:2.2259 aux_loss:0.0244 train_time:3172161ms step_avg:1713.76ms step:1862/6250 train_loss:2.1189 aux_loss:0.0247 train_time:3173367ms step_avg:1713.48ms step:1863/6250 train_loss:2.0811 aux_loss:0.0246 train_time:3174584ms step_avg:1713.21ms step:1864/6250 train_loss:2.1633 aux_loss:0.0242 train_time:3176116ms step_avg:1713.12ms step:1865/6250 train_loss:2.0259 aux_loss:0.0241 train_time:3177662ms step_avg:1713.03ms step:1866/6250 train_loss:2.0410 aux_loss:0.0243 train_time:3178553ms step_avg:1712.58ms step:1867/6250 train_loss:2.1409 aux_loss:0.0243 train_time:3179573ms step_avg:1712.21ms step:1868/6250 train_loss:2.0694 aux_loss:0.0242 train_time:3181267ms step_avg:1712.20ms step:1869/6250 train_loss:2.1036 aux_loss:0.0247 train_time:3182427ms step_avg:1711.90ms step:1870/6250 train_loss:2.1360 aux_loss:0.0249 train_time:3183486ms step_avg:1711.55ms step:1871/6250 train_loss:2.0895 aux_loss:0.0245 train_time:3184390ms step_avg:1711.12ms step:1872/6250 train_loss:2.1900 aux_loss:0.0241 train_time:3185388ms step_avg:1710.73ms step:1873/6250 train_loss:1.9878 aux_loss:0.0241 train_time:3186372ms step_avg:1710.34ms step:1874/6250 train_loss:2.1075 aux_loss:0.0244 train_time:3187269ms step_avg:1709.91ms step:1875/6250 train_loss:1.9630 aux_loss:0.0238 train_time:3188206ms step_avg:1709.49ms step:1876/6250 train_loss:2.1567 aux_loss:0.0243 train_time:3189147ms step_avg:1709.08ms step:1877/6250 train_loss:2.1131 aux_loss:0.0243 train_time:3190114ms step_avg:1708.68ms step:1878/6250 train_loss:2.1265 aux_loss:0.0242 train_time:3191020ms step_avg:1708.25ms step:1879/6250 train_loss:2.1309 aux_loss:0.0246 train_time:3191929ms step_avg:1707.83ms step:1880/6250 train_loss:2.1166 aux_loss:0.0246 train_time:3192870ms step_avg:1707.42ms step:1881/6250 train_loss:2.0804 aux_loss:0.0244 train_time:3193791ms step_avg:1707.00ms step:1882/6250 train_loss:2.0280 aux_loss:0.0239 train_time:3194705ms step_avg:1706.57ms step:1883/6250 train_loss:2.1474 aux_loss:0.0240 train_time:3195635ms step_avg:1706.16ms step:1884/6250 train_loss:2.0760 aux_loss:0.0241 train_time:3196574ms step_avg:1705.75ms step:1885/6250 train_loss:2.0928 aux_loss:0.0239 train_time:3197483ms step_avg:1705.32ms step:1886/6250 train_loss:2.0842 aux_loss:0.0237 train_time:3198377ms step_avg:1704.89ms step:1887/6250 train_loss:2.0647 aux_loss:0.0241 train_time:3199269ms step_avg:1704.46ms step:1888/6250 train_loss:2.0249 aux_loss:0.0244 train_time:3200280ms step_avg:1704.09ms step:1889/6250 train_loss:2.0689 aux_loss:0.0241 train_time:3201222ms step_avg:1703.68ms step:1890/6250 train_loss:2.0625 aux_loss:0.0245 train_time:3202108ms step_avg:1703.25ms step:1891/6250 train_loss:2.0839 aux_loss:0.0241 train_time:3202986ms step_avg:1702.81ms step:1892/6250 train_loss:2.0980 aux_loss:0.0242 train_time:3203870ms step_avg:1702.38ms step:1893/6250 train_loss:2.1354 aux_loss:0.0243 train_time:3204732ms step_avg:1701.93ms step:1894/6250 train_loss:2.1116 aux_loss:0.0243 train_time:3205615ms step_avg:1701.49ms step:1895/6250 train_loss:2.0923 aux_loss:0.0245 train_time:3206609ms step_avg:1701.12ms step:1896/6250 train_loss:2.1495 aux_loss:0.0242 train_time:3207560ms step_avg:1700.72ms step:1897/6250 train_loss:2.0427 aux_loss:0.0241 train_time:3208504ms step_avg:1700.32ms step:1898/6250 train_loss:2.0949 aux_loss:0.0239 train_time:3209457ms step_avg:1699.92ms step:1899/6250 train_loss:2.1785 aux_loss:0.0242 train_time:3210415ms step_avg:1699.53ms step:1900/6250 train_loss:2.0717 aux_loss:0.0239 train_time:3211345ms step_avg:1699.12ms step:1901/6250 train_loss:2.0384 aux_loss:0.0242 train_time:3212285ms step_avg:1698.72ms step:1902/6250 train_loss:2.0695 aux_loss:0.0247 train_time:3213203ms step_avg:1698.31ms step:1903/6250 train_loss:2.0898 aux_loss:0.0241 train_time:3214075ms step_avg:1697.87ms step:1904/6250 train_loss:2.0778 aux_loss:0.0242 train_time:3214955ms step_avg:1697.44ms step:1905/6250 train_loss:2.0828 aux_loss:0.0241 train_time:3215868ms step_avg:1697.03ms step:1906/6250 train_loss:2.0447 aux_loss:0.0234 train_time:3216738ms step_avg:1696.59ms step:1907/6250 train_loss:2.1429 aux_loss:0.0241 train_time:3217646ms step_avg:1696.18ms step:1908/6250 train_loss:2.0924 aux_loss:0.0242 train_time:3218532ms step_avg:1695.75ms step:1909/6250 train_loss:2.0468 aux_loss:0.0238 train_time:3219425ms step_avg:1695.33ms step:1910/6250 train_loss:2.1073 aux_loss:0.0239 train_time:3220375ms step_avg:1694.93ms step:1911/6250 train_loss:2.0721 aux_loss:0.0242 train_time:3221292ms step_avg:1694.52ms step:1912/6250 train_loss:2.2132 aux_loss:0.0239 train_time:3222159ms step_avg:1694.09ms step:1913/6250 train_loss:2.0263 aux_loss:0.0243 train_time:3223050ms step_avg:1693.67ms step:1914/6250 train_loss:2.0548 aux_loss:0.0244 train_time:3224107ms step_avg:1693.33ms step:1915/6250 train_loss:2.0939 aux_loss:0.0239 train_time:3225076ms step_avg:1692.95ms step:1916/6250 train_loss:2.1550 aux_loss:0.0242 train_time:3226034ms step_avg:1692.57ms step:1917/6250 train_loss:1.9943 aux_loss:0.0246 train_time:3236803ms step_avg:1697.33ms step:1918/6250 train_loss:2.1310 aux_loss:0.0243 train_time:3238723ms step_avg:1697.44ms step:1919/6250 train_loss:2.0907 aux_loss:0.0239 train_time:3243797ms step_avg:1699.21ms step:1920/6250 train_loss:2.1267 aux_loss:0.0239 train_time:3244671ms step_avg:1698.78ms step:1921/6250 train_loss:2.0657 aux_loss:0.0241 train_time:3245654ms step_avg:1698.41ms step:1922/6250 train_loss:2.0415 aux_loss:0.0239 train_time:3246669ms step_avg:1698.05ms step:1923/6250 train_loss:2.1401 aux_loss:0.0236 train_time:3247600ms step_avg:1697.65ms step:1924/6250 train_loss:2.1278 aux_loss:0.0239 train_time:3248555ms step_avg:1697.26ms step:1925/6250 train_loss:2.1754 aux_loss:0.0238 train_time:3249473ms step_avg:1696.85ms step:1926/6250 train_loss:2.0427 aux_loss:0.0237 train_time:3250418ms step_avg:1696.46ms step:1927/6250 train_loss:2.0413 aux_loss:0.0237 train_time:3251337ms step_avg:1696.05ms step:1928/6250 train_loss:2.1147 aux_loss:0.0239 train_time:3252359ms step_avg:1695.70ms step:1929/6250 train_loss:2.1178 aux_loss:0.0239 train_time:3253321ms step_avg:1695.32ms step:1930/6250 train_loss:2.0451 aux_loss:0.0241 train_time:3254559ms step_avg:1695.08ms step:1931/6250 train_loss:2.1055 aux_loss:0.0242 train_time:3258729ms step_avg:1696.37ms step:1932/6250 train_loss:2.0410 aux_loss:0.0242 train_time:3259998ms step_avg:1696.15ms step:1933/6250 train_loss:2.0485 aux_loss:0.0238 train_time:3261649ms step_avg:1696.13ms step:1934/6250 train_loss:2.1228 aux_loss:0.0238 train_time:3263218ms step_avg:1696.06ms step:1935/6250 train_loss:2.1064 aux_loss:0.0236 train_time:3264173ms step_avg:1695.67ms step:1936/6250 train_loss:2.1287 aux_loss:0.0234 train_time:3265146ms step_avg:1695.30ms step:1937/6250 train_loss:2.0818 aux_loss:0.0236 train_time:3266133ms step_avg:1694.93ms step:1938/6250 train_loss:2.0578 aux_loss:0.0234 train_time:3267062ms step_avg:1694.53ms step:1939/6250 train_loss:2.0795 aux_loss:0.0235 train_time:3267998ms step_avg:1694.14ms step:1940/6250 train_loss:2.0495 aux_loss:0.0237 train_time:3268875ms step_avg:1693.72ms step:1941/6250 train_loss:2.0736 aux_loss:0.0239 train_time:3269745ms step_avg:1693.29ms step:1942/6250 train_loss:2.1946 aux_loss:0.0240 train_time:3270589ms step_avg:1692.85ms step:1943/6250 train_loss:2.0289 aux_loss:0.0235 train_time:3271524ms step_avg:1692.46ms step:1944/6250 train_loss:2.0968 aux_loss:0.0235 train_time:3272441ms step_avg:1692.06ms step:1945/6250 train_loss:2.0697 aux_loss:0.0242 train_time:3273334ms step_avg:1691.65ms step:1946/6250 train_loss:2.1572 aux_loss:0.0237 train_time:3274217ms step_avg:1691.23ms step:1947/6250 train_loss:2.1531 aux_loss:0.0235 train_time:3275083ms step_avg:1690.80ms step:1948/6250 train_loss:2.0900 aux_loss:0.0236 train_time:3275963ms step_avg:1690.38ms step:1949/6250 train_loss:2.0315 aux_loss:0.0235 train_time:3276882ms step_avg:1689.99ms step:1950/6250 train_loss:2.0045 aux_loss:0.0234 train_time:3277762ms step_avg:1689.57ms step:1951/6250 train_loss:2.0615 aux_loss:0.0236 train_time:3278642ms step_avg:1689.15ms step:1952/6250 train_loss:2.0794 aux_loss:0.0236 train_time:3279501ms step_avg:1688.72ms step:1953/6250 train_loss:1.9887 aux_loss:0.0239 train_time:3280443ms step_avg:1688.34ms step:1954/6250 train_loss:2.0655 aux_loss:0.0238 train_time:3281338ms step_avg:1687.93ms step:1955/6250 train_loss:2.0611 aux_loss:0.0234 train_time:3282206ms step_avg:1687.51ms step:1956/6250 train_loss:2.0178 aux_loss:0.0234 train_time:3283092ms step_avg:1687.10ms step:1957/6250 train_loss:2.0782 aux_loss:0.0237 train_time:3283980ms step_avg:1686.69ms step:1958/6250 train_loss:2.0827 aux_loss:0.0242 train_time:3285746ms step_avg:1686.73ms step:1959/6250 train_loss:2.0726 aux_loss:0.0234 train_time:3286666ms step_avg:1686.33ms step:1960/6250 train_loss:2.1306 aux_loss:0.0239 train_time:3288346ms step_avg:1686.33ms step:1961/6250 train_loss:2.0991 aux_loss:0.0238 train_time:3289440ms step_avg:1686.03ms step:1962/6250 train_loss:2.0265 aux_loss:0.0236 train_time:3290312ms step_avg:1685.61ms step:1963/6250 train_loss:2.1216 aux_loss:0.0236 train_time:3291191ms step_avg:1685.20ms step:1964/6250 train_loss:2.0338 aux_loss:0.0242 train_time:3292604ms step_avg:1685.06ms step:1965/6250 train_loss:2.2451 aux_loss:0.0241 train_time:3293800ms step_avg:1684.81ms step:1966/6250 train_loss:2.0020 aux_loss:0.0235 train_time:3294733ms step_avg:1684.42ms step:1967/6250 train_loss:2.1267 aux_loss:0.0237 train_time:3296757ms step_avg:1684.60ms step:1968/6250 train_loss:2.0774 aux_loss:0.0240 train_time:3298773ms step_avg:1684.77ms step:1969/6250 train_loss:2.0508 aux_loss:0.0242 train_time:3299720ms step_avg:1684.39ms step:1970/6250 train_loss:2.0518 aux_loss:0.0240 train_time:3300602ms step_avg:1683.98ms step:1971/6250 train_loss:2.0673 aux_loss:0.0235 train_time:3301693ms step_avg:1683.68ms step:1972/6250 train_loss:2.0676 aux_loss:0.0237 train_time:3302582ms step_avg:1683.27ms step:1973/6250 train_loss:2.0739 aux_loss:0.0238 train_time:3303530ms step_avg:1682.90ms step:1974/6250 train_loss:2.0288 aux_loss:0.0235 train_time:3305517ms step_avg:1683.05ms step:1975/6250 train_loss:2.0040 aux_loss:0.0235 train_time:3306423ms step_avg:1682.66ms step:1976/6250 train_loss:2.0881 aux_loss:0.0233 train_time:3307316ms step_avg:1682.26ms step:1977/6250 train_loss:2.1323 aux_loss:0.0237 train_time:3308302ms step_avg:1681.90ms step:1978/6250 train_loss:2.0421 aux_loss:0.0236 train_time:3309359ms step_avg:1681.58ms step:1979/6250 train_loss:1.9693 aux_loss:0.0238 train_time:3310292ms step_avg:1681.20ms step:1980/6250 train_loss:2.1416 aux_loss:0.0239 train_time:3311231ms step_avg:1680.83ms step:1981/6250 train_loss:2.0147 aux_loss:0.0239 train_time:3312160ms step_avg:1680.45ms step:1982/6250 train_loss:2.0665 aux_loss:0.0237 train_time:3313658ms step_avg:1680.35ms step:1983/6250 train_loss:2.0243 aux_loss:0.0230 train_time:3314891ms step_avg:1680.13ms step:1984/6250 train_loss:2.0883 aux_loss:0.0232 train_time:3315945ms step_avg:1679.81ms step:1985/6250 train_loss:2.1306 aux_loss:0.0235 train_time:3317157ms step_avg:1679.57ms step:1986/6250 train_loss:2.1554 aux_loss:0.0234 train_time:3318602ms step_avg:1679.45ms step:1987/6250 train_loss:2.0785 aux_loss:0.0236 train_time:3319488ms step_avg:1679.05ms step:1988/6250 train_loss:2.0351 aux_loss:0.0234 train_time:3320507ms step_avg:1678.72ms step:1989/6250 train_loss:2.1281 aux_loss:0.0233 train_time:3322350ms step_avg:1678.80ms step:1990/6250 train_loss:2.1061 aux_loss:0.0235 train_time:3326053ms step_avg:1679.82ms step:1991/6250 train_loss:2.0084 aux_loss:0.0234 train_time:3327731ms step_avg:1679.82ms step:1992/6250 train_loss:2.0163 aux_loss:0.0234 train_time:3329382ms step_avg:1679.81ms step:1993/6250 train_loss:2.1710 aux_loss:0.0234 train_time:3330287ms step_avg:1679.42ms step:1994/6250 train_loss:2.1693 aux_loss:0.0234 train_time:3331187ms step_avg:1679.03ms step:1995/6250 train_loss:1.9929 aux_loss:0.0233 train_time:3332073ms step_avg:1678.63ms step:1996/6250 train_loss:2.0574 aux_loss:0.0233 train_time:3332990ms step_avg:1678.24ms step:1997/6250 train_loss:2.0781 aux_loss:0.0235 train_time:3334186ms step_avg:1678.00ms step:1998/6250 train_loss:2.1306 aux_loss:0.0235 train_time:3335196ms step_avg:1677.66ms step:1999/6250 train_loss:2.1057 aux_loss:0.0235 train_time:3336150ms step_avg:1677.30ms step:2000/6250 train_loss:2.0605 aux_loss:0.0235 train_time:3337018ms step_avg:1676.89ms step:2000/6250 val_loss:2.0636 val_aux_loss:0.0236 train_time:3337018ms step_avg:1676.89ms step:2001/6250 train_loss:2.1324 aux_loss:0.0235 train_time:3337876ms step_avg:1676.48ms step:2002/6250 train_loss:2.0700 aux_loss:0.0237 train_time:3338758ms step_avg:1676.08ms step:2003/6250 train_loss:2.0180 aux_loss:0.0235 train_time:3340597ms step_avg:1676.17ms step:2004/6250 train_loss:2.0793 aux_loss:0.0237 train_time:3341448ms step_avg:1675.75ms step:2005/6250 train_loss:1.9941 aux_loss:0.0233 train_time:3342304ms step_avg:1675.34ms step:2006/6250 train_loss:2.0372 aux_loss:0.0232 train_time:3343983ms step_avg:1675.34ms step:2007/6250 train_loss:2.1438 aux_loss:0.0236 train_time:3345754ms step_avg:1675.39ms step:2008/6250 train_loss:2.0415 aux_loss:0.0234 train_time:3347090ms step_avg:1675.22ms step:2009/6250 train_loss:2.0692 aux_loss:0.0235 train_time:3348645ms step_avg:1675.16ms step:2010/6250 train_loss:2.0261 aux_loss:0.0235 train_time:3349509ms step_avg:1674.75ms step:2011/6250 train_loss:2.0927 aux_loss:0.0240 train_time:3350363ms step_avg:1674.34ms step:2012/6250 train_loss:2.1092 aux_loss:0.0234 train_time:3351783ms step_avg:1674.22ms step:2013/6250 train_loss:2.0237 aux_loss:0.0239 train_time:3353073ms step_avg:1674.03ms step:2014/6250 train_loss:1.9920 aux_loss:0.0240 train_time:3353932ms step_avg:1673.62ms step:2015/6250 train_loss:2.0940 aux_loss:0.0231 train_time:3354802ms step_avg:1673.22ms step:2016/6250 train_loss:2.1074 aux_loss:0.0239 train_time:3355667ms step_avg:1672.82ms step:2017/6250 train_loss:2.0879 aux_loss:0.0235 train_time:3356530ms step_avg:1672.41ms step:2018/6250 train_loss:2.0575 aux_loss:0.0232 train_time:3357391ms step_avg:1672.01ms step:2019/6250 train_loss:2.0650 aux_loss:0.0232 train_time:3358246ms step_avg:1671.60ms step:2020/6250 train_loss:2.0551 aux_loss:0.0234 train_time:3359099ms step_avg:1671.19ms step:2021/6250 train_loss:2.0811 aux_loss:0.0236 train_time:3359956ms step_avg:1670.79ms step:2022/6250 train_loss:2.0615 aux_loss:0.0232 train_time:3361251ms step_avg:1670.60ms step:2023/6250 train_loss:2.0135 aux_loss:0.0231 train_time:3362109ms step_avg:1670.20ms step:2024/6250 train_loss:2.0852 aux_loss:0.0237 train_time:3363383ms step_avg:1670.00ms step:2025/6250 train_loss:2.0735 aux_loss:0.0235 train_time:3364234ms step_avg:1669.60ms step:2026/6250 train_loss:2.0523 aux_loss:0.0236 train_time:3365085ms step_avg:1669.19ms step:2027/6250 train_loss:2.0447 aux_loss:0.0232 train_time:3365938ms step_avg:1668.78ms step:2028/6250 train_loss:2.0162 aux_loss:0.0232 train_time:3367238ms step_avg:1668.60ms step:2029/6250 train_loss:2.0916 aux_loss:0.0234 train_time:3368417ms step_avg:1668.36ms step:2030/6250 train_loss:2.0641 aux_loss:0.0231 train_time:3374310ms step_avg:1670.45ms step:2031/6250 train_loss:2.1166 aux_loss:0.0232 train_time:3376808ms step_avg:1670.86ms step:2032/6250 train_loss:2.0350 aux_loss:0.0231 train_time:3378279ms step_avg:1670.76ms step:2033/6250 train_loss:2.0153 aux_loss:0.0234 train_time:3383844ms step_avg:1672.69ms step:2034/6250 train_loss:2.0952 aux_loss:0.0233 train_time:3386143ms step_avg:1673.00ms step:2035/6250 train_loss:2.0600 aux_loss:0.0230 train_time:3388415ms step_avg:1673.29ms step:2036/6250 train_loss:1.9949 aux_loss:0.0229 train_time:3389869ms step_avg:1673.18ms step:2037/6250 train_loss:2.1306 aux_loss:0.0231 train_time:3390782ms step_avg:1672.81ms step:2038/6250 train_loss:2.0404 aux_loss:0.0235 train_time:3391728ms step_avg:1672.45ms step:2039/6250 train_loss:2.0301 aux_loss:0.0228 train_time:3392882ms step_avg:1672.19ms step:2040/6250 train_loss:1.9919 aux_loss:0.0228 train_time:3393780ms step_avg:1671.81ms step:2041/6250 train_loss:2.0795 aux_loss:0.0232 train_time:3394868ms step_avg:1671.53ms step:2042/6250 train_loss:2.0894 aux_loss:0.0230 train_time:3395968ms step_avg:1671.24ms step:2043/6250 train_loss:2.0209 aux_loss:0.0230 train_time:3396883ms step_avg:1670.87ms step:2044/6250 train_loss:1.9595 aux_loss:0.0226 train_time:3397958ms step_avg:1670.58ms step:2045/6250 train_loss:2.0091 aux_loss:0.0232 train_time:3399048ms step_avg:1670.29ms step:2046/6250 train_loss:2.1713 aux_loss:0.0229 train_time:3400176ms step_avg:1670.03ms step:2047/6250 train_loss:2.1146 aux_loss:0.0232 train_time:3401151ms step_avg:1669.69ms step:2048/6250 train_loss:2.0440 aux_loss:0.0231 train_time:3402109ms step_avg:1669.34ms step:2049/6250 train_loss:2.0683 aux_loss:0.0227 train_time:3403262ms step_avg:1669.08ms step:2050/6250 train_loss:1.9556 aux_loss:0.0229 train_time:3404364ms step_avg:1668.81ms step:2051/6250 train_loss:2.0285 aux_loss:0.0233 train_time:3405278ms step_avg:1668.44ms step:2052/6250 train_loss:2.0322 aux_loss:0.0226 train_time:3406441ms step_avg:1668.19ms step:2053/6250 train_loss:1.9905 aux_loss:0.0224 train_time:3407570ms step_avg:1667.92ms step:2054/6250 train_loss:2.0640 aux_loss:0.0233 train_time:3408537ms step_avg:1667.58ms step:2055/6250 train_loss:1.9419 aux_loss:0.0230 train_time:3409512ms step_avg:1667.24ms step:2056/6250 train_loss:2.1512 aux_loss:0.0231 train_time:3410524ms step_avg:1666.92ms step:2057/6250 train_loss:2.0293 aux_loss:0.0231 train_time:3411713ms step_avg:1666.69ms step:2058/6250 train_loss:2.0225 aux_loss:0.0234 train_time:3412809ms step_avg:1666.41ms step:2059/6250 train_loss:2.1067 aux_loss:0.0231 train_time:3413830ms step_avg:1666.10ms step:2060/6250 train_loss:2.0711 aux_loss:0.0229 train_time:3415059ms step_avg:1665.88ms step:2061/6250 train_loss:1.9806 aux_loss:0.0230 train_time:3416242ms step_avg:1665.65ms step:2062/6250 train_loss:2.1349 aux_loss:0.0229 train_time:3417224ms step_avg:1665.31ms step:2063/6250 train_loss:2.0245 aux_loss:0.0230 train_time:3418215ms step_avg:1664.99ms step:2064/6250 train_loss:2.0000 aux_loss:0.0228 train_time:3419173ms step_avg:1664.64ms step:2065/6250 train_loss:2.1123 aux_loss:0.0233 train_time:3420163ms step_avg:1664.31ms step:2066/6250 train_loss:2.0305 aux_loss:0.0234 train_time:3421124ms step_avg:1663.97ms step:2067/6250 train_loss:2.0434 aux_loss:0.0230 train_time:3422247ms step_avg:1663.71ms step:2068/6250 train_loss:1.9730 aux_loss:0.0229 train_time:3423211ms step_avg:1663.37ms step:2069/6250 train_loss:1.9912 aux_loss:0.0230 train_time:3424136ms step_avg:1663.01ms step:2070/6250 train_loss:2.1109 aux_loss:0.0227 train_time:3425054ms step_avg:1662.65ms step:2071/6250 train_loss:2.0921 aux_loss:0.0229 train_time:3426143ms step_avg:1662.37ms step:2072/6250 train_loss:2.0828 aux_loss:0.0234 train_time:3427272ms step_avg:1662.11ms step:2073/6250 train_loss:2.0079 aux_loss:0.0225 train_time:3428202ms step_avg:1661.76ms step:2074/6250 train_loss:2.0624 aux_loss:0.0227 train_time:3429105ms step_avg:1661.39ms step:2075/6250 train_loss:2.0957 aux_loss:0.0231 train_time:3430171ms step_avg:1661.10ms step:2076/6250 train_loss:2.0536 aux_loss:0.0227 train_time:3431096ms step_avg:1660.74ms step:2077/6250 train_loss:2.0697 aux_loss:0.0230 train_time:3432011ms step_avg:1660.38ms step:2078/6250 train_loss:2.1139 aux_loss:0.0227 train_time:3432930ms step_avg:1660.02ms step:2079/6250 train_loss:2.1067 aux_loss:0.0227 train_time:3433879ms step_avg:1659.68ms step:2080/6250 train_loss:2.0042 aux_loss:0.0229 train_time:3434906ms step_avg:1659.37ms step:2081/6250 train_loss:2.0750 aux_loss:0.0231 train_time:3436160ms step_avg:1659.18ms step:2082/6250 train_loss:2.0362 aux_loss:0.0228 train_time:3437121ms step_avg:1658.84ms step:2083/6250 train_loss:2.1349 aux_loss:0.0228 train_time:3438009ms step_avg:1658.47ms step:2084/6250 train_loss:2.0877 aux_loss:0.0229 train_time:3438921ms step_avg:1658.11ms step:2085/6250 train_loss:1.9689 aux_loss:0.0230 train_time:3439854ms step_avg:1657.76ms step:2086/6250 train_loss:1.9771 aux_loss:0.0224 train_time:3448819ms step_avg:1661.28ms step:2087/6250 train_loss:2.1347 aux_loss:0.0228 train_time:3449767ms step_avg:1660.94ms step:2088/6250 train_loss:2.0852 aux_loss:0.0223 train_time:3450766ms step_avg:1660.62ms step:2089/6250 train_loss:2.1132 aux_loss:0.0226 train_time:3451785ms step_avg:1660.31ms step:2090/6250 train_loss:1.9966 aux_loss:0.0228 train_time:3452705ms step_avg:1659.95ms step:2091/6250 train_loss:2.1104 aux_loss:0.0226 train_time:3453628ms step_avg:1659.60ms step:2092/6250 train_loss:2.0624 aux_loss:0.0228 train_time:3454530ms step_avg:1659.24ms step:2093/6250 train_loss:1.9889 aux_loss:0.0227 train_time:3455607ms step_avg:1658.96ms step:2094/6250 train_loss:1.9194 aux_loss:0.0229 train_time:3456710ms step_avg:1658.69ms step:2095/6250 train_loss:2.0520 aux_loss:0.0227 train_time:3457652ms step_avg:1658.35ms step:2096/6250 train_loss:2.0512 aux_loss:0.0224 train_time:3458641ms step_avg:1658.03ms step:2097/6250 train_loss:2.0076 aux_loss:0.0226 train_time:3459964ms step_avg:1657.86ms step:2098/6250 train_loss:2.0919 aux_loss:0.0228 train_time:3461024ms step_avg:1657.58ms step:2099/6250 train_loss:1.9890 aux_loss:0.0229 train_time:3462017ms step_avg:1657.26ms step:2100/6250 train_loss:2.0494 aux_loss:0.0226 train_time:3462970ms step_avg:1656.92ms step:2101/6250 train_loss:1.9662 aux_loss:0.0224 train_time:3463975ms step_avg:1656.61ms step:2102/6250 train_loss:2.1164 aux_loss:0.0223 train_time:3465006ms step_avg:1656.31ms step:2103/6250 train_loss:1.9686 aux_loss:0.0225 train_time:3466203ms step_avg:1656.09ms step:2104/6250 train_loss:2.0898 aux_loss:0.0225 train_time:3467205ms step_avg:1655.78ms step:2105/6250 train_loss:2.0754 aux_loss:0.0225 train_time:3468134ms step_avg:1655.43ms step:2106/6250 train_loss:1.9937 aux_loss:0.0224 train_time:3469129ms step_avg:1655.12ms step:2107/6250 train_loss:1.9854 aux_loss:0.0224 train_time:3470092ms step_avg:1654.79ms step:2108/6250 train_loss:2.0935 aux_loss:0.0223 train_time:3471080ms step_avg:1654.47ms step:2109/6250 train_loss:2.0479 aux_loss:0.0228 train_time:3472043ms step_avg:1654.14ms step:2110/6250 train_loss:2.0903 aux_loss:0.0226 train_time:3473015ms step_avg:1653.82ms step:2111/6250 train_loss:2.0364 aux_loss:0.0226 train_time:3474010ms step_avg:1653.50ms step:2112/6250 train_loss:2.0178 aux_loss:0.0222 train_time:3475002ms step_avg:1653.19ms step:2113/6250 train_loss:2.0738 aux_loss:0.0220 train_time:3476427ms step_avg:1653.08ms step:2114/6250 train_loss:1.9779 aux_loss:0.0222 train_time:3477468ms step_avg:1652.79ms step:2115/6250 train_loss:2.0674 aux_loss:0.0231 train_time:3478457ms step_avg:1652.47ms step:2116/6250 train_loss:1.9632 aux_loss:0.0224 train_time:3479476ms step_avg:1652.17ms step:2117/6250 train_loss:2.1428 aux_loss:0.0225 train_time:3480405ms step_avg:1651.83ms step:2118/6250 train_loss:2.0270 aux_loss:0.0228 train_time:3481303ms step_avg:1651.47ms step:2119/6250 train_loss:2.0813 aux_loss:0.0223 train_time:3482215ms step_avg:1651.12ms step:2120/6250 train_loss:2.0056 aux_loss:0.0221 train_time:3483161ms step_avg:1650.79ms step:2121/6250 train_loss:2.0939 aux_loss:0.0225 train_time:3484050ms step_avg:1650.43ms step:2122/6250 train_loss:2.0252 aux_loss:0.0223 train_time:3484982ms step_avg:1650.09ms step:2123/6250 train_loss:2.1064 aux_loss:0.0226 train_time:3485887ms step_avg:1649.73ms step:2124/6250 train_loss:2.1052 aux_loss:0.0223 train_time:3486788ms step_avg:1649.38ms step:2125/6250 train_loss:2.0528 aux_loss:0.0222 train_time:3487690ms step_avg:1649.03ms step:2126/6250 train_loss:2.0134 aux_loss:0.0225 train_time:3488612ms step_avg:1648.68ms step:2127/6250 train_loss:2.0677 aux_loss:0.0227 train_time:3489562ms step_avg:1648.35ms step:2128/6250 train_loss:2.0160 aux_loss:0.0222 train_time:3490472ms step_avg:1648.00ms step:2129/6250 train_loss:2.0667 aux_loss:0.0225 train_time:3491387ms step_avg:1647.66ms step:2130/6250 train_loss:2.0002 aux_loss:0.0229 train_time:3492304ms step_avg:1647.31ms step:2131/6250 train_loss:2.0284 aux_loss:0.0222 train_time:3493209ms step_avg:1646.96ms step:2132/6250 train_loss:2.0261 aux_loss:0.0220 train_time:3494107ms step_avg:1646.61ms step:2133/6250 train_loss:2.0989 aux_loss:0.0222 train_time:3495021ms step_avg:1646.27ms step:2134/6250 train_loss:1.9400 aux_loss:0.0219 train_time:3495968ms step_avg:1645.94ms step:2135/6250 train_loss:1.9272 aux_loss:0.0220 train_time:3496886ms step_avg:1645.59ms step:2136/6250 train_loss:1.9045 aux_loss:0.0223 train_time:3497806ms step_avg:1645.25ms step:2137/6250 train_loss:2.0459 aux_loss:0.0223 train_time:3498710ms step_avg:1644.90ms step:2138/6250 train_loss:1.9450 aux_loss:0.0226 train_time:3499648ms step_avg:1644.57ms step:2139/6250 train_loss:2.0591 aux_loss:0.0224 train_time:3500547ms step_avg:1644.22ms step:2140/6250 train_loss:2.0651 aux_loss:0.0222 train_time:3501443ms step_avg:1643.87ms step:2141/6250 train_loss:2.0388 aux_loss:0.0227 train_time:3502359ms step_avg:1643.53ms step:2142/6250 train_loss:2.0244 aux_loss:0.0230 train_time:3503256ms step_avg:1643.18ms step:2143/6250 train_loss:2.0011 aux_loss:0.0224 train_time:3504136ms step_avg:1642.82ms step:2144/6250 train_loss:2.0244 aux_loss:0.0220 train_time:3505066ms step_avg:1642.49ms step:2145/6250 train_loss:2.0779 aux_loss:0.0225 train_time:3506040ms step_avg:1642.17ms step:2146/6250 train_loss:2.1241 aux_loss:0.0224 train_time:3507015ms step_avg:1641.86ms step:2147/6250 train_loss:2.0275 aux_loss:0.0222 train_time:3508166ms step_avg:1641.63ms step:2148/6250 train_loss:2.0848 aux_loss:0.0219 train_time:3509364ms step_avg:1641.42ms step:2149/6250 train_loss:1.9911 aux_loss:0.0219 train_time:3510264ms step_avg:1641.08ms step:2150/6250 train_loss:2.0197 aux_loss:0.0223 train_time:3511584ms step_avg:1640.93ms step:2151/6250 train_loss:2.1143 aux_loss:0.0226 train_time:3512495ms step_avg:1640.59ms step:2152/6250 train_loss:2.0245 aux_loss:0.0223 train_time:3513418ms step_avg:1640.25ms step:2153/6250 train_loss:2.0221 aux_loss:0.0221 train_time:3514338ms step_avg:1639.91ms step:2154/6250 train_loss:2.0295 aux_loss:0.0220 train_time:3515219ms step_avg:1639.56ms step:2155/6250 train_loss:1.9333 aux_loss:0.0223 train_time:3516117ms step_avg:1639.22ms step:2156/6250 train_loss:1.9899 aux_loss:0.0220 train_time:3517011ms step_avg:1638.87ms step:2157/6250 train_loss:1.9432 aux_loss:0.0222 train_time:3517916ms step_avg:1638.53ms step:2158/6250 train_loss:2.0340 aux_loss:0.0221 train_time:3519005ms step_avg:1638.27ms step:2159/6250 train_loss:2.0700 aux_loss:0.0218 train_time:3520053ms step_avg:1638.00ms step:2160/6250 train_loss:2.0108 aux_loss:0.0221 train_time:3521090ms step_avg:1637.72ms step:2161/6250 train_loss:1.9754 aux_loss:0.0219 train_time:3522219ms step_avg:1637.48ms step:2162/6250 train_loss:1.9821 aux_loss:0.0221 train_time:3523161ms step_avg:1637.16ms step:2163/6250 train_loss:2.1403 aux_loss:0.0221 train_time:3524066ms step_avg:1636.82ms step:2164/6250 train_loss:1.9048 aux_loss:0.0219 train_time:3524991ms step_avg:1636.49ms step:2165/6250 train_loss:1.9998 aux_loss:0.0220 train_time:3525895ms step_avg:1636.15ms step:2166/6250 train_loss:2.0119 aux_loss:0.0218 train_time:3526798ms step_avg:1635.81ms step:2167/6250 train_loss:2.0698 aux_loss:0.0217 train_time:3527717ms step_avg:1635.47ms step:2168/6250 train_loss:2.0126 aux_loss:0.0221 train_time:3528628ms step_avg:1635.14ms step:2169/6250 train_loss:2.0038 aux_loss:0.0220 train_time:3529566ms step_avg:1634.82ms step:2170/6250 train_loss:2.0074 aux_loss:0.0217 train_time:3530495ms step_avg:1634.49ms step:2171/6250 train_loss:2.0886 aux_loss:0.0217 train_time:3531410ms step_avg:1634.16ms step:2172/6250 train_loss:2.0830 aux_loss:0.0218 train_time:3532294ms step_avg:1633.81ms step:2173/6250 train_loss:2.0437 aux_loss:0.0218 train_time:3533213ms step_avg:1633.48ms step:2174/6250 train_loss:2.0350 aux_loss:0.0216 train_time:3534136ms step_avg:1633.15ms step:2175/6250 train_loss:1.9666 aux_loss:0.0218 train_time:3535028ms step_avg:1632.81ms step:2176/6250 train_loss:2.0834 aux_loss:0.0220 train_time:3535941ms step_avg:1632.48ms step:2177/6250 train_loss:2.0417 aux_loss:0.0221 train_time:3536840ms step_avg:1632.14ms step:2178/6250 train_loss:2.0542 aux_loss:0.0215 train_time:3537740ms step_avg:1631.80ms step:2179/6250 train_loss:2.0189 aux_loss:0.0219 train_time:3538646ms step_avg:1631.46ms step:2180/6250 train_loss:2.0975 aux_loss:0.0224 train_time:3539663ms step_avg:1631.18ms step:2181/6250 train_loss:1.9617 aux_loss:0.0219 train_time:3540620ms step_avg:1630.87ms step:2182/6250 train_loss:1.9993 aux_loss:0.0220 train_time:3541652ms step_avg:1630.59ms step:2183/6250 train_loss:1.9911 aux_loss:0.0221 train_time:3542632ms step_avg:1630.30ms step:2184/6250 train_loss:2.0711 aux_loss:0.0220 train_time:3543601ms step_avg:1629.99ms step:2185/6250 train_loss:2.0344 aux_loss:0.0216 train_time:3544557ms step_avg:1629.68ms step:2186/6250 train_loss:2.0899 aux_loss:0.0218 train_time:3545517ms step_avg:1629.37ms step:2187/6250 train_loss:2.0554 aux_loss:0.0225 train_time:3546565ms step_avg:1629.11ms step:2188/6250 train_loss:2.0313 aux_loss:0.0220 train_time:3547551ms step_avg:1628.81ms step:2189/6250 train_loss:2.1038 aux_loss:0.0221 train_time:3548513ms step_avg:1628.51ms step:2190/6250 train_loss:2.0857 aux_loss:0.0222 train_time:3549538ms step_avg:1628.23ms step:2191/6250 train_loss:1.9989 aux_loss:0.0219 train_time:3550501ms step_avg:1627.92ms step:2192/6250 train_loss:1.9361 aux_loss:0.0223 train_time:3551504ms step_avg:1627.64ms step:2193/6250 train_loss:2.0788 aux_loss:0.0223 train_time:3552520ms step_avg:1627.36ms step:2194/6250 train_loss:2.1248 aux_loss:0.0223 train_time:3553528ms step_avg:1627.07ms step:2195/6250 train_loss:1.9743 aux_loss:0.0218 train_time:3554446ms step_avg:1626.75ms step:2196/6250 train_loss:2.1443 aux_loss:0.0221 train_time:3555832ms step_avg:1626.64ms step:2197/6250 train_loss:2.0431 aux_loss:0.0223 train_time:3557721ms step_avg:1626.76ms step:2198/6250 train_loss:2.0979 aux_loss:0.0217 train_time:3558695ms step_avg:1626.46ms step:2199/6250 train_loss:2.0321 aux_loss:0.0217 train_time:3559597ms step_avg:1626.13ms step:2200/6250 train_loss:1.9992 aux_loss:0.0222 train_time:3560487ms step_avg:1625.79ms step:2201/6250 train_loss:1.9966 aux_loss:0.0219 train_time:3561405ms step_avg:1625.47ms step:2202/6250 train_loss:1.9795 aux_loss:0.0219 train_time:3563046ms step_avg:1625.48ms step:2203/6250 train_loss:2.0318 aux_loss:0.0218 train_time:3566381ms step_avg:1626.26ms step:2204/6250 train_loss:2.0231 aux_loss:0.0216 train_time:3567560ms step_avg:1626.05ms step:2205/6250 train_loss:2.0772 aux_loss:0.0222 train_time:3569010ms step_avg:1625.97ms step:2206/6250 train_loss:2.0700 aux_loss:0.0218 train_time:3570629ms step_avg:1625.97ms step:2207/6250 train_loss:2.0915 aux_loss:0.0222 train_time:3571909ms step_avg:1625.81ms step:2208/6250 train_loss:1.8962 aux_loss:0.0215 train_time:3572833ms step_avg:1625.49ms step:2209/6250 train_loss:2.0596 aux_loss:0.0219 train_time:3573838ms step_avg:1625.21ms step:2210/6250 train_loss:1.9818 aux_loss:0.0218 train_time:3574872ms step_avg:1624.94ms step:2211/6250 train_loss:2.0146 aux_loss:0.0222 train_time:3575827ms step_avg:1624.64ms step:2212/6250 train_loss:1.8979 aux_loss:0.0218 train_time:3576838ms step_avg:1624.36ms step:2213/6250 train_loss:1.9671 aux_loss:0.0216 train_time:3577844ms step_avg:1624.08ms step:2214/6250 train_loss:1.9604 aux_loss:0.0218 train_time:3578848ms step_avg:1623.80ms step:2215/6250 train_loss:1.9978 aux_loss:0.0220 train_time:3579840ms step_avg:1623.51ms step:2216/6250 train_loss:2.0242 aux_loss:0.0219 train_time:3580800ms step_avg:1623.21ms step:2217/6250 train_loss:2.0255 aux_loss:0.0215 train_time:3581786ms step_avg:1622.92ms step:2218/6250 train_loss:2.0090 aux_loss:0.0219 train_time:3582676ms step_avg:1622.59ms step:2219/6250 train_loss:2.0069 aux_loss:0.0218 train_time:3583624ms step_avg:1622.28ms step:2220/6250 train_loss:1.9360 aux_loss:0.0212 train_time:3584575ms step_avg:1621.98ms step:2221/6250 train_loss:2.0900 aux_loss:0.0217 train_time:3585542ms step_avg:1621.68ms step:2222/6250 train_loss:2.0950 aux_loss:0.0221 train_time:3586485ms step_avg:1621.38ms step:2223/6250 train_loss:1.9642 aux_loss:0.0217 train_time:3587555ms step_avg:1621.13ms step:2224/6250 train_loss:2.0086 aux_loss:0.0215 train_time:3588505ms step_avg:1620.82ms step:2225/6250 train_loss:2.0151 aux_loss:0.0216 train_time:3589479ms step_avg:1620.53ms step:2226/6250 train_loss:2.0714 aux_loss:0.0218 train_time:3590407ms step_avg:1620.22ms step:2227/6250 train_loss:2.0651 aux_loss:0.0219 train_time:3591415ms step_avg:1619.94ms step:2228/6250 train_loss:2.0128 aux_loss:0.0212 train_time:3592401ms step_avg:1619.66ms step:2229/6250 train_loss:2.0510 aux_loss:0.0218 train_time:3593376ms step_avg:1619.37ms step:2230/6250 train_loss:1.9824 aux_loss:0.0214 train_time:3594433ms step_avg:1619.11ms step:2231/6250 train_loss:2.2370 aux_loss:0.0219 train_time:3595479ms step_avg:1618.86ms step:2232/6250 train_loss:2.1070 aux_loss:0.0217 train_time:3596558ms step_avg:1618.61ms step:2233/6250 train_loss:2.0278 aux_loss:0.0214 train_time:3597696ms step_avg:1618.40ms step:2234/6250 train_loss:1.9889 aux_loss:0.0213 train_time:3598654ms step_avg:1618.10ms step:2235/6250 train_loss:2.0492 aux_loss:0.0216 train_time:3599663ms step_avg:1617.83ms step:2236/6250 train_loss:2.0813 aux_loss:0.0219 train_time:3600830ms step_avg:1617.62ms step:2237/6250 train_loss:2.0654 aux_loss:0.0217 train_time:3601849ms step_avg:1617.35ms step:2238/6250 train_loss:2.0584 aux_loss:0.0218 train_time:3602849ms step_avg:1617.08ms step:2239/6250 train_loss:1.9448 aux_loss:0.0219 train_time:3603904ms step_avg:1616.83ms step:2240/6250 train_loss:1.9490 aux_loss:0.0216 train_time:3604880ms step_avg:1616.54ms step:2241/6250 train_loss:1.9507 aux_loss:0.0214 train_time:3605850ms step_avg:1616.25ms step:2242/6250 train_loss:2.0273 aux_loss:0.0213 train_time:3606773ms step_avg:1615.94ms step:2243/6250 train_loss:2.0097 aux_loss:0.0216 train_time:3607819ms step_avg:1615.68ms step:2244/6250 train_loss:2.0857 aux_loss:0.0217 train_time:3608756ms step_avg:1615.38ms step:2245/6250 train_loss:1.9828 aux_loss:0.0215 train_time:3609690ms step_avg:1615.07ms step:2246/6250 train_loss:2.0383 aux_loss:0.0218 train_time:3610584ms step_avg:1614.75ms step:2247/6250 train_loss:2.0196 aux_loss:0.0217 train_time:3611495ms step_avg:1614.44ms step:2248/6250 train_loss:1.9653 aux_loss:0.0214 train_time:3612466ms step_avg:1614.15ms step:2249/6250 train_loss:2.1139 aux_loss:0.0214 train_time:3613500ms step_avg:1613.89ms step:2250/6250 train_loss:1.9556 aux_loss:0.0215 train_time:3614493ms step_avg:1613.61ms step:2251/6250 train_loss:2.0094 aux_loss:0.0214 train_time:3615431ms step_avg:1613.31ms step:2252/6250 train_loss:2.0430 aux_loss:0.0215 train_time:3616447ms step_avg:1613.04ms step:2253/6250 train_loss:2.0802 aux_loss:0.0217 train_time:3617350ms step_avg:1612.73ms step:2254/6250 train_loss:2.0120 aux_loss:0.0219 train_time:3618249ms step_avg:1612.41ms step:2255/6250 train_loss:2.0346 aux_loss:0.0216 train_time:3619164ms step_avg:1612.10ms step:2256/6250 train_loss:1.9939 aux_loss:0.0216 train_time:3620076ms step_avg:1611.79ms step:2257/6250 train_loss:1.9788 aux_loss:0.0215 train_time:3620971ms step_avg:1611.47ms step:2258/6250 train_loss:1.9135 aux_loss:0.0213 train_time:3621862ms step_avg:1611.15ms step:2259/6250 train_loss:2.0508 aux_loss:0.0213 train_time:3622768ms step_avg:1610.84ms step:2260/6250 train_loss:1.9463 aux_loss:0.0215 train_time:3623670ms step_avg:1610.52ms step:2261/6250 train_loss:1.9351 aux_loss:0.0212 train_time:3624598ms step_avg:1610.22ms step:2262/6250 train_loss:1.9951 aux_loss:0.0212 train_time:3625533ms step_avg:1609.92ms step:2263/6250 train_loss:1.9357 aux_loss:0.0218 train_time:3626430ms step_avg:1609.60ms step:2264/6250 train_loss:2.0244 aux_loss:0.0217 train_time:3627354ms step_avg:1609.30ms step:2265/6250 train_loss:2.0266 aux_loss:0.0213 train_time:3628257ms step_avg:1608.98ms step:2266/6250 train_loss:1.9341 aux_loss:0.0209 train_time:3629343ms step_avg:1608.75ms step:2267/6250 train_loss:1.9824 aux_loss:0.0215 train_time:3630334ms step_avg:1608.48ms step:2268/6250 train_loss:2.0179 aux_loss:0.0213 train_time:3631242ms step_avg:1608.17ms step:2269/6250 train_loss:2.2217 aux_loss:0.0211 train_time:3632216ms step_avg:1607.89ms step:2270/6250 train_loss:2.0112 aux_loss:0.0219 train_time:3633403ms step_avg:1607.70ms step:2271/6250 train_loss:2.0378 aux_loss:0.0214 train_time:3634405ms step_avg:1607.43ms step:2272/6250 train_loss:1.9622 aux_loss:0.0211 train_time:3635358ms step_avg:1607.14ms step:2273/6250 train_loss:2.0420 aux_loss:0.0215 train_time:3636243ms step_avg:1606.82ms step:2274/6250 train_loss:2.0115 aux_loss:0.0215 train_time:3637186ms step_avg:1606.53ms step:2275/6250 train_loss:2.0051 aux_loss:0.0214 train_time:3638107ms step_avg:1606.23ms step:2276/6250 train_loss:2.1585 aux_loss:0.0215 train_time:3638989ms step_avg:1605.91ms step:2277/6250 train_loss:1.9896 aux_loss:0.0216 train_time:3639992ms step_avg:1605.64ms step:2278/6250 train_loss:2.0395 aux_loss:0.0213 train_time:3641023ms step_avg:1605.39ms step:2279/6250 train_loss:1.9667 aux_loss:0.0208 train_time:3642043ms step_avg:1605.13ms step:2280/6250 train_loss:1.9941 aux_loss:0.0213 train_time:3642984ms step_avg:1604.84ms step:2281/6250 train_loss:2.0178 aux_loss:0.0214 train_time:3643925ms step_avg:1604.55ms step:2282/6250 train_loss:1.9998 aux_loss:0.0209 train_time:3644871ms step_avg:1604.26ms step:2283/6250 train_loss:1.9271 aux_loss:0.0207 train_time:3645930ms step_avg:1604.02ms step:2284/6250 train_loss:2.1812 aux_loss:0.0211 train_time:3646913ms step_avg:1603.74ms step:2285/6250 train_loss:2.0337 aux_loss:0.0211 train_time:3647911ms step_avg:1603.48ms step:2286/6250 train_loss:2.0335 aux_loss:0.0215 train_time:3648886ms step_avg:1603.20ms step:2287/6250 train_loss:2.0205 aux_loss:0.0213 train_time:3649838ms step_avg:1602.92ms step:2288/6250 train_loss:1.9655 aux_loss:0.0216 train_time:3650767ms step_avg:1602.62ms step:2289/6250 train_loss:1.9523 aux_loss:0.0213 train_time:3651677ms step_avg:1602.32ms step:2290/6250 train_loss:2.0992 aux_loss:0.0213 train_time:3652596ms step_avg:1602.02ms step:2291/6250 train_loss:2.0523 aux_loss:0.0212 train_time:3653511ms step_avg:1601.71ms step:2292/6250 train_loss:2.0347 aux_loss:0.0209 train_time:3654519ms step_avg:1601.45ms step:2293/6250 train_loss:2.0215 aux_loss:0.0212 train_time:3655494ms step_avg:1601.18ms step:2294/6250 train_loss:1.9701 aux_loss:0.0213 train_time:3656386ms step_avg:1600.87ms step:2295/6250 train_loss:2.0650 aux_loss:0.0214 train_time:3657294ms step_avg:1600.57ms step:2296/6250 train_loss:2.0724 aux_loss:0.0215 train_time:3658216ms step_avg:1600.27ms step:2297/6250 train_loss:2.0510 aux_loss:0.0214 train_time:3659190ms step_avg:1600.00ms step:2298/6250 train_loss:2.0204 aux_loss:0.0214 train_time:3660289ms step_avg:1599.78ms step:2299/6250 train_loss:2.0528 aux_loss:0.0215 train_time:3661382ms step_avg:1599.56ms step:2300/6250 train_loss:2.0732 aux_loss:0.0211 train_time:3662407ms step_avg:1599.30ms step:2301/6250 train_loss:2.0205 aux_loss:0.0207 train_time:3663423ms step_avg:1599.05ms step:2302/6250 train_loss:1.9441 aux_loss:0.0214 train_time:3664593ms step_avg:1598.86ms step:2303/6250 train_loss:1.9814 aux_loss:0.0210 train_time:3665572ms step_avg:1598.59ms step:2304/6250 train_loss:1.9685 aux_loss:0.0210 train_time:3666549ms step_avg:1598.32ms step:2305/6250 train_loss:2.0265 aux_loss:0.0224 train_time:3667500ms step_avg:1598.04ms step:2306/6250 train_loss:1.9647 aux_loss:0.0215 train_time:3668545ms step_avg:1597.80ms step:2307/6250 train_loss:1.9605 aux_loss:0.0211 train_time:3669889ms step_avg:1597.69ms step:2308/6250 train_loss:1.9729 aux_loss:0.0211 train_time:3671026ms step_avg:1597.49ms step:2309/6250 train_loss:1.9862 aux_loss:0.0207 train_time:3671960ms step_avg:1597.20ms step:2310/6250 train_loss:1.9683 aux_loss:0.0212 train_time:3672880ms step_avg:1596.90ms step:2311/6250 train_loss:1.9793 aux_loss:0.0213 train_time:3673857ms step_avg:1596.63ms step:2312/6250 train_loss:1.9800 aux_loss:0.0214 train_time:3675013ms step_avg:1596.44ms step:2313/6250 train_loss:2.0316 aux_loss:0.0214 train_time:3676259ms step_avg:1596.29ms step:2314/6250 train_loss:1.9606 aux_loss:0.0211 train_time:3677424ms step_avg:1596.10ms step:2315/6250 train_loss:2.0785 aux_loss:0.0212 train_time:3678335ms step_avg:1595.81ms step:2316/6250 train_loss:2.0376 aux_loss:0.0217 train_time:3679205ms step_avg:1595.49ms step:2317/6250 train_loss:2.0707 aux_loss:0.0214 train_time:3680122ms step_avg:1595.20ms step:2318/6250 train_loss:1.9951 aux_loss:0.0210 train_time:3681038ms step_avg:1594.90ms step:2319/6250 train_loss:2.0184 aux_loss:0.0212 train_time:3681961ms step_avg:1594.61ms step:2320/6250 train_loss:1.9650 aux_loss:0.0211 train_time:3682982ms step_avg:1594.36ms step:2321/6250 train_loss:1.9815 aux_loss:0.0210 train_time:3683895ms step_avg:1594.07ms step:2322/6250 train_loss:1.9878 aux_loss:0.0209 train_time:3684809ms step_avg:1593.78ms step:2323/6250 train_loss:1.9608 aux_loss:0.0211 train_time:3685812ms step_avg:1593.52ms step:2324/6250 train_loss:2.0337 aux_loss:0.0209 train_time:3689185ms step_avg:1594.29ms step:2325/6250 train_loss:2.0444 aux_loss:0.0214 train_time:3690229ms step_avg:1594.05ms step:2326/6250 train_loss:2.0321 aux_loss:0.0212 train_time:3691163ms step_avg:1593.77ms step:2327/6250 train_loss:2.0356 aux_loss:0.0213 train_time:3692113ms step_avg:1593.49ms step:2328/6250 train_loss:1.9788 aux_loss:0.0210 train_time:3693031ms step_avg:1593.20ms step:2329/6250 train_loss:2.0379 aux_loss:0.0206 train_time:3694000ms step_avg:1592.93ms step:2330/6250 train_loss:2.0117 aux_loss:0.0214 train_time:3695026ms step_avg:1592.68ms step:2331/6250 train_loss:1.9665 aux_loss:0.0212 train_time:3695997ms step_avg:1592.42ms step:2332/6250 train_loss:2.0896 aux_loss:0.0209 train_time:3696933ms step_avg:1592.13ms step:2333/6250 train_loss:1.9817 aux_loss:0.0212 train_time:3697864ms step_avg:1591.85ms step:2334/6250 train_loss:1.9485 aux_loss:0.0212 train_time:3698821ms step_avg:1591.58ms step:2335/6250 train_loss:2.0106 aux_loss:0.0213 train_time:3699800ms step_avg:1591.31ms step:2336/6250 train_loss:2.0088 aux_loss:0.0210 train_time:3700834ms step_avg:1591.07ms step:2337/6250 train_loss:2.0279 aux_loss:0.0211 train_time:3701729ms step_avg:1590.77ms step:2338/6250 train_loss:2.0304 aux_loss:0.0211 train_time:3702627ms step_avg:1590.48ms step:2339/6250 train_loss:2.0546 aux_loss:0.0208 train_time:3703497ms step_avg:1590.17ms step:2340/6250 train_loss:2.0148 aux_loss:0.0212 train_time:3704397ms step_avg:1589.87ms step:2341/6250 train_loss:2.0061 aux_loss:0.0211 train_time:3705300ms step_avg:1589.58ms step:2342/6250 train_loss:1.9812 aux_loss:0.0212 train_time:3706184ms step_avg:1589.27ms step:2343/6250 train_loss:2.0423 aux_loss:0.0206 train_time:3707068ms step_avg:1588.97ms step:2344/6250 train_loss:1.9974 aux_loss:0.0210 train_time:3707969ms step_avg:1588.68ms step:2345/6250 train_loss:1.9624 aux_loss:0.0208 train_time:3708835ms step_avg:1588.37ms step:2346/6250 train_loss:2.1393 aux_loss:0.0208 train_time:3709754ms step_avg:1588.08ms step:2347/6250 train_loss:1.9322 aux_loss:0.0207 train_time:3710635ms step_avg:1587.78ms step:2348/6250 train_loss:2.1096 aux_loss:0.0213 train_time:3711543ms step_avg:1587.49ms step:2349/6250 train_loss:2.0656 aux_loss:0.0211 train_time:3712422ms step_avg:1587.18ms step:2350/6250 train_loss:1.9225 aux_loss:0.0208 train_time:3713314ms step_avg:1586.89ms step:2351/6250 train_loss:1.9338 aux_loss:0.0209 train_time:3714196ms step_avg:1586.59ms step:2352/6250 train_loss:2.0588 aux_loss:0.0212 train_time:3715109ms step_avg:1586.30ms step:2353/6250 train_loss:1.9882 aux_loss:0.0211 train_time:3715997ms step_avg:1586.00ms step:2354/6250 train_loss:2.0222 aux_loss:0.0210 train_time:3716913ms step_avg:1585.71ms step:2355/6250 train_loss:1.9876 aux_loss:0.0208 train_time:3717785ms step_avg:1585.41ms step:2356/6250 train_loss:2.0264 aux_loss:0.0207 train_time:3718701ms step_avg:1585.12ms step:2357/6250 train_loss:2.0622 aux_loss:0.0207 train_time:3719598ms step_avg:1584.83ms step:2358/6250 train_loss:1.9433 aux_loss:0.0203 train_time:3720503ms step_avg:1584.54ms step:2359/6250 train_loss:2.0391 aux_loss:0.0206 train_time:3721393ms step_avg:1584.25ms step:2360/6250 train_loss:1.9138 aux_loss:0.0210 train_time:3722272ms step_avg:1583.95ms step:2361/6250 train_loss:2.1111 aux_loss:0.0212 train_time:3723138ms step_avg:1583.64ms step:2362/6250 train_loss:1.9784 aux_loss:0.0209 train_time:3724066ms step_avg:1583.36ms step:2363/6250 train_loss:2.1120 aux_loss:0.0209 train_time:3724965ms step_avg:1583.07ms step:2364/6250 train_loss:1.9534 aux_loss:0.0209 train_time:3725880ms step_avg:1582.79ms step:2365/6250 train_loss:1.9505 aux_loss:0.0206 train_time:3726764ms step_avg:1582.49ms step:2366/6250 train_loss:2.0685 aux_loss:0.0209 train_time:3727863ms step_avg:1582.28ms step:2367/6250 train_loss:2.0151 aux_loss:0.0213 train_time:3728769ms step_avg:1582.00ms step:2368/6250 train_loss:1.9992 aux_loss:0.0207 train_time:3729747ms step_avg:1581.74ms step:2369/6250 train_loss:2.0195 aux_loss:0.0208 train_time:3730788ms step_avg:1581.51ms step:2370/6250 train_loss:1.9443 aux_loss:0.0212 train_time:3731717ms step_avg:1581.24ms step:2371/6250 train_loss:1.9477 aux_loss:0.0210 train_time:3732628ms step_avg:1580.95ms step:2372/6250 train_loss:2.0242 aux_loss:0.0204 train_time:3735072ms step_avg:1581.32ms step:2373/6250 train_loss:2.0624 aux_loss:0.0206 train_time:3736463ms step_avg:1581.24ms step:2374/6250 train_loss:1.9628 aux_loss:0.0210 train_time:3737754ms step_avg:1581.11ms step:2375/6250 train_loss:1.9514 aux_loss:0.0207 train_time:3739149ms step_avg:1581.04ms step:2376/6250 train_loss:1.9713 aux_loss:0.0210 train_time:3740461ms step_avg:1580.92ms step:2377/6250 train_loss:2.0060 aux_loss:0.0210 train_time:3741553ms step_avg:1580.72ms step:2378/6250 train_loss:1.8802 aux_loss:0.0207 train_time:3742718ms step_avg:1580.54ms step:2379/6250 train_loss:2.0394 aux_loss:0.0209 train_time:3743867ms step_avg:1580.36ms step:2380/6250 train_loss:2.0258 aux_loss:0.0207 train_time:3745007ms step_avg:1580.17ms step:2381/6250 train_loss:1.9783 aux_loss:0.0206 train_time:3746321ms step_avg:1580.06ms step:2382/6250 train_loss:1.9660 aux_loss:0.0209 train_time:3747424ms step_avg:1579.86ms step:2383/6250 train_loss:2.0552 aux_loss:0.0213 train_time:3748596ms step_avg:1579.69ms step:2384/6250 train_loss:1.9924 aux_loss:0.0212 train_time:3750033ms step_avg:1579.63ms step:2385/6250 train_loss:2.0394 aux_loss:0.0210 train_time:3751555ms step_avg:1579.60ms step:2386/6250 train_loss:2.0658 aux_loss:0.0208 train_time:3752915ms step_avg:1579.51ms step:2387/6250 train_loss:2.0189 aux_loss:0.0209 train_time:3754074ms step_avg:1579.33ms step:2388/6250 train_loss:1.9747 aux_loss:0.0208 train_time:3755284ms step_avg:1579.18ms step:2389/6250 train_loss:2.0387 aux_loss:0.0210 train_time:3756419ms step_avg:1578.99ms step:2390/6250 train_loss:2.0146 aux_loss:0.0206 train_time:3757850ms step_avg:1578.93ms step:2391/6250 train_loss:1.9540 aux_loss:0.0212 train_time:3759207ms step_avg:1578.84ms step:2392/6250 train_loss:2.0272 aux_loss:0.0212 train_time:3761572ms step_avg:1579.17ms step:2393/6250 train_loss:1.9385 aux_loss:0.0206 train_time:3763621ms step_avg:1579.36ms step:2394/6250 train_loss:2.0045 aux_loss:0.0204 train_time:3765165ms step_avg:1579.35ms step:2395/6250 train_loss:2.0665 aux_loss:0.0212 train_time:3766415ms step_avg:1579.21ms step:2396/6250 train_loss:2.0903 aux_loss:0.0208 train_time:3767544ms step_avg:1579.02ms step:2397/6250 train_loss:1.9344 aux_loss:0.0202 train_time:3768674ms step_avg:1578.83ms step:2398/6250 train_loss:2.0423 aux_loss:0.0208 train_time:3770221ms step_avg:1578.82ms step:2399/6250 train_loss:2.1379 aux_loss:0.0212 train_time:3771970ms step_avg:1578.89ms step:2400/6250 train_loss:2.0322 aux_loss:0.0207 train_time:3774208ms step_avg:1579.17ms step:2401/6250 train_loss:2.0591 aux_loss:0.0207 train_time:3775454ms step_avg:1579.03ms step:2402/6250 train_loss:1.9803 aux_loss:0.0208 train_time:3776597ms step_avg:1578.84ms step:2403/6250 train_loss:2.0005 aux_loss:0.0205 train_time:3777913ms step_avg:1578.73ms step:2404/6250 train_loss:2.0278 aux_loss:0.0207 train_time:3779213ms step_avg:1578.62ms step:2405/6250 train_loss:2.0011 aux_loss:0.0208 train_time:3780602ms step_avg:1578.54ms step:2406/6250 train_loss:2.0323 aux_loss:0.0205 train_time:3781839ms step_avg:1578.40ms step:2407/6250 train_loss:2.0026 aux_loss:0.0203 train_time:3782961ms step_avg:1578.21ms step:2408/6250 train_loss:2.0551 aux_loss:0.0206 train_time:3784100ms step_avg:1578.02ms step:2409/6250 train_loss:1.9372 aux_loss:0.0204 train_time:3785270ms step_avg:1577.85ms step:2410/6250 train_loss:2.0930 aux_loss:0.0209 train_time:3786514ms step_avg:1577.71ms step:2411/6250 train_loss:2.0569 aux_loss:0.0211 train_time:3787833ms step_avg:1577.61ms step:2412/6250 train_loss:2.0702 aux_loss:0.0207 train_time:3789162ms step_avg:1577.50ms step:2413/6250 train_loss:2.0653 aux_loss:0.0203 train_time:3790403ms step_avg:1577.36ms step:2414/6250 train_loss:1.8967 aux_loss:0.0205 train_time:3791535ms step_avg:1577.18ms step:2415/6250 train_loss:1.9954 aux_loss:0.0209 train_time:3792660ms step_avg:1576.99ms step:2416/6250 train_loss:1.9875 aux_loss:0.0204 train_time:3793875ms step_avg:1576.84ms step:2417/6250 train_loss:2.1174 aux_loss:0.0207 train_time:3795078ms step_avg:1576.68ms step:2418/6250 train_loss:1.9750 aux_loss:0.0206 train_time:3796213ms step_avg:1576.50ms step:2419/6250 train_loss:1.9817 aux_loss:0.0201 train_time:3797445ms step_avg:1576.36ms step:2420/6250 train_loss:1.8768 aux_loss:0.0202 train_time:3799701ms step_avg:1576.64ms step:2421/6250 train_loss:1.9752 aux_loss:0.0209 train_time:3803929ms step_avg:1577.74ms step:2422/6250 train_loss:1.9823 aux_loss:0.0206 train_time:3805264ms step_avg:1577.64ms step:2423/6250 train_loss:1.9639 aux_loss:0.0203 train_time:3806793ms step_avg:1577.62ms step:2424/6250 train_loss:2.0079 aux_loss:0.0204 train_time:3808050ms step_avg:1577.49ms step:2425/6250 train_loss:2.0164 aux_loss:0.0208 train_time:3809836ms step_avg:1577.57ms step:2426/6250 train_loss:2.0282 aux_loss:0.0203 train_time:3811421ms step_avg:1577.58ms step:2427/6250 train_loss:1.9700 aux_loss:0.0202 train_time:3812672ms step_avg:1577.44ms step:2428/6250 train_loss:2.0499 aux_loss:0.0209 train_time:3813990ms step_avg:1577.33ms step:2429/6250 train_loss:1.9204 aux_loss:0.0211 train_time:3815465ms step_avg:1577.29ms step:2430/6250 train_loss:1.9485 aux_loss:0.0205 train_time:3816599ms step_avg:1577.11ms step:2431/6250 train_loss:2.0016 aux_loss:0.0203 train_time:3817778ms step_avg:1576.94ms step:2432/6250 train_loss:1.9620 aux_loss:0.0204 train_time:3819544ms step_avg:1577.02ms step:2433/6250 train_loss:1.9861 aux_loss:0.0207 train_time:3821143ms step_avg:1577.03ms step:2434/6250 train_loss:1.9604 aux_loss:0.0201 train_time:3822258ms step_avg:1576.84ms step:2435/6250 train_loss:2.1106 aux_loss:0.0204 train_time:3823661ms step_avg:1576.77ms step:2436/6250 train_loss:2.0184 aux_loss:0.0205 train_time:3825143ms step_avg:1576.73ms step:2437/6250 train_loss:1.9495 aux_loss:0.0206 train_time:3830701ms step_avg:1578.37ms step:2438/6250 train_loss:2.0395 aux_loss:0.0203 train_time:3832608ms step_avg:1578.50ms step:2439/6250 train_loss:2.0103 aux_loss:0.0202 train_time:3833984ms step_avg:1578.42ms step:2440/6250 train_loss:1.9284 aux_loss:0.0198 train_time:3835533ms step_avg:1578.41ms step:2441/6250 train_loss:1.9404 aux_loss:0.0201 train_time:3836672ms step_avg:1578.23ms step:2442/6250 train_loss:1.9328 aux_loss:0.0204 train_time:3837822ms step_avg:1578.05ms step:2443/6250 train_loss:1.9973 aux_loss:0.0201 train_time:3838962ms step_avg:1577.87ms step:2444/6250 train_loss:2.1052 aux_loss:0.0202 train_time:3841961ms step_avg:1578.46ms step:2445/6250 train_loss:1.9503 aux_loss:0.0201 train_time:3845028ms step_avg:1579.07ms step:2446/6250 train_loss:2.0806 aux_loss:0.0203 train_time:3848127ms step_avg:1579.69ms step:2447/6250 train_loss:1.9893 aux_loss:0.0201 train_time:3850734ms step_avg:1580.11ms step:2448/6250 train_loss:2.0403 aux_loss:0.0203 train_time:3853771ms step_avg:1580.71ms step:2449/6250 train_loss:2.0000 aux_loss:0.0203 train_time:3857983ms step_avg:1581.79ms step:2450/6250 train_loss:1.9480 aux_loss:0.0203 train_time:3860588ms step_avg:1582.21ms step:2451/6250 train_loss:1.9574 aux_loss:0.0202 train_time:3864024ms step_avg:1582.97ms step:2452/6250 train_loss:1.9335 aux_loss:0.0203 train_time:3867956ms step_avg:1583.93ms step:2453/6250 train_loss:1.9244 aux_loss:0.0207 train_time:3871475ms step_avg:1584.72ms step:2454/6250 train_loss:2.0486 aux_loss:0.0203 train_time:3872665ms step_avg:1584.56ms step:2455/6250 train_loss:2.0090 aux_loss:0.0202 train_time:3874190ms step_avg:1584.54ms step:2456/6250 train_loss:1.9835 aux_loss:0.0202 train_time:3875490ms step_avg:1584.42ms step:2457/6250 train_loss:1.9460 aux_loss:0.0206 train_time:3876638ms step_avg:1584.24ms step:2458/6250 train_loss:2.0013 aux_loss:0.0203 train_time:3877838ms step_avg:1584.08ms step:2459/6250 train_loss:1.9169 aux_loss:0.0196 train_time:3879078ms step_avg:1583.94ms step:2460/6250 train_loss:2.0292 aux_loss:0.0201 train_time:3880613ms step_avg:1583.92ms step:2461/6250 train_loss:1.9930 aux_loss:0.0203 train_time:3882060ms step_avg:1583.87ms step:2462/6250 train_loss:1.9894 aux_loss:0.0203 train_time:3883316ms step_avg:1583.73ms step:2463/6250 train_loss:2.0196 aux_loss:0.0200 train_time:3884534ms step_avg:1583.58ms step:2464/6250 train_loss:2.0019 aux_loss:0.0200 train_time:3885776ms step_avg:1583.45ms step:2465/6250 train_loss:1.9871 aux_loss:0.0199 train_time:3888481ms step_avg:1583.90ms step:2466/6250 train_loss:1.9900 aux_loss:0.0200 train_time:3890043ms step_avg:1583.89ms step:2467/6250 train_loss:1.9445 aux_loss:0.0198 train_time:3891272ms step_avg:1583.75ms step:2468/6250 train_loss:2.0247 aux_loss:0.0197 train_time:3892514ms step_avg:1583.61ms step:2469/6250 train_loss:1.9774 aux_loss:0.0198 train_time:3893757ms step_avg:1583.47ms step:2470/6250 train_loss:1.9670 aux_loss:0.0199 train_time:3894987ms step_avg:1583.33ms step:2471/6250 train_loss:2.0226 aux_loss:0.0202 train_time:3896973ms step_avg:1583.49ms step:2472/6250 train_loss:2.0633 aux_loss:0.0201 train_time:3898824ms step_avg:1583.60ms step:2473/6250 train_loss:2.0259 aux_loss:0.0199 train_time:3899947ms step_avg:1583.41ms step:2474/6250 train_loss:2.0428 aux_loss:0.0201 train_time:3901087ms step_avg:1583.23ms step:2475/6250 train_loss:1.9623 aux_loss:0.0200 train_time:3903611ms step_avg:1583.61ms step:2476/6250 train_loss:1.9677 aux_loss:0.0195 train_time:3904944ms step_avg:1583.51ms step:2477/6250 train_loss:1.9762 aux_loss:0.0202 train_time:3906107ms step_avg:1583.34ms step:2478/6250 train_loss:2.0312 aux_loss:0.0202 train_time:3907487ms step_avg:1583.26ms step:2479/6250 train_loss:2.0080 aux_loss:0.0203 train_time:3908825ms step_avg:1583.16ms step:2480/6250 train_loss:2.0365 aux_loss:0.0199 train_time:3911436ms step_avg:1583.58ms step:2481/6250 train_loss:2.0437 aux_loss:0.0201 train_time:3913561ms step_avg:1583.80ms step:2482/6250 train_loss:2.0522 aux_loss:0.0202 train_time:3914726ms step_avg:1583.63ms step:2483/6250 train_loss:1.9166 aux_loss:0.0204 train_time:3916049ms step_avg:1583.52ms step:2484/6250 train_loss:2.0347 aux_loss:0.0203 train_time:3917158ms step_avg:1583.33ms step:2485/6250 train_loss:1.9840 aux_loss:0.0202 train_time:3918785ms step_avg:1583.35ms step:2486/6250 train_loss:1.9531 aux_loss:0.0201 train_time:3920006ms step_avg:1583.20ms step:2487/6250 train_loss:1.9196 aux_loss:0.0201 train_time:3921753ms step_avg:1583.27ms step:2488/6250 train_loss:1.9796 aux_loss:0.0209 train_time:3924849ms step_avg:1583.88ms step:2489/6250 train_loss:1.9428 aux_loss:0.0206 train_time:3926763ms step_avg:1584.01ms step:2490/6250 train_loss:2.0008 aux_loss:0.0200 train_time:3927817ms step_avg:1583.80ms step:2491/6250 train_loss:2.0190 aux_loss:0.0199 train_time:3932179ms step_avg:1584.92ms step:2492/6250 train_loss:1.9694 aux_loss:0.0200 train_time:3933311ms step_avg:1584.73ms step:2493/6250 train_loss:2.1123 aux_loss:0.0202 train_time:3934385ms step_avg:1584.53ms step:2494/6250 train_loss:1.9027 aux_loss:0.0202 train_time:3935566ms step_avg:1584.37ms step:2495/6250 train_loss:1.9678 aux_loss:0.0202 train_time:3936961ms step_avg:1584.29ms step:2496/6250 train_loss:1.9466 aux_loss:0.0201 train_time:3938171ms step_avg:1584.14ms step:2497/6250 train_loss:1.9099 aux_loss:0.0201 train_time:3939918ms step_avg:1584.21ms step:2498/6250 train_loss:1.9650 aux_loss:0.0200 train_time:3941164ms step_avg:1584.07ms step:2499/6250 train_loss:1.9815 aux_loss:0.0201 train_time:3942355ms step_avg:1583.91ms step:2500/6250 train_loss:2.0158 aux_loss:0.0200 train_time:3943620ms step_avg:1583.78ms step:2501/6250 train_loss:2.0544 aux_loss:0.0199 train_time:3944781ms step_avg:1583.61ms step:2502/6250 train_loss:1.8942 aux_loss:0.0197 train_time:3946045ms step_avg:1583.49ms step:2503/6250 train_loss:1.9301 aux_loss:0.0197 train_time:3947194ms step_avg:1583.31ms step:2504/6250 train_loss:2.0100 aux_loss:0.0198 train_time:3948370ms step_avg:1583.15ms step:2505/6250 train_loss:1.9595 aux_loss:0.0202 train_time:3949611ms step_avg:1583.01ms step:2506/6250 train_loss:1.9675 aux_loss:0.0202 train_time:3950824ms step_avg:1582.86ms step:2507/6250 train_loss:1.9026 aux_loss:0.0203 train_time:3952228ms step_avg:1582.79ms step:2508/6250 train_loss:1.8974 aux_loss:0.0198 train_time:3953396ms step_avg:1582.62ms step:2509/6250 train_loss:1.9635 aux_loss:0.0199 train_time:3954571ms step_avg:1582.46ms step:2510/6250 train_loss:1.9896 aux_loss:0.0203 train_time:3955793ms step_avg:1582.32ms step:2511/6250 train_loss:1.9738 aux_loss:0.0198 train_time:3956954ms step_avg:1582.15ms step:2512/6250 train_loss:2.0846 aux_loss:0.0199 train_time:3959154ms step_avg:1582.40ms step:2513/6250 train_loss:1.9606 aux_loss:0.0197 train_time:3961188ms step_avg:1582.58ms step:2514/6250 train_loss:2.0469 aux_loss:0.0200 train_time:3962711ms step_avg:1582.55ms step:2515/6250 train_loss:1.9768 aux_loss:0.0200 train_time:3964433ms step_avg:1582.61ms step:2516/6250 train_loss:1.9648 aux_loss:0.0195 train_time:3965602ms step_avg:1582.44ms step:2517/6250 train_loss:1.9591 aux_loss:0.0200 train_time:3966731ms step_avg:1582.26ms step:2518/6250 train_loss:1.9331 aux_loss:0.0198 train_time:3967958ms step_avg:1582.12ms step:2519/6250 train_loss:1.9560 aux_loss:0.0198 train_time:3969083ms step_avg:1581.94ms step:2520/6250 train_loss:1.9537 aux_loss:0.0203 train_time:3976631ms step_avg:1584.32ms step:2521/6250 train_loss:1.9115 aux_loss:0.0201 train_time:3978458ms step_avg:1584.41ms step:2522/6250 train_loss:2.0129 aux_loss:0.0199 train_time:3980008ms step_avg:1584.40ms step:2523/6250 train_loss:2.0460 aux_loss:0.0202 train_time:3982888ms step_avg:1584.91ms step:2524/6250 train_loss:1.9890 aux_loss:0.0197 train_time:3984506ms step_avg:1584.93ms step:2525/6250 train_loss:1.9706 aux_loss:0.0202 train_time:3985894ms step_avg:1584.85ms step:2526/6250 train_loss:1.9678 aux_loss:0.0202 train_time:3987029ms step_avg:1584.67ms step:2527/6250 train_loss:1.9989 aux_loss:0.0194 train_time:3989362ms step_avg:1584.97ms step:2528/6250 train_loss:2.0300 aux_loss:0.0195 train_time:3990497ms step_avg:1584.79ms step:2529/6250 train_loss:1.9299 aux_loss:0.0202 train_time:3991906ms step_avg:1584.72ms step:2530/6250 train_loss:2.1143 aux_loss:0.0202 train_time:3995538ms step_avg:1585.53ms step:2531/6250 train_loss:1.9752 aux_loss:0.0198 train_time:3996690ms step_avg:1585.36ms step:2532/6250 train_loss:2.0353 aux_loss:0.0198 train_time:3997872ms step_avg:1585.20ms step:2533/6250 train_loss:2.0748 aux_loss:0.0204 train_time:3999509ms step_avg:1585.22ms step:2534/6250 train_loss:1.9210 aux_loss:0.0198 train_time:4001172ms step_avg:1585.25ms step:2535/6250 train_loss:2.1163 aux_loss:0.0201 train_time:4002751ms step_avg:1585.25ms step:2536/6250 train_loss:2.0581 aux_loss:0.0202 train_time:4004562ms step_avg:1585.34ms step:2537/6250 train_loss:1.9369 aux_loss:0.0203 train_time:4005980ms step_avg:1585.27ms step:2538/6250 train_loss:1.9560 aux_loss:0.0201 train_time:4007101ms step_avg:1585.09ms step:2539/6250 train_loss:1.9601 aux_loss:0.0200 train_time:4008187ms step_avg:1584.89ms step:2540/6250 train_loss:1.9912 aux_loss:0.0199 train_time:4009859ms step_avg:1584.92ms step:2541/6250 train_loss:1.9290 aux_loss:0.0197 train_time:4010888ms step_avg:1584.70ms step:2542/6250 train_loss:1.9239 aux_loss:0.0201 train_time:4011903ms step_avg:1584.48ms step:2543/6250 train_loss:1.9290 aux_loss:0.0205 train_time:4012872ms step_avg:1584.24ms step:2544/6250 train_loss:1.9198 aux_loss:0.0205 train_time:4013867ms step_avg:1584.00ms step:2545/6250 train_loss:1.9383 aux_loss:0.0202 train_time:4016461ms step_avg:1584.40ms step:2546/6250 train_loss:2.0265 aux_loss:0.0199 train_time:4017584ms step_avg:1584.22ms step:2547/6250 train_loss:1.8871 aux_loss:0.0199 train_time:4018697ms step_avg:1584.03ms step:2548/6250 train_loss:1.9779 aux_loss:0.0199 train_time:4019816ms step_avg:1583.85ms step:2549/6250 train_loss:1.9866 aux_loss:0.0197 train_time:4020974ms step_avg:1583.68ms step:2550/6250 train_loss:1.9984 aux_loss:0.0195 train_time:4022116ms step_avg:1583.51ms step:2551/6250 train_loss:2.0522 aux_loss:0.0197 train_time:4023298ms step_avg:1583.35ms step:2552/6250 train_loss:2.0127 aux_loss:0.0196 train_time:4024666ms step_avg:1583.27ms step:2553/6250 train_loss:1.9629 aux_loss:0.0197 train_time:4025825ms step_avg:1583.10ms step:2554/6250 train_loss:1.9674 aux_loss:0.0194 train_time:4026966ms step_avg:1582.93ms step:2555/6250 train_loss:1.9592 aux_loss:0.0200 train_time:4028541ms step_avg:1582.92ms step:2556/6250 train_loss:2.0000 aux_loss:0.0198 train_time:4029736ms step_avg:1582.77ms step:2557/6250 train_loss:1.9927 aux_loss:0.0199 train_time:4030869ms step_avg:1582.59ms step:2558/6250 train_loss:2.0037 aux_loss:0.0201 train_time:4031989ms step_avg:1582.41ms step:2559/6250 train_loss:1.9283 aux_loss:0.0199 train_time:4033604ms step_avg:1582.43ms step:2560/6250 train_loss:1.9472 aux_loss:0.0201 train_time:4034959ms step_avg:1582.34ms step:2561/6250 train_loss:1.9680 aux_loss:0.0202 train_time:4036066ms step_avg:1582.15ms step:2562/6250 train_loss:1.9405 aux_loss:0.0199 train_time:4037158ms step_avg:1581.96ms step:2563/6250 train_loss:1.8987 aux_loss:0.0195 train_time:4038558ms step_avg:1581.89ms step:2564/6250 train_loss:1.9018 aux_loss:0.0195 train_time:4040504ms step_avg:1582.03ms step:2565/6250 train_loss:2.0252 aux_loss:0.0198 train_time:4041871ms step_avg:1581.95ms step:2566/6250 train_loss:1.9845 aux_loss:0.0198 train_time:4043157ms step_avg:1581.83ms step:2567/6250 train_loss:1.9855 aux_loss:0.0194 train_time:4050713ms step_avg:1584.17ms step:2568/6250 train_loss:1.8873 aux_loss:0.0195 train_time:4051842ms step_avg:1583.99ms step:2569/6250 train_loss:1.9879 aux_loss:0.0198 train_time:4053889ms step_avg:1584.17ms step:2570/6250 train_loss:2.0450 aux_loss:0.0196 train_time:4054748ms step_avg:1583.89ms step:2571/6250 train_loss:2.0458 aux_loss:0.0198 train_time:4056989ms step_avg:1584.14ms step:2572/6250 train_loss:1.9053 aux_loss:0.0198 train_time:4058935ms step_avg:1584.28ms step:2573/6250 train_loss:2.0122 aux_loss:0.0199 train_time:4060580ms step_avg:1584.31ms step:2574/6250 train_loss:1.9492 aux_loss:0.0201 train_time:4061610ms step_avg:1584.09ms step:2575/6250 train_loss:1.9800 aux_loss:0.0200 train_time:4062544ms step_avg:1583.84ms step:2576/6250 train_loss:2.0281 aux_loss:0.0196 train_time:4064282ms step_avg:1583.90ms step:2577/6250 train_loss:1.9971 aux_loss:0.0205 train_time:4065133ms step_avg:1583.61ms step:2578/6250 train_loss:2.0653 aux_loss:0.0197 train_time:4066020ms step_avg:1583.34ms step:2579/6250 train_loss:1.9620 aux_loss:0.0194 train_time:4066876ms step_avg:1583.06ms step:2580/6250 train_loss:1.9825 aux_loss:0.0196 train_time:4067722ms step_avg:1582.77ms step:2581/6250 train_loss:1.9926 aux_loss:0.0199 train_time:4068606ms step_avg:1582.50ms step:2582/6250 train_loss:1.9270 aux_loss:0.0196 train_time:4069472ms step_avg:1582.22ms step:2583/6250 train_loss:1.9957 aux_loss:0.0196 train_time:4070318ms step_avg:1581.93ms step:2584/6250 train_loss:1.9905 aux_loss:0.0196 train_time:4072237ms step_avg:1582.07ms step:2585/6250 train_loss:1.9995 aux_loss:0.0197 train_time:4073552ms step_avg:1581.96ms step:2586/6250 train_loss:1.9521 aux_loss:0.0196 train_time:4074915ms step_avg:1581.88ms step:2587/6250 train_loss:1.9918 aux_loss:0.0199 train_time:4076343ms step_avg:1581.82ms step:2588/6250 train_loss:1.9553 aux_loss:0.0198 train_time:4077319ms step_avg:1581.58ms step:2589/6250 train_loss:2.0804 aux_loss:0.0198 train_time:4078405ms step_avg:1581.39ms step:2590/6250 train_loss:1.9059 aux_loss:0.0192 train_time:4079269ms step_avg:1581.11ms step:2591/6250 train_loss:1.9741 aux_loss:0.0197 train_time:4080145ms step_avg:1580.84ms step:2592/6250 train_loss:2.0281 aux_loss:0.0198 train_time:4083656ms step_avg:1581.59ms step:2593/6250 train_loss:1.9546 aux_loss:0.0195 train_time:4085824ms step_avg:1581.81ms step:2594/6250 train_loss:1.9800 aux_loss:0.0196 train_time:4088500ms step_avg:1582.24ms step:2595/6250 train_loss:1.9818 aux_loss:0.0198 train_time:4089711ms step_avg:1582.09ms step:2596/6250 train_loss:1.9833 aux_loss:0.0199 train_time:4091259ms step_avg:1582.08ms step:2597/6250 train_loss:2.0887 aux_loss:0.0199 train_time:4096650ms step_avg:1583.55ms step:2598/6250 train_loss:1.9434 aux_loss:0.0199 train_time:4099496ms step_avg:1584.04ms step:2599/6250 train_loss:1.9285 aux_loss:0.0193 train_time:4103622ms step_avg:1585.02ms step:2600/6250 train_loss:1.8943 aux_loss:0.0194 train_time:4105908ms step_avg:1585.29ms step:2601/6250 train_loss:1.9523 aux_loss:0.0196 train_time:4108414ms step_avg:1585.65ms step:2602/6250 train_loss:1.9357 aux_loss:0.0198 train_time:4110247ms step_avg:1585.74ms step:2603/6250 train_loss:2.0134 aux_loss:0.0196 train_time:4111851ms step_avg:1585.75ms step:2604/6250 train_loss:1.9779 aux_loss:0.0198 train_time:4113574ms step_avg:1585.80ms step:2605/6250 train_loss:2.0480 aux_loss:0.0195 train_time:4115125ms step_avg:1585.79ms step:2606/6250 train_loss:1.9860 aux_loss:0.0193 train_time:4127519ms step_avg:1589.95ms step:2607/6250 train_loss:2.0989 aux_loss:0.0197 train_time:4161472ms step_avg:1602.42ms step:2608/6250 train_loss:1.9407 aux_loss:0.0193 train_time:4202016ms step_avg:1617.40ms step:2609/6250 train_loss:1.9936 aux_loss:0.0191 train_time:4238449ms step_avg:1630.80ms step:2610/6250 train_loss:2.0068 aux_loss:0.0192 train_time:4264234ms step_avg:1640.09ms step:2611/6250 train_loss:1.9791 aux_loss:0.0196 train_time:4295906ms step_avg:1651.64ms step:2612/6250 train_loss:1.9819 aux_loss:0.0195 train_time:4300408ms step_avg:1652.73ms step:2613/6250 train_loss:1.9748 aux_loss:0.0192 train_time:4302075ms step_avg:1652.74ms step:2614/6250 train_loss:2.0714 aux_loss:0.0192 train_time:4303518ms step_avg:1652.66ms step:2615/6250 train_loss:1.9288 aux_loss:0.0191 train_time:4305666ms step_avg:1652.85ms step:2616/6250 train_loss:1.9947 aux_loss:0.0192 train_time:4306794ms step_avg:1652.65ms step:2617/6250 train_loss:2.0117 aux_loss:0.0196 train_time:4307972ms step_avg:1652.46ms step:2618/6250 train_loss:2.0110 aux_loss:0.0200 train_time:4310217ms step_avg:1652.69ms step:2619/6250 train_loss:1.9271 aux_loss:0.0194 train_time:4312218ms step_avg:1652.82ms step:2620/6250 train_loss:1.9499 aux_loss:0.0195 train_time:4316100ms step_avg:1653.68ms step:2621/6250 train_loss:1.9599 aux_loss:0.0194 train_time:4317155ms step_avg:1653.45ms step:2622/6250 train_loss:1.9902 aux_loss:0.0195 train_time:4319444ms step_avg:1653.69ms step:2623/6250 train_loss:1.8557 aux_loss:0.0189 train_time:4321149ms step_avg:1653.71ms step:2624/6250 train_loss:1.9596 aux_loss:0.0190 train_time:4322470ms step_avg:1653.58ms step:2625/6250 train_loss:2.0174 aux_loss:0.0198 train_time:4325133ms step_avg:1653.97ms step:2626/6250 train_loss:1.9440 aux_loss:0.0195 train_time:4326953ms step_avg:1654.03ms step:2627/6250 train_loss:1.9445 aux_loss:0.0190 train_time:4328156ms step_avg:1653.86ms step:2628/6250 train_loss:1.9528 aux_loss:0.0195 train_time:4329710ms step_avg:1653.82ms step:2629/6250 train_loss:2.0246 aux_loss:0.0197 train_time:4330882ms step_avg:1653.64ms step:2630/6250 train_loss:1.9814 aux_loss:0.0194 train_time:4332470ms step_avg:1653.61ms step:2631/6250 train_loss:1.9203 aux_loss:0.0192 train_time:4334488ms step_avg:1653.75ms step:2632/6250 train_loss:1.9631 aux_loss:0.0192 train_time:4335657ms step_avg:1653.57ms step:2633/6250 train_loss:1.9233 aux_loss:0.0195 train_time:4336874ms step_avg:1653.40ms step:2634/6250 train_loss:2.0242 aux_loss:0.0196 train_time:4337833ms step_avg:1653.14ms step:2635/6250 train_loss:2.0183 aux_loss:0.0197 train_time:4338750ms step_avg:1652.86ms step:2636/6250 train_loss:1.9542 aux_loss:0.0190 train_time:4339624ms step_avg:1652.56ms step:2637/6250 train_loss:1.9907 aux_loss:0.0190 train_time:4340475ms step_avg:1652.26ms step:2638/6250 train_loss:1.9477 aux_loss:0.0194 train_time:4341357ms step_avg:1651.96ms step:2639/6250 train_loss:1.9839 aux_loss:0.0196 train_time:4344180ms step_avg:1652.41ms step:2640/6250 train_loss:2.0665 aux_loss:0.0195 train_time:4345724ms step_avg:1652.37ms step:2641/6250 train_loss:1.9397 aux_loss:0.0192 train_time:4346795ms step_avg:1652.15ms step:2642/6250 train_loss:1.9400 aux_loss:0.0193 train_time:4347653ms step_avg:1651.84ms step:2643/6250 train_loss:2.0180 aux_loss:0.0199 train_time:4348506ms step_avg:1651.54ms step:2644/6250 train_loss:1.8859 aux_loss:0.0194 train_time:4349352ms step_avg:1651.23ms step:2645/6250 train_loss:1.9180 aux_loss:0.0194 train_time:4350194ms step_avg:1650.93ms step:2646/6250 train_loss:1.8966 aux_loss:0.0188 train_time:4351061ms step_avg:1650.63ms step:2647/6250 train_loss:2.0711 aux_loss:0.0194 train_time:4351925ms step_avg:1650.33ms step:2648/6250 train_loss:2.0233 aux_loss:0.0195 train_time:4352792ms step_avg:1650.03ms step:2649/6250 train_loss:1.9734 aux_loss:0.0193 train_time:4353662ms step_avg:1649.74ms step:2650/6250 train_loss:2.0166 aux_loss:0.0197 train_time:4354531ms step_avg:1649.44ms step:2651/6250 train_loss:2.0484 aux_loss:0.0199 train_time:4355607ms step_avg:1649.23ms step:2652/6250 train_loss:2.0907 aux_loss:0.0196 train_time:4356536ms step_avg:1648.95ms step:2653/6250 train_loss:1.9699 aux_loss:0.0193 train_time:4357395ms step_avg:1648.65ms step:2654/6250 train_loss:1.9734 aux_loss:0.0191 train_time:4358234ms step_avg:1648.35ms step:2655/6250 train_loss:1.9560 aux_loss:0.0195 train_time:4359094ms step_avg:1648.05ms step:2656/6250 train_loss:2.0412 aux_loss:0.0194 train_time:4359937ms step_avg:1647.75ms step:2657/6250 train_loss:2.0369 aux_loss:0.0195 train_time:4360824ms step_avg:1647.46ms step:2658/6250 train_loss:1.9904 aux_loss:0.0191 train_time:4361785ms step_avg:1647.20ms step:2659/6250 train_loss:1.8542 aux_loss:0.0191 train_time:4362883ms step_avg:1646.99ms step:2660/6250 train_loss:2.0208 aux_loss:0.0193 train_time:4364094ms step_avg:1646.83ms step:2661/6250 train_loss:1.8971 aux_loss:0.0192 train_time:4364941ms step_avg:1646.53ms step:2662/6250 train_loss:1.9597 aux_loss:0.0191 train_time:4366456ms step_avg:1646.48ms step:2663/6250 train_loss:1.9950 aux_loss:0.0191 train_time:4367542ms step_avg:1646.27ms step:2664/6250 train_loss:1.9293 aux_loss:0.0188 train_time:4368404ms step_avg:1645.97ms step:2665/6250 train_loss:1.8981 aux_loss:0.0189 train_time:4369246ms step_avg:1645.67ms step:2666/6250 train_loss:1.9959 aux_loss:0.0194 train_time:4370118ms step_avg:1645.38ms step:2667/6250 train_loss:2.0567 aux_loss:0.0201 train_time:4370975ms step_avg:1645.08ms step:2668/6250 train_loss:1.9701 aux_loss:0.0193 train_time:4371835ms step_avg:1644.78ms step:2669/6250 train_loss:1.9387 aux_loss:0.0189 train_time:4372705ms step_avg:1644.49ms step:2670/6250 train_loss:1.9979 aux_loss:0.0195 train_time:4373563ms step_avg:1644.20ms step:2671/6250 train_loss:1.9485 aux_loss:0.0194 train_time:4374423ms step_avg:1643.90ms step:2672/6250 train_loss:1.9446 aux_loss:0.0192 train_time:4375587ms step_avg:1643.72ms step:2673/6250 train_loss:1.9379 aux_loss:0.0189 train_time:4376461ms step_avg:1643.43ms step:2674/6250 train_loss:2.1119 aux_loss:0.0194 train_time:4377337ms step_avg:1643.14ms step:2675/6250 train_loss:1.9447 aux_loss:0.0195 train_time:4378223ms step_avg:1642.86ms step:2676/6250 train_loss:1.8696 aux_loss:0.0190 train_time:4379336ms step_avg:1642.66ms step:2677/6250 train_loss:2.0284 aux_loss:0.0193 train_time:4380581ms step_avg:1642.51ms step:2678/6250 train_loss:1.9797 aux_loss:0.0193 train_time:4381490ms step_avg:1642.24ms step:2679/6250 train_loss:1.9654 aux_loss:0.0193 train_time:4382399ms step_avg:1641.96ms step:2680/6250 train_loss:2.0085 aux_loss:0.0191 train_time:4383252ms step_avg:1641.67ms step:2681/6250 train_loss:2.0149 aux_loss:0.0194 train_time:4384302ms step_avg:1641.45ms step:2682/6250 train_loss:1.9283 aux_loss:0.0190 train_time:4385977ms step_avg:1641.46ms step:2683/6250 train_loss:1.9212 aux_loss:0.0188 train_time:4387501ms step_avg:1641.41ms step:2684/6250 train_loss:1.8943 aux_loss:0.0193 train_time:4388410ms step_avg:1641.14ms step:2685/6250 train_loss:1.9530 aux_loss:0.0190 train_time:4389256ms step_avg:1640.84ms step:2686/6250 train_loss:1.9890 aux_loss:0.0191 train_time:4390100ms step_avg:1640.55ms step:2687/6250 train_loss:2.0171 aux_loss:0.0192 train_time:4390959ms step_avg:1640.25ms step:2688/6250 train_loss:1.9192 aux_loss:0.0189 train_time:4391818ms step_avg:1639.96ms step:2689/6250 train_loss:1.9487 aux_loss:0.0190 train_time:4392681ms step_avg:1639.67ms step:2690/6250 train_loss:1.9373 aux_loss:0.0188 train_time:4393544ms step_avg:1639.38ms step:2691/6250 train_loss:1.9618 aux_loss:0.0192 train_time:4394399ms step_avg:1639.09ms step:2692/6250 train_loss:2.0049 aux_loss:0.0191 train_time:4395236ms step_avg:1638.79ms step:2693/6250 train_loss:1.9847 aux_loss:0.0188 train_time:4396109ms step_avg:1638.51ms step:2694/6250 train_loss:1.9290 aux_loss:0.0187 train_time:4396971ms step_avg:1638.22ms step:2695/6250 train_loss:1.9993 aux_loss:0.0190 train_time:4397844ms step_avg:1637.93ms step:2696/6250 train_loss:2.0274 aux_loss:0.0192 train_time:4398716ms step_avg:1637.65ms step:2697/6250 train_loss:1.9409 aux_loss:0.0191 train_time:4399580ms step_avg:1637.36ms step:2698/6250 train_loss:1.9016 aux_loss:0.0189 train_time:4400438ms step_avg:1637.07ms step:2699/6250 train_loss:1.9174 aux_loss:0.0189 train_time:4401285ms step_avg:1636.77ms step:2700/6250 train_loss:2.0140 aux_loss:0.0191 train_time:4402127ms step_avg:1636.48ms step:2701/6250 train_loss:1.9899 aux_loss:0.0192 train_time:4403025ms step_avg:1636.20ms step:2702/6250 train_loss:1.9353 aux_loss:0.0194 train_time:4403919ms step_avg:1635.93ms step:2703/6250 train_loss:1.9478 aux_loss:0.0195 train_time:4404853ms step_avg:1635.67ms step:2704/6250 train_loss:1.8509 aux_loss:0.0189 train_time:4405735ms step_avg:1635.39ms step:2705/6250 train_loss:1.8934 aux_loss:0.0191 train_time:4406684ms step_avg:1635.13ms step:2706/6250 train_loss:1.9317 aux_loss:0.0192 train_time:4407624ms step_avg:1634.88ms step:2707/6250 train_loss:1.9349 aux_loss:0.0191 train_time:4408564ms step_avg:1634.62ms step:2708/6250 train_loss:1.9652 aux_loss:0.0190 train_time:4409563ms step_avg:1634.38ms step:2709/6250 train_loss:1.8795 aux_loss:0.0189 train_time:4410436ms step_avg:1634.10ms step:2710/6250 train_loss:1.9913 aux_loss:0.0189 train_time:4411342ms step_avg:1633.83ms step:2711/6250 train_loss:1.9476 aux_loss:0.0193 train_time:4412236ms step_avg:1633.56ms step:2712/6250 train_loss:1.9929 aux_loss:0.0193 train_time:4413183ms step_avg:1633.30ms step:2713/6250 train_loss:2.0517 aux_loss:0.0191 train_time:4414090ms step_avg:1633.03ms step:2714/6250 train_loss:2.0182 aux_loss:0.0189 train_time:4414975ms step_avg:1632.76ms step:2715/6250 train_loss:1.9277 aux_loss:0.0191 train_time:4415913ms step_avg:1632.50ms step:2716/6250 train_loss:1.8499 aux_loss:0.0187 train_time:4416894ms step_avg:1632.26ms step:2717/6250 train_loss:2.1183 aux_loss:0.0186 train_time:4417876ms step_avg:1632.02ms step:2718/6250 train_loss:1.8562 aux_loss:0.0189 train_time:4418793ms step_avg:1631.76ms step:2719/6250 train_loss:2.0260 aux_loss:0.0191 train_time:4419669ms step_avg:1631.48ms step:2720/6250 train_loss:2.0051 aux_loss:0.0189 train_time:4421038ms step_avg:1631.38ms step:2721/6250 train_loss:1.9670 aux_loss:0.0193 train_time:4422760ms step_avg:1631.41ms step:2722/6250 train_loss:1.9533 aux_loss:0.0187 train_time:4424159ms step_avg:1631.33ms step:2723/6250 train_loss:1.8755 aux_loss:0.0189 train_time:4425522ms step_avg:1631.23ms step:2724/6250 train_loss:1.9528 aux_loss:0.0190 train_time:4426991ms step_avg:1631.17ms step:2725/6250 train_loss:1.8718 aux_loss:0.0188 train_time:4428008ms step_avg:1630.94ms step:2726/6250 train_loss:1.9601 aux_loss:0.0191 train_time:4428907ms step_avg:1630.67ms step:2727/6250 train_loss:1.9736 aux_loss:0.0190 train_time:4429768ms step_avg:1630.39ms step:2728/6250 train_loss:1.8738 aux_loss:0.0195 train_time:4430637ms step_avg:1630.11ms step:2729/6250 train_loss:2.1082 aux_loss:0.0188 train_time:4431517ms step_avg:1629.83ms step:2730/6250 train_loss:1.9474 aux_loss:0.0189 train_time:4432389ms step_avg:1629.55ms step:2731/6250 train_loss:1.9965 aux_loss:0.0189 train_time:4433251ms step_avg:1629.27ms step:2732/6250 train_loss:1.9518 aux_loss:0.0190 train_time:4434131ms step_avg:1629.00ms step:2733/6250 train_loss:2.0125 aux_loss:0.0189 train_time:4435009ms step_avg:1628.72ms step:2734/6250 train_loss:1.9965 aux_loss:0.0191 train_time:4435881ms step_avg:1628.44ms step:2735/6250 train_loss:1.9349 aux_loss:0.0190 train_time:4436762ms step_avg:1628.17ms step:2736/6250 train_loss:1.9791 aux_loss:0.0187 train_time:4437629ms step_avg:1627.89ms step:2737/6250 train_loss:1.9208 aux_loss:0.0187 train_time:4438517ms step_avg:1627.62ms step:2738/6250 train_loss:1.9344 aux_loss:0.0189 train_time:4439400ms step_avg:1627.35ms step:2739/6250 train_loss:1.9313 aux_loss:0.0189 train_time:4440279ms step_avg:1627.07ms step:2740/6250 train_loss:1.9938 aux_loss:0.0190 train_time:4442456ms step_avg:1627.27ms step:2741/6250 train_loss:1.9621 aux_loss:0.0188 train_time:4444270ms step_avg:1627.34ms step:2742/6250 train_loss:1.9718 aux_loss:0.0192 train_time:4445618ms step_avg:1627.24ms step:2743/6250 train_loss:1.9411 aux_loss:0.0190 train_time:4447879ms step_avg:1627.47ms step:2744/6250 train_loss:2.0202 aux_loss:0.0190 train_time:4448984ms step_avg:1627.28ms step:2745/6250 train_loss:1.9628 aux_loss:0.0187 train_time:4450079ms step_avg:1627.09ms step:2746/6250 train_loss:1.9557 aux_loss:0.0190 train_time:4451205ms step_avg:1626.90ms step:2747/6250 train_loss:2.0232 aux_loss:0.0188 train_time:4452316ms step_avg:1626.71ms step:2748/6250 train_loss:1.9002 aux_loss:0.0189 train_time:4453411ms step_avg:1626.52ms step:2749/6250 train_loss:1.9606 aux_loss:0.0186 train_time:4455253ms step_avg:1626.60ms step:2750/6250 train_loss:1.9804 aux_loss:0.0189 train_time:4456369ms step_avg:1626.41ms step:2751/6250 train_loss:1.9775 aux_loss:0.0193 train_time:4457480ms step_avg:1626.22ms step:2752/6250 train_loss:1.9612 aux_loss:0.0187 train_time:4458937ms step_avg:1626.16ms step:2753/6250 train_loss:1.9637 aux_loss:0.0182 train_time:4460346ms step_avg:1626.08ms step:2754/6250 train_loss:1.9381 aux_loss:0.0187 train_time:4461674ms step_avg:1625.97ms step:2755/6250 train_loss:1.9612 aux_loss:0.0190 train_time:4462953ms step_avg:1625.85ms step:2756/6250 train_loss:1.9377 aux_loss:0.0192 train_time:4464091ms step_avg:1625.67ms step:2757/6250 train_loss:1.8751 aux_loss:0.0190 train_time:4465220ms step_avg:1625.49ms step:2758/6250 train_loss:1.8609 aux_loss:0.0189 train_time:4466246ms step_avg:1625.27ms step:2759/6250 train_loss:1.8805 aux_loss:0.0186 train_time:4467643ms step_avg:1625.19ms step:2760/6250 train_loss:1.8690 aux_loss:0.0189 train_time:4469590ms step_avg:1625.31ms step:2761/6250 train_loss:1.9424 aux_loss:0.0189 train_time:4471300ms step_avg:1625.34ms step:2762/6250 train_loss:2.0279 aux_loss:0.0190 train_time:4474263ms step_avg:1625.82ms step:2763/6250 train_loss:1.8864 aux_loss:0.0194 train_time:4477060ms step_avg:1626.25ms step:2764/6250 train_loss:1.9318 aux_loss:0.0189 train_time:4478156ms step_avg:1626.06ms step:2765/6250 train_loss:1.9330 aux_loss:0.0183 train_time:4480454ms step_avg:1626.30ms step:2766/6250 train_loss:1.9316 aux_loss:0.0187 train_time:4481938ms step_avg:1626.25ms step:2767/6250 train_loss:1.9610 aux_loss:0.0190 train_time:4483018ms step_avg:1626.05ms step:2768/6250 train_loss:1.9549 aux_loss:0.0188 train_time:4483892ms step_avg:1625.78ms step:2769/6250 train_loss:1.9803 aux_loss:0.0192 train_time:4484770ms step_avg:1625.51ms step:2770/6250 train_loss:2.1007 aux_loss:0.0190 train_time:4485649ms step_avg:1625.24ms step:2771/6250 train_loss:1.9744 aux_loss:0.0189 train_time:4486523ms step_avg:1624.96ms step:2772/6250 train_loss:2.0058 aux_loss:0.0189 train_time:4487391ms step_avg:1624.69ms step:2773/6250 train_loss:1.9422 aux_loss:0.0189 train_time:4488260ms step_avg:1624.42ms step:2774/6250 train_loss:1.9567 aux_loss:0.0190 train_time:4489483ms step_avg:1624.27ms step:2775/6250 train_loss:1.9971 aux_loss:0.0190 train_time:4490516ms step_avg:1624.06ms step:2776/6250 train_loss:1.8931 aux_loss:0.0191 train_time:4491534ms step_avg:1623.84ms step:2777/6250 train_loss:2.0797 aux_loss:0.0193 train_time:4492427ms step_avg:1623.57ms step:2778/6250 train_loss:1.9354 aux_loss:0.0192 train_time:4493314ms step_avg:1623.31ms step:2779/6250 train_loss:1.9336 aux_loss:0.0186 train_time:4494185ms step_avg:1623.04ms step:2780/6250 train_loss:2.0783 aux_loss:0.0186 train_time:4495072ms step_avg:1622.77ms step:2781/6250 train_loss:1.9387 aux_loss:0.0191 train_time:4496134ms step_avg:1622.57ms step:2782/6250 train_loss:2.0041 aux_loss:0.0190 train_time:4497012ms step_avg:1622.30ms step:2783/6250 train_loss:1.9372 aux_loss:0.0186 train_time:4497900ms step_avg:1622.03ms step:2784/6250 train_loss:1.7923 aux_loss:0.0191 train_time:4498766ms step_avg:1621.76ms step:2785/6250 train_loss:1.8395 aux_loss:0.0184 train_time:4499649ms step_avg:1621.50ms step:2786/6250 train_loss:1.9662 aux_loss:0.0188 train_time:4500512ms step_avg:1621.22ms step:2787/6250 train_loss:1.9831 aux_loss:0.0191 train_time:4501386ms step_avg:1620.95ms step:2788/6250 train_loss:2.0274 aux_loss:0.0188 train_time:4502235ms step_avg:1620.67ms step:2789/6250 train_loss:1.9728 aux_loss:0.0184 train_time:4503122ms step_avg:1620.41ms step:2790/6250 train_loss:1.9561 aux_loss:0.0191 train_time:4504461ms step_avg:1620.31ms step:2791/6250 train_loss:2.0153 aux_loss:0.0184 train_time:4505586ms step_avg:1620.13ms step:2792/6250 train_loss:1.9445 aux_loss:0.0188 train_time:4506668ms step_avg:1619.94ms step:2793/6250 train_loss:1.9199 aux_loss:0.0190 train_time:4508199ms step_avg:1619.91ms step:2794/6250 train_loss:1.8322 aux_loss:0.0184 train_time:4509385ms step_avg:1619.75ms step:2795/6250 train_loss:1.9798 aux_loss:0.0189 train_time:4510220ms step_avg:1619.47ms step:2796/6250 train_loss:1.9110 aux_loss:0.0191 train_time:4511543ms step_avg:1619.36ms step:2797/6250 train_loss:2.0208 aux_loss:0.0181 train_time:4512429ms step_avg:1619.10ms step:2798/6250 train_loss:1.8602 aux_loss:0.0182 train_time:4513873ms step_avg:1619.04ms step:2799/6250 train_loss:1.9433 aux_loss:0.0182 train_time:4514799ms step_avg:1618.79ms step:2800/6250 train_loss:2.0661 aux_loss:0.0184 train_time:4515671ms step_avg:1618.52ms step:2801/6250 train_loss:1.9203 aux_loss:0.0186 train_time:4516556ms step_avg:1618.26ms step:2802/6250 train_loss:1.9469 aux_loss:0.0184 train_time:4517601ms step_avg:1618.05ms step:2803/6250 train_loss:2.0140 aux_loss:0.0184 train_time:4518492ms step_avg:1617.79ms step:2804/6250 train_loss:1.8697 aux_loss:0.0186 train_time:4519369ms step_avg:1617.53ms step:2805/6250 train_loss:1.9227 aux_loss:0.0185 train_time:4520217ms step_avg:1617.25ms step:2806/6250 train_loss:2.0477 aux_loss:0.0184 train_time:4521084ms step_avg:1616.98ms step:2807/6250 train_loss:1.9960 aux_loss:0.0185 train_time:4521940ms step_avg:1616.71ms step:2808/6250 train_loss:1.8881 aux_loss:0.0184 train_time:4522810ms step_avg:1616.44ms step:2809/6250 train_loss:1.8588 aux_loss:0.0183 train_time:4523681ms step_avg:1616.18ms step:2810/6250 train_loss:1.9117 aux_loss:0.0189 train_time:4524548ms step_avg:1615.91ms step:2811/6250 train_loss:1.8854 aux_loss:0.0186 train_time:4526066ms step_avg:1615.88ms step:2812/6250 train_loss:1.9489 aux_loss:0.0188 train_time:4526927ms step_avg:1615.61ms step:2813/6250 train_loss:2.1164 aux_loss:0.0187 train_time:4527807ms step_avg:1615.34ms step:2814/6250 train_loss:2.0715 aux_loss:0.0185 train_time:4528672ms step_avg:1615.08ms step:2815/6250 train_loss:2.0432 aux_loss:0.0187 train_time:4529558ms step_avg:1614.82ms step:2816/6250 train_loss:1.9124 aux_loss:0.0183 train_time:4530732ms step_avg:1614.66ms step:2817/6250 train_loss:1.9036 aux_loss:0.0183 train_time:4532508ms step_avg:1614.72ms step:2818/6250 train_loss:1.9739 aux_loss:0.0190 train_time:4533498ms step_avg:1614.49ms step:2819/6250 train_loss:1.9930 aux_loss:0.0187 train_time:4534388ms step_avg:1614.24ms step:2820/6250 train_loss:1.9005 aux_loss:0.0182 train_time:4535238ms step_avg:1613.96ms step:2821/6250 train_loss:1.9316 aux_loss:0.0190 train_time:4536088ms step_avg:1613.69ms step:2822/6250 train_loss:1.9876 aux_loss:0.0186 train_time:4537249ms step_avg:1613.53ms step:2823/6250 train_loss:2.0780 aux_loss:0.0190 train_time:4538165ms step_avg:1613.28ms step:2824/6250 train_loss:1.9480 aux_loss:0.0182 train_time:4539652ms step_avg:1613.24ms step:2825/6250 train_loss:1.9873 aux_loss:0.0183 train_time:4542017ms step_avg:1613.51ms step:2826/6250 train_loss:2.0307 aux_loss:0.0185 train_time:4543077ms step_avg:1613.31ms step:2827/6250 train_loss:1.9572 aux_loss:0.0189 train_time:4543992ms step_avg:1613.06ms step:2828/6250 train_loss:1.9060 aux_loss:0.0180 train_time:4544859ms step_avg:1612.80ms step:2829/6250 train_loss:2.0337 aux_loss:0.0187 train_time:4545704ms step_avg:1612.52ms step:2830/6250 train_loss:1.8863 aux_loss:0.0186 train_time:4546563ms step_avg:1612.26ms step:2831/6250 train_loss:1.9989 aux_loss:0.0183 train_time:4547419ms step_avg:1611.99ms step:2832/6250 train_loss:1.9284 aux_loss:0.0187 train_time:4548276ms step_avg:1611.72ms step:2833/6250 train_loss:1.8974 aux_loss:0.0184 train_time:4549189ms step_avg:1611.47ms step:2834/6250 train_loss:1.9466 aux_loss:0.0183 train_time:4550965ms step_avg:1611.53ms step:2835/6250 train_loss:2.0297 aux_loss:0.0185 train_time:4552199ms step_avg:1611.40ms step:2836/6250 train_loss:2.0051 aux_loss:0.0184 train_time:4553957ms step_avg:1611.45ms step:2837/6250 train_loss:2.0057 aux_loss:0.0183 train_time:4555920ms step_avg:1611.57ms step:2838/6250 train_loss:1.9570 aux_loss:0.0183 train_time:4557416ms step_avg:1611.53ms step:2839/6250 train_loss:1.8308 aux_loss:0.0182 train_time:4558420ms step_avg:1611.32ms step:2840/6250 train_loss:2.0015 aux_loss:0.0184 train_time:4559791ms step_avg:1611.23ms step:2841/6250 train_loss:1.9040 aux_loss:0.0180 train_time:4560900ms step_avg:1611.06ms step:2842/6250 train_loss:1.9806 aux_loss:0.0184 train_time:4561992ms step_avg:1610.87ms step:2843/6250 train_loss:1.9918 aux_loss:0.0184 train_time:4562995ms step_avg:1610.66ms step:2844/6250 train_loss:1.8726 aux_loss:0.0185 train_time:4563950ms step_avg:1610.43ms step:2845/6250 train_loss:1.9394 aux_loss:0.0182 train_time:4564875ms step_avg:1610.19ms step:2846/6250 train_loss:1.9227 aux_loss:0.0186 train_time:4566121ms step_avg:1610.06ms step:2847/6250 train_loss:1.8890 aux_loss:0.0179 train_time:4567364ms step_avg:1609.93ms step:2848/6250 train_loss:1.9931 aux_loss:0.0177 train_time:4568491ms step_avg:1609.76ms step:2849/6250 train_loss:1.9267 aux_loss:0.0182 train_time:4569688ms step_avg:1609.61ms step:2850/6250 train_loss:1.9267 aux_loss:0.0185 train_time:4570674ms step_avg:1609.39ms step:2851/6250 train_loss:1.9647 aux_loss:0.0180 train_time:4571651ms step_avg:1609.17ms step:2852/6250 train_loss:2.0276 aux_loss:0.0181 train_time:4572558ms step_avg:1608.92ms step:2853/6250 train_loss:1.9189 aux_loss:0.0181 train_time:4573496ms step_avg:1608.69ms step:2854/6250 train_loss:1.9746 aux_loss:0.0185 train_time:4574446ms step_avg:1608.45ms step:2855/6250 train_loss:1.9877 aux_loss:0.0181 train_time:4575494ms step_avg:1608.26ms step:2856/6250 train_loss:1.9135 aux_loss:0.0180 train_time:4576420ms step_avg:1608.02ms step:2857/6250 train_loss:1.9043 aux_loss:0.0182 train_time:4577274ms step_avg:1607.75ms step:2858/6250 train_loss:1.9189 aux_loss:0.0186 train_time:4578136ms step_avg:1607.49ms step:2859/6250 train_loss:1.9477 aux_loss:0.0182 train_time:4579320ms step_avg:1607.34ms step:2860/6250 train_loss:1.9459 aux_loss:0.0186 train_time:4580457ms step_avg:1607.18ms step:2861/6250 train_loss:1.9547 aux_loss:0.0183 train_time:4581334ms step_avg:1606.92ms step:2862/6250 train_loss:1.9060 aux_loss:0.0181 train_time:4583156ms step_avg:1607.00ms step:2863/6250 train_loss:1.9075 aux_loss:0.0181 train_time:4585254ms step_avg:1607.17ms step:2864/6250 train_loss:2.0765 aux_loss:0.0187 train_time:4586103ms step_avg:1606.90ms step:2865/6250 train_loss:1.9253 aux_loss:0.0183 train_time:4587697ms step_avg:1606.90ms step:2866/6250 train_loss:1.9873 aux_loss:0.0181 train_time:4588707ms step_avg:1606.69ms step:2867/6250 train_loss:1.9207 aux_loss:0.0186 train_time:4589646ms step_avg:1606.46ms step:2868/6250 train_loss:1.9443 aux_loss:0.0185 train_time:4590552ms step_avg:1606.21ms step:2869/6250 train_loss:1.9501 aux_loss:0.0180 train_time:4591734ms step_avg:1606.06ms step:2870/6250 train_loss:1.9753 aux_loss:0.0181 train_time:4593319ms step_avg:1606.06ms step:2871/6250 train_loss:1.9849 aux_loss:0.0192 train_time:4594313ms step_avg:1605.84ms step:2872/6250 train_loss:1.9189 aux_loss:0.0188 train_time:4595196ms step_avg:1605.59ms step:2873/6250 train_loss:1.9677 aux_loss:0.0187 train_time:4596105ms step_avg:1605.35ms step:2874/6250 train_loss:1.9126 aux_loss:0.0182 train_time:4596990ms step_avg:1605.09ms step:2875/6250 train_loss:1.9754 aux_loss:0.0179 train_time:4598025ms step_avg:1604.90ms step:2876/6250 train_loss:1.9411 aux_loss:0.0183 train_time:4599137ms step_avg:1604.72ms step:2877/6250 train_loss:1.9509 aux_loss:0.0184 train_time:4600450ms step_avg:1604.62ms step:2878/6250 train_loss:2.0019 aux_loss:0.0183 train_time:4602467ms step_avg:1604.77ms step:2879/6250 train_loss:2.0278 aux_loss:0.0180 train_time:4603942ms step_avg:1604.72ms step:2880/6250 train_loss:1.8776 aux_loss:0.0181 train_time:4605027ms step_avg:1604.54ms step:2881/6250 train_loss:1.9491 aux_loss:0.0179 train_time:4606321ms step_avg:1604.43ms step:2882/6250 train_loss:1.9039 aux_loss:0.0180 train_time:4607627ms step_avg:1604.33ms step:2883/6250 train_loss:1.9105 aux_loss:0.0180 train_time:4608827ms step_avg:1604.19ms step:2884/6250 train_loss:1.8555 aux_loss:0.0180 train_time:4610421ms step_avg:1604.18ms step:2885/6250 train_loss:2.0364 aux_loss:0.0181 train_time:4611691ms step_avg:1604.07ms step:2886/6250 train_loss:1.8977 aux_loss:0.0180 train_time:4614050ms step_avg:1604.33ms step:2887/6250 train_loss:1.9703 aux_loss:0.0188 train_time:4615084ms step_avg:1604.13ms step:2888/6250 train_loss:1.9681 aux_loss:0.0176 train_time:4616561ms step_avg:1604.09ms step:2889/6250 train_loss:2.0276 aux_loss:0.0182 train_time:4617470ms step_avg:1603.84ms step:2890/6250 train_loss:1.9024 aux_loss:0.0184 train_time:4618565ms step_avg:1603.67ms step:2891/6250 train_loss:1.9008 aux_loss:0.0180 train_time:4619768ms step_avg:1603.53ms step:2892/6250 train_loss:1.9084 aux_loss:0.0176 train_time:4620665ms step_avg:1603.28ms step:2893/6250 train_loss:1.9398 aux_loss:0.0180 train_time:4621517ms step_avg:1603.02ms step:2894/6250 train_loss:1.9807 aux_loss:0.0182 train_time:4622412ms step_avg:1602.78ms step:2895/6250 train_loss:2.0048 aux_loss:0.0180 train_time:4623257ms step_avg:1602.52ms step:2896/6250 train_loss:1.8650 aux_loss:0.0178 train_time:4624177ms step_avg:1602.28ms step:2897/6250 train_loss:2.0192 aux_loss:0.0181 train_time:4625170ms step_avg:1602.07ms step:2898/6250 train_loss:1.9889 aux_loss:0.0184 train_time:4626435ms step_avg:1601.95ms step:2899/6250 train_loss:1.9196 aux_loss:0.0183 train_time:4627543ms step_avg:1601.78ms step:2900/6250 train_loss:2.0422 aux_loss:0.0184 train_time:4628579ms step_avg:1601.58ms step:2901/6250 train_loss:1.9610 aux_loss:0.0180 train_time:4629994ms step_avg:1601.52ms step:2902/6250 train_loss:1.9857 aux_loss:0.0181 train_time:4632352ms step_avg:1601.78ms step:2903/6250 train_loss:1.8899 aux_loss:0.0180 train_time:4635709ms step_avg:1602.39ms step:2904/6250 train_loss:1.9478 aux_loss:0.0180 train_time:4636549ms step_avg:1602.12ms step:2905/6250 train_loss:1.8586 aux_loss:0.0178 train_time:4637429ms step_avg:1601.88ms step:2906/6250 train_loss:1.9337 aux_loss:0.0182 train_time:4639540ms step_avg:1602.05ms step:2907/6250 train_loss:1.9625 aux_loss:0.0179 train_time:4640399ms step_avg:1601.79ms step:2908/6250 train_loss:1.9284 aux_loss:0.0178 train_time:4641967ms step_avg:1601.78ms step:2909/6250 train_loss:1.9569 aux_loss:0.0179 train_time:4643223ms step_avg:1601.66ms step:2910/6250 train_loss:1.9232 aux_loss:0.0182 train_time:4645109ms step_avg:1601.76ms step:2911/6250 train_loss:1.9511 aux_loss:0.0180 train_time:4647574ms step_avg:1602.06ms step:2912/6250 train_loss:1.9854 aux_loss:0.0179 train_time:4648492ms step_avg:1601.82ms step:2913/6250 train_loss:1.9488 aux_loss:0.0183 train_time:4649376ms step_avg:1601.58ms step:2914/6250 train_loss:1.9720 aux_loss:0.0181 train_time:4650217ms step_avg:1601.31ms step:2915/6250 train_loss:2.0584 aux_loss:0.0181 train_time:4651079ms step_avg:1601.06ms step:2916/6250 train_loss:1.9007 aux_loss:0.0180 train_time:4651925ms step_avg:1600.80ms step:2917/6250 train_loss:1.8143 aux_loss:0.0177 train_time:4654887ms step_avg:1601.27ms step:2918/6250 train_loss:1.9034 aux_loss:0.0175 train_time:4657136ms step_avg:1601.49ms step:2919/6250 train_loss:1.8375 aux_loss:0.0180 train_time:4658204ms step_avg:1601.31ms step:2920/6250 train_loss:1.9130 aux_loss:0.0183 train_time:4660033ms step_avg:1601.39ms step:2921/6250 train_loss:1.9688 aux_loss:0.0179 train_time:4661130ms step_avg:1601.21ms step:2922/6250 train_loss:1.9516 aux_loss:0.0179 train_time:4662335ms step_avg:1601.08ms step:2923/6250 train_loss:1.8521 aux_loss:0.0182 train_time:4663326ms step_avg:1600.87ms step:2924/6250 train_loss:1.9728 aux_loss:0.0182 train_time:4664342ms step_avg:1600.67ms step:2925/6250 train_loss:1.9217 aux_loss:0.0180 train_time:4665216ms step_avg:1600.42ms step:2926/6250 train_loss:1.9587 aux_loss:0.0179 train_time:4666107ms step_avg:1600.17ms step:2927/6250 train_loss:1.8825 aux_loss:0.0179 train_time:4666960ms step_avg:1599.92ms step:2928/6250 train_loss:1.9154 aux_loss:0.0181 train_time:4667996ms step_avg:1599.72ms step:2929/6250 train_loss:1.8767 aux_loss:0.0181 train_time:4669233ms step_avg:1599.60ms step:2930/6250 train_loss:2.0347 aux_loss:0.0179 train_time:4670130ms step_avg:1599.36ms step:2931/6250 train_loss:1.8933 aux_loss:0.0177 train_time:4670985ms step_avg:1599.10ms step:2932/6250 train_loss:1.9686 aux_loss:0.0180 train_time:4671943ms step_avg:1598.89ms step:2933/6250 train_loss:1.9397 aux_loss:0.0178 train_time:4673128ms step_avg:1598.74ms step:2934/6250 train_loss:2.1054 aux_loss:0.0178 train_time:4674920ms step_avg:1598.81ms step:2935/6250 train_loss:2.0152 aux_loss:0.0177 train_time:4675894ms step_avg:1598.60ms step:2936/6250 train_loss:1.8383 aux_loss:0.0177 train_time:4676920ms step_avg:1598.40ms step:2937/6250 train_loss:2.0243 aux_loss:0.0178 train_time:4678836ms step_avg:1598.51ms step:2938/6250 train_loss:2.0091 aux_loss:0.0181 train_time:4679783ms step_avg:1598.29ms step:2939/6250 train_loss:2.0024 aux_loss:0.0183 train_time:4680665ms step_avg:1598.04ms step:2940/6250 train_loss:1.8177 aux_loss:0.0181 train_time:4681525ms step_avg:1597.79ms step:2941/6250 train_loss:1.8791 aux_loss:0.0177 train_time:4682623ms step_avg:1597.62ms step:2942/6250 train_loss:1.9454 aux_loss:0.0182 train_time:4683488ms step_avg:1597.37ms step:2943/6250 train_loss:1.9257 aux_loss:0.0180 train_time:4684545ms step_avg:1597.19ms step:2944/6250 train_loss:1.9511 aux_loss:0.0177 train_time:4685437ms step_avg:1596.95ms step:2945/6250 train_loss:2.0535 aux_loss:0.0182 train_time:4686479ms step_avg:1596.76ms step:2946/6250 train_loss:1.8114 aux_loss:0.0181 train_time:4687344ms step_avg:1596.51ms step:2947/6250 train_loss:1.9916 aux_loss:0.0182 train_time:4688443ms step_avg:1596.34ms step:2948/6250 train_loss:1.8445 aux_loss:0.0178 train_time:4690954ms step_avg:1596.65ms step:2949/6250 train_loss:2.0296 aux_loss:0.0180 train_time:4693138ms step_avg:1596.85ms step:2950/6250 train_loss:1.9202 aux_loss:0.0181 train_time:4694007ms step_avg:1596.60ms step:2951/6250 train_loss:1.9511 aux_loss:0.0178 train_time:4695205ms step_avg:1596.47ms step:2952/6250 train_loss:1.9487 aux_loss:0.0182 train_time:4696447ms step_avg:1596.34ms step:2953/6250 train_loss:2.0408 aux_loss:0.0182 train_time:4699205ms step_avg:1596.74ms step:2954/6250 train_loss:1.9137 aux_loss:0.0180 train_time:4701972ms step_avg:1597.14ms step:2955/6250 train_loss:1.9466 aux_loss:0.0181 train_time:4702835ms step_avg:1596.89ms step:2956/6250 train_loss:1.9197 aux_loss:0.0183 train_time:4703716ms step_avg:1596.65ms step:2957/6250 train_loss:1.8669 aux_loss:0.0178 train_time:4704554ms step_avg:1596.39ms step:2958/6250 train_loss:1.8200 aux_loss:0.0175 train_time:4705428ms step_avg:1596.14ms step:2959/6250 train_loss:1.9156 aux_loss:0.0178 train_time:4706269ms step_avg:1595.89ms step:2960/6250 train_loss:1.9846 aux_loss:0.0179 train_time:4707159ms step_avg:1595.65ms step:2961/6250 train_loss:1.9356 aux_loss:0.0179 train_time:4708187ms step_avg:1595.45ms step:2962/6250 train_loss:1.8961 aux_loss:0.0175 train_time:4709105ms step_avg:1595.23ms step:2963/6250 train_loss:1.9356 aux_loss:0.0178 train_time:4710417ms step_avg:1595.13ms step:2964/6250 train_loss:1.9394 aux_loss:0.0175 train_time:4713687ms step_avg:1595.70ms step:2965/6250 train_loss:1.9696 aux_loss:0.0176 train_time:4715301ms step_avg:1595.70ms step:2966/6250 train_loss:1.8842 aux_loss:0.0175 train_time:4717938ms step_avg:1596.05ms step:2967/6250 train_loss:1.9131 aux_loss:0.0177 train_time:4718840ms step_avg:1595.82ms step:2968/6250 train_loss:1.9446 aux_loss:0.0177 train_time:4719880ms step_avg:1595.63ms step:2969/6250 train_loss:1.8957 aux_loss:0.0181 train_time:4721539ms step_avg:1595.65ms step:2970/6250 train_loss:1.9134 aux_loss:0.0185 train_time:4723143ms step_avg:1595.66ms step:2971/6250 train_loss:1.8611 aux_loss:0.0177 train_time:4724415ms step_avg:1595.55ms step:2972/6250 train_loss:1.8358 aux_loss:0.0172 train_time:4725314ms step_avg:1595.31ms step:2973/6250 train_loss:1.9887 aux_loss:0.0175 train_time:4726157ms step_avg:1595.06ms step:2974/6250 train_loss:2.0050 aux_loss:0.0178 train_time:4727046ms step_avg:1594.82ms step:2975/6250 train_loss:1.9406 aux_loss:0.0173 train_time:4727911ms step_avg:1594.57ms step:2976/6250 train_loss:1.8856 aux_loss:0.0174 train_time:4728804ms step_avg:1594.34ms step:2977/6250 train_loss:1.8956 aux_loss:0.0178 train_time:4729689ms step_avg:1594.10ms step:2978/6250 train_loss:1.8655 aux_loss:0.0176 train_time:4730575ms step_avg:1593.86ms step:2979/6250 train_loss:1.9669 aux_loss:0.0181 train_time:4731431ms step_avg:1593.61ms step:2980/6250 train_loss:1.9943 aux_loss:0.0178 train_time:4732339ms step_avg:1593.38ms step:2981/6250 train_loss:1.9423 aux_loss:0.0180 train_time:4733200ms step_avg:1593.13ms step:2982/6250 train_loss:1.9635 aux_loss:0.0177 train_time:4734116ms step_avg:1592.91ms step:2983/6250 train_loss:2.0303 aux_loss:0.0178 train_time:4734972ms step_avg:1592.66ms step:2984/6250 train_loss:1.9461 aux_loss:0.0178 train_time:4735883ms step_avg:1592.43ms step:2985/6250 train_loss:1.8848 aux_loss:0.0182 train_time:4736755ms step_avg:1592.19ms step:2986/6250 train_loss:2.0776 aux_loss:0.0180 train_time:4738040ms step_avg:1592.08ms step:2987/6250 train_loss:1.9670 aux_loss:0.0178 train_time:4740293ms step_avg:1592.31ms step:2988/6250 train_loss:1.9245 aux_loss:0.0174 train_time:4741489ms step_avg:1592.17ms step:2989/6250 train_loss:1.9357 aux_loss:0.0175 train_time:4742346ms step_avg:1591.93ms step:2990/6250 train_loss:1.9424 aux_loss:0.0182 train_time:4743194ms step_avg:1591.68ms step:2991/6250 train_loss:1.9646 aux_loss:0.0181 train_time:4744092ms step_avg:1591.44ms step:2992/6250 train_loss:2.0410 aux_loss:0.0176 train_time:4745104ms step_avg:1591.25ms step:2993/6250 train_loss:1.8794 aux_loss:0.0175 train_time:4746315ms step_avg:1591.12ms step:2994/6250 train_loss:2.0224 aux_loss:0.0183 train_time:4747232ms step_avg:1590.90ms step:2995/6250 train_loss:1.9981 aux_loss:0.0180 train_time:4748098ms step_avg:1590.65ms step:2996/6250 train_loss:2.0311 aux_loss:0.0177 train_time:4748967ms step_avg:1590.41ms step:2997/6250 train_loss:1.9607 aux_loss:0.0177 train_time:4749849ms step_avg:1590.17ms step:2998/6250 train_loss:1.9366 aux_loss:0.0177 train_time:4750704ms step_avg:1589.93ms step:2999/6250 train_loss:1.9682 aux_loss:0.0175 train_time:4751595ms step_avg:1589.69ms step:3000/6250 train_loss:1.9281 aux_loss:0.0176 train_time:4752455ms step_avg:1589.45ms step:3001/6250 train_loss:2.0035 aux_loss:0.0177 train_time:4753318ms step_avg:1589.21ms step:3002/6250 train_loss:1.9141 aux_loss:0.0176 train_time:4754170ms step_avg:1588.96ms step:3003/6250 train_loss:1.8788 aux_loss:0.0173 train_time:4755042ms step_avg:1588.72ms step:3004/6250 train_loss:1.9237 aux_loss:0.0181 train_time:4755898ms step_avg:1588.48ms step:3005/6250 train_loss:1.8443 aux_loss:0.0179 train_time:4756796ms step_avg:1588.25ms step:3006/6250 train_loss:1.8306 aux_loss:0.0170 train_time:4757652ms step_avg:1588.00ms step:3007/6250 train_loss:1.9972 aux_loss:0.0175 train_time:4758512ms step_avg:1587.76ms step:3008/6250 train_loss:1.9240 aux_loss:0.0176 train_time:4760411ms step_avg:1587.86ms step:3009/6250 train_loss:1.8832 aux_loss:0.0175 train_time:4761702ms step_avg:1587.76ms step:3010/6250 train_loss:1.9085 aux_loss:0.0172 train_time:4762584ms step_avg:1587.53ms step:3011/6250 train_loss:1.9264 aux_loss:0.0174 train_time:4763757ms step_avg:1587.39ms step:3012/6250 train_loss:1.9834 aux_loss:0.0177 train_time:4765759ms step_avg:1587.53ms step:3013/6250 train_loss:1.8615 aux_loss:0.0174 train_time:4767520ms step_avg:1587.59ms step:3014/6250 train_loss:1.9895 aux_loss:0.0173 train_time:4768450ms step_avg:1587.37ms step:3015/6250 train_loss:1.9991 aux_loss:0.0173 train_time:4769362ms step_avg:1587.14ms step:3016/6250 train_loss:1.9925 aux_loss:0.0175 train_time:4770211ms step_avg:1586.90ms step:3017/6250 train_loss:1.9084 aux_loss:0.0175 train_time:4771180ms step_avg:1586.69ms step:3018/6250 train_loss:2.0260 aux_loss:0.0176 train_time:4772889ms step_avg:1586.73ms step:3019/6250 train_loss:1.8942 aux_loss:0.0177 train_time:4774530ms step_avg:1586.75ms step:3020/6250 train_loss:1.9272 aux_loss:0.0175 train_time:4779668ms step_avg:1587.93ms step:3021/6250 train_loss:1.9739 aux_loss:0.0175 train_time:4781799ms step_avg:1588.11ms step:3022/6250 train_loss:1.9192 aux_loss:0.0179 train_time:4784023ms step_avg:1588.32ms step:3023/6250 train_loss:1.9607 aux_loss:0.0177 train_time:4784883ms step_avg:1588.08ms step:3024/6250 train_loss:1.8759 aux_loss:0.0169 train_time:4785765ms step_avg:1587.84ms step:3025/6250 train_loss:1.9536 aux_loss:0.0171 train_time:4786617ms step_avg:1587.60ms step:3026/6250 train_loss:1.9482 aux_loss:0.0174 train_time:4787697ms step_avg:1587.43ms step:3027/6250 train_loss:1.9754 aux_loss:0.0174 train_time:4789225ms step_avg:1587.41ms step:3028/6250 train_loss:1.8508 aux_loss:0.0173 train_time:4791946ms step_avg:1587.79ms step:3029/6250 train_loss:1.8689 aux_loss:0.0176 train_time:4792805ms step_avg:1587.55ms step:3030/6250 train_loss:1.8685 aux_loss:0.0177 train_time:4793707ms step_avg:1587.32ms step:3031/6250 train_loss:1.9842 aux_loss:0.0175 train_time:4794561ms step_avg:1587.08ms step:3032/6250 train_loss:1.9587 aux_loss:0.0174 train_time:4795439ms step_avg:1586.84ms step:3033/6250 train_loss:1.8424 aux_loss:0.0175 train_time:4796296ms step_avg:1586.60ms step:3034/6250 train_loss:1.9236 aux_loss:0.0176 train_time:4797153ms step_avg:1586.36ms step:3035/6250 train_loss:1.9184 aux_loss:0.0176 train_time:4798013ms step_avg:1586.12ms step:3036/6250 train_loss:1.9373 aux_loss:0.0175 train_time:4798895ms step_avg:1585.89ms step:3037/6250 train_loss:1.8653 aux_loss:0.0178 train_time:4799758ms step_avg:1585.65ms step:3038/6250 train_loss:2.0643 aux_loss:0.0175 train_time:4800648ms step_avg:1585.42ms step:3039/6250 train_loss:1.9431 aux_loss:0.0171 train_time:4801498ms step_avg:1585.18ms step:3040/6250 train_loss:1.9181 aux_loss:0.0174 train_time:4802378ms step_avg:1584.94ms step:3041/6250 train_loss:1.9235 aux_loss:0.0175 train_time:4803222ms step_avg:1584.70ms step:3042/6250 train_loss:1.9940 aux_loss:0.0172 train_time:4804112ms step_avg:1584.47ms step:3043/6250 train_loss:1.9176 aux_loss:0.0175 train_time:4804969ms step_avg:1584.23ms step:3044/6250 train_loss:1.9642 aux_loss:0.0174 train_time:4805827ms step_avg:1583.99ms step:3045/6250 train_loss:1.9296 aux_loss:0.0175 train_time:4807016ms step_avg:1583.86ms step:3046/6250 train_loss:1.8944 aux_loss:0.0176 train_time:4808462ms step_avg:1583.81ms step:3047/6250 train_loss:2.0189 aux_loss:0.0176 train_time:4809537ms step_avg:1583.65ms step:3048/6250 train_loss:1.9145 aux_loss:0.0174 train_time:4810646ms step_avg:1583.49ms step:3049/6250 train_loss:1.9669 aux_loss:0.0175 train_time:4811709ms step_avg:1583.32ms step:3050/6250 train_loss:1.8829 aux_loss:0.0174 train_time:4813106ms step_avg:1583.26ms step:3051/6250 train_loss:1.9974 aux_loss:0.0174 train_time:4814464ms step_avg:1583.18ms step:3052/6250 train_loss:1.9372 aux_loss:0.0174 train_time:4815544ms step_avg:1583.02ms step:3053/6250 train_loss:1.8922 aux_loss:0.0173 train_time:4816638ms step_avg:1582.86ms step:3054/6250 train_loss:1.8678 aux_loss:0.0168 train_time:4817895ms step_avg:1582.75ms step:3055/6250 train_loss:1.8888 aux_loss:0.0172 train_time:4818763ms step_avg:1582.52ms step:3056/6250 train_loss:1.8277 aux_loss:0.0173 train_time:4820677ms step_avg:1582.63ms step:3057/6250 train_loss:1.9541 aux_loss:0.0174 train_time:4821528ms step_avg:1582.39ms step:3058/6250 train_loss:1.8616 aux_loss:0.0170 train_time:4822898ms step_avg:1582.32ms step:3059/6250 train_loss:1.9415 aux_loss:0.0175 train_time:4824789ms step_avg:1582.42ms step:3060/6250 train_loss:1.9271 aux_loss:0.0176 train_time:4825687ms step_avg:1582.19ms step:3061/6250 train_loss:1.9129 aux_loss:0.0174 train_time:4826565ms step_avg:1581.96ms step:3062/6250 train_loss:1.8924 aux_loss:0.0172 train_time:4827412ms step_avg:1581.72ms step:3063/6250 train_loss:1.9279 aux_loss:0.0174 train_time:4828291ms step_avg:1581.49ms step:3064/6250 train_loss:1.9395 aux_loss:0.0174 train_time:4829137ms step_avg:1581.25ms step:3065/6250 train_loss:1.9056 aux_loss:0.0174 train_time:4830018ms step_avg:1581.02ms step:3066/6250 train_loss:1.9272 aux_loss:0.0174 train_time:4830881ms step_avg:1580.79ms step:3067/6250 train_loss:1.9365 aux_loss:0.0169 train_time:4831753ms step_avg:1580.55ms step:3068/6250 train_loss:1.9287 aux_loss:0.0171 train_time:4832615ms step_avg:1580.32ms step:3069/6250 train_loss:1.8668 aux_loss:0.0174 train_time:4833478ms step_avg:1580.08ms step:3070/6250 train_loss:1.9697 aux_loss:0.0174 train_time:4834337ms step_avg:1579.85ms step:3071/6250 train_loss:1.9054 aux_loss:0.0172 train_time:4835835ms step_avg:1579.82ms step:3072/6250 train_loss:1.8595 aux_loss:0.0173 train_time:4837261ms step_avg:1579.77ms step:3073/6250 train_loss:1.9698 aux_loss:0.0176 train_time:4839409ms step_avg:1579.96ms step:3074/6250 train_loss:1.9285 aux_loss:0.0177 train_time:4840694ms step_avg:1579.86ms step:3075/6250 train_loss:1.9047 aux_loss:0.0171 train_time:4841586ms step_avg:1579.64ms step:3076/6250 train_loss:1.9090 aux_loss:0.0167 train_time:4842442ms step_avg:1579.40ms step:3077/6250 train_loss:1.9186 aux_loss:0.0169 train_time:4843323ms step_avg:1579.17ms step:3078/6250 train_loss:1.8408 aux_loss:0.0168 train_time:4844187ms step_avg:1578.94ms step:3079/6250 train_loss:1.9182 aux_loss:0.0170 train_time:4845070ms step_avg:1578.71ms step:3080/6250 train_loss:1.9479 aux_loss:0.0175 train_time:4845931ms step_avg:1578.48ms step:3081/6250 train_loss:1.8445 aux_loss:0.0170 train_time:4847048ms step_avg:1578.33ms step:3082/6250 train_loss:2.0179 aux_loss:0.0175 train_time:4848689ms step_avg:1578.35ms step:3083/6250 train_loss:1.9092 aux_loss:0.0170 train_time:4849737ms step_avg:1578.18ms step:3084/6250 train_loss:1.9817 aux_loss:0.0169 train_time:4850856ms step_avg:1578.03ms step:3085/6250 train_loss:1.8975 aux_loss:0.0171 train_time:4852684ms step_avg:1578.11ms step:3086/6250 train_loss:1.8876 aux_loss:0.0174 train_time:4855892ms step_avg:1578.64ms step:3087/6250 train_loss:1.8737 aux_loss:0.0173 train_time:4857827ms step_avg:1578.75ms step:3088/6250 train_loss:1.9322 aux_loss:0.0173 train_time:4859032ms step_avg:1578.63ms step:3089/6250 train_loss:1.9557 aux_loss:0.0169 train_time:4859892ms step_avg:1578.40ms step:3090/6250 train_loss:1.9437 aux_loss:0.0171 train_time:4860766ms step_avg:1578.17ms step:3091/6250 train_loss:1.8485 aux_loss:0.0170 train_time:4861666ms step_avg:1577.95ms step:3092/6250 train_loss:1.9641 aux_loss:0.0168 train_time:4862510ms step_avg:1577.71ms step:3093/6250 train_loss:1.9300 aux_loss:0.0172 train_time:4863437ms step_avg:1577.50ms step:3094/6250 train_loss:1.9334 aux_loss:0.0172 train_time:4864356ms step_avg:1577.29ms step:3095/6250 train_loss:1.9102 aux_loss:0.0171 train_time:4865965ms step_avg:1577.30ms step:3096/6250 train_loss:1.9285 aux_loss:0.0174 train_time:4866822ms step_avg:1577.06ms step:3097/6250 train_loss:1.9294 aux_loss:0.0172 train_time:4867668ms step_avg:1576.83ms step:3098/6250 train_loss:1.8135 aux_loss:0.0171 train_time:4868512ms step_avg:1576.59ms step:3099/6250 train_loss:1.9364 aux_loss:0.0170 train_time:4869381ms step_avg:1576.36ms step:3100/6250 train_loss:1.8760 aux_loss:0.0166 train_time:4870232ms step_avg:1576.13ms step:3101/6250 train_loss:1.9751 aux_loss:0.0170 train_time:4871521ms step_avg:1576.03ms step:3102/6250 train_loss:1.9124 aux_loss:0.0171 train_time:4872587ms step_avg:1575.87ms step:3103/6250 train_loss:1.8656 aux_loss:0.0165 train_time:4873491ms step_avg:1575.65ms step:3104/6250 train_loss:1.9589 aux_loss:0.0163 train_time:4874673ms step_avg:1575.52ms step:3105/6250 train_loss:1.9317 aux_loss:0.0171 train_time:4875655ms step_avg:1575.33ms step:3106/6250 train_loss:1.9371 aux_loss:0.0170 train_time:4876539ms step_avg:1575.11ms step:3107/6250 train_loss:1.9797 aux_loss:0.0168 train_time:4877406ms step_avg:1574.88ms step:3108/6250 train_loss:1.9353 aux_loss:0.0171 train_time:4878249ms step_avg:1574.64ms step:3109/6250 train_loss:1.8899 aux_loss:0.0171 train_time:4879115ms step_avg:1574.42ms step:3110/6250 train_loss:1.9306 aux_loss:0.0173 train_time:4879982ms step_avg:1574.19ms step:3111/6250 train_loss:1.9410 aux_loss:0.0170 train_time:4881383ms step_avg:1574.13ms step:3112/6250 train_loss:1.9064 aux_loss:0.0167 train_time:4882259ms step_avg:1573.91ms step:3113/6250 train_loss:1.9573 aux_loss:0.0170 train_time:4883116ms step_avg:1573.68ms step:3114/6250 train_loss:1.8575 aux_loss:0.0167 train_time:4884006ms step_avg:1573.46ms step:3115/6250 train_loss:1.7996 aux_loss:0.0168 train_time:4885295ms step_avg:1573.36ms step:3116/6250 train_loss:1.9474 aux_loss:0.0169 train_time:4886468ms step_avg:1573.24ms step:3117/6250 train_loss:2.0250 aux_loss:0.0168 train_time:4887974ms step_avg:1573.21ms step:3118/6250 train_loss:1.7999 aux_loss:0.0168 train_time:4888839ms step_avg:1572.99ms step:3119/6250 train_loss:1.9593 aux_loss:0.0172 train_time:4889699ms step_avg:1572.76ms step:3120/6250 train_loss:1.9540 aux_loss:0.0172 train_time:4890566ms step_avg:1572.53ms step:3121/6250 train_loss:1.9446 aux_loss:0.0169 train_time:4891438ms step_avg:1572.30ms step:3122/6250 train_loss:1.9150 aux_loss:0.0175 train_time:4892297ms step_avg:1572.07ms step:3123/6250 train_loss:1.9177 aux_loss:0.0171 train_time:4893322ms step_avg:1571.90ms step:3124/6250 train_loss:1.9385 aux_loss:0.0169 train_time:4894638ms step_avg:1571.82ms step:3125/6250 train_loss:1.9803 aux_loss:0.0172 train_time:4895525ms step_avg:1571.60ms step:3126/6250 train_loss:1.9226 aux_loss:0.0170 train_time:4896384ms step_avg:1571.37ms step:3127/6250 train_loss:1.9332 aux_loss:0.0170 train_time:4897235ms step_avg:1571.14ms step:3128/6250 train_loss:1.8959 aux_loss:0.0172 train_time:4898107ms step_avg:1570.91ms step:3129/6250 train_loss:1.8785 aux_loss:0.0172 train_time:4898983ms step_avg:1570.69ms step:3130/6250 train_loss:1.8608 aux_loss:0.0172 train_time:4901403ms step_avg:1570.96ms step:3131/6250 train_loss:1.9324 aux_loss:0.0175 train_time:4902481ms step_avg:1570.80ms step:3132/6250 train_loss:2.0177 aux_loss:0.0177 train_time:4903585ms step_avg:1570.66ms step:3133/6250 train_loss:1.9852 aux_loss:0.0167 train_time:4905091ms step_avg:1570.63ms step:3134/6250 train_loss:1.8109 aux_loss:0.0168 train_time:4906102ms step_avg:1570.46ms step:3135/6250 train_loss:1.9039 aux_loss:0.0175 train_time:4907629ms step_avg:1570.44ms step:3136/6250 train_loss:1.8919 aux_loss:0.0172 train_time:4908482ms step_avg:1570.21ms step:3137/6250 train_loss:1.9227 aux_loss:0.0165 train_time:4910395ms step_avg:1570.32ms step:3138/6250 train_loss:1.9460 aux_loss:0.0170 train_time:4911914ms step_avg:1570.30ms step:3139/6250 train_loss:1.9721 aux_loss:0.0170 train_time:4914015ms step_avg:1570.47ms step:3140/6250 train_loss:1.9252 aux_loss:0.0167 train_time:4916327ms step_avg:1570.71ms step:3141/6250 train_loss:1.9395 aux_loss:0.0169 train_time:4917765ms step_avg:1570.67ms step:3142/6250 train_loss:1.8465 aux_loss:0.0169 train_time:4918768ms step_avg:1570.49ms step:3143/6250 train_loss:1.9040 aux_loss:0.0167 train_time:4920185ms step_avg:1570.44ms step:3144/6250 train_loss:1.9099 aux_loss:0.0169 train_time:4921494ms step_avg:1570.36ms step:3145/6250 train_loss:1.9278 aux_loss:0.0169 train_time:4922349ms step_avg:1570.13ms step:3146/6250 train_loss:2.0209 aux_loss:0.0171 train_time:4923386ms step_avg:1569.96ms step:3147/6250 train_loss:1.9015 aux_loss:0.0166 train_time:4924236ms step_avg:1569.73ms step:3148/6250 train_loss:1.8807 aux_loss:0.0168 train_time:4925109ms step_avg:1569.51ms step:3149/6250 train_loss:1.9837 aux_loss:0.0170 train_time:4925945ms step_avg:1569.27ms step:3150/6250 train_loss:1.9202 aux_loss:0.0172 train_time:4926830ms step_avg:1569.05ms step:3151/6250 train_loss:1.9767 aux_loss:0.0170 train_time:4928523ms step_avg:1569.09ms step:3152/6250 train_loss:1.8822 aux_loss:0.0168 train_time:4929402ms step_avg:1568.87ms step:3153/6250 train_loss:1.9110 aux_loss:0.0168 train_time:4930485ms step_avg:1568.72ms step:3154/6250 train_loss:2.0342 aux_loss:0.0168 train_time:4934872ms step_avg:1569.62ms step:3155/6250 train_loss:1.8994 aux_loss:0.0166 train_time:4936155ms step_avg:1569.52ms step:3156/6250 train_loss:1.8782 aux_loss:0.0167 train_time:4940502ms step_avg:1570.41ms step:3157/6250 train_loss:1.8807 aux_loss:0.0167 train_time:4942804ms step_avg:1570.64ms step:3158/6250 train_loss:2.0216 aux_loss:0.0171 train_time:4943782ms step_avg:1570.45ms step:3159/6250 train_loss:1.9339 aux_loss:0.0165 train_time:4944747ms step_avg:1570.26ms step:3160/6250 train_loss:1.9313 aux_loss:0.0162 train_time:4946109ms step_avg:1570.19ms step:3161/6250 train_loss:1.9449 aux_loss:0.0168 train_time:4947752ms step_avg:1570.22ms step:3162/6250 train_loss:1.9495 aux_loss:0.0169 train_time:4949342ms step_avg:1570.22ms step:3163/6250 train_loss:1.8612 aux_loss:0.0166 train_time:4951063ms step_avg:1570.27ms step:3164/6250 train_loss:1.9268 aux_loss:0.0168 train_time:4952339ms step_avg:1570.18ms step:3165/6250 train_loss:1.9271 aux_loss:0.0166 train_time:4953581ms step_avg:1570.07ms step:3166/6250 train_loss:1.8935 aux_loss:0.0171 train_time:4954905ms step_avg:1570.00ms step:3167/6250 train_loss:1.9298 aux_loss:0.0171 train_time:4956385ms step_avg:1569.97ms step:3168/6250 train_loss:1.9522 aux_loss:0.0168 train_time:4957832ms step_avg:1569.93ms step:3169/6250 train_loss:2.0225 aux_loss:0.0167 train_time:4959312ms step_avg:1569.90ms step:3170/6250 train_loss:1.9850 aux_loss:0.0168 train_time:4960692ms step_avg:1569.84ms step:3171/6250 train_loss:1.8714 aux_loss:0.0170 train_time:4962232ms step_avg:1569.83ms step:3172/6250 train_loss:1.9347 aux_loss:0.0170 train_time:4963837ms step_avg:1569.84ms step:3173/6250 train_loss:1.8958 aux_loss:0.0166 train_time:4965351ms step_avg:1569.82ms step:3174/6250 train_loss:1.9233 aux_loss:0.0169 train_time:4966974ms step_avg:1569.84ms step:3175/6250 train_loss:1.9419 aux_loss:0.0167 train_time:4968349ms step_avg:1569.78ms step:3176/6250 train_loss:1.9554 aux_loss:0.0171 train_time:4969590ms step_avg:1569.67ms step:3177/6250 train_loss:1.8426 aux_loss:0.0171 train_time:4970836ms step_avg:1569.57ms step:3178/6250 train_loss:1.9287 aux_loss:0.0173 train_time:4972110ms step_avg:1569.48ms step:3179/6250 train_loss:1.9209 aux_loss:0.0171 train_time:4973371ms step_avg:1569.38ms step:3180/6250 train_loss:1.9024 aux_loss:0.0172 train_time:4974606ms step_avg:1569.28ms step:3181/6250 train_loss:1.9237 aux_loss:0.0169 train_time:4975860ms step_avg:1569.18ms step:3182/6250 train_loss:1.8566 aux_loss:0.0164 train_time:4977433ms step_avg:1569.18ms step:3183/6250 train_loss:1.8502 aux_loss:0.0165 train_time:4978717ms step_avg:1569.09ms step:3184/6250 train_loss:1.9495 aux_loss:0.0167 train_time:4979914ms step_avg:1568.97ms step:3185/6250 train_loss:1.9097 aux_loss:0.0165 train_time:4981209ms step_avg:1568.88ms step:3186/6250 train_loss:1.9232 aux_loss:0.0165 train_time:4982535ms step_avg:1568.81ms step:3187/6250 train_loss:1.9691 aux_loss:0.0169 train_time:4983797ms step_avg:1568.71ms step:3188/6250 train_loss:1.8780 aux_loss:0.0168 train_time:4985227ms step_avg:1568.67ms step:3189/6250 train_loss:1.9613 aux_loss:0.0172 train_time:4987049ms step_avg:1568.75ms step:3190/6250 train_loss:1.9985 aux_loss:0.0172 train_time:4988856ms step_avg:1568.82ms step:3191/6250 train_loss:1.8766 aux_loss:0.0168 train_time:4990106ms step_avg:1568.72ms step:3192/6250 train_loss:1.9292 aux_loss:0.0168 train_time:4991759ms step_avg:1568.75ms step:3193/6250 train_loss:1.9018 aux_loss:0.0174 train_time:4993101ms step_avg:1568.68ms step:3194/6250 train_loss:1.8653 aux_loss:0.0166 train_time:4994429ms step_avg:1568.60ms step:3195/6250 train_loss:1.9485 aux_loss:0.0170 train_time:4995826ms step_avg:1568.55ms step:3196/6250 train_loss:1.8834 aux_loss:0.0167 train_time:4997368ms step_avg:1568.54ms step:3197/6250 train_loss:2.0323 aux_loss:0.0168 train_time:4998584ms step_avg:1568.43ms step:3198/6250 train_loss:1.8645 aux_loss:0.0169 train_time:4999837ms step_avg:1568.33ms step:3199/6250 train_loss:1.9013 aux_loss:0.0168 train_time:5001068ms step_avg:1568.22ms step:3200/6250 train_loss:1.8334 aux_loss:0.0163 train_time:5003707ms step_avg:1568.56ms step:3201/6250 train_loss:1.9516 aux_loss:0.0168 train_time:5005112ms step_avg:1568.51ms step:3202/6250 train_loss:1.8548 aux_loss:0.0166 train_time:5006391ms step_avg:1568.42ms step:3203/6250 train_loss:1.8960 aux_loss:0.0165 train_time:5015536ms step_avg:1570.79ms step:3204/6250 train_loss:1.9441 aux_loss:0.0165 train_time:5018056ms step_avg:1571.09ms step:3205/6250 train_loss:2.0625 aux_loss:0.0169 train_time:5020455ms step_avg:1571.35ms step:3206/6250 train_loss:1.8300 aux_loss:0.0163 train_time:5026142ms step_avg:1572.64ms step:3207/6250 train_loss:1.9261 aux_loss:0.0168 train_time:5033489ms step_avg:1574.44ms step:3208/6250 train_loss:1.8742 aux_loss:0.0167 train_time:5035450ms step_avg:1574.56ms step:3209/6250 train_loss:1.8912 aux_loss:0.0167 train_time:5036589ms step_avg:1574.43ms step:3210/6250 train_loss:1.9808 aux_loss:0.0167 train_time:5037448ms step_avg:1574.20ms step:3211/6250 train_loss:1.8851 aux_loss:0.0162 train_time:5039311ms step_avg:1574.29ms step:3212/6250 train_loss:1.9449 aux_loss:0.0170 train_time:5040288ms step_avg:1574.11ms step:3213/6250 train_loss:1.8736 aux_loss:0.0170 train_time:5050410ms step_avg:1576.77ms step:3214/6250 train_loss:1.9470 aux_loss:0.0164 train_time:5051800ms step_avg:1576.72ms step:3215/6250 train_loss:1.9952 aux_loss:0.0168 train_time:5052843ms step_avg:1576.55ms step:3216/6250 train_loss:1.9499 aux_loss:0.0170 train_time:5053985ms step_avg:1576.41ms step:3217/6250 train_loss:1.8581 aux_loss:0.0164 train_time:5054911ms step_avg:1576.21ms step:3218/6250 train_loss:2.0982 aux_loss:0.0166 train_time:5055931ms step_avg:1576.04ms step:3219/6250 train_loss:1.8899 aux_loss:0.0170 train_time:5057048ms step_avg:1575.90ms step:3220/6250 train_loss:1.9377 aux_loss:0.0166 train_time:5058178ms step_avg:1575.76ms step:3221/6250 train_loss:1.9239 aux_loss:0.0168 train_time:5059075ms step_avg:1575.54ms step:3222/6250 train_loss:1.9128 aux_loss:0.0165 train_time:5060179ms step_avg:1575.40ms step:3223/6250 train_loss:1.9138 aux_loss:0.0165 train_time:5061892ms step_avg:1575.44ms step:3224/6250 train_loss:1.9713 aux_loss:0.0166 train_time:5063214ms step_avg:1575.36ms step:3225/6250 train_loss:1.9629 aux_loss:0.0170 train_time:5065312ms step_avg:1575.52ms step:3226/6250 train_loss:1.9888 aux_loss:0.0171 train_time:5070672ms step_avg:1576.70ms step:3227/6250 train_loss:1.8389 aux_loss:0.0169 train_time:5072131ms step_avg:1576.66ms step:3228/6250 train_loss:1.9207 aux_loss:0.0166 train_time:5073018ms step_avg:1576.45ms step:3229/6250 train_loss:1.8736 aux_loss:0.0166 train_time:5074066ms step_avg:1576.29ms step:3230/6250 train_loss:1.9673 aux_loss:0.0168 train_time:5075468ms step_avg:1576.23ms step:3231/6250 train_loss:1.8141 aux_loss:0.0162 train_time:5076322ms step_avg:1576.01ms step:3232/6250 train_loss:1.8207 aux_loss:0.0159 train_time:5077183ms step_avg:1575.79ms step:3233/6250 train_loss:1.9006 aux_loss:0.0170 train_time:5078044ms step_avg:1575.56ms step:3234/6250 train_loss:1.8917 aux_loss:0.0166 train_time:5078937ms step_avg:1575.35ms step:3235/6250 train_loss:1.9068 aux_loss:0.0166 train_time:5079801ms step_avg:1575.13ms step:3236/6250 train_loss:1.9475 aux_loss:0.0166 train_time:5081552ms step_avg:1575.19ms step:3237/6250 train_loss:1.9357 aux_loss:0.0165 train_time:5083100ms step_avg:1575.18ms step:3238/6250 train_loss:1.8690 aux_loss:0.0161 train_time:5084170ms step_avg:1575.02ms step:3239/6250 train_loss:1.9200 aux_loss:0.0162 train_time:5085492ms step_avg:1574.94ms step:3240/6250 train_loss:1.8679 aux_loss:0.0165 train_time:5086388ms step_avg:1574.73ms step:3241/6250 train_loss:1.8933 aux_loss:0.0165 train_time:5087246ms step_avg:1574.51ms step:3242/6250 train_loss:1.9327 aux_loss:0.0166 train_time:5088088ms step_avg:1574.28ms step:3243/6250 train_loss:1.9297 aux_loss:0.0166 train_time:5088955ms step_avg:1574.07ms step:3244/6250 train_loss:1.9319 aux_loss:0.0166 train_time:5089808ms step_avg:1573.84ms step:3245/6250 train_loss:1.8662 aux_loss:0.0166 train_time:5090899ms step_avg:1573.69ms step:3246/6250 train_loss:1.9980 aux_loss:0.0169 train_time:5091962ms step_avg:1573.54ms step:3247/6250 train_loss:1.9371 aux_loss:0.0163 train_time:5093047ms step_avg:1573.38ms step:3248/6250 train_loss:1.8381 aux_loss:0.0163 train_time:5093923ms step_avg:1573.17ms step:3249/6250 train_loss:1.8293 aux_loss:0.0165 train_time:5094805ms step_avg:1572.96ms step:3250/6250 train_loss:1.9001 aux_loss:0.0166 train_time:5095675ms step_avg:1572.74ms step:3251/6250 train_loss:1.9687 aux_loss:0.0164 train_time:5096541ms step_avg:1572.52ms step:3252/6250 train_loss:1.8996 aux_loss:0.0164 train_time:5097399ms step_avg:1572.30ms step:3253/6250 train_loss:1.9039 aux_loss:0.0162 train_time:5098243ms step_avg:1572.08ms step:3254/6250 train_loss:1.8254 aux_loss:0.0165 train_time:5099096ms step_avg:1571.85ms step:3255/6250 train_loss:1.8763 aux_loss:0.0166 train_time:5099972ms step_avg:1571.64ms step:3256/6250 train_loss:1.8812 aux_loss:0.0164 train_time:5101285ms step_avg:1571.56ms step:3257/6250 train_loss:1.8679 aux_loss:0.0163 train_time:5102610ms step_avg:1571.48ms step:3258/6250 train_loss:1.9632 aux_loss:0.0165 train_time:5104078ms step_avg:1571.45ms step:3259/6250 train_loss:1.8792 aux_loss:0.0163 train_time:5105258ms step_avg:1571.33ms step:3260/6250 train_loss:1.9739 aux_loss:0.0166 train_time:5106114ms step_avg:1571.11ms step:3261/6250 train_loss:2.0366 aux_loss:0.0168 train_time:5106985ms step_avg:1570.90ms step:3262/6250 train_loss:1.9621 aux_loss:0.0167 train_time:5107855ms step_avg:1570.68ms step:3263/6250 train_loss:1.9358 aux_loss:0.0163 train_time:5108721ms step_avg:1570.46ms step:3264/6250 train_loss:1.8854 aux_loss:0.0164 train_time:5109585ms step_avg:1570.25ms step:3265/6250 train_loss:1.9767 aux_loss:0.0171 train_time:5110454ms step_avg:1570.03ms step:3266/6250 train_loss:1.9050 aux_loss:0.0170 train_time:5111315ms step_avg:1569.81ms step:3267/6250 train_loss:1.9122 aux_loss:0.0167 train_time:5112898ms step_avg:1569.82ms step:3268/6250 train_loss:1.9393 aux_loss:0.0165 train_time:5115836ms step_avg:1570.24ms step:3269/6250 train_loss:1.8898 aux_loss:0.0161 train_time:5117293ms step_avg:1570.20ms step:3270/6250 train_loss:1.9053 aux_loss:0.0163 train_time:5118230ms step_avg:1570.01ms step:3271/6250 train_loss:1.9550 aux_loss:0.0166 train_time:5120390ms step_avg:1570.19ms step:3272/6250 train_loss:1.8996 aux_loss:0.0164 train_time:5121501ms step_avg:1570.05ms step:3273/6250 train_loss:1.8864 aux_loss:0.0164 train_time:5122367ms step_avg:1569.83ms step:3274/6250 train_loss:1.9038 aux_loss:0.0168 train_time:5123968ms step_avg:1569.84ms step:3275/6250 train_loss:1.8706 aux_loss:0.0164 train_time:5124835ms step_avg:1569.63ms step:3276/6250 train_loss:1.9881 aux_loss:0.0167 train_time:5125707ms step_avg:1569.41ms step:3277/6250 train_loss:1.8415 aux_loss:0.0163 train_time:5126564ms step_avg:1569.20ms step:3278/6250 train_loss:1.9136 aux_loss:0.0163 train_time:5127450ms step_avg:1568.99ms step:3279/6250 train_loss:1.9135 aux_loss:0.0160 train_time:5128285ms step_avg:1568.76ms step:3280/6250 train_loss:2.0740 aux_loss:0.0163 train_time:5130631ms step_avg:1569.00ms step:3281/6250 train_loss:1.9698 aux_loss:0.0167 train_time:5132105ms step_avg:1568.97ms step:3282/6250 train_loss:1.8252 aux_loss:0.0162 train_time:5132963ms step_avg:1568.75ms step:3283/6250 train_loss:1.9086 aux_loss:0.0160 train_time:5133815ms step_avg:1568.53ms step:3284/6250 train_loss:1.8365 aux_loss:0.0166 train_time:5134710ms step_avg:1568.33ms step:3285/6250 train_loss:1.9942 aux_loss:0.0169 train_time:5135571ms step_avg:1568.11ms step:3286/6250 train_loss:1.9847 aux_loss:0.0164 train_time:5136448ms step_avg:1567.90ms step:3287/6250 train_loss:1.8841 aux_loss:0.0161 train_time:5137780ms step_avg:1567.83ms step:3288/6250 train_loss:1.9249 aux_loss:0.0165 train_time:5139781ms step_avg:1567.96ms step:3289/6250 train_loss:1.8355 aux_loss:0.0163 train_time:5140983ms step_avg:1567.85ms step:3290/6250 train_loss:1.8937 aux_loss:0.0163 train_time:5142474ms step_avg:1567.83ms step:3291/6250 train_loss:1.9310 aux_loss:0.0164 train_time:5143318ms step_avg:1567.61ms step:3292/6250 train_loss:1.9697 aux_loss:0.0165 train_time:5144216ms step_avg:1567.40ms step:3293/6250 train_loss:1.9103 aux_loss:0.0165 train_time:5145153ms step_avg:1567.21ms step:3294/6250 train_loss:1.8469 aux_loss:0.0163 train_time:5146370ms step_avg:1567.10ms step:3295/6250 train_loss:1.8616 aux_loss:0.0166 train_time:5147214ms step_avg:1566.88ms step:3296/6250 train_loss:1.8388 aux_loss:0.0166 train_time:5148413ms step_avg:1566.77ms step:3297/6250 train_loss:1.8958 aux_loss:0.0165 train_time:5149651ms step_avg:1566.67ms step:3298/6250 train_loss:1.8761 aux_loss:0.0162 train_time:5150555ms step_avg:1566.47ms step:3299/6250 train_loss:1.8918 aux_loss:0.0175 train_time:5151416ms step_avg:1566.26ms step:3300/6250 train_loss:1.8344 aux_loss:0.0165 train_time:5152526ms step_avg:1566.12ms step:3301/6250 train_loss:1.9006 aux_loss:0.0162 train_time:5153392ms step_avg:1565.90ms step:3302/6250 train_loss:1.8571 aux_loss:0.0161 train_time:5154263ms step_avg:1565.69ms step:3303/6250 train_loss:2.0225 aux_loss:0.0166 train_time:5155125ms step_avg:1565.48ms step:3304/6250 train_loss:1.9618 aux_loss:0.0167 train_time:5156021ms step_avg:1565.28ms step:3305/6250 train_loss:1.8398 aux_loss:0.0165 train_time:5157686ms step_avg:1565.31ms step:3306/6250 train_loss:1.9524 aux_loss:0.0169 train_time:5158572ms step_avg:1565.10ms step:3307/6250 train_loss:1.8655 aux_loss:0.0169 train_time:5159836ms step_avg:1565.01ms step:3308/6250 train_loss:1.9773 aux_loss:0.0167 train_time:5160718ms step_avg:1564.80ms step:3309/6250 train_loss:1.9194 aux_loss:0.0163 train_time:5161599ms step_avg:1564.60ms step:3310/6250 train_loss:1.9445 aux_loss:0.0162 train_time:5162468ms step_avg:1564.38ms step:3311/6250 train_loss:1.8955 aux_loss:0.0164 train_time:5163314ms step_avg:1564.17ms step:3312/6250 train_loss:1.9521 aux_loss:0.0162 train_time:5164161ms step_avg:1563.95ms step:3313/6250 train_loss:1.9235 aux_loss:0.0162 train_time:5165481ms step_avg:1563.88ms step:3314/6250 train_loss:1.9301 aux_loss:0.0164 train_time:5166332ms step_avg:1563.66ms step:3315/6250 train_loss:1.9376 aux_loss:0.0167 train_time:5167195ms step_avg:1563.45ms step:3316/6250 train_loss:2.0462 aux_loss:0.0165 train_time:5168056ms step_avg:1563.24ms step:3317/6250 train_loss:1.9464 aux_loss:0.0168 train_time:5168928ms step_avg:1563.03ms step:3318/6250 train_loss:1.8446 aux_loss:0.0161 train_time:5169795ms step_avg:1562.82ms step:3319/6250 train_loss:1.9292 aux_loss:0.0165 train_time:5170684ms step_avg:1562.61ms step:3320/6250 train_loss:1.8806 aux_loss:0.0165 train_time:5171526ms step_avg:1562.39ms step:3321/6250 train_loss:1.8768 aux_loss:0.0165 train_time:5172401ms step_avg:1562.19ms step:3322/6250 train_loss:1.9001 aux_loss:0.0162 train_time:5173562ms step_avg:1562.07ms step:3323/6250 train_loss:1.8673 aux_loss:0.0163 train_time:5174466ms step_avg:1561.87ms step:3324/6250 train_loss:1.9381 aux_loss:0.0165 train_time:5175330ms step_avg:1561.66ms step:3325/6250 train_loss:1.8396 aux_loss:0.0160 train_time:5176185ms step_avg:1561.44ms step:3326/6250 train_loss:1.8647 aux_loss:0.0158 train_time:5177054ms step_avg:1561.23ms step:3327/6250 train_loss:1.8686 aux_loss:0.0164 train_time:5177897ms step_avg:1561.02ms step:3328/6250 train_loss:1.9486 aux_loss:0.0165 train_time:5179145ms step_avg:1560.92ms step:3329/6250 train_loss:1.8809 aux_loss:0.0163 train_time:5180028ms step_avg:1560.72ms step:3330/6250 train_loss:1.9233 aux_loss:0.0166 train_time:5181026ms step_avg:1560.55ms step:3331/6250 train_loss:1.8869 aux_loss:0.0159 train_time:5181915ms step_avg:1560.35ms step:3332/6250 train_loss:1.9154 aux_loss:0.0158 train_time:5182969ms step_avg:1560.20ms step:3333/6250 train_loss:1.9302 aux_loss:0.0161 train_time:5184085ms step_avg:1560.06ms step:3334/6250 train_loss:1.9368 aux_loss:0.0166 train_time:5184937ms step_avg:1559.85ms step:3335/6250 train_loss:1.8495 aux_loss:0.0161 train_time:5186330ms step_avg:1559.80ms step:3336/6250 train_loss:1.8620 aux_loss:0.0164 train_time:5187168ms step_avg:1559.58ms step:3337/6250 train_loss:1.8710 aux_loss:0.0164 train_time:5188050ms step_avg:1559.38ms step:3338/6250 train_loss:1.8328 aux_loss:0.0160 train_time:5188885ms step_avg:1559.16ms step:3339/6250 train_loss:1.9120 aux_loss:0.0158 train_time:5189753ms step_avg:1558.95ms step:3340/6250 train_loss:1.8517 aux_loss:0.0162 train_time:5190612ms step_avg:1558.74ms step:3341/6250 train_loss:1.9379 aux_loss:0.0163 train_time:5191468ms step_avg:1558.53ms step:3342/6250 train_loss:1.9596 aux_loss:0.0162 train_time:5192331ms step_avg:1558.32ms step:3343/6250 train_loss:1.9289 aux_loss:0.0160 train_time:5193196ms step_avg:1558.11ms step:3344/6250 train_loss:1.9579 aux_loss:0.0161 train_time:5194352ms step_avg:1557.99ms step:3345/6250 train_loss:1.9145 aux_loss:0.0163 train_time:5195467ms step_avg:1557.86ms step:3346/6250 train_loss:1.8351 aux_loss:0.0158 train_time:5196329ms step_avg:1557.65ms step:3347/6250 train_loss:1.8733 aux_loss:0.0161 train_time:5197203ms step_avg:1557.45ms step:3348/6250 train_loss:1.9978 aux_loss:0.0165 train_time:5198068ms step_avg:1557.24ms step:3349/6250 train_loss:1.9067 aux_loss:0.0164 train_time:5199105ms step_avg:1557.08ms step:3350/6250 train_loss:1.8766 aux_loss:0.0160 train_time:5201533ms step_avg:1557.35ms step:3351/6250 train_loss:1.9400 aux_loss:0.0165 train_time:5202397ms step_avg:1557.14ms step:3352/6250 train_loss:2.0118 aux_loss:0.0163 train_time:5203262ms step_avg:1556.93ms step:3353/6250 train_loss:1.8600 aux_loss:0.0159 train_time:5216059ms step_avg:1560.29ms step:3354/6250 train_loss:1.8413 aux_loss:0.0156 train_time:5237646ms step_avg:1566.28ms step:3355/6250 train_loss:1.9128 aux_loss:0.0160 train_time:5238503ms step_avg:1566.07ms step:3356/6250 train_loss:1.9327 aux_loss:0.0159 train_time:5250149ms step_avg:1569.08ms step:3357/6250 train_loss:1.9059 aux_loss:0.0158 train_time:5251791ms step_avg:1569.10ms step:3358/6250 train_loss:1.8857 aux_loss:0.0160 train_time:5253385ms step_avg:1569.11ms step:3359/6250 train_loss:1.8425 aux_loss:0.0160 train_time:5261648ms step_avg:1571.11ms step:3360/6250 train_loss:1.8546 aux_loss:0.0156 train_time:5268802ms step_avg:1572.78ms step:3361/6250 train_loss:1.9624 aux_loss:0.0155 train_time:5272230ms step_avg:1573.33ms step:3362/6250 train_loss:1.8843 aux_loss:0.0160 train_time:5274207ms step_avg:1573.45ms step:3363/6250 train_loss:1.9315 aux_loss:0.0160 train_time:5275876ms step_avg:1573.48ms step:3364/6250 train_loss:1.9369 aux_loss:0.0156 train_time:5277652ms step_avg:1573.54ms step:3365/6250 train_loss:1.9567 aux_loss:0.0161 train_time:5278972ms step_avg:1573.46ms step:3366/6250 train_loss:1.7967 aux_loss:0.0161 train_time:5279874ms step_avg:1573.26ms step:3367/6250 train_loss:1.9240 aux_loss:0.0159 train_time:5294058ms step_avg:1577.02ms step:3368/6250 train_loss:1.9803 aux_loss:0.0160 train_time:5300205ms step_avg:1578.38ms step:3369/6250 train_loss:1.9114 aux_loss:0.0163 train_time:5301393ms step_avg:1578.27ms step:3370/6250 train_loss:1.8867 aux_loss:0.0163 train_time:5302920ms step_avg:1578.25ms step:3371/6250 train_loss:1.8917 aux_loss:0.0160 train_time:5304097ms step_avg:1578.13ms step:3372/6250 train_loss:2.0057 aux_loss:0.0160 train_time:5306052ms step_avg:1578.24ms step:3373/6250 train_loss:1.9295 aux_loss:0.0161 train_time:5308402ms step_avg:1578.47ms step:3374/6250 train_loss:1.9097 aux_loss:0.0163 train_time:5309731ms step_avg:1578.40ms step:3375/6250 train_loss:1.9342 aux_loss:0.0163 train_time:5311284ms step_avg:1578.39ms step:3376/6250 train_loss:1.8773 aux_loss:0.0156 train_time:5313496ms step_avg:1578.58ms step:3377/6250 train_loss:1.8774 aux_loss:0.0162 train_time:5319245ms step_avg:1579.82ms step:3378/6250 train_loss:1.9440 aux_loss:0.0163 train_time:5321373ms step_avg:1579.98ms step:3379/6250 train_loss:1.9730 aux_loss:0.0159 train_time:5326609ms step_avg:1581.07ms step:3380/6250 train_loss:1.9897 aux_loss:0.0156 train_time:5327799ms step_avg:1580.95ms step:3381/6250 train_loss:1.9868 aux_loss:0.0159 train_time:5328845ms step_avg:1580.79ms step:3382/6250 train_loss:1.8920 aux_loss:0.0159 train_time:5330813ms step_avg:1580.91ms step:3383/6250 train_loss:1.8439 aux_loss:0.0156 train_time:5331698ms step_avg:1580.70ms step:3384/6250 train_loss:1.8279 aux_loss:0.0152 train_time:5332624ms step_avg:1580.50ms step:3385/6250 train_loss:1.8566 aux_loss:0.0158 train_time:5334513ms step_avg:1580.60ms step:3386/6250 train_loss:1.9238 aux_loss:0.0165 train_time:5336250ms step_avg:1580.64ms step:3387/6250 train_loss:1.9464 aux_loss:0.0159 train_time:5337160ms step_avg:1580.44ms step:3388/6250 train_loss:1.8868 aux_loss:0.0159 train_time:5339152ms step_avg:1580.57ms step:3389/6250 train_loss:2.0073 aux_loss:0.0160 train_time:5342984ms step_avg:1581.23ms step:3390/6250 train_loss:1.9182 aux_loss:0.0160 train_time:5343864ms step_avg:1581.02ms step:3391/6250 train_loss:2.0040 aux_loss:0.0161 train_time:5345884ms step_avg:1581.15ms step:3392/6250 train_loss:1.8455 aux_loss:0.0159 train_time:5347324ms step_avg:1581.11ms step:3393/6250 train_loss:1.8628 aux_loss:0.0158 train_time:5348397ms step_avg:1580.96ms step:3394/6250 train_loss:1.9307 aux_loss:0.0161 train_time:5349801ms step_avg:1580.91ms step:3395/6250 train_loss:1.8977 aux_loss:0.0160 train_time:5352126ms step_avg:1581.13ms step:3396/6250 train_loss:1.9445 aux_loss:0.0160 train_time:5353832ms step_avg:1581.17ms step:3397/6250 train_loss:1.8957 aux_loss:0.0160 train_time:5357444ms step_avg:1581.77ms step:3398/6250 train_loss:1.9311 aux_loss:0.0163 train_time:5359288ms step_avg:1581.84ms step:3399/6250 train_loss:1.9090 aux_loss:0.0163 train_time:5361055ms step_avg:1581.90ms step:3400/6250 train_loss:1.9172 aux_loss:0.0160 train_time:5362371ms step_avg:1581.82ms step:3401/6250 train_loss:1.8744 aux_loss:0.0161 train_time:5363826ms step_avg:1581.78ms step:3402/6250 train_loss:1.9472 aux_loss:0.0163 train_time:5365496ms step_avg:1581.81ms step:3403/6250 train_loss:1.9266 aux_loss:0.0157 train_time:5368406ms step_avg:1582.20ms step:3404/6250 train_loss:1.8365 aux_loss:0.0156 train_time:5371947ms step_avg:1582.78ms step:3405/6250 train_loss:1.8658 aux_loss:0.0159 train_time:5373224ms step_avg:1582.69ms step:3406/6250 train_loss:1.9628 aux_loss:0.0165 train_time:5376444ms step_avg:1583.17ms step:3407/6250 train_loss:1.9164 aux_loss:0.0160 train_time:5378438ms step_avg:1583.29ms step:3408/6250 train_loss:1.7873 aux_loss:0.0153 train_time:5379344ms step_avg:1583.09ms step:3409/6250 train_loss:1.8102 aux_loss:0.0158 train_time:5380254ms step_avg:1582.89ms step:3410/6250 train_loss:1.8757 aux_loss:0.0159 train_time:5381312ms step_avg:1582.74ms step:3411/6250 train_loss:1.8839 aux_loss:0.0163 train_time:5382360ms step_avg:1582.58ms step:3412/6250 train_loss:1.9199 aux_loss:0.0161 train_time:5383495ms step_avg:1582.45ms step:3413/6250 train_loss:2.0346 aux_loss:0.0158 train_time:5384421ms step_avg:1582.26ms step:3414/6250 train_loss:1.9245 aux_loss:0.0160 train_time:5385499ms step_avg:1582.11ms step:3415/6250 train_loss:1.8622 aux_loss:0.0160 train_time:5386412ms step_avg:1581.91ms step:3416/6250 train_loss:1.8647 aux_loss:0.0159 train_time:5387343ms step_avg:1581.72ms step:3417/6250 train_loss:1.9412 aux_loss:0.0158 train_time:5388214ms step_avg:1581.51ms step:3418/6250 train_loss:1.8940 aux_loss:0.0158 train_time:5389144ms step_avg:1581.32ms step:3419/6250 train_loss:1.8817 aux_loss:0.0158 train_time:5390239ms step_avg:1581.18ms step:3420/6250 train_loss:1.9128 aux_loss:0.0159 train_time:5391341ms step_avg:1581.04ms step:3421/6250 train_loss:1.9570 aux_loss:0.0161 train_time:5392826ms step_avg:1581.01ms step:3422/6250 train_loss:1.9081 aux_loss:0.0164 train_time:5394733ms step_avg:1581.11ms step:3423/6250 train_loss:1.9044 aux_loss:0.0156 train_time:5396568ms step_avg:1581.18ms step:3424/6250 train_loss:1.9164 aux_loss:0.0158 train_time:5398481ms step_avg:1581.28ms step:3425/6250 train_loss:1.9206 aux_loss:0.0158 train_time:5400296ms step_avg:1581.35ms step:3426/6250 train_loss:2.0064 aux_loss:0.0160 train_time:5401421ms step_avg:1581.21ms step:3427/6250 train_loss:1.8546 aux_loss:0.0158 train_time:5402351ms step_avg:1581.02ms step:3428/6250 train_loss:1.8981 aux_loss:0.0157 train_time:5403779ms step_avg:1580.98ms step:3429/6250 train_loss:1.8419 aux_loss:0.0158 train_time:5404716ms step_avg:1580.79ms step:3430/6250 train_loss:1.9486 aux_loss:0.0157 train_time:5405606ms step_avg:1580.59ms step:3431/6250 train_loss:1.9327 aux_loss:0.0158 train_time:5406482ms step_avg:1580.38ms step:3432/6250 train_loss:1.9232 aux_loss:0.0159 train_time:5407380ms step_avg:1580.18ms step:3433/6250 train_loss:1.9642 aux_loss:0.0161 train_time:5408342ms step_avg:1580.00ms step:3434/6250 train_loss:1.8966 aux_loss:0.0159 train_time:5411126ms step_avg:1580.35ms step:3435/6250 train_loss:1.9110 aux_loss:0.0157 train_time:5415891ms step_avg:1581.28ms step:3436/6250 train_loss:1.9006 aux_loss:0.0158 train_time:5417653ms step_avg:1581.33ms step:3437/6250 train_loss:1.8994 aux_loss:0.0159 train_time:5418671ms step_avg:1581.17ms step:3438/6250 train_loss:1.8587 aux_loss:0.0160 train_time:5420733ms step_avg:1581.31ms step:3439/6250 train_loss:1.8368 aux_loss:0.0162 train_time:5424380ms step_avg:1581.91ms step:3440/6250 train_loss:1.9795 aux_loss:0.0158 train_time:5426178ms step_avg:1581.98ms step:3441/6250 train_loss:1.9042 aux_loss:0.0155 train_time:5427472ms step_avg:1581.89ms step:3442/6250 train_loss:1.9060 aux_loss:0.0156 train_time:5428426ms step_avg:1581.71ms step:3443/6250 train_loss:1.9011 aux_loss:0.0156 train_time:5429849ms step_avg:1581.66ms step:3444/6250 train_loss:1.9563 aux_loss:0.0160 train_time:5431202ms step_avg:1581.60ms step:3445/6250 train_loss:1.8014 aux_loss:0.0158 train_time:5434548ms step_avg:1582.11ms step:3446/6250 train_loss:1.9411 aux_loss:0.0158 train_time:5435582ms step_avg:1581.95ms step:3447/6250 train_loss:1.9386 aux_loss:0.0159 train_time:5436762ms step_avg:1581.83ms step:3448/6250 train_loss:1.7607 aux_loss:0.0156 train_time:5437918ms step_avg:1581.71ms step:3449/6250 train_loss:1.8591 aux_loss:0.0159 train_time:5439137ms step_avg:1581.60ms step:3450/6250 train_loss:1.8947 aux_loss:0.0159 train_time:5440357ms step_avg:1581.50ms step:3451/6250 train_loss:1.7887 aux_loss:0.0160 train_time:5441496ms step_avg:1581.37ms step:3452/6250 train_loss:1.8754 aux_loss:0.0167 train_time:5442729ms step_avg:1581.27ms step:3453/6250 train_loss:1.8774 aux_loss:0.0157 train_time:5443993ms step_avg:1581.18ms step:3454/6250 train_loss:1.8799 aux_loss:0.0155 train_time:5445223ms step_avg:1581.08ms step:3455/6250 train_loss:2.0309 aux_loss:0.0155 train_time:5446373ms step_avg:1580.95ms step:3456/6250 train_loss:1.9275 aux_loss:0.0156 train_time:5447640ms step_avg:1580.86ms step:3457/6250 train_loss:1.9485 aux_loss:0.0158 train_time:5448926ms step_avg:1580.77ms step:3458/6250 train_loss:2.0193 aux_loss:0.0160 train_time:5450080ms step_avg:1580.65ms step:3459/6250 train_loss:1.8796 aux_loss:0.0157 train_time:5451402ms step_avg:1580.57ms step:3460/6250 train_loss:1.9503 aux_loss:0.0158 train_time:5452703ms step_avg:1580.49ms step:3461/6250 train_loss:1.8538 aux_loss:0.0158 train_time:5453962ms step_avg:1580.40ms step:3462/6250 train_loss:1.8255 aux_loss:0.0153 train_time:5455219ms step_avg:1580.31ms step:3463/6250 train_loss:1.9419 aux_loss:0.0155 train_time:5456352ms step_avg:1580.18ms step:3464/6250 train_loss:1.8910 aux_loss:0.0159 train_time:5457568ms step_avg:1580.07ms step:3465/6250 train_loss:1.9094 aux_loss:0.0159 train_time:5458755ms step_avg:1579.96ms step:3466/6250 train_loss:1.8697 aux_loss:0.0163 train_time:5460006ms step_avg:1579.86ms step:3467/6250 train_loss:1.8828 aux_loss:0.0157 train_time:5461153ms step_avg:1579.74ms step:3468/6250 train_loss:1.9061 aux_loss:0.0157 train_time:5462399ms step_avg:1579.64ms step:3469/6250 train_loss:1.9870 aux_loss:0.0159 train_time:5463595ms step_avg:1579.53ms step:3470/6250 train_loss:1.8444 aux_loss:0.0158 train_time:5464753ms step_avg:1579.41ms step:3471/6250 train_loss:1.8479 aux_loss:0.0157 train_time:5465857ms step_avg:1579.27ms step:3472/6250 train_loss:1.8139 aux_loss:0.0156 train_time:5466891ms step_avg:1579.11ms step:3473/6250 train_loss:1.8766 aux_loss:0.0158 train_time:5468031ms step_avg:1578.99ms step:3474/6250 train_loss:1.8885 aux_loss:0.0153 train_time:5469158ms step_avg:1578.86ms step:3475/6250 train_loss:1.8067 aux_loss:0.0154 train_time:5470154ms step_avg:1578.69ms step:3476/6250 train_loss:1.8735 aux_loss:0.0156 train_time:5471022ms step_avg:1578.48ms step:3477/6250 train_loss:2.0047 aux_loss:0.0158 train_time:5471901ms step_avg:1578.28ms step:3478/6250 train_loss:1.8495 aux_loss:0.0155 train_time:5472754ms step_avg:1578.07ms step:3479/6250 train_loss:1.7729 aux_loss:0.0155 train_time:5473643ms step_avg:1577.87ms step:3480/6250 train_loss:1.8073 aux_loss:0.0155 train_time:5474518ms step_avg:1577.67ms step:3481/6250 train_loss:2.0045 aux_loss:0.0164 train_time:5475395ms step_avg:1577.47ms step:3482/6250 train_loss:1.8317 aux_loss:0.0155 train_time:5476271ms step_avg:1577.27ms step:3483/6250 train_loss:1.9587 aux_loss:0.0153 train_time:5477188ms step_avg:1577.08ms step:3484/6250 train_loss:2.0116 aux_loss:0.0154 train_time:5478117ms step_avg:1576.89ms step:3485/6250 train_loss:1.9363 aux_loss:0.0154 train_time:5479091ms step_avg:1576.72ms step:3486/6250 train_loss:1.8436 aux_loss:0.0154 train_time:5480090ms step_avg:1576.55ms step:3487/6250 train_loss:2.0754 aux_loss:0.0164 train_time:5481057ms step_avg:1576.38ms step:3488/6250 train_loss:1.9048 aux_loss:0.0160 train_time:5481992ms step_avg:1576.19ms step:3489/6250 train_loss:1.9101 aux_loss:0.0156 train_time:5483334ms step_avg:1576.12ms step:3490/6250 train_loss:1.8839 aux_loss:0.0156 train_time:5485021ms step_avg:1576.16ms step:3491/6250 train_loss:1.9978 aux_loss:0.0156 train_time:5486625ms step_avg:1576.16ms step:3492/6250 train_loss:1.9239 aux_loss:0.0156 train_time:5488194ms step_avg:1576.16ms step:3493/6250 train_loss:1.9659 aux_loss:0.0160 train_time:5489745ms step_avg:1576.15ms step:3494/6250 train_loss:1.8272 aux_loss:0.0155 train_time:5491361ms step_avg:1576.17ms step:3495/6250 train_loss:1.8753 aux_loss:0.0157 train_time:5492859ms step_avg:1576.14ms step:3496/6250 train_loss:1.8384 aux_loss:0.0156 train_time:5494083ms step_avg:1576.04ms step:3497/6250 train_loss:1.9252 aux_loss:0.0154 train_time:5495211ms step_avg:1575.91ms step:3498/6250 train_loss:1.9472 aux_loss:0.0155 train_time:5496106ms step_avg:1575.72ms step:3499/6250 train_loss:1.8355 aux_loss:0.0154 train_time:5496969ms step_avg:1575.51ms step:3500/6250 train_loss:1.8729 aux_loss:0.0154 train_time:5498145ms step_avg:1575.40ms step:3501/6250 train_loss:1.8620 aux_loss:0.0153 train_time:5499103ms step_avg:1575.22ms step:3502/6250 train_loss:1.8794 aux_loss:0.0157 train_time:5500030ms step_avg:1575.04ms step:3503/6250 train_loss:1.8746 aux_loss:0.0158 train_time:5501200ms step_avg:1574.92ms step:3504/6250 train_loss:1.9186 aux_loss:0.0159 train_time:5502127ms step_avg:1574.74ms step:3505/6250 train_loss:1.9388 aux_loss:0.0156 train_time:5503017ms step_avg:1574.54ms step:3506/6250 train_loss:2.0359 aux_loss:0.0158 train_time:5503988ms step_avg:1574.37ms step:3507/6250 train_loss:1.8876 aux_loss:0.0158 train_time:5504977ms step_avg:1574.20ms step:3508/6250 train_loss:1.8812 aux_loss:0.0160 train_time:5505982ms step_avg:1574.04ms step:3509/6250 train_loss:1.9357 aux_loss:0.0155 train_time:5506901ms step_avg:1573.85ms step:3510/6250 train_loss:1.9593 aux_loss:0.0157 train_time:5507802ms step_avg:1573.66ms step:3511/6250 train_loss:1.9055 aux_loss:0.0158 train_time:5508688ms step_avg:1573.46ms step:3512/6250 train_loss:1.9309 aux_loss:0.0155 train_time:5509585ms step_avg:1573.27ms step:3513/6250 train_loss:1.9730 aux_loss:0.0154 train_time:5510588ms step_avg:1573.11ms step:3514/6250 train_loss:1.9440 aux_loss:0.0152 train_time:5511536ms step_avg:1572.93ms step:3515/6250 train_loss:1.8873 aux_loss:0.0156 train_time:5512481ms step_avg:1572.75ms step:3516/6250 train_loss:1.8695 aux_loss:0.0154 train_time:5513373ms step_avg:1572.55ms step:3517/6250 train_loss:1.9606 aux_loss:0.0155 train_time:5514371ms step_avg:1572.39ms step:3518/6250 train_loss:1.8915 aux_loss:0.0155 train_time:5515448ms step_avg:1572.25ms step:3519/6250 train_loss:1.8680 aux_loss:0.0156 train_time:5516627ms step_avg:1572.14ms step:3520/6250 train_loss:1.8446 aux_loss:0.0153 train_time:5517537ms step_avg:1571.95ms step:3521/6250 train_loss:1.8569 aux_loss:0.0154 train_time:5518456ms step_avg:1571.76ms step:3522/6250 train_loss:1.9575 aux_loss:0.0157 train_time:5519419ms step_avg:1571.59ms step:3523/6250 train_loss:1.8902 aux_loss:0.0154 train_time:5520413ms step_avg:1571.42ms step:3524/6250 train_loss:1.8789 aux_loss:0.0155 train_time:5521387ms step_avg:1571.25ms step:3525/6250 train_loss:1.9067 aux_loss:0.0156 train_time:5522252ms step_avg:1571.05ms step:3526/6250 train_loss:1.9173 aux_loss:0.0155 train_time:5523113ms step_avg:1570.85ms step:3527/6250 train_loss:1.8992 aux_loss:0.0153 train_time:5524095ms step_avg:1570.68ms step:3528/6250 train_loss:1.8865 aux_loss:0.0156 train_time:5525013ms step_avg:1570.50ms step:3529/6250 train_loss:2.0025 aux_loss:0.0157 train_time:5526226ms step_avg:1570.40ms step:3530/6250 train_loss:1.8164 aux_loss:0.0151 train_time:5527068ms step_avg:1570.19ms step:3531/6250 train_loss:1.9419 aux_loss:0.0149 train_time:5527936ms step_avg:1569.99ms step:3532/6250 train_loss:1.8992 aux_loss:0.0151 train_time:5528802ms step_avg:1569.79ms step:3533/6250 train_loss:1.8327 aux_loss:0.0150 train_time:5529707ms step_avg:1569.60ms step:3534/6250 train_loss:1.9441 aux_loss:0.0153 train_time:5530720ms step_avg:1569.44ms step:3535/6250 train_loss:1.8592 aux_loss:0.0155 train_time:5531985ms step_avg:1569.36ms step:3536/6250 train_loss:1.9333 aux_loss:0.0158 train_time:5533215ms step_avg:1569.26ms step:3537/6250 train_loss:1.7886 aux_loss:0.0158 train_time:5534209ms step_avg:1569.10ms step:3538/6250 train_loss:1.8977 aux_loss:0.0158 train_time:5535086ms step_avg:1568.90ms step:3539/6250 train_loss:1.8968 aux_loss:0.0158 train_time:5536023ms step_avg:1568.72ms step:3540/6250 train_loss:1.8907 aux_loss:0.0154 train_time:5536955ms step_avg:1568.54ms step:3541/6250 train_loss:1.9673 aux_loss:0.0157 train_time:5537956ms step_avg:1568.38ms step:3542/6250 train_loss:1.8887 aux_loss:0.0152 train_time:5538911ms step_avg:1568.21ms step:3543/6250 train_loss:1.9587 aux_loss:0.0155 train_time:5539933ms step_avg:1568.05ms step:3544/6250 train_loss:1.8401 aux_loss:0.0153 train_time:5540858ms step_avg:1567.87ms step:3545/6250 train_loss:1.8763 aux_loss:0.0149 train_time:5541735ms step_avg:1567.68ms step:3546/6250 train_loss:1.9012 aux_loss:0.0153 train_time:5542729ms step_avg:1567.51ms step:3547/6250 train_loss:1.9036 aux_loss:0.0156 train_time:5543679ms step_avg:1567.34ms step:3548/6250 train_loss:1.9449 aux_loss:0.0154 train_time:5544695ms step_avg:1567.18ms step:3549/6250 train_loss:2.0047 aux_loss:0.0158 train_time:5545950ms step_avg:1567.10ms step:3550/6250 train_loss:1.9052 aux_loss:0.0153 train_time:5547259ms step_avg:1567.02ms step:3551/6250 train_loss:1.8998 aux_loss:0.0154 train_time:5548971ms step_avg:1567.06ms step:3552/6250 train_loss:1.9161 aux_loss:0.0155 train_time:5550265ms step_avg:1566.99ms step:3553/6250 train_loss:1.9369 aux_loss:0.0156 train_time:5551822ms step_avg:1566.98ms step:3554/6250 train_loss:1.8555 aux_loss:0.0151 train_time:5553355ms step_avg:1566.97ms step:3555/6250 train_loss:1.8611 aux_loss:0.0151 train_time:5554551ms step_avg:1566.87ms step:3556/6250 train_loss:1.9163 aux_loss:0.0153 train_time:5555851ms step_avg:1566.79ms step:3557/6250 train_loss:1.8417 aux_loss:0.0152 train_time:5557332ms step_avg:1566.77ms step:3558/6250 train_loss:1.9357 aux_loss:0.0154 train_time:5558708ms step_avg:1566.72ms step:3559/6250 train_loss:1.8113 aux_loss:0.0151 train_time:5560167ms step_avg:1566.69ms step:3560/6250 train_loss:1.8561 aux_loss:0.0152 train_time:5561450ms step_avg:1566.61ms step:3561/6250 train_loss:1.9044 aux_loss:0.0153 train_time:5563081ms step_avg:1566.62ms step:3562/6250 train_loss:1.8519 aux_loss:0.0151 train_time:5564452ms step_avg:1566.57ms step:3563/6250 train_loss:1.8986 aux_loss:0.0152 train_time:5566052ms step_avg:1566.58ms step:3564/6250 train_loss:1.8541 aux_loss:0.0153 train_time:5567576ms step_avg:1566.57ms step:3565/6250 train_loss:1.8834 aux_loss:0.0150 train_time:5569160ms step_avg:1566.57ms step:3566/6250 train_loss:1.9451 aux_loss:0.0152 train_time:5570746ms step_avg:1566.58ms step:3567/6250 train_loss:1.8655 aux_loss:0.0152 train_time:5572382ms step_avg:1566.60ms step:3568/6250 train_loss:1.8002 aux_loss:0.0205 train_time:5573929ms step_avg:1566.59ms step:3569/6250 train_loss:1.8311 aux_loss:0.0149 train_time:5575590ms step_avg:1566.62ms step:3570/6250 train_loss:1.8006 aux_loss:0.0154 train_time:5577295ms step_avg:1566.66ms step:3571/6250 train_loss:1.8857 aux_loss:0.0153 train_time:5579810ms step_avg:1566.92ms step:3572/6250 train_loss:2.0232 aux_loss:0.0154 train_time:5581732ms step_avg:1567.02ms step:3573/6250 train_loss:1.8897 aux_loss:0.0154 train_time:5582641ms step_avg:1566.84ms step:3574/6250 train_loss:1.9194 aux_loss:0.0155 train_time:5583985ms step_avg:1566.77ms step:3575/6250 train_loss:1.8413 aux_loss:0.0155 train_time:5586380ms step_avg:1567.01ms step:3576/6250 train_loss:1.9143 aux_loss:0.0155 train_time:5587920ms step_avg:1567.00ms step:3577/6250 train_loss:1.9362 aux_loss:0.0154 train_time:5591062ms step_avg:1567.44ms step:3578/6250 train_loss:1.8585 aux_loss:0.0148 train_time:5596132ms step_avg:1568.42ms step:3579/6250 train_loss:1.9123 aux_loss:0.0152 train_time:5599251ms step_avg:1568.86ms step:3580/6250 train_loss:1.8750 aux_loss:0.0153 train_time:5600146ms step_avg:1568.67ms step:3581/6250 train_loss:1.9294 aux_loss:0.0149 train_time:5601096ms step_avg:1568.50ms step:3582/6250 train_loss:1.8258 aux_loss:0.0151 train_time:5601985ms step_avg:1568.30ms step:3583/6250 train_loss:1.9273 aux_loss:0.0157 train_time:5603089ms step_avg:1568.17ms step:3584/6250 train_loss:1.8425 aux_loss:0.0150 train_time:5603989ms step_avg:1567.99ms step:3585/6250 train_loss:1.7584 aux_loss:0.0150 train_time:5606195ms step_avg:1568.17ms step:3586/6250 train_loss:1.9156 aux_loss:0.0152 train_time:5607863ms step_avg:1568.19ms step:3587/6250 train_loss:1.8365 aux_loss:0.0154 train_time:5610947ms step_avg:1568.62ms step:3588/6250 train_loss:1.7856 aux_loss:0.0154 train_time:5612646ms step_avg:1568.65ms step:3589/6250 train_loss:1.8973 aux_loss:0.0148 train_time:5615161ms step_avg:1568.92ms step:3590/6250 train_loss:1.9202 aux_loss:0.0147 train_time:5617781ms step_avg:1569.21ms step:3591/6250 train_loss:1.8761 aux_loss:0.0153 train_time:5619116ms step_avg:1569.15ms step:3592/6250 train_loss:1.9546 aux_loss:0.0155 train_time:5620184ms step_avg:1569.01ms step:3593/6250 train_loss:1.8572 aux_loss:0.0149 train_time:5621096ms step_avg:1568.82ms step:3594/6250 train_loss:1.8357 aux_loss:0.0146 train_time:5638774ms step_avg:1573.32ms step:3595/6250 train_loss:1.8737 aux_loss:0.0152 train_time:5668930ms step_avg:1581.29ms step:3596/6250 train_loss:1.8447 aux_loss:0.0153 train_time:5699573ms step_avg:1589.40ms step:3597/6250 train_loss:1.8511 aux_loss:0.0151 train_time:5742793ms step_avg:1601.00ms step:3598/6250 train_loss:1.8183 aux_loss:0.0150 train_time:5784688ms step_avg:1612.23ms step:3599/6250 train_loss:1.7942 aux_loss:0.0147 train_time:5832640ms step_avg:1625.14ms step:3600/6250 train_loss:1.7881 aux_loss:0.0152 train_time:5881196ms step_avg:1638.22ms step:3601/6250 train_loss:1.9091 aux_loss:0.0154 train_time:5918305ms step_avg:1648.09ms step:3602/6250 train_loss:1.8590 aux_loss:0.0149 train_time:5955959ms step_avg:1658.12ms step:3603/6250 train_loss:1.8241 aux_loss:0.0148 train_time:5988931ms step_avg:1666.83ms step:3604/6250 train_loss:1.8301 aux_loss:0.0148 train_time:6029167ms step_avg:1677.56ms step:3605/6250 train_loss:1.9071 aux_loss:0.0152 train_time:6062428ms step_avg:1686.35ms step:3606/6250 train_loss:1.8515 aux_loss:0.0151 train_time:6096133ms step_avg:1695.25ms step:3607/6250 train_loss:1.8531 aux_loss:0.0152 train_time:6142458ms step_avg:1707.66ms step:3608/6250 train_loss:1.9845 aux_loss:0.0153 train_time:6183178ms step_avg:1718.50ms step:3609/6250 train_loss:1.8640 aux_loss:0.0147 train_time:6227168ms step_avg:1730.25ms step:3610/6250 train_loss:1.8629 aux_loss:0.0149 train_time:6265164ms step_avg:1740.32ms step:3611/6250 train_loss:1.9444 aux_loss:0.0152 train_time:6298995ms step_avg:1749.23ms step:3612/6250 train_loss:1.9747 aux_loss:0.0151 train_time:6342238ms step_avg:1760.75ms step:3613/6250 train_loss:1.8325 aux_loss:0.0148 train_time:6378692ms step_avg:1770.38ms step:3614/6250 train_loss:1.7940 aux_loss:0.0151 train_time:6424203ms step_avg:1782.52ms step:3615/6250 train_loss:1.8188 aux_loss:0.0152 train_time:6476632ms step_avg:1796.57ms step:3616/6250 train_loss:1.9008 aux_loss:0.0154 train_time:6535356ms step_avg:1812.36ms step:3617/6250 train_loss:1.8549 aux_loss:0.0152 train_time:6585895ms step_avg:1825.86ms step:3618/6250 train_loss:1.8718 aux_loss:0.0149 train_time:6614058ms step_avg:1833.16ms step:3619/6250 train_loss:1.9455 aux_loss:0.0152 train_time:6616596ms step_avg:1833.36ms step:3620/6250 train_loss:1.8813 aux_loss:0.0151 train_time:6618822ms step_avg:1833.47ms step:3621/6250 train_loss:1.9404 aux_loss:0.0153 train_time:6620450ms step_avg:1833.41ms step:3622/6250 train_loss:1.8708 aux_loss:0.0154 train_time:6621779ms step_avg:1833.27ms step:3623/6250 train_loss:1.9342 aux_loss:0.0150 train_time:6623692ms step_avg:1833.29ms step:3624/6250 train_loss:1.8470 aux_loss:0.0151 train_time:6625201ms step_avg:1833.20ms step:3625/6250 train_loss:1.9705 aux_loss:0.0155 train_time:6627025ms step_avg:1833.20ms step:3626/6250 train_loss:2.0284 aux_loss:0.0155 train_time:6628354ms step_avg:1833.06ms step:3627/6250 train_loss:1.9011 aux_loss:0.0152 train_time:6629763ms step_avg:1832.95ms step:3628/6250 train_loss:1.8668 aux_loss:0.0149 train_time:6631087ms step_avg:1832.80ms step:3629/6250 train_loss:1.7970 aux_loss:0.0148 train_time:6632837ms step_avg:1832.78ms step:3630/6250 train_loss:1.9103 aux_loss:0.0152 train_time:6634184ms step_avg:1832.65ms step:3631/6250 train_loss:1.9272 aux_loss:0.0149 train_time:6637143ms step_avg:1832.96ms step:3632/6250 train_loss:1.8587 aux_loss:0.0152 train_time:6640435ms step_avg:1833.36ms step:3633/6250 train_loss:1.8052 aux_loss:0.0151 train_time:6642448ms step_avg:1833.41ms step:3634/6250 train_loss:1.8869 aux_loss:0.0151 train_time:6645083ms step_avg:1833.63ms step:3635/6250 train_loss:1.8634 aux_loss:0.0149 train_time:6647378ms step_avg:1833.76ms step:3636/6250 train_loss:1.9307 aux_loss:0.0153 train_time:6649386ms step_avg:1833.81ms step:3637/6250 train_loss:1.9495 aux_loss:0.0150 train_time:6651996ms step_avg:1834.02ms step:3638/6250 train_loss:1.9104 aux_loss:0.0152 train_time:6654454ms step_avg:1834.19ms step:3639/6250 train_loss:1.9688 aux_loss:0.0150 train_time:6656491ms step_avg:1834.25ms step:3640/6250 train_loss:1.8696 aux_loss:0.0146 train_time:6694640ms step_avg:1844.25ms step:3641/6250 train_loss:1.9357 aux_loss:0.0151 train_time:6735628ms step_avg:1855.03ms step:3642/6250 train_loss:1.8600 aux_loss:0.0150 train_time:6767474ms step_avg:1863.29ms step:3643/6250 train_loss:2.0210 aux_loss:0.0152 train_time:6770857ms step_avg:1863.71ms step:3644/6250 train_loss:1.8717 aux_loss:0.0152 train_time:6772944ms step_avg:1863.77ms step:3645/6250 train_loss:1.8709 aux_loss:0.0151 train_time:6779173ms step_avg:1864.97ms step:3646/6250 train_loss:1.9045 aux_loss:0.0149 train_time:6781719ms step_avg:1865.16ms step:3647/6250 train_loss:1.8830 aux_loss:0.0149 train_time:6784614ms step_avg:1865.44ms step:3648/6250 train_loss:1.8778 aux_loss:0.0156 train_time:6787232ms step_avg:1865.65ms step:3649/6250 train_loss:1.9142 aux_loss:0.0155 train_time:6788846ms step_avg:1865.58ms step:3650/6250 train_loss:1.8198 aux_loss:0.0149 train_time:6790232ms step_avg:1865.45ms step:3651/6250 train_loss:1.9417 aux_loss:0.0151 train_time:6793301ms step_avg:1865.78ms step:3652/6250 train_loss:1.8192 aux_loss:0.0149 train_time:6795284ms step_avg:1865.81ms step:3653/6250 train_loss:1.8219 aux_loss:0.0152 train_time:6797148ms step_avg:1865.81ms step:3654/6250 train_loss:1.8503 aux_loss:0.0148 train_time:6799741ms step_avg:1866.01ms step:3655/6250 train_loss:1.9056 aux_loss:0.0147 train_time:6801143ms step_avg:1865.88ms step:3656/6250 train_loss:1.9296 aux_loss:0.0151 train_time:6802458ms step_avg:1865.73ms step:3657/6250 train_loss:1.9065 aux_loss:0.0150 train_time:6803918ms step_avg:1865.62ms step:3658/6250 train_loss:1.8920 aux_loss:0.0150 train_time:6806135ms step_avg:1865.72ms step:3659/6250 train_loss:1.9730 aux_loss:0.0151 train_time:6807695ms step_avg:1865.63ms step:3660/6250 train_loss:1.7798 aux_loss:0.0148 train_time:6808983ms step_avg:1865.47ms step:3661/6250 train_loss:1.8629 aux_loss:0.0150 train_time:6810859ms step_avg:1865.48ms step:3662/6250 train_loss:1.8185 aux_loss:0.0149 train_time:6812238ms step_avg:1865.34ms step:3663/6250 train_loss:1.8239 aux_loss:0.0144 train_time:6813609ms step_avg:1865.21ms step:3664/6250 train_loss:1.8694 aux_loss:0.0150 train_time:6815017ms step_avg:1865.08ms step:3665/6250 train_loss:1.8716 aux_loss:0.0151 train_time:6816376ms step_avg:1864.95ms step:3666/6250 train_loss:1.8825 aux_loss:0.0151 train_time:6817659ms step_avg:1864.79ms step:3667/6250 train_loss:1.8868 aux_loss:0.0150 train_time:6818988ms step_avg:1864.64ms step:3668/6250 train_loss:1.8999 aux_loss:0.0150 train_time:6820297ms step_avg:1864.49ms step:3669/6250 train_loss:1.8605 aux_loss:0.0152 train_time:6821662ms step_avg:1864.35ms step:3670/6250 train_loss:1.9120 aux_loss:0.0149 train_time:6822923ms step_avg:1864.19ms step:3671/6250 train_loss:1.8623 aux_loss:0.0150 train_time:6824837ms step_avg:1864.20ms step:3672/6250 train_loss:1.8706 aux_loss:0.0154 train_time:6826870ms step_avg:1864.25ms step:3673/6250 train_loss:1.7923 aux_loss:0.0148 train_time:6829176ms step_avg:1864.37ms step:3674/6250 train_loss:1.9156 aux_loss:0.0148 train_time:6830556ms step_avg:1864.23ms step:3675/6250 train_loss:1.9287 aux_loss:0.0151 train_time:6831900ms step_avg:1864.09ms step:3676/6250 train_loss:1.9236 aux_loss:0.0150 train_time:6833153ms step_avg:1863.93ms step:3677/6250 train_loss:1.9143 aux_loss:0.0148 train_time:6834490ms step_avg:1863.78ms step:3678/6250 train_loss:1.8799 aux_loss:0.0150 train_time:6835834ms step_avg:1863.64ms step:3679/6250 train_loss:1.8648 aux_loss:0.0149 train_time:6837148ms step_avg:1863.49ms step:3680/6250 train_loss:1.9220 aux_loss:0.0150 train_time:6838524ms step_avg:1863.36ms step:3681/6250 train_loss:1.9289 aux_loss:0.0147 train_time:6839896ms step_avg:1863.22ms step:3682/6250 train_loss:1.8657 aux_loss:0.0152 train_time:6841232ms step_avg:1863.08ms step:3683/6250 train_loss:1.9060 aux_loss:0.0156 train_time:6842649ms step_avg:1862.96ms step:3684/6250 train_loss:1.8403 aux_loss:0.0148 train_time:6843960ms step_avg:1862.81ms step:3685/6250 train_loss:1.9208 aux_loss:0.0147 train_time:6845343ms step_avg:1862.68ms step:3686/6250 train_loss:1.8195 aux_loss:0.0146 train_time:6847419ms step_avg:1862.74ms step:3687/6250 train_loss:1.8885 aux_loss:0.0145 train_time:6848857ms step_avg:1862.62ms step:3688/6250 train_loss:1.9091 aux_loss:0.0147 train_time:6850294ms step_avg:1862.51ms step:3689/6250 train_loss:1.8436 aux_loss:0.0144 train_time:6851705ms step_avg:1862.38ms step:3690/6250 train_loss:1.7839 aux_loss:0.0147 train_time:6853101ms step_avg:1862.26ms step:3691/6250 train_loss:1.7826 aux_loss:0.0149 train_time:6854415ms step_avg:1862.11ms step:3692/6250 train_loss:1.9090 aux_loss:0.0149 train_time:6855691ms step_avg:1861.95ms step:3693/6250 train_loss:1.8945 aux_loss:0.0149 train_time:6857090ms step_avg:1861.82ms step:3694/6250 train_loss:1.8595 aux_loss:0.0149 train_time:6858748ms step_avg:1861.77ms step:3695/6250 train_loss:1.8102 aux_loss:0.0151 train_time:6860263ms step_avg:1861.67ms step:3696/6250 train_loss:1.8653 aux_loss:0.0149 train_time:6861584ms step_avg:1861.53ms step:3697/6250 train_loss:1.9801 aux_loss:0.0150 train_time:6862950ms step_avg:1861.39ms step:3698/6250 train_loss:1.8972 aux_loss:0.0147 train_time:6864255ms step_avg:1861.24ms step:3699/6250 train_loss:1.9381 aux_loss:0.0148 train_time:6865685ms step_avg:1861.12ms step:3700/6250 train_loss:1.8690 aux_loss:0.0145 train_time:6867026ms step_avg:1860.98ms step:3701/6250 train_loss:1.8190 aux_loss:0.0145 train_time:6868393ms step_avg:1860.85ms step:3702/6250 train_loss:1.8705 aux_loss:0.0148 train_time:6869728ms step_avg:1860.71ms step:3703/6250 train_loss:1.7968 aux_loss:0.0147 train_time:6871076ms step_avg:1860.57ms step:3704/6250 train_loss:1.8337 aux_loss:0.0148 train_time:6872466ms step_avg:1860.44ms step:3705/6250 train_loss:1.9200 aux_loss:0.0150 train_time:6873701ms step_avg:1860.27ms step:3706/6250 train_loss:1.8526 aux_loss:0.0152 train_time:6875224ms step_avg:1860.18ms step:3707/6250 train_loss:1.8235 aux_loss:0.0148 train_time:6876536ms step_avg:1860.03ms step:3708/6250 train_loss:1.8698 aux_loss:0.0148 train_time:6878469ms step_avg:1860.05ms step:3709/6250 train_loss:1.8730 aux_loss:0.0148 train_time:6879815ms step_avg:1859.91ms step:3710/6250 train_loss:1.9103 aux_loss:0.0150 train_time:6881109ms step_avg:1859.76ms step:3711/6250 train_loss:1.8862 aux_loss:0.0147 train_time:6882423ms step_avg:1859.61ms step:3712/6250 train_loss:1.8793 aux_loss:0.0147 train_time:6883675ms step_avg:1859.45ms step:3713/6250 train_loss:1.8981 aux_loss:0.0150 train_time:6884957ms step_avg:1859.29ms step:3714/6250 train_loss:1.8917 aux_loss:0.0143 train_time:6886423ms step_avg:1859.19ms step:3715/6250 train_loss:1.8902 aux_loss:0.0142 train_time:6889441ms step_avg:1859.50ms step:3716/6250 train_loss:1.9625 aux_loss:0.0148 train_time:6890747ms step_avg:1859.35ms step:3717/6250 train_loss:1.9083 aux_loss:0.0153 train_time:6892055ms step_avg:1859.20ms step:3718/6250 train_loss:1.8613 aux_loss:0.0149 train_time:6893561ms step_avg:1859.10ms step:3719/6250 train_loss:1.8735 aux_loss:0.0148 train_time:6897105ms step_avg:1859.56ms step:3720/6250 train_loss:1.9235 aux_loss:0.0152 train_time:6898487ms step_avg:1859.43ms step:3721/6250 train_loss:1.8440 aux_loss:0.0150 train_time:6899707ms step_avg:1859.26ms step:3722/6250 train_loss:1.9224 aux_loss:0.0147 train_time:6901487ms step_avg:1859.24ms step:3723/6250 train_loss:1.8503 aux_loss:0.0147 train_time:6902777ms step_avg:1859.08ms step:3724/6250 train_loss:1.8478 aux_loss:0.0148 train_time:6904150ms step_avg:1858.95ms step:3725/6250 train_loss:1.9449 aux_loss:0.0150 train_time:6905445ms step_avg:1858.80ms step:3726/6250 train_loss:1.8325 aux_loss:0.0146 train_time:6906716ms step_avg:1858.64ms step:3727/6250 train_loss:1.9410 aux_loss:0.0145 train_time:6908093ms step_avg:1858.51ms step:3728/6250 train_loss:1.8452 aux_loss:0.0145 train_time:6909417ms step_avg:1858.37ms step:3729/6250 train_loss:1.8363 aux_loss:0.0148 train_time:6910802ms step_avg:1858.24ms step:3730/6250 train_loss:1.8745 aux_loss:0.0146 train_time:6912102ms step_avg:1858.09ms step:3731/6250 train_loss:1.8715 aux_loss:0.0145 train_time:6916804ms step_avg:1858.86ms step:3732/6250 train_loss:1.9995 aux_loss:0.0150 train_time:6918088ms step_avg:1858.70ms step:3733/6250 train_loss:1.8414 aux_loss:0.0150 train_time:6921000ms step_avg:1858.98ms step:3734/6250 train_loss:1.9295 aux_loss:0.0149 train_time:6922665ms step_avg:1858.93ms step:3735/6250 train_loss:1.9323 aux_loss:0.0149 train_time:6925033ms step_avg:1859.07ms step:3736/6250 train_loss:1.8863 aux_loss:0.0149 train_time:6928745ms step_avg:1859.57ms step:3737/6250 train_loss:1.8799 aux_loss:0.0148 train_time:6930376ms step_avg:1859.51ms step:3738/6250 train_loss:1.9310 aux_loss:0.0151 train_time:6933154ms step_avg:1859.75ms step:3739/6250 train_loss:1.8245 aux_loss:0.0148 train_time:6935042ms step_avg:1859.76ms step:3740/6250 train_loss:1.8698 aux_loss:0.0147 train_time:6936902ms step_avg:1859.76ms step:3741/6250 train_loss:1.8898 aux_loss:0.0146 train_time:6939375ms step_avg:1859.92ms step:3742/6250 train_loss:1.8630 aux_loss:0.0149 train_time:6941939ms step_avg:1860.11ms step:3743/6250 train_loss:1.8229 aux_loss:0.0146 train_time:6945236ms step_avg:1860.50ms step:3744/6250 train_loss:1.9113 aux_loss:0.0144 train_time:6947047ms step_avg:1860.48ms step:3745/6250 train_loss:1.8783 aux_loss:0.0142 train_time:6948414ms step_avg:1860.35ms step:3746/6250 train_loss:1.9011 aux_loss:0.0149 train_time:6949790ms step_avg:1860.22ms step:3747/6250 train_loss:1.8969 aux_loss:0.0147 train_time:6951356ms step_avg:1860.14ms step:3748/6250 train_loss:1.8598 aux_loss:0.0145 train_time:6952668ms step_avg:1860.00ms step:3749/6250 train_loss:1.7974 aux_loss:0.0149 train_time:6954013ms step_avg:1859.86ms step:3750/6250 train_loss:1.9333 aux_loss:0.0150 train_time:6955471ms step_avg:1859.75ms step:3751/6250 train_loss:1.9093 aux_loss:0.0146 train_time:6956879ms step_avg:1859.63ms step:3752/6250 train_loss:1.8261 aux_loss:0.0142 train_time:6962272ms step_avg:1860.58ms step:3753/6250 train_loss:1.8679 aux_loss:0.0143 train_time:6963631ms step_avg:1860.44ms step:3754/6250 train_loss:1.8213 aux_loss:0.0142 train_time:6964890ms step_avg:1860.28ms step:3755/6250 train_loss:1.8958 aux_loss:0.0143 train_time:6966215ms step_avg:1860.14ms step:3756/6250 train_loss:2.0004 aux_loss:0.0146 train_time:6967625ms step_avg:1860.02ms step:3757/6250 train_loss:1.8941 aux_loss:0.0141 train_time:6969003ms step_avg:1859.89ms step:3758/6250 train_loss:1.8738 aux_loss:0.0145 train_time:6970185ms step_avg:1859.71ms step:3759/6250 train_loss:1.8016 aux_loss:0.0144 train_time:6971571ms step_avg:1859.58ms step:3760/6250 train_loss:1.8815 aux_loss:0.0141 train_time:6972928ms step_avg:1859.45ms step:3761/6250 train_loss:1.8322 aux_loss:0.0139 train_time:6974264ms step_avg:1859.31ms step:3762/6250 train_loss:1.9060 aux_loss:0.0144 train_time:6975634ms step_avg:1859.18ms step:3763/6250 train_loss:1.9787 aux_loss:0.0144 train_time:6976917ms step_avg:1859.02ms step:3764/6250 train_loss:1.8855 aux_loss:0.0142 train_time:6978293ms step_avg:1858.90ms step:3765/6250 train_loss:1.8274 aux_loss:0.0146 train_time:6979583ms step_avg:1858.74ms step:3766/6250 train_loss:1.8497 aux_loss:0.0144 train_time:6980943ms step_avg:1858.61ms step:3767/6250 train_loss:1.8365 aux_loss:0.0144 train_time:6982365ms step_avg:1858.49ms step:3768/6250 train_loss:1.8999 aux_loss:0.0144 train_time:6984465ms step_avg:1858.56ms step:3769/6250 train_loss:1.9340 aux_loss:0.0142 train_time:6988487ms step_avg:1859.13ms step:3770/6250 train_loss:1.9936 aux_loss:0.0146 train_time:6990156ms step_avg:1859.08ms step:3771/6250 train_loss:1.8339 aux_loss:0.0145 train_time:6991442ms step_avg:1858.93ms step:3772/6250 train_loss:1.8706 aux_loss:0.0146 train_time:6993028ms step_avg:1858.86ms step:3773/6250 train_loss:1.9197 aux_loss:0.0148 train_time:6995666ms step_avg:1859.07ms step:3774/6250 train_loss:1.8744 aux_loss:0.0146 train_time:6997731ms step_avg:1859.12ms step:3775/6250 train_loss:1.8608 aux_loss:0.0146 train_time:6999166ms step_avg:1859.01ms step:3776/6250 train_loss:1.8453 aux_loss:0.0144 train_time:7002661ms step_avg:1859.44ms step:3777/6250 train_loss:1.8477 aux_loss:0.0143 train_time:7006854ms step_avg:1860.06ms step:3778/6250 train_loss:1.8913 aux_loss:0.0149 train_time:7009036ms step_avg:1860.15ms step:3779/6250 train_loss:1.9374 aux_loss:0.0146 train_time:7011497ms step_avg:1860.31ms step:3780/6250 train_loss:1.8406 aux_loss:0.0141 train_time:7012841ms step_avg:1860.17ms step:3781/6250 train_loss:1.8238 aux_loss:0.0141 train_time:7014246ms step_avg:1860.05ms step:3782/6250 train_loss:1.9543 aux_loss:0.0144 train_time:7015579ms step_avg:1859.91ms step:3783/6250 train_loss:1.8000 aux_loss:0.0143 train_time:7016872ms step_avg:1859.76ms step:3784/6250 train_loss:1.8101 aux_loss:0.0147 train_time:7019167ms step_avg:1859.87ms step:3785/6250 train_loss:1.9275 aux_loss:0.0148 train_time:7020469ms step_avg:1859.73ms step:3786/6250 train_loss:1.8773 aux_loss:0.0143 train_time:7023407ms step_avg:1860.01ms step:3787/6250 train_loss:1.8366 aux_loss:0.0145 train_time:7026681ms step_avg:1860.39ms step:3788/6250 train_loss:1.8825 aux_loss:0.0142 train_time:7028653ms step_avg:1860.42ms step:3789/6250 train_loss:1.8875 aux_loss:0.0140 train_time:7030899ms step_avg:1860.52ms step:3790/6250 train_loss:1.8433 aux_loss:0.0141 train_time:7033518ms step_avg:1860.72ms step:3791/6250 train_loss:1.8939 aux_loss:0.0145 train_time:7034897ms step_avg:1860.59ms step:3792/6250 train_loss:1.9101 aux_loss:0.0142 train_time:7036252ms step_avg:1860.46ms step:3793/6250 train_loss:1.8805 aux_loss:0.0143 train_time:7038396ms step_avg:1860.53ms step:3794/6250 train_loss:1.8648 aux_loss:0.0145 train_time:7040765ms step_avg:1860.67ms step:3795/6250 train_loss:1.8752 aux_loss:0.0143 train_time:7043173ms step_avg:1860.81ms step:3796/6250 train_loss:1.9190 aux_loss:0.0142 train_time:7044585ms step_avg:1860.69ms step:3797/6250 train_loss:1.9556 aux_loss:0.0143 train_time:7045913ms step_avg:1860.55ms step:3798/6250 train_loss:1.8691 aux_loss:0.0145 train_time:7047303ms step_avg:1860.43ms step:3799/6250 train_loss:1.9158 aux_loss:0.0143 train_time:7049091ms step_avg:1860.41ms step:3800/6250 train_loss:1.8242 aux_loss:0.0142 train_time:7050446ms step_avg:1860.28ms step:3801/6250 train_loss:1.9025 aux_loss:0.0144 train_time:7051786ms step_avg:1860.14ms step:3802/6250 train_loss:1.9208 aux_loss:0.0141 train_time:7053099ms step_avg:1859.99ms step:3803/6250 train_loss:1.8653 aux_loss:0.0141 train_time:7054630ms step_avg:1859.91ms step:3804/6250 train_loss:1.7898 aux_loss:0.0141 train_time:7055975ms step_avg:1859.77ms step:3805/6250 train_loss:1.8131 aux_loss:0.0143 train_time:7057330ms step_avg:1859.64ms step:3806/6250 train_loss:1.9142 aux_loss:0.0142 train_time:7059009ms step_avg:1859.59ms step:3807/6250 train_loss:1.8747 aux_loss:0.0145 train_time:7060439ms step_avg:1859.48ms step:3808/6250 train_loss:1.8918 aux_loss:0.0143 train_time:7061787ms step_avg:1859.34ms step:3809/6250 train_loss:1.8759 aux_loss:0.0143 train_time:7063156ms step_avg:1859.21ms step:3810/6250 train_loss:1.8961 aux_loss:0.0145 train_time:7064658ms step_avg:1859.12ms step:3811/6250 train_loss:1.9462 aux_loss:0.0143 train_time:7066012ms step_avg:1858.99ms step:3812/6250 train_loss:1.9039 aux_loss:0.0144 train_time:7067441ms step_avg:1858.87ms step:3813/6250 train_loss:1.8722 aux_loss:0.0146 train_time:7068728ms step_avg:1858.72ms step:3814/6250 train_loss:1.8511 aux_loss:0.0143 train_time:7070063ms step_avg:1858.59ms step:3815/6250 train_loss:1.8448 aux_loss:0.0142 train_time:7071360ms step_avg:1858.44ms step:3816/6250 train_loss:1.8076 aux_loss:0.0142 train_time:7072646ms step_avg:1858.29ms step:3817/6250 train_loss:1.9190 aux_loss:0.0146 train_time:7073995ms step_avg:1858.15ms step:3818/6250 train_loss:1.8153 aux_loss:0.0143 train_time:7075284ms step_avg:1858.01ms step:3819/6250 train_loss:1.8641 aux_loss:0.0141 train_time:7076649ms step_avg:1857.88ms step:3820/6250 train_loss:1.8207 aux_loss:0.0142 train_time:7077968ms step_avg:1857.73ms step:3821/6250 train_loss:1.8841 aux_loss:0.0142 train_time:7079344ms step_avg:1857.61ms step:3822/6250 train_loss:1.8976 aux_loss:0.0144 train_time:7080698ms step_avg:1857.48ms step:3823/6250 train_loss:1.8160 aux_loss:0.0144 train_time:7081963ms step_avg:1857.32ms step:3824/6250 train_loss:1.8975 aux_loss:0.0146 train_time:7083568ms step_avg:1857.25ms step:3825/6250 train_loss:1.8674 aux_loss:0.0141 train_time:7085035ms step_avg:1857.15ms step:3826/6250 train_loss:1.8597 aux_loss:0.0142 train_time:7086393ms step_avg:1857.02ms step:3827/6250 train_loss:1.9372 aux_loss:0.0142 train_time:7087699ms step_avg:1856.88ms step:3828/6250 train_loss:1.8137 aux_loss:0.0140 train_time:7089653ms step_avg:1856.90ms step:3829/6250 train_loss:1.7945 aux_loss:0.0141 train_time:7090934ms step_avg:1856.75ms step:3830/6250 train_loss:1.9066 aux_loss:0.0144 train_time:7092274ms step_avg:1856.62ms step:3831/6250 train_loss:1.8668 aux_loss:0.0142 train_time:7093555ms step_avg:1856.47ms step:3832/6250 train_loss:1.8939 aux_loss:0.0142 train_time:7094997ms step_avg:1856.36ms step:3833/6250 train_loss:1.7423 aux_loss:0.0143 train_time:7096368ms step_avg:1856.23ms step:3834/6250 train_loss:1.9456 aux_loss:0.0141 train_time:7097695ms step_avg:1856.09ms step:3835/6250 train_loss:1.8763 aux_loss:0.0138 train_time:7098887ms step_avg:1855.92ms step:3836/6250 train_loss:1.8297 aux_loss:0.0140 train_time:7100205ms step_avg:1855.78ms step:3837/6250 train_loss:1.8934 aux_loss:0.0139 train_time:7101585ms step_avg:1855.65ms step:3838/6250 train_loss:1.8662 aux_loss:0.0144 train_time:7103231ms step_avg:1855.60ms step:3839/6250 train_loss:1.9412 aux_loss:0.0146 train_time:7104477ms step_avg:1855.44ms step:3840/6250 train_loss:1.7890 aux_loss:0.0140 train_time:7105794ms step_avg:1855.30ms step:3841/6250 train_loss:1.8314 aux_loss:0.0139 train_time:7107310ms step_avg:1855.21ms step:3842/6250 train_loss:1.9507 aux_loss:0.0144 train_time:7108559ms step_avg:1855.05ms step:3843/6250 train_loss:1.9003 aux_loss:0.0145 train_time:7109907ms step_avg:1854.92ms step:3844/6250 train_loss:1.9176 aux_loss:0.0143 train_time:7111136ms step_avg:1854.76ms step:3845/6250 train_loss:1.9212 aux_loss:0.0142 train_time:7112487ms step_avg:1854.63ms step:3846/6250 train_loss:1.8739 aux_loss:0.0145 train_time:7113815ms step_avg:1854.49ms step:3847/6250 train_loss:1.8709 aux_loss:0.0144 train_time:7115092ms step_avg:1854.34ms step:3848/6250 train_loss:1.8181 aux_loss:0.0143 train_time:7116426ms step_avg:1854.20ms step:3849/6250 train_loss:1.8826 aux_loss:0.0142 train_time:7118146ms step_avg:1854.17ms step:3850/6250 train_loss:1.9736 aux_loss:0.0139 train_time:7119492ms step_avg:1854.03ms step:3851/6250 train_loss:1.9472 aux_loss:0.0143 train_time:7120902ms step_avg:1853.92ms step:3852/6250 train_loss:1.9319 aux_loss:0.0145 train_time:7122220ms step_avg:1853.78ms step:3853/6250 train_loss:1.9771 aux_loss:0.0140 train_time:7123737ms step_avg:1853.69ms step:3854/6250 train_loss:1.8992 aux_loss:0.0144 train_time:7125033ms step_avg:1853.55ms step:3855/6250 train_loss:1.9027 aux_loss:0.0143 train_time:7126377ms step_avg:1853.41ms step:3856/6250 train_loss:1.8593 aux_loss:0.0141 train_time:7128021ms step_avg:1853.36ms step:3857/6250 train_loss:1.7984 aux_loss:0.0140 train_time:7129679ms step_avg:1853.31ms step:3858/6250 train_loss:1.8913 aux_loss:0.0137 train_time:7131072ms step_avg:1853.19ms step:3859/6250 train_loss:1.8645 aux_loss:0.0140 train_time:7132624ms step_avg:1853.11ms step:3860/6250 train_loss:1.8873 aux_loss:0.0146 train_time:7134010ms step_avg:1852.99ms step:3861/6250 train_loss:1.9195 aux_loss:0.0144 train_time:7135447ms step_avg:1852.88ms step:3862/6250 train_loss:1.8275 aux_loss:0.0139 train_time:7136731ms step_avg:1852.73ms step:3863/6250 train_loss:1.8695 aux_loss:0.0139 train_time:7138055ms step_avg:1852.60ms step:3864/6250 train_loss:1.7775 aux_loss:0.0142 train_time:7139356ms step_avg:1852.45ms step:3865/6250 train_loss:1.8388 aux_loss:0.0143 train_time:7140704ms step_avg:1852.32ms step:3866/6250 train_loss:1.8999 aux_loss:0.0141 train_time:7141996ms step_avg:1852.18ms step:3867/6250 train_loss:1.8298 aux_loss:0.0138 train_time:7143553ms step_avg:1852.10ms step:3868/6250 train_loss:1.8223 aux_loss:0.0139 train_time:7145220ms step_avg:1852.05ms step:3869/6250 train_loss:1.9232 aux_loss:0.0139 train_time:7146607ms step_avg:1851.93ms step:3870/6250 train_loss:1.7679 aux_loss:0.0143 train_time:7148130ms step_avg:1851.85ms step:3871/6250 train_loss:1.9240 aux_loss:0.0143 train_time:7149624ms step_avg:1851.75ms step:3872/6250 train_loss:1.8406 aux_loss:0.0139 train_time:7150875ms step_avg:1851.60ms step:3873/6250 train_loss:1.8902 aux_loss:0.0142 train_time:7152230ms step_avg:1851.47ms step:3874/6250 train_loss:1.8331 aux_loss:0.0141 train_time:7153537ms step_avg:1851.33ms step:3875/6250 train_loss:1.8578 aux_loss:0.0139 train_time:7154871ms step_avg:1851.20ms step:3876/6250 train_loss:1.8297 aux_loss:0.0142 train_time:7156167ms step_avg:1851.05ms step:3877/6250 train_loss:1.8791 aux_loss:0.0145 train_time:7157879ms step_avg:1851.02ms step:3878/6250 train_loss:1.9257 aux_loss:0.0145 train_time:7159367ms step_avg:1850.92ms step:3879/6250 train_loss:1.8717 aux_loss:0.0136 train_time:7161322ms step_avg:1850.95ms step:3880/6250 train_loss:1.8228 aux_loss:0.0136 train_time:7162624ms step_avg:1850.81ms step:3881/6250 train_loss:1.9421 aux_loss:0.0139 train_time:7164372ms step_avg:1850.78ms step:3882/6250 train_loss:1.9844 aux_loss:0.0142 train_time:7166218ms step_avg:1850.78ms step:3883/6250 train_loss:1.8322 aux_loss:0.0138 train_time:7167387ms step_avg:1850.60ms step:3884/6250 train_loss:1.8326 aux_loss:0.0137 train_time:7169365ms step_avg:1850.64ms step:3885/6250 train_loss:1.8398 aux_loss:0.0139 train_time:7171811ms step_avg:1850.79ms step:3886/6250 train_loss:1.9337 aux_loss:0.0145 train_time:7173794ms step_avg:1850.82ms step:3887/6250 train_loss:1.9059 aux_loss:0.0143 train_time:7176104ms step_avg:1850.94ms step:3888/6250 train_loss:1.8910 aux_loss:0.0140 train_time:7177543ms step_avg:1850.84ms step:3889/6250 train_loss:1.9445 aux_loss:0.0141 train_time:7179334ms step_avg:1850.82ms step:3890/6250 train_loss:1.9128 aux_loss:0.0146 train_time:7180724ms step_avg:1850.70ms step:3891/6250 train_loss:1.8625 aux_loss:0.0143 train_time:7181977ms step_avg:1850.55ms step:3892/6250 train_loss:1.8388 aux_loss:0.0137 train_time:7183440ms step_avg:1850.45ms step:3893/6250 train_loss:1.8497 aux_loss:0.0138 train_time:7185020ms step_avg:1850.38ms step:3894/6250 train_loss:1.8407 aux_loss:0.0142 train_time:7186643ms step_avg:1850.32ms step:3895/6250 train_loss:1.8279 aux_loss:0.0140 train_time:7188085ms step_avg:1850.21ms step:3896/6250 train_loss:1.8053 aux_loss:0.0137 train_time:7189471ms step_avg:1850.10ms step:3897/6250 train_loss:1.8263 aux_loss:0.0141 train_time:7190758ms step_avg:1849.95ms step:3898/6250 train_loss:1.8426 aux_loss:0.0142 train_time:7192016ms step_avg:1849.80ms step:3899/6250 train_loss:1.9926 aux_loss:0.0140 train_time:7193636ms step_avg:1849.74ms step:3900/6250 train_loss:1.9386 aux_loss:0.0140 train_time:7194982ms step_avg:1849.61ms step:3901/6250 train_loss:1.8897 aux_loss:0.0139 train_time:7196214ms step_avg:1849.45ms step:3902/6250 train_loss:1.9125 aux_loss:0.0141 train_time:7197580ms step_avg:1849.33ms step:3903/6250 train_loss:1.8785 aux_loss:0.0138 train_time:7198904ms step_avg:1849.19ms step:3904/6250 train_loss:1.8669 aux_loss:0.0139 train_time:7200242ms step_avg:1849.06ms step:3905/6250 train_loss:1.9241 aux_loss:0.0138 train_time:7201529ms step_avg:1848.92ms step:3906/6250 train_loss:1.8660 aux_loss:0.0143 train_time:7202840ms step_avg:1848.78ms step:3907/6250 train_loss:1.8458 aux_loss:0.0142 train_time:7204164ms step_avg:1848.64ms step:3908/6250 train_loss:1.9198 aux_loss:0.0142 train_time:7205505ms step_avg:1848.51ms step:3909/6250 train_loss:1.8328 aux_loss:0.0140 train_time:7206881ms step_avg:1848.39ms step:3910/6250 train_loss:1.8925 aux_loss:0.0143 train_time:7208223ms step_avg:1848.26ms step:3911/6250 train_loss:1.8080 aux_loss:0.0140 train_time:7209594ms step_avg:1848.14ms step:3912/6250 train_loss:1.8076 aux_loss:0.0142 train_time:7211202ms step_avg:1848.08ms step:3913/6250 train_loss:1.7963 aux_loss:0.0142 train_time:7212640ms step_avg:1847.97ms step:3914/6250 train_loss:1.8629 aux_loss:0.0141 train_time:7214667ms step_avg:1848.02ms step:3915/6250 train_loss:1.8939 aux_loss:0.0143 train_time:7216735ms step_avg:1848.08ms step:3916/6250 train_loss:1.8862 aux_loss:0.0143 train_time:7218574ms step_avg:1848.07ms step:3917/6250 train_loss:1.8903 aux_loss:0.0148 train_time:7219904ms step_avg:1847.94ms step:3918/6250 train_loss:1.8905 aux_loss:0.0143 train_time:7221281ms step_avg:1847.82ms step:3919/6250 train_loss:1.7682 aux_loss:0.0141 train_time:7222579ms step_avg:1847.68ms step:3920/6250 train_loss:1.9046 aux_loss:0.0142 train_time:7223950ms step_avg:1847.56ms step:3921/6250 train_loss:1.8479 aux_loss:0.0140 train_time:7225586ms step_avg:1847.50ms step:3922/6250 train_loss:1.8864 aux_loss:0.0148 train_time:7227187ms step_avg:1847.44ms step:3923/6250 train_loss:1.8910 aux_loss:0.0142 train_time:7228625ms step_avg:1847.34ms step:3924/6250 train_loss:1.9001 aux_loss:0.0140 train_time:7229958ms step_avg:1847.20ms step:3925/6250 train_loss:1.7758 aux_loss:0.0140 train_time:7231472ms step_avg:1847.12ms step:3926/6250 train_loss:1.8969 aux_loss:0.0142 train_time:7232832ms step_avg:1846.99ms step:3927/6250 train_loss:1.8167 aux_loss:0.0142 train_time:7234154ms step_avg:1846.86ms step:3928/6250 train_loss:1.8812 aux_loss:0.0144 train_time:7235565ms step_avg:1846.75ms step:3929/6250 train_loss:1.9522 aux_loss:0.0147 train_time:7236986ms step_avg:1846.64ms step:3930/6250 train_loss:1.9142 aux_loss:0.0141 train_time:7238351ms step_avg:1846.52ms step:3931/6250 train_loss:1.9087 aux_loss:0.0140 train_time:7239922ms step_avg:1846.45ms step:3932/6250 train_loss:1.9463 aux_loss:0.0139 train_time:7241247ms step_avg:1846.31ms step:3933/6250 train_loss:1.9522 aux_loss:0.0141 train_time:7242559ms step_avg:1846.18ms step:3934/6250 train_loss:1.8116 aux_loss:0.0139 train_time:7243910ms step_avg:1846.05ms step:3935/6250 train_loss:1.8539 aux_loss:0.0143 train_time:7245210ms step_avg:1845.91ms step:3936/6250 train_loss:1.9177 aux_loss:0.0143 train_time:7246498ms step_avg:1845.77ms step:3937/6250 train_loss:1.8711 aux_loss:0.0143 train_time:7247838ms step_avg:1845.64ms step:3938/6250 train_loss:1.9100 aux_loss:0.0138 train_time:7249561ms step_avg:1845.61ms step:3939/6250 train_loss:1.8377 aux_loss:0.0137 train_time:7251124ms step_avg:1845.54ms step:3940/6250 train_loss:1.8742 aux_loss:0.0139 train_time:7252508ms step_avg:1845.42ms step:3941/6250 train_loss:1.9452 aux_loss:0.0141 train_time:7253872ms step_avg:1845.30ms step:3942/6250 train_loss:1.8704 aux_loss:0.0138 train_time:7255377ms step_avg:1845.21ms step:3943/6250 train_loss:1.8474 aux_loss:0.0141 train_time:7256756ms step_avg:1845.09ms step:3944/6250 train_loss:1.8951 aux_loss:0.0143 train_time:7258352ms step_avg:1845.03ms step:3945/6250 train_loss:1.9083 aux_loss:0.0143 train_time:7259737ms step_avg:1844.91ms step:3946/6250 train_loss:1.8754 aux_loss:0.0138 train_time:7261074ms step_avg:1844.79ms step:3947/6250 train_loss:1.8684 aux_loss:0.0140 train_time:7262743ms step_avg:1844.74ms step:3948/6250 train_loss:1.9023 aux_loss:0.0139 train_time:7264130ms step_avg:1844.62ms step:3949/6250 train_loss:1.8281 aux_loss:0.0142 train_time:7265535ms step_avg:1844.51ms step:3950/6250 train_loss:1.9136 aux_loss:0.0141 train_time:7266894ms step_avg:1844.39ms step:3951/6250 train_loss:1.8343 aux_loss:0.0141 train_time:7268329ms step_avg:1844.29ms step:3952/6250 train_loss:1.8468 aux_loss:0.0139 train_time:7269617ms step_avg:1844.14ms step:3953/6250 train_loss:1.8908 aux_loss:0.0138 train_time:7270986ms step_avg:1844.02ms step:3954/6250 train_loss:1.9070 aux_loss:0.0141 train_time:7272367ms step_avg:1843.91ms step:3955/6250 train_loss:1.8735 aux_loss:0.0138 train_time:7273932ms step_avg:1843.84ms step:3956/6250 train_loss:1.8731 aux_loss:0.0136 train_time:7275342ms step_avg:1843.73ms step:3957/6250 train_loss:1.8732 aux_loss:0.0141 train_time:7277055ms step_avg:1843.69ms step:3958/6250 train_loss:1.7581 aux_loss:0.0138 train_time:7278503ms step_avg:1843.59ms step:3959/6250 train_loss:1.8382 aux_loss:0.0139 train_time:7279879ms step_avg:1843.47ms step:3960/6250 train_loss:1.9196 aux_loss:0.0136 train_time:7281228ms step_avg:1843.35ms step:3961/6250 train_loss:1.8401 aux_loss:0.0139 train_time:7283062ms step_avg:1843.35ms step:3962/6250 train_loss:1.8151 aux_loss:0.0136 train_time:7284712ms step_avg:1843.30ms step:3963/6250 train_loss:1.8442 aux_loss:0.0133 train_time:7286322ms step_avg:1843.24ms step:3964/6250 train_loss:1.8842 aux_loss:0.0139 train_time:7287888ms step_avg:1843.17ms step:3965/6250 train_loss:1.8406 aux_loss:0.0142 train_time:7289382ms step_avg:1843.08ms step:3966/6250 train_loss:1.8235 aux_loss:0.0137 train_time:7290988ms step_avg:1843.02ms step:3967/6250 train_loss:1.8392 aux_loss:0.0135 train_time:7292592ms step_avg:1842.96ms step:3968/6250 train_loss:1.8950 aux_loss:0.0138 train_time:7294741ms step_avg:1843.04ms step:3969/6250 train_loss:1.7918 aux_loss:0.0138 train_time:7296384ms step_avg:1842.99ms step:3970/6250 train_loss:1.8938 aux_loss:0.0133 train_time:7297948ms step_avg:1842.92ms step:3971/6250 train_loss:1.8911 aux_loss:0.0135 train_time:7299456ms step_avg:1842.83ms step:3972/6250 train_loss:1.9314 aux_loss:0.0140 train_time:7301064ms step_avg:1842.77ms step:3973/6250 train_loss:1.8760 aux_loss:0.0145 train_time:7303035ms step_avg:1842.80ms step:3974/6250 train_loss:1.9055 aux_loss:0.0138 train_time:7304799ms step_avg:1842.78ms step:3975/6250 train_loss:1.9093 aux_loss:0.0136 train_time:7306579ms step_avg:1842.77ms step:3976/6250 train_loss:1.8817 aux_loss:0.0137 train_time:7308114ms step_avg:1842.69ms step:3977/6250 train_loss:1.9063 aux_loss:0.0136 train_time:7309680ms step_avg:1842.62ms step:3978/6250 train_loss:1.8769 aux_loss:0.0137 train_time:7311206ms step_avg:1842.54ms step:3979/6250 train_loss:1.8097 aux_loss:0.0132 train_time:7312873ms step_avg:1842.50ms step:3980/6250 train_loss:1.9036 aux_loss:0.0138 train_time:7314758ms step_avg:1842.51ms step:3981/6250 train_loss:1.8676 aux_loss:0.0137 train_time:7316787ms step_avg:1842.56ms step:3982/6250 train_loss:1.8495 aux_loss:0.0137 train_time:7318374ms step_avg:1842.49ms step:3983/6250 train_loss:1.8084 aux_loss:0.0136 train_time:7319837ms step_avg:1842.40ms step:3984/6250 train_loss:1.9279 aux_loss:0.0133 train_time:7321404ms step_avg:1842.33ms step:3985/6250 train_loss:1.8338 aux_loss:0.0140 train_time:7323107ms step_avg:1842.29ms step:3986/6250 train_loss:1.8919 aux_loss:0.0142 train_time:7325131ms step_avg:1842.34ms step:3987/6250 train_loss:1.8719 aux_loss:0.0134 train_time:7326699ms step_avg:1842.27ms step:3988/6250 train_loss:1.9127 aux_loss:0.0137 train_time:7328284ms step_avg:1842.20ms step:3989/6250 train_loss:1.8800 aux_loss:0.0140 train_time:7330180ms step_avg:1842.22ms step:3990/6250 train_loss:1.9140 aux_loss:0.0139 train_time:7331772ms step_avg:1842.15ms step:3991/6250 train_loss:1.9641 aux_loss:0.0139 train_time:7333262ms step_avg:1842.07ms step:3992/6250 train_loss:1.8987 aux_loss:0.0138 train_time:7334751ms step_avg:1841.98ms step:3993/6250 train_loss:1.8637 aux_loss:0.0137 train_time:7336282ms step_avg:1841.90ms step:3994/6250 train_loss:1.8495 aux_loss:0.0134 train_time:7338375ms step_avg:1841.96ms step:3995/6250 train_loss:1.8569 aux_loss:0.0137 train_time:7340007ms step_avg:1841.91ms step:3996/6250 train_loss:1.7797 aux_loss:0.0137 train_time:7341640ms step_avg:1841.86ms step:3997/6250 train_loss:1.8481 aux_loss:0.0134 train_time:7343262ms step_avg:1841.80ms step:3998/6250 train_loss:1.8562 aux_loss:0.0134 train_time:7344715ms step_avg:1841.70ms step:3999/6250 train_loss:1.9288 aux_loss:0.0138 train_time:7346348ms step_avg:1841.65ms step:4000/6250 train_loss:1.8924 aux_loss:0.0137 train_time:7347933ms step_avg:1841.59ms step:4000/6250 val_loss:1.8566 val_aux_loss:0.0139 train_time:7347934ms step_avg:1841.59ms step:4001/6250 train_loss:1.8224 aux_loss:0.0138 train_time:7349285ms step_avg:1841.46ms step:4002/6250 train_loss:1.8112 aux_loss:0.0138 train_time:7351033ms step_avg:1841.44ms step:4003/6250 train_loss:1.8693 aux_loss:0.0140 train_time:7352332ms step_avg:1841.31ms step:4004/6250 train_loss:1.8657 aux_loss:0.0133 train_time:7353668ms step_avg:1841.18ms step:4005/6250 train_loss:1.8618 aux_loss:0.0135 train_time:7355053ms step_avg:1841.06ms step:4006/6250 train_loss:1.7685 aux_loss:0.0136 train_time:7356747ms step_avg:1841.03ms step:4007/6250 train_loss:1.9378 aux_loss:0.0138 train_time:7358129ms step_avg:1840.91ms step:4008/6250 train_loss:1.8670 aux_loss:0.0135 train_time:7359538ms step_avg:1840.80ms step:4009/6250 train_loss:1.9423 aux_loss:0.0137 train_time:7360871ms step_avg:1840.68ms step:4010/6250 train_loss:1.8625 aux_loss:0.0137 train_time:7362225ms step_avg:1840.56ms step:4011/6250 train_loss:1.8960 aux_loss:0.0138 train_time:7363604ms step_avg:1840.44ms step:4012/6250 train_loss:1.8508 aux_loss:0.0142 train_time:7365270ms step_avg:1840.40ms step:4013/6250 train_loss:1.9768 aux_loss:0.0141 train_time:7366802ms step_avg:1840.32ms step:4014/6250 train_loss:1.8087 aux_loss:0.0135 train_time:7368097ms step_avg:1840.18ms step:4015/6250 train_loss:1.8219 aux_loss:0.0136 train_time:7369554ms step_avg:1840.09ms step:4016/6250 train_loss:1.8598 aux_loss:0.0136 train_time:7370859ms step_avg:1839.95ms step:4017/6250 train_loss:1.8868 aux_loss:0.0134 train_time:7375982ms step_avg:1840.77ms step:4018/6250 train_loss:1.8395 aux_loss:0.0137 train_time:7378533ms step_avg:1840.95ms step:4019/6250 train_loss:1.8013 aux_loss:0.0136 train_time:7381238ms step_avg:1841.17ms step:4020/6250 train_loss:1.8877 aux_loss:0.0136 train_time:7383592ms step_avg:1841.29ms step:4021/6250 train_loss:1.9301 aux_loss:0.0135 train_time:7386073ms step_avg:1841.45ms step:4022/6250 train_loss:1.8531 aux_loss:0.0135 train_time:7388037ms step_avg:1841.48ms step:4023/6250 train_loss:1.8137 aux_loss:0.0137 train_time:7390524ms step_avg:1841.65ms step:4024/6250 train_loss:1.8497 aux_loss:0.0139 train_time:7392650ms step_avg:1841.72ms step:4025/6250 train_loss:1.8660 aux_loss:0.0138 train_time:7395172ms step_avg:1841.89ms step:4026/6250 train_loss:1.9181 aux_loss:0.0135 train_time:7397721ms step_avg:1842.06ms step:4027/6250 train_loss:1.7956 aux_loss:0.0140 train_time:7399654ms step_avg:1842.08ms step:4028/6250 train_loss:1.7789 aux_loss:0.0137 train_time:7401760ms step_avg:1842.15ms step:4029/6250 train_loss:1.8792 aux_loss:0.0138 train_time:7404269ms step_avg:1842.32ms step:4030/6250 train_loss:1.7785 aux_loss:0.0135 train_time:7406606ms step_avg:1842.44ms step:4031/6250 train_loss:1.9252 aux_loss:0.0137 train_time:7408839ms step_avg:1842.54ms step:4032/6250 train_loss:1.9176 aux_loss:0.0135 train_time:7410629ms step_avg:1842.52ms step:4033/6250 train_loss:1.7452 aux_loss:0.0132 train_time:7412565ms step_avg:1842.55ms step:4034/6250 train_loss:1.7947 aux_loss:0.0133 train_time:7414830ms step_avg:1842.65ms step:4035/6250 train_loss:1.8573 aux_loss:0.0136 train_time:7416518ms step_avg:1842.61ms step:4036/6250 train_loss:1.9012 aux_loss:0.0132 train_time:7418644ms step_avg:1842.68ms step:4037/6250 train_loss:1.9000 aux_loss:0.0137 train_time:7420789ms step_avg:1842.76ms step:4038/6250 train_loss:1.8973 aux_loss:0.0135 train_time:7422832ms step_avg:1842.81ms step:4039/6250 train_loss:1.8154 aux_loss:0.0137 train_time:7425293ms step_avg:1842.96ms step:4040/6250 train_loss:1.8850 aux_loss:0.0134 train_time:7427006ms step_avg:1842.93ms step:4041/6250 train_loss:1.8870 aux_loss:0.0131 train_time:7429494ms step_avg:1843.09ms step:4042/6250 train_loss:1.8644 aux_loss:0.0133 train_time:7431328ms step_avg:1843.09ms step:4043/6250 train_loss:1.8424 aux_loss:0.0133 train_time:7433133ms step_avg:1843.08ms step:4044/6250 train_loss:1.9270 aux_loss:0.0136 train_time:7435241ms step_avg:1843.14ms step:4045/6250 train_loss:1.8052 aux_loss:0.0131 train_time:7437090ms step_avg:1843.15ms step:4046/6250 train_loss:1.8446 aux_loss:0.0133 train_time:7439135ms step_avg:1843.20ms step:4047/6250 train_loss:1.7887 aux_loss:0.0136 train_time:7441198ms step_avg:1843.25ms step:4048/6250 train_loss:1.8593 aux_loss:0.0132 train_time:7443142ms step_avg:1843.27ms step:4049/6250 train_loss:1.8696 aux_loss:0.0133 train_time:7445583ms step_avg:1843.42ms step:4050/6250 train_loss:2.0063 aux_loss:0.0135 train_time:7448456ms step_avg:1843.68ms step:4051/6250 train_loss:1.9070 aux_loss:0.0133 train_time:7450311ms step_avg:1843.68ms step:4052/6250 train_loss:1.9115 aux_loss:0.0134 train_time:7451936ms step_avg:1843.63ms step:4053/6250 train_loss:1.8140 aux_loss:0.0132 train_time:7453771ms step_avg:1843.62ms step:4054/6250 train_loss:1.8835 aux_loss:0.0134 train_time:7455594ms step_avg:1843.62ms step:4055/6250 train_loss:1.8499 aux_loss:0.0131 train_time:7457696ms step_avg:1843.68ms step:4056/6250 train_loss:1.8867 aux_loss:0.0128 train_time:7460090ms step_avg:1843.82ms step:4057/6250 train_loss:1.8894 aux_loss:0.0136 train_time:7462236ms step_avg:1843.89ms step:4058/6250 train_loss:1.7940 aux_loss:0.0132 train_time:7464405ms step_avg:1843.97ms step:4059/6250 train_loss:1.9683 aux_loss:0.0134 train_time:7466764ms step_avg:1844.10ms step:4060/6250 train_loss:1.8772 aux_loss:0.0132 train_time:7469035ms step_avg:1844.21ms step:4061/6250 train_loss:1.8853 aux_loss:0.0133 train_time:7471053ms step_avg:1844.25ms step:4062/6250 train_loss:1.8280 aux_loss:0.0134 train_time:7472893ms step_avg:1844.25ms step:4063/6250 train_loss:1.8301 aux_loss:0.0133 train_time:7474693ms step_avg:1844.24ms step:4064/6250 train_loss:1.8026 aux_loss:0.0133 train_time:7477186ms step_avg:1844.40ms step:4065/6250 train_loss:1.8763 aux_loss:0.0133 train_time:7479755ms step_avg:1844.58ms step:4066/6250 train_loss:1.8013 aux_loss:0.0134 train_time:7481881ms step_avg:1844.65ms step:4067/6250 train_loss:1.7733 aux_loss:0.0133 train_time:7488163ms step_avg:1845.74ms step:4068/6250 train_loss:1.9165 aux_loss:0.0136 train_time:7490914ms step_avg:1845.96ms step:4069/6250 train_loss:1.9152 aux_loss:0.0133 train_time:7493477ms step_avg:1846.14ms step:4070/6250 train_loss:1.8762 aux_loss:0.0133 train_time:7495779ms step_avg:1846.25ms step:4071/6250 train_loss:1.8252 aux_loss:0.0132 train_time:7498072ms step_avg:1846.36ms step:4072/6250 train_loss:1.7913 aux_loss:0.0133 train_time:7501078ms step_avg:1846.65ms step:4073/6250 train_loss:1.8236 aux_loss:0.0135 train_time:7503082ms step_avg:1846.69ms step:4074/6250 train_loss:1.8249 aux_loss:0.0135 train_time:7505731ms step_avg:1846.88ms step:4075/6250 train_loss:1.8594 aux_loss:0.0132 train_time:7508517ms step_avg:1847.11ms step:4076/6250 train_loss:1.8641 aux_loss:0.0131 train_time:7516771ms step_avg:1848.69ms step:4077/6250 train_loss:1.9280 aux_loss:0.0130 train_time:7518789ms step_avg:1848.73ms step:4078/6250 train_loss:1.8159 aux_loss:0.0129 train_time:7520437ms step_avg:1848.68ms step:4079/6250 train_loss:1.9126 aux_loss:0.0135 train_time:7522007ms step_avg:1848.61ms step:4080/6250 train_loss:1.8134 aux_loss:0.0129 train_time:7524671ms step_avg:1848.81ms step:4081/6250 train_loss:1.8596 aux_loss:0.0133 train_time:7527037ms step_avg:1848.94ms step:4082/6250 train_loss:1.8538 aux_loss:0.0135 train_time:7528496ms step_avg:1848.84ms step:4083/6250 train_loss:1.8054 aux_loss:0.0137 train_time:7529799ms step_avg:1848.71ms step:4084/6250 train_loss:1.8388 aux_loss:0.0131 train_time:7531913ms step_avg:1848.78ms step:4085/6250 train_loss:2.0004 aux_loss:0.0136 train_time:7534387ms step_avg:1848.93ms step:4086/6250 train_loss:1.9812 aux_loss:0.0140 train_time:7536992ms step_avg:1849.11ms step:4087/6250 train_loss:1.8615 aux_loss:0.0136 train_time:7539655ms step_avg:1849.31ms step:4088/6250 train_loss:1.8414 aux_loss:0.0131 train_time:7541541ms step_avg:1849.32ms step:4089/6250 train_loss:1.8322 aux_loss:0.0129 train_time:7542822ms step_avg:1849.18ms step:4090/6250 train_loss:1.8902 aux_loss:0.0135 train_time:7551415ms step_avg:1850.84ms step:4091/6250 train_loss:1.8433 aux_loss:0.0137 train_time:7554420ms step_avg:1851.12ms step:4092/6250 train_loss:1.8030 aux_loss:0.0134 train_time:7557021ms step_avg:1851.30ms step:4093/6250 train_loss:1.8922 aux_loss:0.0133 train_time:7559534ms step_avg:1851.47ms step:4094/6250 train_loss:1.8161 aux_loss:0.0133 train_time:7561645ms step_avg:1851.53ms step:4095/6250 train_loss:1.8385 aux_loss:0.0136 train_time:7564031ms step_avg:1851.66ms step:4096/6250 train_loss:1.8131 aux_loss:0.0132 train_time:7567117ms step_avg:1851.96ms step:4097/6250 train_loss:1.8990 aux_loss:0.0132 train_time:7569410ms step_avg:1852.07ms step:4098/6250 train_loss:1.8737 aux_loss:0.0137 train_time:7571993ms step_avg:1852.25ms step:4099/6250 train_loss:1.8738 aux_loss:0.0136 train_time:7574211ms step_avg:1852.34ms step:4100/6250 train_loss:1.9128 aux_loss:0.0138 train_time:7576603ms step_avg:1852.47ms step:4101/6250 train_loss:1.8111 aux_loss:0.0130 train_time:7579002ms step_avg:1852.60ms step:4102/6250 train_loss:1.8364 aux_loss:0.0131 train_time:7581009ms step_avg:1852.64ms step:4103/6250 train_loss:1.8605 aux_loss:0.0137 train_time:7583277ms step_avg:1852.74ms step:4104/6250 train_loss:1.8685 aux_loss:0.0134 train_time:7585428ms step_avg:1852.82ms step:4105/6250 train_loss:1.8729 aux_loss:0.0131 train_time:7587532ms step_avg:1852.88ms step:4106/6250 train_loss:1.7780 aux_loss:0.0128 train_time:7589677ms step_avg:1852.95ms step:4107/6250 train_loss:1.8006 aux_loss:0.0131 train_time:7592049ms step_avg:1853.08ms step:4108/6250 train_loss:1.9572 aux_loss:0.0139 train_time:7593932ms step_avg:1853.08ms step:4109/6250 train_loss:1.7867 aux_loss:0.0134 train_time:7596135ms step_avg:1853.17ms step:4110/6250 train_loss:1.7950 aux_loss:0.0130 train_time:7598412ms step_avg:1853.27ms step:4111/6250 train_loss:1.8815 aux_loss:0.0133 train_time:7600678ms step_avg:1853.37ms step:4112/6250 train_loss:1.8698 aux_loss:0.0136 train_time:7602945ms step_avg:1853.47ms step:4113/6250 train_loss:1.8535 aux_loss:0.0135 train_time:7605313ms step_avg:1853.60ms step:4114/6250 train_loss:1.8476 aux_loss:0.0132 train_time:7607971ms step_avg:1853.79ms step:4115/6250 train_loss:1.7971 aux_loss:0.0133 train_time:7610056ms step_avg:1853.85ms step:4116/6250 train_loss:1.8809 aux_loss:0.0134 train_time:7612769ms step_avg:1854.06ms step:4117/6250 train_loss:1.8732 aux_loss:0.0133 train_time:7615054ms step_avg:1854.16ms step:4118/6250 train_loss:1.8827 aux_loss:0.0133 train_time:7617566ms step_avg:1854.32ms step:4119/6250 train_loss:1.8582 aux_loss:0.0131 train_time:7618973ms step_avg:1854.22ms step:4120/6250 train_loss:1.8720 aux_loss:0.0132 train_time:7620339ms step_avg:1854.10ms step:4121/6250 train_loss:1.7686 aux_loss:0.0132 train_time:7621663ms step_avg:1853.97ms step:4122/6250 train_loss:1.9488 aux_loss:0.0131 train_time:7623109ms step_avg:1853.87ms step:4123/6250 train_loss:1.9121 aux_loss:0.0128 train_time:7625023ms step_avg:1853.88ms step:4124/6250 train_loss:1.8311 aux_loss:0.0129 train_time:7626472ms step_avg:1853.79ms step:4125/6250 train_loss:1.8746 aux_loss:0.0131 train_time:7631267ms step_avg:1854.50ms step:4126/6250 train_loss:1.8565 aux_loss:0.0137 train_time:7634230ms step_avg:1854.77ms step:4127/6250 train_loss:1.8486 aux_loss:0.0136 train_time:7635705ms step_avg:1854.68ms step:4128/6250 train_loss:1.8597 aux_loss:0.0133 train_time:7637119ms step_avg:1854.57ms step:4129/6250 train_loss:1.8343 aux_loss:0.0134 train_time:7639497ms step_avg:1854.70ms step:4130/6250 train_loss:1.8523 aux_loss:0.0136 train_time:7641501ms step_avg:1854.73ms step:4131/6250 train_loss:1.8676 aux_loss:0.0132 train_time:7644686ms step_avg:1855.06ms step:4132/6250 train_loss:1.8249 aux_loss:0.0131 train_time:7647119ms step_avg:1855.20ms step:4133/6250 train_loss:1.8557 aux_loss:0.0133 train_time:7650466ms step_avg:1855.56ms step:4134/6250 train_loss:1.9287 aux_loss:0.0134 train_time:7656151ms step_avg:1856.49ms step:4135/6250 train_loss:1.8092 aux_loss:0.0130 train_time:7657602ms step_avg:1856.39ms step:4136/6250 train_loss:1.8950 aux_loss:0.0129 train_time:7659172ms step_avg:1856.32ms step:4137/6250 train_loss:1.9303 aux_loss:0.0133 train_time:7660884ms step_avg:1856.28ms step:4138/6250 train_loss:1.8619 aux_loss:0.0136 train_time:7662481ms step_avg:1856.22ms step:4139/6250 train_loss:1.8643 aux_loss:0.0130 train_time:7663980ms step_avg:1856.13ms step:4140/6250 train_loss:1.8576 aux_loss:0.0131 train_time:7665571ms step_avg:1856.07ms step:4141/6250 train_loss:1.8002 aux_loss:0.0129 train_time:7667265ms step_avg:1856.03ms step:4142/6250 train_loss:1.7188 aux_loss:0.0127 train_time:7668821ms step_avg:1855.96ms step:4143/6250 train_loss:1.8343 aux_loss:0.0130 train_time:7672473ms step_avg:1856.39ms step:4144/6250 train_loss:1.8491 aux_loss:0.0131 train_time:7674853ms step_avg:1856.52ms step:4145/6250 train_loss:1.8626 aux_loss:0.0131 train_time:7676416ms step_avg:1856.45ms step:4146/6250 train_loss:1.8279 aux_loss:0.0132 train_time:7677699ms step_avg:1856.31ms step:4147/6250 train_loss:1.9053 aux_loss:0.0135 train_time:7679334ms step_avg:1856.26ms step:4148/6250 train_loss:1.8316 aux_loss:0.0132 train_time:7680653ms step_avg:1856.13ms step:4149/6250 train_loss:1.8155 aux_loss:0.0132 train_time:7682257ms step_avg:1856.07ms step:4150/6250 train_loss:1.8696 aux_loss:0.0132 train_time:7684850ms step_avg:1856.24ms step:4151/6250 train_loss:1.9921 aux_loss:0.0133 train_time:7686658ms step_avg:1856.23ms step:4152/6250 train_loss:1.8597 aux_loss:0.0133 train_time:7688268ms step_avg:1856.17ms step:4153/6250 train_loss:1.9085 aux_loss:0.0133 train_time:7690939ms step_avg:1856.37ms step:4154/6250 train_loss:1.7815 aux_loss:0.0130 train_time:7692519ms step_avg:1856.30ms step:4155/6250 train_loss:1.8844 aux_loss:0.0132 train_time:7695902ms step_avg:1856.67ms step:4156/6250 train_loss:1.7826 aux_loss:0.0129 train_time:7698974ms step_avg:1856.96ms step:4157/6250 train_loss:1.8761 aux_loss:0.0127 train_time:7702415ms step_avg:1857.35ms step:4158/6250 train_loss:1.8485 aux_loss:0.0129 train_time:7705417ms step_avg:1857.62ms step:4159/6250 train_loss:1.7761 aux_loss:0.0132 train_time:7707720ms step_avg:1857.73ms step:4160/6250 train_loss:1.8160 aux_loss:0.0130 train_time:7710009ms step_avg:1857.83ms step:4161/6250 train_loss:1.9849 aux_loss:0.0130 train_time:7712294ms step_avg:1857.94ms step:4162/6250 train_loss:1.8159 aux_loss:0.0135 train_time:7715188ms step_avg:1858.19ms step:4163/6250 train_loss:1.8594 aux_loss:0.0134 train_time:7717467ms step_avg:1858.29ms step:4164/6250 train_loss:1.8630 aux_loss:0.0128 train_time:7720127ms step_avg:1858.48ms step:4165/6250 train_loss:1.8156 aux_loss:0.0128 train_time:7722326ms step_avg:1858.56ms step:4166/6250 train_loss:1.8419 aux_loss:0.0130 train_time:7724508ms step_avg:1858.64ms step:4167/6250 train_loss:1.9123 aux_loss:0.0137 train_time:7726757ms step_avg:1858.73ms step:4168/6250 train_loss:1.8183 aux_loss:0.0130 train_time:7728677ms step_avg:1858.75ms step:4169/6250 train_loss:1.8384 aux_loss:0.0132 train_time:7730617ms step_avg:1858.77ms step:4170/6250 train_loss:1.8996 aux_loss:0.0128 train_time:7732451ms step_avg:1858.76ms step:4171/6250 train_loss:1.8964 aux_loss:0.0128 train_time:7734717ms step_avg:1858.86ms step:4172/6250 train_loss:1.7431 aux_loss:0.0125 train_time:7736797ms step_avg:1858.91ms step:4173/6250 train_loss:1.8979 aux_loss:0.0128 train_time:7739265ms step_avg:1859.06ms step:4174/6250 train_loss:1.8527 aux_loss:0.0128 train_time:7741300ms step_avg:1859.10ms step:4175/6250 train_loss:1.8306 aux_loss:0.0126 train_time:7743631ms step_avg:1859.22ms step:4176/6250 train_loss:1.8453 aux_loss:0.0129 train_time:7745893ms step_avg:1859.31ms step:4177/6250 train_loss:1.8361 aux_loss:0.0135 train_time:7748307ms step_avg:1859.44ms step:4178/6250 train_loss:1.8303 aux_loss:0.0131 train_time:7751202ms step_avg:1859.69ms step:4179/6250 train_loss:1.9031 aux_loss:0.0129 train_time:7753784ms step_avg:1859.87ms step:4180/6250 train_loss:1.9013 aux_loss:0.0130 train_time:7756118ms step_avg:1859.98ms step:4181/6250 train_loss:1.8950 aux_loss:0.0133 train_time:7758523ms step_avg:1860.11ms step:4182/6250 train_loss:1.8598 aux_loss:0.0136 train_time:7761095ms step_avg:1860.28ms step:4183/6250 train_loss:1.8261 aux_loss:0.0130 train_time:7763326ms step_avg:1860.37ms step:4184/6250 train_loss:1.8581 aux_loss:0.0127 train_time:7765314ms step_avg:1860.40ms step:4185/6250 train_loss:1.7777 aux_loss:0.0126 train_time:7767756ms step_avg:1860.54ms step:4186/6250 train_loss:1.8278 aux_loss:0.0128 train_time:7769844ms step_avg:1860.59ms step:4187/6250 train_loss:1.8629 aux_loss:0.0126 train_time:7772130ms step_avg:1860.70ms step:4188/6250 train_loss:1.7899 aux_loss:0.0128 train_time:7775016ms step_avg:1860.94ms step:4189/6250 train_loss:1.8189 aux_loss:0.0127 train_time:7777060ms step_avg:1860.99ms step:4190/6250 train_loss:1.8231 aux_loss:0.0130 train_time:7779553ms step_avg:1861.14ms step:4191/6250 train_loss:1.9203 aux_loss:0.0130 train_time:7782422ms step_avg:1861.38ms step:4192/6250 train_loss:1.8463 aux_loss:0.0125 train_time:7784503ms step_avg:1861.43ms step:4193/6250 train_loss:1.8141 aux_loss:0.0126 train_time:7786857ms step_avg:1861.55ms step:4194/6250 train_loss:1.8790 aux_loss:0.0128 train_time:7789446ms step_avg:1861.72ms step:4195/6250 train_loss:1.8431 aux_loss:0.0129 train_time:7791503ms step_avg:1861.77ms step:4196/6250 train_loss:1.8571 aux_loss:0.0128 train_time:7793983ms step_avg:1861.92ms step:4197/6250 train_loss:1.8808 aux_loss:0.0128 train_time:7796069ms step_avg:1861.97ms step:4198/6250 train_loss:1.8245 aux_loss:0.0133 train_time:7798549ms step_avg:1862.12ms step:4199/6250 train_loss:1.8336 aux_loss:0.0129 train_time:7800850ms step_avg:1862.22ms step:4200/6250 train_loss:1.8891 aux_loss:0.0129 train_time:7803405ms step_avg:1862.39ms step:4201/6250 train_loss:1.7691 aux_loss:0.0130 train_time:7805582ms step_avg:1862.46ms step:4202/6250 train_loss:1.9287 aux_loss:0.0131 train_time:7807437ms step_avg:1862.46ms step:4203/6250 train_loss:1.8123 aux_loss:0.0126 train_time:7809707ms step_avg:1862.56ms step:4204/6250 train_loss:1.7800 aux_loss:0.0128 train_time:7811960ms step_avg:1862.65ms step:4205/6250 train_loss:1.8495 aux_loss:0.0125 train_time:7814447ms step_avg:1862.80ms step:4206/6250 train_loss:1.8581 aux_loss:0.0121 train_time:7816956ms step_avg:1862.95ms step:4207/6250 train_loss:1.8946 aux_loss:0.0124 train_time:7819338ms step_avg:1863.08ms step:4208/6250 train_loss:1.7375 aux_loss:0.0127 train_time:7823444ms step_avg:1863.61ms step:4209/6250 train_loss:1.8181 aux_loss:0.0128 train_time:7825075ms step_avg:1863.56ms step:4210/6250 train_loss:1.8865 aux_loss:0.0129 train_time:7826724ms step_avg:1863.51ms step:4211/6250 train_loss:1.8363 aux_loss:0.0130 train_time:7828161ms step_avg:1863.40ms step:4212/6250 train_loss:1.7400 aux_loss:0.0127 train_time:7829556ms step_avg:1863.29ms step:4213/6250 train_loss:1.9068 aux_loss:0.0130 train_time:7831344ms step_avg:1863.27ms step:4214/6250 train_loss:1.8756 aux_loss:0.0132 train_time:7832806ms step_avg:1863.18ms step:4215/6250 train_loss:1.8784 aux_loss:0.0132 train_time:7835098ms step_avg:1863.28ms step:4216/6250 train_loss:1.8774 aux_loss:0.0126 train_time:7837310ms step_avg:1863.36ms step:4217/6250 train_loss:1.8299 aux_loss:0.0127 train_time:7838726ms step_avg:1863.26ms step:4218/6250 train_loss:1.8962 aux_loss:0.0129 train_time:7840478ms step_avg:1863.23ms step:4219/6250 train_loss:1.7956 aux_loss:0.0124 train_time:7841683ms step_avg:1863.08ms step:4220/6250 train_loss:1.8648 aux_loss:0.0126 train_time:7843045ms step_avg:1862.96ms step:4221/6250 train_loss:1.8290 aux_loss:0.0127 train_time:7845307ms step_avg:1863.05ms step:4222/6250 train_loss:1.9887 aux_loss:0.0128 train_time:7847757ms step_avg:1863.19ms step:4223/6250 train_loss:1.8763 aux_loss:0.0129 train_time:7850345ms step_avg:1863.36ms step:4224/6250 train_loss:1.8415 aux_loss:0.0126 train_time:7853033ms step_avg:1863.56ms step:4225/6250 train_loss:1.9153 aux_loss:0.0131 train_time:7857601ms step_avg:1864.20ms step:4226/6250 train_loss:1.8074 aux_loss:0.0130 train_time:7859014ms step_avg:1864.09ms step:4227/6250 train_loss:1.8893 aux_loss:0.0126 train_time:7860589ms step_avg:1864.02ms step:4228/6250 train_loss:1.9045 aux_loss:0.0131 train_time:7862511ms step_avg:1864.04ms step:4229/6250 train_loss:1.9072 aux_loss:0.0126 train_time:7863978ms step_avg:1863.94ms step:4230/6250 train_loss:1.9360 aux_loss:0.0126 train_time:7867448ms step_avg:1864.32ms step:4231/6250 train_loss:1.8348 aux_loss:0.0125 train_time:7869329ms step_avg:1864.33ms step:4232/6250 train_loss:1.8254 aux_loss:0.0129 train_time:7870874ms step_avg:1864.25ms step:4233/6250 train_loss:1.8962 aux_loss:0.0128 train_time:7872667ms step_avg:1864.24ms step:4234/6250 train_loss:1.8658 aux_loss:0.0129 train_time:7874136ms step_avg:1864.14ms step:4235/6250 train_loss:1.8835 aux_loss:0.0130 train_time:7875363ms step_avg:1863.99ms step:4236/6250 train_loss:1.8151 aux_loss:0.0129 train_time:7876817ms step_avg:1863.89ms step:4237/6250 train_loss:1.8184 aux_loss:0.0125 train_time:7878429ms step_avg:1863.83ms step:4238/6250 train_loss:1.8700 aux_loss:0.0124 train_time:7880167ms step_avg:1863.80ms step:4239/6250 train_loss:1.9016 aux_loss:0.0126 train_time:7881834ms step_avg:1863.76ms step:4240/6250 train_loss:1.8373 aux_loss:0.0125 train_time:7883453ms step_avg:1863.70ms step:4241/6250 train_loss:1.9286 aux_loss:0.0127 train_time:7885041ms step_avg:1863.64ms step:4242/6250 train_loss:1.8307 aux_loss:0.0121 train_time:7886541ms step_avg:1863.55ms step:4243/6250 train_loss:1.8563 aux_loss:0.0124 train_time:7888009ms step_avg:1863.46ms step:4244/6250 train_loss:1.8017 aux_loss:0.0127 train_time:7889601ms step_avg:1863.39ms step:4245/6250 train_loss:1.9076 aux_loss:0.0127 train_time:7891867ms step_avg:1863.49ms step:4246/6250 train_loss:1.8517 aux_loss:0.0124 train_time:7893247ms step_avg:1863.37ms step:4247/6250 train_loss:1.9409 aux_loss:0.0126 train_time:7894779ms step_avg:1863.29ms step:4248/6250 train_loss:1.9131 aux_loss:0.0130 train_time:7897509ms step_avg:1863.50ms step:4249/6250 train_loss:1.8778 aux_loss:0.0127 train_time:7899011ms step_avg:1863.41ms step:4250/6250 train_loss:1.8113 aux_loss:0.0127 train_time:7900316ms step_avg:1863.28ms step:4251/6250 train_loss:1.8785 aux_loss:0.0130 train_time:7901701ms step_avg:1863.17ms step:4252/6250 train_loss:1.8857 aux_loss:0.0130 train_time:7903023ms step_avg:1863.04ms step:4253/6250 train_loss:1.8725 aux_loss:0.0128 train_time:7904332ms step_avg:1862.91ms step:4254/6250 train_loss:1.8678 aux_loss:0.0128 train_time:7906219ms step_avg:1862.92ms step:4255/6250 train_loss:1.8472 aux_loss:0.0132 train_time:7907477ms step_avg:1862.77ms step:4256/6250 train_loss:1.8647 aux_loss:0.0130 train_time:7908825ms step_avg:1862.65ms step:4257/6250 train_loss:1.8069 aux_loss:0.0125 train_time:7910400ms step_avg:1862.59ms step:4258/6250 train_loss:1.8627 aux_loss:0.0124 train_time:7911831ms step_avg:1862.48ms step:4259/6250 train_loss:1.8931 aux_loss:0.0124 train_time:7913235ms step_avg:1862.38ms step:4260/6250 train_loss:1.9262 aux_loss:0.0126 train_time:7914682ms step_avg:1862.28ms step:4261/6250 train_loss:1.7663 aux_loss:0.0126 train_time:7916018ms step_avg:1862.15ms step:4262/6250 train_loss:1.8375 aux_loss:0.0125 train_time:7917359ms step_avg:1862.03ms step:4263/6250 train_loss:1.7379 aux_loss:0.0122 train_time:7918720ms step_avg:1861.91ms step:4264/6250 train_loss:1.8998 aux_loss:0.0124 train_time:7920040ms step_avg:1861.79ms step:4265/6250 train_loss:1.8502 aux_loss:0.0126 train_time:7921340ms step_avg:1861.65ms step:4266/6250 train_loss:1.8388 aux_loss:0.0125 train_time:7922590ms step_avg:1861.51ms step:4267/6250 train_loss:1.9154 aux_loss:0.0126 train_time:7923991ms step_avg:1861.40ms step:4268/6250 train_loss:1.9114 aux_loss:0.0129 train_time:7925366ms step_avg:1861.29ms step:4269/6250 train_loss:1.8677 aux_loss:0.0129 train_time:7926714ms step_avg:1861.17ms step:4270/6250 train_loss:1.9413 aux_loss:0.0126 train_time:7928031ms step_avg:1861.04ms step:4271/6250 train_loss:1.8255 aux_loss:0.0128 train_time:7929371ms step_avg:1860.92ms step:4272/6250 train_loss:1.9182 aux_loss:0.0130 train_time:7930657ms step_avg:1860.78ms step:4273/6250 train_loss:1.7773 aux_loss:0.0127 train_time:7932247ms step_avg:1860.72ms step:4274/6250 train_loss:1.8942 aux_loss:0.0126 train_time:7933594ms step_avg:1860.60ms step:4275/6250 train_loss:1.7706 aux_loss:0.0122 train_time:7935002ms step_avg:1860.49ms step:4276/6250 train_loss:1.8616 aux_loss:0.0128 train_time:7936166ms step_avg:1860.33ms step:4277/6250 train_loss:1.7702 aux_loss:0.0129 train_time:7937569ms step_avg:1860.22ms step:4278/6250 train_loss:1.9150 aux_loss:0.0129 train_time:7938944ms step_avg:1860.11ms step:4279/6250 train_loss:1.8873 aux_loss:0.0122 train_time:7940356ms step_avg:1860.00ms step:4280/6250 train_loss:1.8254 aux_loss:0.0127 train_time:7941678ms step_avg:1859.88ms step:4281/6250 train_loss:1.8305 aux_loss:0.0126 train_time:7943495ms step_avg:1859.87ms step:4282/6250 train_loss:1.8630 aux_loss:0.0127 train_time:7944825ms step_avg:1859.74ms step:4283/6250 train_loss:1.8814 aux_loss:0.0127 train_time:7946154ms step_avg:1859.62ms step:4284/6250 train_loss:1.8618 aux_loss:0.0128 train_time:7948100ms step_avg:1859.64ms step:4285/6250 train_loss:1.7805 aux_loss:0.0126 train_time:7949440ms step_avg:1859.52ms step:4286/6250 train_loss:1.8455 aux_loss:0.0128 train_time:7950880ms step_avg:1859.42ms step:4287/6250 train_loss:1.8470 aux_loss:0.0130 train_time:7952158ms step_avg:1859.28ms step:4288/6250 train_loss:1.8885 aux_loss:0.0132 train_time:7953508ms step_avg:1859.17ms step:4289/6250 train_loss:1.7627 aux_loss:0.0124 train_time:7954754ms step_avg:1859.02ms step:4290/6250 train_loss:1.8500 aux_loss:0.0127 train_time:7956314ms step_avg:1858.95ms step:4291/6250 train_loss:1.8608 aux_loss:0.0126 train_time:7957810ms step_avg:1858.87ms step:4292/6250 train_loss:1.8208 aux_loss:0.0128 train_time:7959233ms step_avg:1858.77ms step:4293/6250 train_loss:1.8025 aux_loss:0.0130 train_time:7960852ms step_avg:1858.71ms step:4294/6250 train_loss:1.8125 aux_loss:0.0121 train_time:7962261ms step_avg:1858.60ms step:4295/6250 train_loss:1.8384 aux_loss:0.0121 train_time:7963824ms step_avg:1858.54ms step:4296/6250 train_loss:1.9085 aux_loss:0.0122 train_time:7965133ms step_avg:1858.41ms step:4297/6250 train_loss:1.8164 aux_loss:0.0121 train_time:7966808ms step_avg:1858.36ms step:4298/6250 train_loss:1.8691 aux_loss:0.0123 train_time:7968588ms step_avg:1858.35ms step:4299/6250 train_loss:1.8079 aux_loss:0.0126 train_time:7969953ms step_avg:1858.23ms step:4300/6250 train_loss:1.9345 aux_loss:0.0125 train_time:7971288ms step_avg:1858.11ms step:4301/6250 train_loss:1.8094 aux_loss:0.0126 train_time:7972583ms step_avg:1857.98ms step:4302/6250 train_loss:1.8648 aux_loss:0.0123 train_time:7973909ms step_avg:1857.85ms step:4303/6250 train_loss:1.8865 aux_loss:0.0126 train_time:7975213ms step_avg:1857.72ms step:4304/6250 train_loss:1.8019 aux_loss:0.0127 train_time:7976450ms step_avg:1857.58ms step:4305/6250 train_loss:1.8626 aux_loss:0.0123 train_time:7977798ms step_avg:1857.46ms step:4306/6250 train_loss:1.7816 aux_loss:0.0119 train_time:7979169ms step_avg:1857.35ms step:4307/6250 train_loss:1.8721 aux_loss:0.0121 train_time:7980559ms step_avg:1857.24ms step:4308/6250 train_loss:1.8896 aux_loss:0.0124 train_time:7981931ms step_avg:1857.13ms step:4309/6250 train_loss:1.8509 aux_loss:0.0123 train_time:7983972ms step_avg:1857.17ms step:4310/6250 train_loss:1.8630 aux_loss:0.0121 train_time:7986382ms step_avg:1857.30ms step:4311/6250 train_loss:1.7682 aux_loss:0.0122 train_time:7989973ms step_avg:1857.70ms step:4312/6250 train_loss:1.8385 aux_loss:0.0124 train_time:7993559ms step_avg:1858.10ms step:4313/6250 train_loss:1.8878 aux_loss:0.0125 train_time:8000705ms step_avg:1859.33ms step:4314/6250 train_loss:1.8591 aux_loss:0.0132 train_time:8005486ms step_avg:1860.01ms step:4315/6250 train_loss:1.9878 aux_loss:0.0130 train_time:8007991ms step_avg:1860.16ms step:4316/6250 train_loss:1.8382 aux_loss:0.0126 train_time:8009716ms step_avg:1860.13ms step:4317/6250 train_loss:1.9200 aux_loss:0.0123 train_time:8013014ms step_avg:1860.46ms step:4318/6250 train_loss:1.7375 aux_loss:0.0125 train_time:8015727ms step_avg:1860.66ms step:4319/6250 train_loss:1.8781 aux_loss:0.0135 train_time:8018164ms step_avg:1860.79ms step:4320/6250 train_loss:1.8765 aux_loss:0.0122 train_time:8020257ms step_avg:1860.85ms step:4321/6250 train_loss:1.7726 aux_loss:0.0120 train_time:8023415ms step_avg:1861.15ms step:4322/6250 train_loss:1.8584 aux_loss:0.0126 train_time:8032055ms step_avg:1862.72ms step:4323/6250 train_loss:1.7796 aux_loss:0.0124 train_time:8034960ms step_avg:1862.96ms step:4324/6250 train_loss:1.9208 aux_loss:0.0123 train_time:8039078ms step_avg:1863.49ms step:4325/6250 train_loss:1.8992 aux_loss:0.0126 train_time:8040465ms step_avg:1863.38ms step:4326/6250 train_loss:1.8763 aux_loss:0.0129 train_time:8042396ms step_avg:1863.39ms step:4327/6250 train_loss:1.9246 aux_loss:0.0127 train_time:8045114ms step_avg:1863.59ms step:4328/6250 train_loss:1.8542 aux_loss:0.0124 train_time:8048308ms step_avg:1863.90ms step:4329/6250 train_loss:1.7809 aux_loss:0.0123 train_time:8051063ms step_avg:1864.10ms step:4330/6250 train_loss:1.8516 aux_loss:0.0126 train_time:8054047ms step_avg:1864.36ms step:4331/6250 train_loss:1.8901 aux_loss:0.0126 train_time:8056502ms step_avg:1864.50ms step:4332/6250 train_loss:1.8356 aux_loss:0.0123 train_time:8059202ms step_avg:1864.69ms step:4333/6250 train_loss:1.8384 aux_loss:0.0121 train_time:8060658ms step_avg:1864.60ms step:4334/6250 train_loss:1.7620 aux_loss:0.0120 train_time:8063173ms step_avg:1864.75ms step:4335/6250 train_loss:1.9533 aux_loss:0.0125 train_time:8065055ms step_avg:1864.75ms step:4336/6250 train_loss:1.8322 aux_loss:0.0122 train_time:8066554ms step_avg:1864.67ms step:4337/6250 train_loss:1.8078 aux_loss:0.0120 train_time:8067914ms step_avg:1864.55ms step:4338/6250 train_loss:1.9405 aux_loss:0.0123 train_time:8069231ms step_avg:1864.43ms step:4339/6250 train_loss:1.9281 aux_loss:0.0123 train_time:8070478ms step_avg:1864.28ms step:4340/6250 train_loss:1.8582 aux_loss:0.0123 train_time:8071863ms step_avg:1864.17ms step:4341/6250 train_loss:1.8343 aux_loss:0.0121 train_time:8073272ms step_avg:1864.07ms step:4342/6250 train_loss:1.8626 aux_loss:0.0121 train_time:8074677ms step_avg:1863.96ms step:4343/6250 train_loss:1.8885 aux_loss:0.0122 train_time:8076111ms step_avg:1863.86ms step:4344/6250 train_loss:1.8482 aux_loss:0.0121 train_time:8077472ms step_avg:1863.75ms step:4345/6250 train_loss:1.8530 aux_loss:0.0119 train_time:8078748ms step_avg:1863.61ms step:4346/6250 train_loss:1.8313 aux_loss:0.0119 train_time:8080085ms step_avg:1863.49ms step:4347/6250 train_loss:1.9176 aux_loss:0.0127 train_time:8081518ms step_avg:1863.39ms step:4348/6250 train_loss:1.7859 aux_loss:0.0124 train_time:8082975ms step_avg:1863.30ms step:4349/6250 train_loss:1.8523 aux_loss:0.0120 train_time:8084443ms step_avg:1863.20ms step:4350/6250 train_loss:1.9540 aux_loss:0.0118 train_time:8085907ms step_avg:1863.11ms step:4351/6250 train_loss:1.8550 aux_loss:0.0121 train_time:8087245ms step_avg:1862.99ms step:4352/6250 train_loss:1.8510 aux_loss:0.0122 train_time:8088527ms step_avg:1862.86ms step:4353/6250 train_loss:1.8116 aux_loss:0.0121 train_time:8090017ms step_avg:1862.77ms step:4354/6250 train_loss:1.7815 aux_loss:0.0123 train_time:8091413ms step_avg:1862.66ms step:4355/6250 train_loss:1.8608 aux_loss:0.0122 train_time:8092771ms step_avg:1862.55ms step:4356/6250 train_loss:1.7359 aux_loss:0.0123 train_time:8094252ms step_avg:1862.46ms step:4357/6250 train_loss:1.8184 aux_loss:0.0119 train_time:8095711ms step_avg:1862.37ms step:4358/6250 train_loss:1.9215 aux_loss:0.0122 train_time:8097027ms step_avg:1862.24ms step:4359/6250 train_loss:1.9182 aux_loss:0.0125 train_time:8099479ms step_avg:1862.38ms step:4360/6250 train_loss:1.8280 aux_loss:0.0123 train_time:8102006ms step_avg:1862.53ms step:4361/6250 train_loss:1.8871 aux_loss:0.0119 train_time:8104247ms step_avg:1862.62ms step:4362/6250 train_loss:1.8215 aux_loss:0.0119 train_time:8105476ms step_avg:1862.47ms step:4363/6250 train_loss:1.8610 aux_loss:0.0125 train_time:8106814ms step_avg:1862.35ms step:4364/6250 train_loss:1.9025 aux_loss:0.0127 train_time:8108187ms step_avg:1862.24ms step:4365/6250 train_loss:1.8397 aux_loss:0.0124 train_time:8109562ms step_avg:1862.13ms step:4366/6250 train_loss:1.9052 aux_loss:0.0120 train_time:8110904ms step_avg:1862.01ms step:4367/6250 train_loss:1.7691 aux_loss:0.0122 train_time:8112258ms step_avg:1861.89ms step:4368/6250 train_loss:1.8714 aux_loss:0.0121 train_time:8114358ms step_avg:1861.95ms step:4369/6250 train_loss:1.7373 aux_loss:0.0120 train_time:8115726ms step_avg:1861.83ms step:4370/6250 train_loss:1.8184 aux_loss:0.0122 train_time:8117306ms step_avg:1861.77ms step:4371/6250 train_loss:1.9120 aux_loss:0.0122 train_time:8118593ms step_avg:1861.64ms step:4372/6250 train_loss:1.8485 aux_loss:0.0121 train_time:8119939ms step_avg:1861.52ms step:4373/6250 train_loss:1.9079 aux_loss:0.0123 train_time:8121799ms step_avg:1861.52ms step:4374/6250 train_loss:1.8063 aux_loss:0.0123 train_time:8125653ms step_avg:1861.97ms step:4375/6250 train_loss:1.7798 aux_loss:0.0123 train_time:8129997ms step_avg:1862.54ms step:4376/6250 train_loss:1.7906 aux_loss:0.0124 train_time:8131778ms step_avg:1862.52ms step:4377/6250 train_loss:1.8542 aux_loss:0.0122 train_time:8133990ms step_avg:1862.60ms step:4378/6250 train_loss:1.8001 aux_loss:0.0118 train_time:8137611ms step_avg:1863.01ms step:4379/6250 train_loss:1.8333 aux_loss:0.0124 train_time:8140653ms step_avg:1863.28ms step:4380/6250 train_loss:1.7971 aux_loss:0.0126 train_time:8146909ms step_avg:1864.28ms step:4381/6250 train_loss:1.8115 aux_loss:0.0123 train_time:8152058ms step_avg:1865.03ms step:4382/6250 train_loss:1.8062 aux_loss:0.0123 train_time:8156184ms step_avg:1865.55ms step:4383/6250 train_loss:1.7012 aux_loss:0.0121 train_time:8161207ms step_avg:1866.27ms step:4384/6250 train_loss:1.8407 aux_loss:0.0126 train_time:8163092ms step_avg:1866.28ms step:4385/6250 train_loss:1.8377 aux_loss:0.0122 train_time:8164901ms step_avg:1866.26ms step:4386/6250 train_loss:1.8210 aux_loss:0.0120 train_time:8167601ms step_avg:1866.45ms step:4387/6250 train_loss:1.7335 aux_loss:0.0120 train_time:8171101ms step_avg:1866.83ms step:4388/6250 train_loss:1.9027 aux_loss:0.0124 train_time:8173971ms step_avg:1867.06ms step:4389/6250 train_loss:1.8389 aux_loss:0.0121 train_time:8179821ms step_avg:1867.97ms step:4390/6250 train_loss:1.8322 aux_loss:0.0121 train_time:8182352ms step_avg:1868.12ms step:4391/6250 train_loss:1.8731 aux_loss:0.0119 train_time:8186001ms step_avg:1868.52ms step:4392/6250 train_loss:1.8700 aux_loss:0.0123 train_time:8188047ms step_avg:1868.56ms step:4393/6250 train_loss:1.8306 aux_loss:0.0122 train_time:8191469ms step_avg:1868.92ms step:4394/6250 train_loss:1.9409 aux_loss:0.0124 train_time:8193222ms step_avg:1868.89ms step:4395/6250 train_loss:1.8517 aux_loss:0.0122 train_time:8195222ms step_avg:1868.92ms step:4396/6250 train_loss:1.8343 aux_loss:0.0122 train_time:8197444ms step_avg:1869.00ms step:4397/6250 train_loss:1.8130 aux_loss:0.0120 train_time:8199089ms step_avg:1868.95ms step:4398/6250 train_loss:1.8514 aux_loss:0.0128 train_time:8200651ms step_avg:1868.88ms step:4399/6250 train_loss:1.7970 aux_loss:0.0124 train_time:8202188ms step_avg:1868.81ms step:4400/6250 train_loss:1.8562 aux_loss:0.0124 train_time:8203721ms step_avg:1868.73ms step:4401/6250 train_loss:1.9095 aux_loss:0.0123 train_time:8205177ms step_avg:1868.64ms step:4402/6250 train_loss:1.8658 aux_loss:0.0123 train_time:8206760ms step_avg:1868.57ms step:4403/6250 train_loss:1.7986 aux_loss:0.0119 train_time:8208162ms step_avg:1868.46ms step:4404/6250 train_loss:1.8016 aux_loss:0.0121 train_time:8209757ms step_avg:1868.40ms step:4405/6250 train_loss:1.9079 aux_loss:0.0124 train_time:8211454ms step_avg:1868.36ms step:4406/6250 train_loss:1.9470 aux_loss:0.0121 train_time:8213148ms step_avg:1868.32ms step:4407/6250 train_loss:1.8510 aux_loss:0.0119 train_time:8214718ms step_avg:1868.26ms step:4408/6250 train_loss:1.8196 aux_loss:0.0122 train_time:8216307ms step_avg:1868.19ms step:4409/6250 train_loss:1.8111 aux_loss:0.0120 train_time:8217735ms step_avg:1868.09ms step:4410/6250 train_loss:1.8697 aux_loss:0.0120 train_time:8219778ms step_avg:1868.13ms step:4411/6250 train_loss:1.8616 aux_loss:0.0120 train_time:8221401ms step_avg:1868.08ms step:4412/6250 train_loss:1.9044 aux_loss:0.0121 train_time:8222934ms step_avg:1868.00ms step:4413/6250 train_loss:1.7429 aux_loss:0.0121 train_time:8224535ms step_avg:1867.94ms step:4414/6250 train_loss:1.8151 aux_loss:0.0120 train_time:8226369ms step_avg:1867.93ms step:4415/6250 train_loss:1.8868 aux_loss:0.0119 train_time:8228048ms step_avg:1867.89ms step:4416/6250 train_loss:1.7936 aux_loss:0.0117 train_time:8229569ms step_avg:1867.81ms step:4417/6250 train_loss:1.8611 aux_loss:0.0121 train_time:8231749ms step_avg:1867.88ms step:4418/6250 train_loss:1.7807 aux_loss:0.0120 train_time:8233413ms step_avg:1867.83ms step:4419/6250 train_loss:1.8125 aux_loss:0.0119 train_time:8234820ms step_avg:1867.73ms step:4420/6250 train_loss:1.9367 aux_loss:0.0122 train_time:8236433ms step_avg:1867.67ms step:4421/6250 train_loss:1.9425 aux_loss:0.0121 train_time:8237814ms step_avg:1867.56ms step:4422/6250 train_loss:1.8945 aux_loss:0.0121 train_time:8239210ms step_avg:1867.45ms step:4423/6250 train_loss:1.9187 aux_loss:0.0122 train_time:8240646ms step_avg:1867.36ms step:4424/6250 train_loss:1.8156 aux_loss:0.0118 train_time:8241980ms step_avg:1867.24ms step:4425/6250 train_loss:1.8209 aux_loss:0.0124 train_time:8243560ms step_avg:1867.17ms step:4426/6250 train_loss:1.7295 aux_loss:0.0118 train_time:8245371ms step_avg:1867.16ms step:4427/6250 train_loss:1.8760 aux_loss:0.0120 train_time:8246894ms step_avg:1867.08ms step:4428/6250 train_loss:1.8039 aux_loss:0.0121 train_time:8248600ms step_avg:1867.04ms step:4429/6250 train_loss:1.9374 aux_loss:0.0119 train_time:8250115ms step_avg:1866.96ms step:4430/6250 train_loss:1.8198 aux_loss:0.0119 train_time:8251667ms step_avg:1866.89ms step:4431/6250 train_loss:1.7701 aux_loss:0.0118 train_time:8253179ms step_avg:1866.81ms step:4432/6250 train_loss:1.9566 aux_loss:0.0117 train_time:8254861ms step_avg:1866.77ms step:4433/6250 train_loss:1.8039 aux_loss:0.0117 train_time:8256211ms step_avg:1866.65ms step:4434/6250 train_loss:1.8854 aux_loss:0.0121 train_time:8257816ms step_avg:1866.59ms step:4435/6250 train_loss:1.8643 aux_loss:0.0117 train_time:8260609ms step_avg:1866.80ms step:4436/6250 train_loss:1.8277 aux_loss:0.0123 train_time:8262212ms step_avg:1866.74ms step:4437/6250 train_loss:1.7804 aux_loss:0.0116 train_time:8264438ms step_avg:1866.83ms step:4438/6250 train_loss:1.7968 aux_loss:0.0118 train_time:8266617ms step_avg:1866.90ms step:4439/6250 train_loss:1.7432 aux_loss:0.0119 train_time:8268434ms step_avg:1866.89ms step:4440/6250 train_loss:1.9260 aux_loss:0.0122 train_time:8270085ms step_avg:1866.84ms step:4441/6250 train_loss:1.7886 aux_loss:0.0118 train_time:8271765ms step_avg:1866.79ms step:4442/6250 train_loss:1.7627 aux_loss:0.0114 train_time:8273699ms step_avg:1866.81ms step:4443/6250 train_loss:1.8882 aux_loss:0.0118 train_time:8275683ms step_avg:1866.84ms step:4444/6250 train_loss:1.9180 aux_loss:0.0119 train_time:8277807ms step_avg:1866.89ms step:4445/6250 train_loss:1.8298 aux_loss:0.0121 train_time:8281723ms step_avg:1867.36ms step:4446/6250 train_loss:1.7952 aux_loss:0.0117 train_time:8284570ms step_avg:1867.58ms step:4447/6250 train_loss:1.9261 aux_loss:0.0120 train_time:8287878ms step_avg:1867.90ms step:4448/6250 train_loss:1.7586 aux_loss:0.0115 train_time:8292174ms step_avg:1868.45ms step:4449/6250 train_loss:1.7168 aux_loss:0.0115 train_time:8294886ms step_avg:1868.64ms step:4450/6250 train_loss:1.7616 aux_loss:0.0114 train_time:8298536ms step_avg:1869.04ms step:4451/6250 train_loss:1.7741 aux_loss:0.0116 train_time:8303420ms step_avg:1869.72ms step:4452/6250 train_loss:2.0402 aux_loss:0.0118 train_time:8306568ms step_avg:1870.01ms step:4453/6250 train_loss:1.9641 aux_loss:0.0121 train_time:8308457ms step_avg:1870.01ms step:4454/6250 train_loss:1.8717 aux_loss:0.0120 train_time:8310767ms step_avg:1870.11ms step:4455/6250 train_loss:1.8346 aux_loss:0.0119 train_time:8312907ms step_avg:1870.17ms step:4456/6250 train_loss:1.8961 aux_loss:0.0119 train_time:8317204ms step_avg:1870.72ms step:4457/6250 train_loss:1.8245 aux_loss:0.0117 train_time:8320050ms step_avg:1870.94ms step:4458/6250 train_loss:1.9281 aux_loss:0.0126 train_time:8322475ms step_avg:1871.06ms step:4459/6250 train_loss:1.8392 aux_loss:0.0118 train_time:8324333ms step_avg:1871.06ms step:4460/6250 train_loss:1.8334 aux_loss:0.0116 train_time:8327105ms step_avg:1871.26ms step:4461/6250 train_loss:1.7493 aux_loss:0.0120 train_time:8329327ms step_avg:1871.34ms step:4462/6250 train_loss:1.9140 aux_loss:0.0121 train_time:8332213ms step_avg:1871.57ms step:4463/6250 train_loss:1.8549 aux_loss:0.0119 train_time:8367931ms step_avg:1879.17ms step:4464/6250 train_loss:1.8068 aux_loss:0.0113 train_time:8399062ms step_avg:1885.73ms step:4465/6250 train_loss:1.9544 aux_loss:0.0117 train_time:8400799ms step_avg:1885.70ms step:4466/6250 train_loss:1.9088 aux_loss:0.0120 train_time:8402153ms step_avg:1885.58ms step:4467/6250 train_loss:1.7961 aux_loss:0.0120 train_time:8403532ms step_avg:1885.47ms step:4468/6250 train_loss:1.8426 aux_loss:0.0119 train_time:8404869ms step_avg:1885.35ms step:4469/6250 train_loss:1.8483 aux_loss:0.0116 train_time:8406178ms step_avg:1885.22ms step:4470/6250 train_loss:1.8389 aux_loss:0.0117 train_time:8407541ms step_avg:1885.10ms step:4471/6250 train_loss:1.8048 aux_loss:0.0117 train_time:8408865ms step_avg:1884.97ms step:4472/6250 train_loss:1.8693 aux_loss:0.0120 train_time:8410485ms step_avg:1884.91ms step:4473/6250 train_loss:1.8642 aux_loss:0.0124 train_time:8412102ms step_avg:1884.85ms step:4474/6250 train_loss:1.8713 aux_loss:0.0119 train_time:8413426ms step_avg:1884.73ms step:4475/6250 train_loss:1.8061 aux_loss:0.0119 train_time:8415207ms step_avg:1884.70ms step:4476/6250 train_loss:1.8800 aux_loss:0.0120 train_time:8416905ms step_avg:1884.66ms step:4477/6250 train_loss:1.8475 aux_loss:0.0118 train_time:8418382ms step_avg:1884.57ms step:4478/6250 train_loss:1.9200 aux_loss:0.0122 train_time:8419733ms step_avg:1884.45ms step:4479/6250 train_loss:1.7950 aux_loss:0.0117 train_time:8421299ms step_avg:1884.38ms step:4480/6250 train_loss:1.8177 aux_loss:0.0119 train_time:8422865ms step_avg:1884.31ms step:4481/6250 train_loss:1.8136 aux_loss:0.0118 train_time:8424699ms step_avg:1884.30ms step:4482/6250 train_loss:1.8682 aux_loss:0.0117 train_time:8427279ms step_avg:1884.45ms step:4483/6250 train_loss:1.8704 aux_loss:0.0119 train_time:8428726ms step_avg:1884.36ms step:4484/6250 train_loss:1.8826 aux_loss:0.0119 train_time:8430001ms step_avg:1884.22ms step:4485/6250 train_loss:1.8380 aux_loss:0.0115 train_time:8431397ms step_avg:1884.11ms step:4486/6250 train_loss:1.7862 aux_loss:0.0117 train_time:8433373ms step_avg:1884.13ms step:4487/6250 train_loss:1.9553 aux_loss:0.0122 train_time:8434993ms step_avg:1884.07ms step:4488/6250 train_loss:1.9196 aux_loss:0.0120 train_time:8436313ms step_avg:1883.95ms step:4489/6250 train_loss:1.8909 aux_loss:0.0117 train_time:8437629ms step_avg:1883.82ms step:4490/6250 train_loss:1.8222 aux_loss:0.0114 train_time:8438956ms step_avg:1883.70ms step:4491/6250 train_loss:1.7912 aux_loss:0.0115 train_time:8440602ms step_avg:1883.64ms step:4492/6250 train_loss:1.9013 aux_loss:0.0118 train_time:8442116ms step_avg:1883.56ms step:4493/6250 train_loss:1.8602 aux_loss:0.0115 train_time:8443544ms step_avg:1883.46ms step:4494/6250 train_loss:1.8686 aux_loss:0.0119 train_time:8444969ms step_avg:1883.36ms step:4495/6250 train_loss:1.8778 aux_loss:0.0115 train_time:8446249ms step_avg:1883.22ms step:4496/6250 train_loss:1.8984 aux_loss:0.0117 train_time:8447578ms step_avg:1883.10ms step:4497/6250 train_loss:1.9027 aux_loss:0.0119 train_time:8448996ms step_avg:1882.99ms step:4498/6250 train_loss:1.8695 aux_loss:0.0120 train_time:8450364ms step_avg:1882.88ms step:4499/6250 train_loss:1.9039 aux_loss:0.0114 train_time:8451709ms step_avg:1882.76ms step:4500/6250 train_loss:1.8363 aux_loss:0.0117 train_time:8453135ms step_avg:1882.66ms step:4501/6250 train_loss:1.8244 aux_loss:0.0123 train_time:8454474ms step_avg:1882.54ms step:4502/6250 train_loss:1.7684 aux_loss:0.0120 train_time:8455904ms step_avg:1882.44ms step:4503/6250 train_loss:1.7590 aux_loss:0.0116 train_time:8457306ms step_avg:1882.33ms step:4504/6250 train_loss:1.8506 aux_loss:0.0120 train_time:8458649ms step_avg:1882.21ms step:4505/6250 train_loss:1.8664 aux_loss:0.0121 train_time:8460086ms step_avg:1882.11ms step:4506/6250 train_loss:1.7818 aux_loss:0.0118 train_time:8461491ms step_avg:1882.00ms step:4507/6250 train_loss:1.8625 aux_loss:0.0125 train_time:8462818ms step_avg:1881.88ms step:4508/6250 train_loss:1.8029 aux_loss:0.0121 train_time:8464271ms step_avg:1881.79ms step:4509/6250 train_loss:1.8144 aux_loss:0.0119 train_time:8465688ms step_avg:1881.68ms step:4510/6250 train_loss:1.8365 aux_loss:0.0118 train_time:8467054ms step_avg:1881.57ms step:4511/6250 train_loss:1.8558 aux_loss:0.0117 train_time:8468377ms step_avg:1881.44ms step:4512/6250 train_loss:1.7998 aux_loss:0.0118 train_time:8469669ms step_avg:1881.31ms step:4513/6250 train_loss:1.9164 aux_loss:0.0116 train_time:8471068ms step_avg:1881.21ms step:4514/6250 train_loss:1.7759 aux_loss:0.0115 train_time:8472428ms step_avg:1881.09ms step:4515/6250 train_loss:1.9209 aux_loss:0.0117 train_time:8473707ms step_avg:1880.96ms step:4516/6250 train_loss:1.8694 aux_loss:0.0114 train_time:8474995ms step_avg:1880.82ms step:4517/6250 train_loss:1.8869 aux_loss:0.0115 train_time:8476355ms step_avg:1880.71ms step:4518/6250 train_loss:1.7880 aux_loss:0.0115 train_time:8477666ms step_avg:1880.58ms step:4519/6250 train_loss:1.7854 aux_loss:0.0116 train_time:8479065ms step_avg:1880.48ms step:4520/6250 train_loss:1.8481 aux_loss:0.0114 train_time:8480409ms step_avg:1880.36ms step:4521/6250 train_loss:1.8912 aux_loss:0.0114 train_time:8481762ms step_avg:1880.24ms step:4522/6250 train_loss:1.8730 aux_loss:0.0121 train_time:8483161ms step_avg:1880.13ms step:4523/6250 train_loss:1.9305 aux_loss:0.0119 train_time:8484489ms step_avg:1880.01ms step:4524/6250 train_loss:1.9478 aux_loss:0.0124 train_time:8485901ms step_avg:1879.91ms step:4525/6250 train_loss:1.8172 aux_loss:0.0116 train_time:8487246ms step_avg:1879.79ms step:4526/6250 train_loss:1.8303 aux_loss:0.0117 train_time:8488547ms step_avg:1879.66ms step:4527/6250 train_loss:1.8857 aux_loss:0.0120 train_time:8489838ms step_avg:1879.53ms step:4528/6250 train_loss:1.8230 aux_loss:0.0118 train_time:8491204ms step_avg:1879.42ms step:4529/6250 train_loss:1.8345 aux_loss:0.0117 train_time:8492812ms step_avg:1879.36ms step:4530/6250 train_loss:1.7843 aux_loss:0.0117 train_time:8494271ms step_avg:1879.26ms step:4531/6250 train_loss:1.8377 aux_loss:0.0117 train_time:8495806ms step_avg:1879.19ms step:4532/6250 train_loss:1.8220 aux_loss:0.0117 train_time:8497082ms step_avg:1879.05ms step:4533/6250 train_loss:1.8115 aux_loss:0.0114 train_time:8498431ms step_avg:1878.94ms step:4534/6250 train_loss:1.7997 aux_loss:0.0116 train_time:8499777ms step_avg:1878.82ms step:4535/6250 train_loss:1.8001 aux_loss:0.0119 train_time:8501146ms step_avg:1878.71ms step:4536/6250 train_loss:1.8189 aux_loss:0.0119 train_time:8503235ms step_avg:1878.75ms step:4537/6250 train_loss:1.8247 aux_loss:0.0119 train_time:8505540ms step_avg:1878.85ms step:4538/6250 train_loss:1.8875 aux_loss:0.0117 train_time:8506859ms step_avg:1878.72ms step:4539/6250 train_loss:1.8888 aux_loss:0.0115 train_time:8508194ms step_avg:1878.60ms step:4540/6250 train_loss:1.8799 aux_loss:0.0115 train_time:8509679ms step_avg:1878.52ms step:4541/6250 train_loss:1.7775 aux_loss:0.0115 train_time:8511312ms step_avg:1878.46ms step:4542/6250 train_loss:1.9124 aux_loss:0.0114 train_time:8513359ms step_avg:1878.50ms step:4543/6250 train_loss:1.8954 aux_loss:0.0119 train_time:8514760ms step_avg:1878.39ms step:4544/6250 train_loss:1.8749 aux_loss:0.0116 train_time:8516780ms step_avg:1878.43ms step:4545/6250 train_loss:1.9169 aux_loss:0.0117 train_time:8518221ms step_avg:1878.33ms step:4546/6250 train_loss:1.8467 aux_loss:0.0113 train_time:8519614ms step_avg:1878.22ms step:4547/6250 train_loss:1.8841 aux_loss:0.0114 train_time:8520982ms step_avg:1878.11ms step:4548/6250 train_loss:1.8111 aux_loss:0.0112 train_time:8522360ms step_avg:1878.00ms step:4549/6250 train_loss:1.8716 aux_loss:0.0116 train_time:8523598ms step_avg:1877.86ms step:4550/6250 train_loss:1.8547 aux_loss:0.0113 train_time:8525104ms step_avg:1877.78ms step:4551/6250 train_loss:1.8708 aux_loss:0.0117 train_time:8526516ms step_avg:1877.67ms step:4552/6250 train_loss:1.8626 aux_loss:0.0116 train_time:8527905ms step_avg:1877.57ms step:4553/6250 train_loss:1.7689 aux_loss:0.0115 train_time:8529206ms step_avg:1877.44ms step:4554/6250 train_loss:1.7827 aux_loss:0.0115 train_time:8530959ms step_avg:1877.41ms step:4555/6250 train_loss:1.8963 aux_loss:0.0118 train_time:8532930ms step_avg:1877.43ms step:4556/6250 train_loss:1.9171 aux_loss:0.0118 train_time:8534238ms step_avg:1877.31ms step:4557/6250 train_loss:1.7347 aux_loss:0.0118 train_time:8535652ms step_avg:1877.21ms step:4558/6250 train_loss:1.7231 aux_loss:0.0117 train_time:8537005ms step_avg:1877.09ms step:4559/6250 train_loss:1.9015 aux_loss:0.0119 train_time:8538297ms step_avg:1876.96ms step:4560/6250 train_loss:1.8601 aux_loss:0.0117 train_time:8539525ms step_avg:1876.82ms step:4561/6250 train_loss:1.8394 aux_loss:0.0119 train_time:8540907ms step_avg:1876.71ms step:4562/6250 train_loss:1.8358 aux_loss:0.0118 train_time:8542352ms step_avg:1876.62ms step:4563/6250 train_loss:1.8148 aux_loss:0.0115 train_time:8544601ms step_avg:1876.70ms step:4564/6250 train_loss:1.8866 aux_loss:0.0125 train_time:8546180ms step_avg:1876.63ms step:4565/6250 train_loss:1.8505 aux_loss:0.0118 train_time:8547581ms step_avg:1876.53ms step:4566/6250 train_loss:1.8887 aux_loss:0.0115 train_time:8548919ms step_avg:1876.41ms step:4567/6250 train_loss:1.9210 aux_loss:0.0117 train_time:8550222ms step_avg:1876.28ms step:4568/6250 train_loss:1.7837 aux_loss:0.0115 train_time:8551511ms step_avg:1876.15ms step:4569/6250 train_loss:1.8845 aux_loss:0.0114 train_time:8552831ms step_avg:1876.03ms step:4570/6250 train_loss:1.8514 aux_loss:0.0111 train_time:8554296ms step_avg:1875.94ms step:4571/6250 train_loss:1.8253 aux_loss:0.0113 train_time:8555644ms step_avg:1875.83ms step:4572/6250 train_loss:1.7732 aux_loss:0.0116 train_time:8556978ms step_avg:1875.71ms step:4573/6250 train_loss:1.8569 aux_loss:0.0112 train_time:8558337ms step_avg:1875.59ms step:4574/6250 train_loss:1.8443 aux_loss:0.0112 train_time:8559747ms step_avg:1875.49ms step:4575/6250 train_loss:1.7739 aux_loss:0.0116 train_time:8561373ms step_avg:1875.44ms step:4576/6250 train_loss:1.8912 aux_loss:0.0119 train_time:8563000ms step_avg:1875.38ms step:4577/6250 train_loss:1.8188 aux_loss:0.0115 train_time:8564560ms step_avg:1875.31ms step:4578/6250 train_loss:1.7886 aux_loss:0.0114 train_time:8566105ms step_avg:1875.24ms step:4579/6250 train_loss:1.7669 aux_loss:0.0117 train_time:8568293ms step_avg:1875.31ms step:4580/6250 train_loss:1.8462 aux_loss:0.0118 train_time:8569732ms step_avg:1875.21ms step:4581/6250 train_loss:1.7309 aux_loss:0.0112 train_time:8571107ms step_avg:1875.11ms step:4582/6250 train_loss:1.8966 aux_loss:0.0115 train_time:8572873ms step_avg:1875.08ms step:4583/6250 train_loss:1.9407 aux_loss:0.0116 train_time:8574327ms step_avg:1874.99ms step:4584/6250 train_loss:1.7763 aux_loss:0.0114 train_time:8576035ms step_avg:1874.95ms step:4585/6250 train_loss:1.8129 aux_loss:0.0113 train_time:8577338ms step_avg:1874.83ms step:4586/6250 train_loss:1.8488 aux_loss:0.0111 train_time:8578962ms step_avg:1874.77ms step:4587/6250 train_loss:1.8125 aux_loss:0.0113 train_time:8580858ms step_avg:1874.78ms step:4588/6250 train_loss:1.8569 aux_loss:0.0118 train_time:8582170ms step_avg:1874.65ms step:4589/6250 train_loss:1.8676 aux_loss:0.0112 train_time:8584647ms step_avg:1874.79ms step:4590/6250 train_loss:1.8765 aux_loss:0.0112 train_time:8585997ms step_avg:1874.67ms step:4591/6250 train_loss:1.7982 aux_loss:0.0112 train_time:8587615ms step_avg:1874.62ms step:4592/6250 train_loss:1.8207 aux_loss:0.0114 train_time:8589204ms step_avg:1874.55ms step:4593/6250 train_loss:1.8432 aux_loss:0.0113 train_time:8590912ms step_avg:1874.52ms step:4594/6250 train_loss:1.7813 aux_loss:0.0112 train_time:8593447ms step_avg:1874.66ms step:4595/6250 train_loss:1.8715 aux_loss:0.0111 train_time:8595274ms step_avg:1874.65ms step:4596/6250 train_loss:1.7566 aux_loss:0.0110 train_time:8597413ms step_avg:1874.71ms step:4597/6250 train_loss:1.8038 aux_loss:0.0111 train_time:8599170ms step_avg:1874.68ms step:4598/6250 train_loss:1.9196 aux_loss:0.0112 train_time:8601496ms step_avg:1874.78ms step:4599/6250 train_loss:1.8383 aux_loss:0.0111 train_time:8603309ms step_avg:1874.77ms step:4600/6250 train_loss:1.9123 aux_loss:0.0115 train_time:8604758ms step_avg:1874.68ms step:4601/6250 train_loss:1.8904 aux_loss:0.0117 train_time:8606235ms step_avg:1874.59ms step:4602/6250 train_loss:1.8719 aux_loss:0.0108 train_time:8607573ms step_avg:1874.47ms step:4603/6250 train_loss:1.8531 aux_loss:0.0114 train_time:8609129ms step_avg:1874.40ms step:4604/6250 train_loss:1.8082 aux_loss:0.0115 train_time:8610752ms step_avg:1874.35ms step:4605/6250 train_loss:1.8960 aux_loss:0.0112 train_time:8612941ms step_avg:1874.42ms step:4606/6250 train_loss:1.9236 aux_loss:0.0113 train_time:8614795ms step_avg:1874.41ms step:4607/6250 train_loss:1.7963 aux_loss:0.0114 train_time:8616205ms step_avg:1874.31ms step:4608/6250 train_loss:1.7993 aux_loss:0.0115 train_time:8617592ms step_avg:1874.20ms step:4609/6250 train_loss:1.8078 aux_loss:0.0111 train_time:8618945ms step_avg:1874.09ms step:4610/6250 train_loss:1.8495 aux_loss:0.0113 train_time:8620394ms step_avg:1874.00ms step:4611/6250 train_loss:1.8232 aux_loss:0.0118 train_time:8621664ms step_avg:1873.87ms step:4612/6250 train_loss:1.8599 aux_loss:0.0117 train_time:8622935ms step_avg:1873.74ms step:4613/6250 train_loss:1.8495 aux_loss:0.0113 train_time:8624414ms step_avg:1873.65ms step:4614/6250 train_loss:1.7734 aux_loss:0.0110 train_time:8626262ms step_avg:1873.65ms step:4615/6250 train_loss:1.9704 aux_loss:0.0116 train_time:8627943ms step_avg:1873.60ms step:4616/6250 train_loss:1.8463 aux_loss:0.0115 train_time:8629737ms step_avg:1873.59ms step:4617/6250 train_loss:1.8126 aux_loss:0.0116 train_time:8631530ms step_avg:1873.57ms step:4618/6250 train_loss:1.8392 aux_loss:0.0113 train_time:8633196ms step_avg:1873.52ms step:4619/6250 train_loss:1.8156 aux_loss:0.0114 train_time:8635487ms step_avg:1873.61ms step:4620/6250 train_loss:1.9104 aux_loss:0.0114 train_time:8637597ms step_avg:1873.67ms step:4621/6250 train_loss:1.8358 aux_loss:0.0114 train_time:8639027ms step_avg:1873.57ms step:4622/6250 train_loss:1.8545 aux_loss:0.0112 train_time:8641197ms step_avg:1873.63ms step:4623/6250 train_loss:1.8164 aux_loss:0.0111 train_time:8643298ms step_avg:1873.68ms step:4624/6250 train_loss:1.8371 aux_loss:0.0114 train_time:8644728ms step_avg:1873.59ms step:4625/6250 train_loss:1.7582 aux_loss:0.0115 train_time:8646543ms step_avg:1873.57ms step:4626/6250 train_loss:1.8051 aux_loss:0.0112 train_time:8647977ms step_avg:1873.48ms step:4627/6250 train_loss:1.8247 aux_loss:0.0111 train_time:8649392ms step_avg:1873.38ms step:4628/6250 train_loss:1.8518 aux_loss:0.0115 train_time:8650803ms step_avg:1873.28ms step:4629/6250 train_loss:1.8490 aux_loss:0.0124 train_time:8652207ms step_avg:1873.18ms step:4630/6250 train_loss:1.9357 aux_loss:0.0114 train_time:8653687ms step_avg:1873.09ms step:4631/6250 train_loss:1.7842 aux_loss:0.0108 train_time:8655034ms step_avg:1872.98ms step:4632/6250 train_loss:1.8501 aux_loss:0.0113 train_time:8656426ms step_avg:1872.87ms step:4633/6250 train_loss:1.7311 aux_loss:0.0110 train_time:8657956ms step_avg:1872.80ms step:4634/6250 train_loss:1.8330 aux_loss:0.0110 train_time:8659720ms step_avg:1872.78ms step:4635/6250 train_loss:1.9044 aux_loss:0.0115 train_time:8661006ms step_avg:1872.65ms step:4636/6250 train_loss:1.8003 aux_loss:0.0116 train_time:8662342ms step_avg:1872.53ms step:4637/6250 train_loss:1.7796 aux_loss:0.0115 train_time:8663711ms step_avg:1872.43ms step:4638/6250 train_loss:1.8044 aux_loss:0.0113 train_time:8665067ms step_avg:1872.31ms step:4639/6250 train_loss:1.8296 aux_loss:0.0112 train_time:8666293ms step_avg:1872.17ms step:4640/6250 train_loss:1.8458 aux_loss:0.0114 train_time:8667578ms step_avg:1872.05ms step:4641/6250 train_loss:1.6563 aux_loss:0.0112 train_time:8669089ms step_avg:1871.97ms step:4642/6250 train_loss:1.9627 aux_loss:0.0114 train_time:8670427ms step_avg:1871.85ms step:4643/6250 train_loss:1.7632 aux_loss:0.0111 train_time:8672251ms step_avg:1871.84ms step:4644/6250 train_loss:1.7837 aux_loss:0.0109 train_time:8673580ms step_avg:1871.73ms step:4645/6250 train_loss:1.8754 aux_loss:0.0115 train_time:8675001ms step_avg:1871.63ms step:4646/6250 train_loss:1.7844 aux_loss:0.0117 train_time:8676905ms step_avg:1871.64ms step:4647/6250 train_loss:1.7565 aux_loss:0.0113 train_time:8679477ms step_avg:1871.79ms step:4648/6250 train_loss:1.8603 aux_loss:0.0114 train_time:8682062ms step_avg:1871.94ms step:4649/6250 train_loss:1.8435 aux_loss:0.0109 train_time:8683563ms step_avg:1871.86ms step:4650/6250 train_loss:1.7876 aux_loss:0.0110 train_time:8684963ms step_avg:1871.76ms step:4651/6250 train_loss:1.7663 aux_loss:0.0108 train_time:8686229ms step_avg:1871.63ms step:4652/6250 train_loss:1.9036 aux_loss:0.0112 train_time:8687585ms step_avg:1871.52ms step:4653/6250 train_loss:1.8738 aux_loss:0.0111 train_time:8689075ms step_avg:1871.44ms step:4654/6250 train_loss:1.8861 aux_loss:0.0113 train_time:8690341ms step_avg:1871.31ms step:4655/6250 train_loss:1.8515 aux_loss:0.0115 train_time:8691679ms step_avg:1871.19ms step:4656/6250 train_loss:1.9021 aux_loss:0.0112 train_time:8693322ms step_avg:1871.14ms step:4657/6250 train_loss:1.8448 aux_loss:0.0108 train_time:8695688ms step_avg:1871.25ms step:4658/6250 train_loss:1.8521 aux_loss:0.0110 train_time:8697237ms step_avg:1871.18ms step:4659/6250 train_loss:1.8206 aux_loss:0.0114 train_time:8698505ms step_avg:1871.05ms step:4660/6250 train_loss:1.8012 aux_loss:0.0110 train_time:8700393ms step_avg:1871.05ms step:4661/6250 train_loss:1.7977 aux_loss:0.0111 train_time:8702358ms step_avg:1871.07ms step:4662/6250 train_loss:1.8133 aux_loss:0.0112 train_time:8703667ms step_avg:1870.95ms step:4663/6250 train_loss:1.9216 aux_loss:0.0118 train_time:8706250ms step_avg:1871.10ms step:4664/6250 train_loss:1.8428 aux_loss:0.0117 train_time:8707856ms step_avg:1871.05ms step:4665/6250 train_loss:1.7718 aux_loss:0.0115 train_time:8709634ms step_avg:1871.03ms step:4666/6250 train_loss:1.8289 aux_loss:0.0115 train_time:8711169ms step_avg:1870.96ms step:4667/6250 train_loss:1.7716 aux_loss:0.0112 train_time:8713525ms step_avg:1871.06ms step:4668/6250 train_loss:1.8253 aux_loss:0.0111 train_time:8715955ms step_avg:1871.18ms step:4669/6250 train_loss:1.8370 aux_loss:0.0113 train_time:8718384ms step_avg:1871.30ms step:4670/6250 train_loss:1.7764 aux_loss:0.0112 train_time:8719867ms step_avg:1871.22ms step:4671/6250 train_loss:1.8261 aux_loss:0.0115 train_time:8721185ms step_avg:1871.10ms step:4672/6250 train_loss:1.7698 aux_loss:0.0110 train_time:8723482ms step_avg:1871.19ms step:4673/6250 train_loss:1.8772 aux_loss:0.0114 train_time:8725328ms step_avg:1871.18ms step:4674/6250 train_loss:1.8192 aux_loss:0.0114 train_time:8727011ms step_avg:1871.14ms step:4675/6250 train_loss:1.8930 aux_loss:0.0109 train_time:8728541ms step_avg:1871.07ms step:4676/6250 train_loss:1.8880 aux_loss:0.0111 train_time:8729965ms step_avg:1870.97ms step:4677/6250 train_loss:1.9309 aux_loss:0.0114 train_time:8731816ms step_avg:1870.97ms step:4678/6250 train_loss:1.7802 aux_loss:0.0110 train_time:8733307ms step_avg:1870.89ms step:4679/6250 train_loss:1.7750 aux_loss:0.0107 train_time:8734707ms step_avg:1870.79ms step:4680/6250 train_loss:1.8660 aux_loss:0.0113 train_time:8736327ms step_avg:1870.73ms step:4681/6250 train_loss:1.8409 aux_loss:0.0112 train_time:8737823ms step_avg:1870.65ms step:4682/6250 train_loss:1.8066 aux_loss:0.0111 train_time:8739277ms step_avg:1870.56ms step:4683/6250 train_loss:1.8269 aux_loss:0.0111 train_time:8740661ms step_avg:1870.46ms step:4684/6250 train_loss:1.7628 aux_loss:0.0110 train_time:8742072ms step_avg:1870.36ms step:4685/6250 train_loss:1.8594 aux_loss:0.0118 train_time:8744094ms step_avg:1870.39ms step:4686/6250 train_loss:1.8656 aux_loss:0.0110 train_time:8746261ms step_avg:1870.46ms step:4687/6250 train_loss:1.8573 aux_loss:0.0109 train_time:8747680ms step_avg:1870.36ms step:4688/6250 train_loss:1.8792 aux_loss:0.0113 train_time:8749745ms step_avg:1870.40ms step:4689/6250 train_loss:1.8809 aux_loss:0.0112 train_time:8751750ms step_avg:1870.43ms step:4690/6250 train_loss:1.8043 aux_loss:0.0113 train_time:8753287ms step_avg:1870.36ms step:4691/6250 train_loss:1.8127 aux_loss:0.0111 train_time:8754569ms step_avg:1870.23ms step:4692/6250 train_loss:1.8572 aux_loss:0.0113 train_time:8755882ms step_avg:1870.12ms step:4693/6250 train_loss:1.8261 aux_loss:0.0110 train_time:8757243ms step_avg:1870.01ms step:4694/6250 train_loss:1.8187 aux_loss:0.0112 train_time:8758543ms step_avg:1869.89ms step:4695/6250 train_loss:1.8985 aux_loss:0.0113 train_time:8759857ms step_avg:1869.77ms step:4696/6250 train_loss:1.8542 aux_loss:0.0115 train_time:8761695ms step_avg:1869.76ms step:4697/6250 train_loss:1.8432 aux_loss:0.0113 train_time:8763455ms step_avg:1869.74ms step:4698/6250 train_loss:1.9454 aux_loss:0.0108 train_time:8765150ms step_avg:1869.70ms step:4699/6250 train_loss:1.8770 aux_loss:0.0110 train_time:8767179ms step_avg:1869.73ms step:4700/6250 train_loss:1.8949 aux_loss:0.0114 train_time:8769110ms step_avg:1869.75ms step:4701/6250 train_loss:1.8513 aux_loss:0.0112 train_time:8770674ms step_avg:1869.68ms step:4702/6250 train_loss:1.7358 aux_loss:0.0109 train_time:8773111ms step_avg:1869.80ms step:4703/6250 train_loss:1.8648 aux_loss:0.0111 train_time:8774608ms step_avg:1869.72ms step:4704/6250 train_loss:1.8812 aux_loss:0.0112 train_time:8775960ms step_avg:1869.61ms step:4705/6250 train_loss:1.8621 aux_loss:0.0111 train_time:8777345ms step_avg:1869.51ms step:4706/6250 train_loss:2.0147 aux_loss:0.0111 train_time:8778653ms step_avg:1869.39ms step:4707/6250 train_loss:1.8972 aux_loss:0.0112 train_time:8779991ms step_avg:1869.28ms step:4708/6250 train_loss:1.9252 aux_loss:0.0113 train_time:8781955ms step_avg:1869.30ms step:4709/6250 train_loss:1.7843 aux_loss:0.0109 train_time:8783497ms step_avg:1869.23ms step:4710/6250 train_loss:1.7295 aux_loss:0.0107 train_time:8785030ms step_avg:1869.16ms step:4711/6250 train_loss:1.7619 aux_loss:0.0111 train_time:8787138ms step_avg:1869.21ms step:4712/6250 train_loss:1.7965 aux_loss:0.0109 train_time:8789043ms step_avg:1869.21ms step:4713/6250 train_loss:1.8866 aux_loss:0.0112 train_time:8791044ms step_avg:1869.24ms step:4714/6250 train_loss:1.8263 aux_loss:0.0113 train_time:8793035ms step_avg:1869.27ms step:4715/6250 train_loss:1.9153 aux_loss:0.0114 train_time:8795756ms step_avg:1869.45ms step:4716/6250 train_loss:1.8202 aux_loss:0.0108 train_time:8797659ms step_avg:1869.46ms step:4717/6250 train_loss:1.8474 aux_loss:0.0114 train_time:8798974ms step_avg:1869.34ms step:4718/6250 train_loss:1.7585 aux_loss:0.0113 train_time:8800334ms step_avg:1869.23ms step:4719/6250 train_loss:1.8722 aux_loss:0.0110 train_time:8801674ms step_avg:1869.12ms step:4720/6250 train_loss:1.9407 aux_loss:0.0112 train_time:8803034ms step_avg:1869.01ms step:4721/6250 train_loss:1.8542 aux_loss:0.0109 train_time:8804334ms step_avg:1868.89ms step:4722/6250 train_loss:1.9367 aux_loss:0.0110 train_time:8805639ms step_avg:1868.77ms step:4723/6250 train_loss:1.8964 aux_loss:0.0109 train_time:8806950ms step_avg:1868.65ms step:4724/6250 train_loss:1.7929 aux_loss:0.0110 train_time:8808783ms step_avg:1868.64ms step:4725/6250 train_loss:1.7934 aux_loss:0.0114 train_time:8815945ms step_avg:1869.77ms step:4726/6250 train_loss:1.7717 aux_loss:0.0110 train_time:8817527ms step_avg:1869.70ms step:4727/6250 train_loss:1.8203 aux_loss:0.0111 train_time:8818818ms step_avg:1869.58ms step:4728/6250 train_loss:1.8307 aux_loss:0.0109 train_time:8820212ms step_avg:1869.48ms step:4729/6250 train_loss:1.7662 aux_loss:0.0111 train_time:8822476ms step_avg:1869.56ms step:4730/6250 train_loss:1.8533 aux_loss:0.0114 train_time:8824289ms step_avg:1869.55ms step:4731/6250 train_loss:1.7894 aux_loss:0.0111 train_time:8825643ms step_avg:1869.44ms step:4732/6250 train_loss:1.7193 aux_loss:0.0108 train_time:8827484ms step_avg:1869.44ms step:4733/6250 train_loss:1.9066 aux_loss:0.0111 train_time:8828840ms step_avg:1869.33ms step:4734/6250 train_loss:1.7269 aux_loss:0.0110 train_time:8830210ms step_avg:1869.22ms step:4735/6250 train_loss:1.8708 aux_loss:0.0110 train_time:8832007ms step_avg:1869.21ms step:4736/6250 train_loss:1.7386 aux_loss:0.0109 train_time:8833666ms step_avg:1869.16ms step:4737/6250 train_loss:1.8687 aux_loss:0.0114 train_time:8835357ms step_avg:1869.13ms step:4738/6250 train_loss:1.8180 aux_loss:0.0109 train_time:8836772ms step_avg:1869.03ms step:4739/6250 train_loss:1.8179 aux_loss:0.0113 train_time:8838957ms step_avg:1869.10ms step:4740/6250 train_loss:1.7524 aux_loss:0.0111 train_time:8840630ms step_avg:1869.05ms step:4741/6250 train_loss:1.7944 aux_loss:0.0110 train_time:8841978ms step_avg:1868.94ms step:4742/6250 train_loss:1.8805 aux_loss:0.0113 train_time:8844193ms step_avg:1869.02ms step:4743/6250 train_loss:1.8209 aux_loss:0.0112 train_time:8845827ms step_avg:1868.97ms step:4744/6250 train_loss:1.9671 aux_loss:0.0115 train_time:8847599ms step_avg:1868.95ms step:4745/6250 train_loss:1.9939 aux_loss:0.0114 train_time:8849664ms step_avg:1868.99ms step:4746/6250 train_loss:1.7913 aux_loss:0.0118 train_time:8852194ms step_avg:1869.13ms step:4747/6250 train_loss:1.8534 aux_loss:0.0113 train_time:8854693ms step_avg:1869.26ms step:4748/6250 train_loss:1.8410 aux_loss:0.0109 train_time:8856515ms step_avg:1869.25ms step:4749/6250 train_loss:1.9096 aux_loss:0.0108 train_time:8858445ms step_avg:1869.26ms step:4750/6250 train_loss:1.7960 aux_loss:0.0115 train_time:8859840ms step_avg:1869.16ms step:4751/6250 train_loss:1.8165 aux_loss:0.0111 train_time:8861264ms step_avg:1869.07ms step:4752/6250 train_loss:1.8040 aux_loss:0.0108 train_time:8862631ms step_avg:1868.96ms step:4753/6250 train_loss:1.8384 aux_loss:0.0109 train_time:8863945ms step_avg:1868.85ms step:4754/6250 train_loss:1.8180 aux_loss:0.0110 train_time:8865786ms step_avg:1868.84ms step:4755/6250 train_loss:1.7748 aux_loss:0.0109 train_time:8867222ms step_avg:1868.75ms step:4756/6250 train_loss:1.8180 aux_loss:0.0110 train_time:8869120ms step_avg:1868.76ms step:4757/6250 train_loss:1.8689 aux_loss:0.0111 train_time:8871227ms step_avg:1868.81ms step:4758/6250 train_loss:1.7969 aux_loss:0.0110 train_time:8872746ms step_avg:1868.73ms step:4759/6250 train_loss:1.7798 aux_loss:0.0109 train_time:8874119ms step_avg:1868.63ms step:4760/6250 train_loss:1.8442 aux_loss:0.0112 train_time:8875738ms step_avg:1868.58ms step:4761/6250 train_loss:1.7335 aux_loss:0.0108 train_time:8878213ms step_avg:1868.70ms step:4762/6250 train_loss:1.7919 aux_loss:0.0106 train_time:8880122ms step_avg:1868.71ms step:4763/6250 train_loss:1.7898 aux_loss:0.0113 train_time:8882295ms step_avg:1868.78ms step:4764/6250 train_loss:1.8712 aux_loss:0.0109 train_time:8883958ms step_avg:1868.73ms step:4765/6250 train_loss:1.8091 aux_loss:0.0105 train_time:8886049ms step_avg:1868.78ms step:4766/6250 train_loss:1.7188 aux_loss:0.0105 train_time:8888088ms step_avg:1868.82ms step:4767/6250 train_loss:1.8383 aux_loss:0.0108 train_time:8889425ms step_avg:1868.70ms step:4768/6250 train_loss:1.8659 aux_loss:0.0106 train_time:8891358ms step_avg:1868.72ms step:4769/6250 train_loss:1.8336 aux_loss:0.0106 train_time:8892942ms step_avg:1868.66ms step:4770/6250 train_loss:1.7612 aux_loss:0.0105 train_time:8895278ms step_avg:1868.76ms step:4771/6250 train_loss:1.8340 aux_loss:0.0105 train_time:8896639ms step_avg:1868.65ms step:4772/6250 train_loss:1.7673 aux_loss:0.0106 train_time:8898392ms step_avg:1868.62ms step:4773/6250 train_loss:1.8889 aux_loss:0.0110 train_time:8899979ms step_avg:1868.57ms step:4774/6250 train_loss:1.8470 aux_loss:0.0111 train_time:8901338ms step_avg:1868.46ms step:4775/6250 train_loss:1.7746 aux_loss:0.0110 train_time:8902650ms step_avg:1868.34ms step:4776/6250 train_loss:1.7493 aux_loss:0.0111 train_time:8904486ms step_avg:1868.34ms step:4777/6250 train_loss:1.7761 aux_loss:0.0108 train_time:8906063ms step_avg:1868.27ms step:4778/6250 train_loss:1.8501 aux_loss:0.0110 train_time:8909027ms step_avg:1868.50ms step:4779/6250 train_loss:1.7940 aux_loss:0.0112 train_time:8911856ms step_avg:1868.71ms step:4780/6250 train_loss:1.7959 aux_loss:0.0108 train_time:8913624ms step_avg:1868.68ms step:4781/6250 train_loss:1.7972 aux_loss:0.0107 train_time:8915065ms step_avg:1868.59ms step:4782/6250 train_loss:1.8084 aux_loss:0.0109 train_time:8916337ms step_avg:1868.47ms step:4783/6250 train_loss:1.7826 aux_loss:0.0108 train_time:8918018ms step_avg:1868.43ms step:4784/6250 train_loss:1.7803 aux_loss:0.0108 train_time:8919350ms step_avg:1868.32ms step:4785/6250 train_loss:1.7843 aux_loss:0.0107 train_time:8920670ms step_avg:1868.20ms step:4786/6250 train_loss:1.7611 aux_loss:0.0109 train_time:8922031ms step_avg:1868.10ms step:4787/6250 train_loss:1.7792 aux_loss:0.0107 train_time:8924007ms step_avg:1868.12ms step:4788/6250 train_loss:1.8183 aux_loss:0.0105 train_time:8925662ms step_avg:1868.08ms step:4789/6250 train_loss:1.7809 aux_loss:0.0107 train_time:8927475ms step_avg:1868.06ms step:4790/6250 train_loss:1.8117 aux_loss:0.0110 train_time:8929450ms step_avg:1868.09ms step:4791/6250 train_loss:1.8908 aux_loss:0.0110 train_time:8931138ms step_avg:1868.05ms step:4792/6250 train_loss:1.9042 aux_loss:0.0109 train_time:8932418ms step_avg:1867.93ms step:4793/6250 train_loss:1.8768 aux_loss:0.0109 train_time:8933805ms step_avg:1867.82ms step:4794/6250 train_loss:1.8466 aux_loss:0.0107 train_time:8935161ms step_avg:1867.72ms step:4795/6250 train_loss:1.8094 aux_loss:0.0108 train_time:8936476ms step_avg:1867.60ms step:4796/6250 train_loss:1.7317 aux_loss:0.0105 train_time:8937755ms step_avg:1867.48ms step:4797/6250 train_loss:1.9021 aux_loss:0.0109 train_time:8939094ms step_avg:1867.37ms step:4798/6250 train_loss:1.8797 aux_loss:0.0113 train_time:8940433ms step_avg:1867.26ms step:4799/6250 train_loss:1.8450 aux_loss:0.0110 train_time:8941739ms step_avg:1867.14ms step:4800/6250 train_loss:1.7835 aux_loss:0.0106 train_time:8943039ms step_avg:1867.02ms step:4801/6250 train_loss:1.8152 aux_loss:0.0106 train_time:8944429ms step_avg:1866.92ms step:4802/6250 train_loss:1.7870 aux_loss:0.0107 train_time:8945784ms step_avg:1866.82ms step:4803/6250 train_loss:1.8272 aux_loss:0.0107 train_time:8947914ms step_avg:1866.87ms step:4804/6250 train_loss:1.7584 aux_loss:0.0104 train_time:8949228ms step_avg:1866.76ms step:4805/6250 train_loss:1.9035 aux_loss:0.0107 train_time:8950845ms step_avg:1866.70ms step:4806/6250 train_loss:1.8587 aux_loss:0.0108 train_time:8952274ms step_avg:1866.61ms step:4807/6250 train_loss:1.8492 aux_loss:0.0109 train_time:8953587ms step_avg:1866.50ms step:4808/6250 train_loss:1.8225 aux_loss:0.0103 train_time:8955226ms step_avg:1866.45ms step:4809/6250 train_loss:1.8576 aux_loss:0.0109 train_time:8957185ms step_avg:1866.47ms step:4810/6250 train_loss:1.8782 aux_loss:0.0113 train_time:8958672ms step_avg:1866.39ms step:4811/6250 train_loss:1.8456 aux_loss:0.0108 train_time:8960394ms step_avg:1866.36ms step:4812/6250 train_loss:1.7649 aux_loss:0.0107 train_time:8961821ms step_avg:1866.27ms step:4813/6250 train_loss:1.8303 aux_loss:0.0110 train_time:8963369ms step_avg:1866.20ms step:4814/6250 train_loss:1.8236 aux_loss:0.0110 train_time:8965101ms step_avg:1866.17ms step:4815/6250 train_loss:1.8686 aux_loss:0.0110 train_time:8967249ms step_avg:1866.23ms step:4816/6250 train_loss:1.8628 aux_loss:0.0108 train_time:8968547ms step_avg:1866.11ms step:4817/6250 train_loss:1.8489 aux_loss:0.0111 train_time:8970132ms step_avg:1866.06ms step:4818/6250 train_loss:1.8223 aux_loss:0.0109 train_time:8971716ms step_avg:1866.00ms step:4819/6250 train_loss:1.8816 aux_loss:0.0111 train_time:8973111ms step_avg:1865.90ms step:4820/6250 train_loss:1.8767 aux_loss:0.0108 train_time:8974798ms step_avg:1865.86ms step:4821/6250 train_loss:1.8745 aux_loss:0.0107 train_time:8976104ms step_avg:1865.75ms step:4822/6250 train_loss:1.8386 aux_loss:0.0107 train_time:8977524ms step_avg:1865.65ms step:4823/6250 train_loss:1.8185 aux_loss:0.0108 train_time:8979456ms step_avg:1865.67ms step:4824/6250 train_loss:1.8104 aux_loss:0.0108 train_time:8980809ms step_avg:1865.56ms step:4825/6250 train_loss:1.8436 aux_loss:0.0107 train_time:8982135ms step_avg:1865.45ms step:4826/6250 train_loss:1.7813 aux_loss:0.0106 train_time:8983890ms step_avg:1865.43ms step:4827/6250 train_loss:1.7826 aux_loss:0.0108 train_time:8986119ms step_avg:1865.50ms step:4828/6250 train_loss:1.7955 aux_loss:0.0108 train_time:8987584ms step_avg:1865.42ms step:4829/6250 train_loss:1.8210 aux_loss:0.0108 train_time:8989583ms step_avg:1865.45ms step:4830/6250 train_loss:1.8850 aux_loss:0.0110 train_time:8990981ms step_avg:1865.35ms step:4831/6250 train_loss:1.8578 aux_loss:0.0112 train_time:8992307ms step_avg:1865.24ms step:4832/6250 train_loss:1.9318 aux_loss:0.0113 train_time:8993909ms step_avg:1865.18ms step:4833/6250 train_loss:1.8473 aux_loss:0.0107 train_time:8995223ms step_avg:1865.07ms step:4834/6250 train_loss:1.8064 aux_loss:0.0108 train_time:8996618ms step_avg:1864.97ms step:4835/6250 train_loss:1.8915 aux_loss:0.0109 train_time:8998075ms step_avg:1864.89ms step:4836/6250 train_loss:1.8182 aux_loss:0.0107 train_time:9000109ms step_avg:1864.92ms step:4837/6250 train_loss:1.8452 aux_loss:0.0104 train_time:9002166ms step_avg:1864.96ms step:4838/6250 train_loss:1.8187 aux_loss:0.0101 train_time:9003491ms step_avg:1864.85ms step:4839/6250 train_loss:1.7829 aux_loss:0.0105 train_time:9004819ms step_avg:1864.74ms step:4840/6250 train_loss:1.8002 aux_loss:0.0105 train_time:9006179ms step_avg:1864.63ms step:4841/6250 train_loss:1.8260 aux_loss:0.0108 train_time:9007820ms step_avg:1864.59ms step:4842/6250 train_loss:1.8847 aux_loss:0.0106 train_time:9009402ms step_avg:1864.53ms step:4843/6250 train_loss:1.8219 aux_loss:0.0105 train_time:9010992ms step_avg:1864.47ms step:4844/6250 train_loss:1.9011 aux_loss:0.0111 train_time:9012418ms step_avg:1864.38ms step:4845/6250 train_loss:1.8027 aux_loss:0.0114 train_time:9014364ms step_avg:1864.40ms step:4846/6250 train_loss:1.7139 aux_loss:0.0103 train_time:9017401ms step_avg:1864.64ms step:4847/6250 train_loss:1.8033 aux_loss:0.0106 train_time:9019241ms step_avg:1864.64ms step:4848/6250 train_loss:1.7572 aux_loss:0.0107 train_time:9021181ms step_avg:1864.65ms step:4849/6250 train_loss:1.8543 aux_loss:0.0110 train_time:9022914ms step_avg:1864.62ms step:4850/6250 train_loss:1.7528 aux_loss:0.0105 train_time:9024553ms step_avg:1864.58ms step:4851/6250 train_loss:1.9184 aux_loss:0.0108 train_time:9026034ms step_avg:1864.50ms step:4852/6250 train_loss:1.8429 aux_loss:0.0111 train_time:9028297ms step_avg:1864.58ms step:4853/6250 train_loss:1.7585 aux_loss:0.0110 train_time:9030014ms step_avg:1864.55ms step:4854/6250 train_loss:1.8022 aux_loss:0.0106 train_time:9032418ms step_avg:1864.66ms step:4855/6250 train_loss:1.8462 aux_loss:0.0106 train_time:9034890ms step_avg:1864.79ms step:4856/6250 train_loss:1.8425 aux_loss:0.0103 train_time:9036247ms step_avg:1864.68ms step:4857/6250 train_loss:1.8628 aux_loss:0.0106 train_time:9037609ms step_avg:1864.58ms step:4858/6250 train_loss:1.7365 aux_loss:0.0105 train_time:9038956ms step_avg:1864.47ms step:4859/6250 train_loss:1.9098 aux_loss:0.0109 train_time:9040331ms step_avg:1864.37ms step:4860/6250 train_loss:1.8597 aux_loss:0.0108 train_time:9041684ms step_avg:1864.26ms step:4861/6250 train_loss:1.7320 aux_loss:0.0102 train_time:9043144ms step_avg:1864.18ms step:4862/6250 train_loss:1.7551 aux_loss:0.0104 train_time:9045174ms step_avg:1864.22ms step:4863/6250 train_loss:1.8392 aux_loss:0.0104 train_time:9046783ms step_avg:1864.16ms step:4864/6250 train_loss:1.7984 aux_loss:0.0103 train_time:9048541ms step_avg:1864.14ms step:4865/6250 train_loss:1.8741 aux_loss:0.0106 train_time:9049890ms step_avg:1864.04ms step:4866/6250 train_loss:1.8262 aux_loss:0.0108 train_time:9052260ms step_avg:1864.14ms step:4867/6250 train_loss:1.8700 aux_loss:0.0110 train_time:9054153ms step_avg:1864.15ms step:4868/6250 train_loss:1.8576 aux_loss:0.0110 train_time:9057772ms step_avg:1864.51ms step:4869/6250 train_loss:1.9588 aux_loss:0.0112 train_time:9059394ms step_avg:1864.46ms step:4870/6250 train_loss:1.8267 aux_loss:0.0109 train_time:9060703ms step_avg:1864.34ms step:4871/6250 train_loss:1.8068 aux_loss:0.0108 train_time:9062155ms step_avg:1864.26ms step:4872/6250 train_loss:1.7584 aux_loss:0.0109 train_time:9063505ms step_avg:1864.15ms step:4873/6250 train_loss:1.8568 aux_loss:0.0108 train_time:9064867ms step_avg:1864.05ms step:4874/6250 train_loss:1.8659 aux_loss:0.0107 train_time:9066435ms step_avg:1863.99ms step:4875/6250 train_loss:1.7832 aux_loss:0.0106 train_time:9067674ms step_avg:1863.86ms step:4876/6250 train_loss:1.8286 aux_loss:0.0112 train_time:9069047ms step_avg:1863.76ms step:4877/6250 train_loss:1.7831 aux_loss:0.0108 train_time:9070622ms step_avg:1863.70ms step:4878/6250 train_loss:1.8143 aux_loss:0.0108 train_time:9072087ms step_avg:1863.62ms step:4879/6250 train_loss:1.9037 aux_loss:0.0107 train_time:9074172ms step_avg:1863.66ms step:4880/6250 train_loss:1.8003 aux_loss:0.0106 train_time:9076287ms step_avg:1863.71ms step:4881/6250 train_loss:1.9136 aux_loss:0.0105 train_time:9077794ms step_avg:1863.64ms step:4882/6250 train_loss:1.8302 aux_loss:0.0107 train_time:9079271ms step_avg:1863.56ms step:4883/6250 train_loss:1.8480 aux_loss:0.0108 train_time:9080831ms step_avg:1863.50ms step:4884/6250 train_loss:1.9008 aux_loss:0.0105 train_time:9082126ms step_avg:1863.38ms step:4885/6250 train_loss:1.7594 aux_loss:0.0107 train_time:9083733ms step_avg:1863.33ms step:4886/6250 train_loss:1.7472 aux_loss:0.0107 train_time:9085622ms step_avg:1863.34ms step:4887/6250 train_loss:1.8286 aux_loss:0.0108 train_time:9086969ms step_avg:1863.23ms step:4888/6250 train_loss:1.7954 aux_loss:0.0107 train_time:9088295ms step_avg:1863.12ms step:4889/6250 train_loss:1.7877 aux_loss:0.0105 train_time:9089627ms step_avg:1863.01ms step:4890/6250 train_loss:1.7697 aux_loss:0.0106 train_time:9091193ms step_avg:1862.95ms step:4891/6250 train_loss:1.7595 aux_loss:0.0108 train_time:9092597ms step_avg:1862.86ms step:4892/6250 train_loss:1.8996 aux_loss:0.0109 train_time:9094415ms step_avg:1862.85ms step:4893/6250 train_loss:1.8204 aux_loss:0.0107 train_time:9095927ms step_avg:1862.77ms step:4894/6250 train_loss:1.8155 aux_loss:0.0106 train_time:9097239ms step_avg:1862.66ms step:4895/6250 train_loss:1.7830 aux_loss:0.0106 train_time:9100610ms step_avg:1862.97ms step:4896/6250 train_loss:1.7560 aux_loss:0.0108 train_time:9103309ms step_avg:1863.14ms step:4897/6250 train_loss:1.8641 aux_loss:0.0107 train_time:9107234ms step_avg:1863.56ms step:4898/6250 train_loss:1.8486 aux_loss:0.0106 train_time:9109235ms step_avg:1863.59ms step:4899/6250 train_loss:1.8790 aux_loss:0.0106 train_time:9110668ms step_avg:1863.50ms step:4900/6250 train_loss:1.7998 aux_loss:0.0105 train_time:9112572ms step_avg:1863.51ms step:4901/6250 train_loss:1.9083 aux_loss:0.0109 train_time:9113894ms step_avg:1863.40ms step:4902/6250 train_loss:1.8309 aux_loss:0.0107 train_time:9115188ms step_avg:1863.28ms step:4903/6250 train_loss:1.7933 aux_loss:0.0103 train_time:9118079ms step_avg:1863.49ms step:4904/6250 train_loss:1.8968 aux_loss:0.0112 train_time:9120422ms step_avg:1863.59ms step:4905/6250 train_loss:1.7784 aux_loss:0.0108 train_time:9122523ms step_avg:1863.64ms step:4906/6250 train_loss:1.8208 aux_loss:0.0109 train_time:9123969ms step_avg:1863.56ms step:4907/6250 train_loss:1.8205 aux_loss:0.0106 train_time:9125925ms step_avg:1863.57ms step:4908/6250 train_loss:1.8227 aux_loss:0.0111 train_time:9128358ms step_avg:1863.69ms step:4909/6250 train_loss:1.7948 aux_loss:0.0112 train_time:9131583ms step_avg:1863.97ms step:4910/6250 train_loss:1.8292 aux_loss:0.0106 train_time:9133868ms step_avg:1864.05ms step:4911/6250 train_loss:1.8222 aux_loss:0.0108 train_time:9136885ms step_avg:1864.29ms step:4912/6250 train_loss:1.7711 aux_loss:0.0109 train_time:9138832ms step_avg:1864.31ms step:4913/6250 train_loss:1.8566 aux_loss:0.0106 train_time:9142181ms step_avg:1864.61ms step:4914/6250 train_loss:1.7742 aux_loss:0.0104 train_time:9144480ms step_avg:1864.70ms step:4915/6250 train_loss:1.8659 aux_loss:0.0105 train_time:9146966ms step_avg:1864.82ms step:4916/6250 train_loss:1.7761 aux_loss:0.0105 train_time:9148771ms step_avg:1864.81ms step:4917/6250 train_loss:1.7939 aux_loss:0.0105 train_time:9151301ms step_avg:1864.95ms step:4918/6250 train_loss:1.8351 aux_loss:0.0106 train_time:9153348ms step_avg:1864.99ms step:4919/6250 train_loss:1.9219 aux_loss:0.0109 train_time:9157019ms step_avg:1865.35ms step:4920/6250 train_loss:1.7617 aux_loss:0.0108 train_time:9159281ms step_avg:1865.43ms step:4921/6250 train_loss:1.8052 aux_loss:0.0105 train_time:9161891ms step_avg:1865.59ms step:4922/6250 train_loss:1.8486 aux_loss:0.0103 train_time:9164239ms step_avg:1865.68ms step:4923/6250 train_loss:1.8132 aux_loss:0.0102 train_time:9166858ms step_avg:1865.84ms step:4924/6250 train_loss:1.8007 aux_loss:0.0108 train_time:9169100ms step_avg:1865.91ms step:4925/6250 train_loss:1.8359 aux_loss:0.0107 train_time:9171004ms step_avg:1865.92ms step:4926/6250 train_loss:1.7908 aux_loss:0.0108 train_time:9174162ms step_avg:1866.18ms step:4927/6250 train_loss:1.8030 aux_loss:0.0107 train_time:9176283ms step_avg:1866.24ms step:4928/6250 train_loss:1.8159 aux_loss:0.0110 train_time:9178703ms step_avg:1866.35ms step:4929/6250 train_loss:1.8401 aux_loss:0.0107 train_time:9180615ms step_avg:1866.36ms step:4930/6250 train_loss:1.8694 aux_loss:0.0112 train_time:9182697ms step_avg:1866.40ms step:4931/6250 train_loss:1.9007 aux_loss:0.0108 train_time:9185431ms step_avg:1866.58ms step:4932/6250 train_loss:1.9301 aux_loss:0.0110 train_time:9187514ms step_avg:1866.62ms step:4933/6250 train_loss:1.8249 aux_loss:0.0104 train_time:9193423ms step_avg:1867.44ms step:4934/6250 train_loss:1.7724 aux_loss:0.0103 train_time:9194743ms step_avg:1867.33ms step:4935/6250 train_loss:1.7721 aux_loss:0.0106 train_time:9196679ms step_avg:1867.35ms step:4936/6250 train_loss:1.8094 aux_loss:0.0106 train_time:9199233ms step_avg:1867.49ms step:4937/6250 train_loss:1.8235 aux_loss:0.0106 train_time:9200558ms step_avg:1867.38ms step:4938/6250 train_loss:1.8825 aux_loss:0.0107 train_time:9201824ms step_avg:1867.25ms step:4939/6250 train_loss:1.8180 aux_loss:0.0107 train_time:9203089ms step_avg:1867.13ms step:4940/6250 train_loss:1.7431 aux_loss:0.0106 train_time:9204617ms step_avg:1867.06ms step:4941/6250 train_loss:1.7768 aux_loss:0.0105 train_time:9206140ms step_avg:1866.99ms step:4942/6250 train_loss:1.7946 aux_loss:0.0105 train_time:9208380ms step_avg:1867.07ms step:4943/6250 train_loss:1.8312 aux_loss:0.0108 train_time:9211054ms step_avg:1867.23ms step:4944/6250 train_loss:1.7309 aux_loss:0.0105 train_time:9216721ms step_avg:1868.00ms step:4945/6250 train_loss:1.9725 aux_loss:0.0109 train_time:9220381ms step_avg:1868.37ms step:4946/6250 train_loss:1.8555 aux_loss:0.0107 train_time:9223450ms step_avg:1868.61ms step:4947/6250 train_loss:1.8672 aux_loss:0.0111 train_time:9232002ms step_avg:1869.96ms step:4948/6250 train_loss:1.8237 aux_loss:0.0107 train_time:9235528ms step_avg:1870.30ms step:4949/6250 train_loss:1.7570 aux_loss:0.0110 train_time:9237690ms step_avg:1870.36ms step:4950/6250 train_loss:1.7499 aux_loss:0.0107 train_time:9239892ms step_avg:1870.42ms step:4951/6250 train_loss:1.8819 aux_loss:0.0108 train_time:9242082ms step_avg:1870.49ms step:4952/6250 train_loss:1.9103 aux_loss:0.0108 train_time:9243554ms step_avg:1870.41ms step:4953/6250 train_loss:1.8245 aux_loss:0.0108 train_time:9245627ms step_avg:1870.45ms step:4954/6250 train_loss:1.8864 aux_loss:0.0109 train_time:9247388ms step_avg:1870.43ms step:4955/6250 train_loss:1.8352 aux_loss:0.0106 train_time:9249240ms step_avg:1870.42ms step:4956/6250 train_loss:1.8889 aux_loss:0.0107 train_time:9250668ms step_avg:1870.33ms step:4957/6250 train_loss:1.8485 aux_loss:0.0108 train_time:9252047ms step_avg:1870.23ms step:4958/6250 train_loss:1.7959 aux_loss:0.0107 train_time:9253291ms step_avg:1870.11ms step:4959/6250 train_loss:1.8453 aux_loss:0.0107 train_time:9254559ms step_avg:1869.99ms step:4960/6250 train_loss:1.8241 aux_loss:0.0107 train_time:9255868ms step_avg:1869.87ms step:4961/6250 train_loss:1.8694 aux_loss:0.0106 train_time:9257245ms step_avg:1869.77ms step:4962/6250 train_loss:1.8319 aux_loss:0.0104 train_time:9258599ms step_avg:1869.67ms step:4963/6250 train_loss:1.8290 aux_loss:0.0104 train_time:9260014ms step_avg:1869.58ms step:4964/6250 train_loss:1.8148 aux_loss:0.0103 train_time:9261317ms step_avg:1869.46ms step:4965/6250 train_loss:1.8299 aux_loss:0.0106 train_time:9262923ms step_avg:1869.41ms step:4966/6250 train_loss:1.8311 aux_loss:0.0110 train_time:9264632ms step_avg:1869.38ms step:4967/6250 train_loss:1.7758 aux_loss:0.0105 train_time:9266498ms step_avg:1869.38ms step:4968/6250 train_loss:1.8849 aux_loss:0.0108 train_time:9269491ms step_avg:1869.60ms step:4969/6250 train_loss:1.9155 aux_loss:0.0108 train_time:9271594ms step_avg:1869.65ms step:4970/6250 train_loss:1.8863 aux_loss:0.0109 train_time:9272951ms step_avg:1869.55ms step:4971/6250 train_loss:1.7594 aux_loss:0.0109 train_time:9274970ms step_avg:1869.58ms step:4972/6250 train_loss:1.8413 aux_loss:0.0105 train_time:9276336ms step_avg:1869.48ms step:4973/6250 train_loss:1.7844 aux_loss:0.0106 train_time:9277659ms step_avg:1869.37ms step:4974/6250 train_loss:1.7928 aux_loss:0.0105 train_time:9279079ms step_avg:1869.27ms step:4975/6250 train_loss:1.8690 aux_loss:0.0106 train_time:9280408ms step_avg:1869.17ms step:4976/6250 train_loss:1.9591 aux_loss:0.0108 train_time:9281777ms step_avg:1869.07ms step:4977/6250 train_loss:1.8195 aux_loss:0.0113 train_time:9283115ms step_avg:1868.96ms step:4978/6250 train_loss:1.8534 aux_loss:0.0109 train_time:9284446ms step_avg:1868.85ms step:4979/6250 train_loss:1.8382 aux_loss:0.0107 train_time:9285756ms step_avg:1868.74ms step:4980/6250 train_loss:1.8275 aux_loss:0.0109 train_time:9287141ms step_avg:1868.64ms step:4981/6250 train_loss:1.7466 aux_loss:0.0105 train_time:9288420ms step_avg:1868.52ms step:4982/6250 train_loss:1.7803 aux_loss:0.0104 train_time:9289771ms step_avg:1868.42ms step:4983/6250 train_loss:1.8365 aux_loss:0.0105 train_time:9291081ms step_avg:1868.31ms step:4984/6250 train_loss:1.7910 aux_loss:0.0104 train_time:9292417ms step_avg:1868.20ms step:4985/6250 train_loss:1.7634 aux_loss:0.0106 train_time:9293878ms step_avg:1868.12ms step:4986/6250 train_loss:1.8941 aux_loss:0.0108 train_time:9295538ms step_avg:1868.07ms step:4987/6250 train_loss:1.7920 aux_loss:0.0104 train_time:9297891ms step_avg:1868.17ms step:4988/6250 train_loss:1.7972 aux_loss:0.0101 train_time:9300311ms step_avg:1868.28ms step:4989/6250 train_loss:1.7864 aux_loss:0.0103 train_time:9301894ms step_avg:1868.23ms step:4990/6250 train_loss:1.8447 aux_loss:0.0107 train_time:9303443ms step_avg:1868.16ms step:4991/6250 train_loss:1.8372 aux_loss:0.0106 train_time:9305147ms step_avg:1868.13ms step:4992/6250 train_loss:1.7905 aux_loss:0.0104 train_time:9306395ms step_avg:1868.00ms step:4993/6250 train_loss:1.8162 aux_loss:0.0105 train_time:9307756ms step_avg:1867.90ms step:4994/6250 train_loss:1.7454 aux_loss:0.0104 train_time:9309142ms step_avg:1867.81ms step:4995/6250 train_loss:1.8602 aux_loss:0.0105 train_time:9310547ms step_avg:1867.71ms step:4996/6250 train_loss:1.9086 aux_loss:0.0105 train_time:9311875ms step_avg:1867.60ms step:4997/6250 train_loss:1.8930 aux_loss:0.0106 train_time:9313246ms step_avg:1867.50ms step:4998/6250 train_loss:1.8588 aux_loss:0.0108 train_time:9314769ms step_avg:1867.44ms step:4999/6250 train_loss:1.8247 aux_loss:0.0106 train_time:9316194ms step_avg:1867.35ms step:5000/6250 train_loss:1.8084 aux_loss:0.0102 train_time:9317469ms step_avg:1867.23ms step:5001/6250 train_loss:1.7661 aux_loss:0.0103 train_time:9319568ms step_avg:1867.27ms step:5002/6250 train_loss:1.9109 aux_loss:0.0106 train_time:9322284ms step_avg:1867.44ms step:5003/6250 train_loss:1.8551 aux_loss:0.0106 train_time:9324871ms step_avg:1867.59ms step:5004/6250 train_loss:1.8170 aux_loss:0.0104 train_time:9326220ms step_avg:1867.48ms step:5005/6250 train_loss:1.8498 aux_loss:0.0105 train_time:9327543ms step_avg:1867.38ms step:5006/6250 train_loss:1.9069 aux_loss:0.0105 train_time:9329010ms step_avg:1867.30ms step:5007/6250 train_loss:1.7984 aux_loss:0.0103 train_time:9330402ms step_avg:1867.20ms step:5008/6250 train_loss:1.7873 aux_loss:0.0103 train_time:9332129ms step_avg:1867.17ms step:5009/6250 train_loss:1.8113 aux_loss:0.0105 train_time:9333692ms step_avg:1867.11ms step:5010/6250 train_loss:1.8807 aux_loss:0.0106 train_time:9335030ms step_avg:1867.01ms step:5011/6250 train_loss:1.7587 aux_loss:0.0107 train_time:9337051ms step_avg:1867.04ms step:5012/6250 train_loss:1.7897 aux_loss:0.0107 train_time:9338990ms step_avg:1867.05ms step:5013/6250 train_loss:1.8324 aux_loss:0.0104 train_time:9340565ms step_avg:1866.99ms step:5014/6250 train_loss:1.8169 aux_loss:0.0107 train_time:9341915ms step_avg:1866.89ms step:5015/6250 train_loss:1.8947 aux_loss:0.0105 train_time:9343098ms step_avg:1866.75ms step:5016/6250 train_loss:1.7960 aux_loss:0.0106 train_time:9344455ms step_avg:1866.65ms step:5017/6250 train_loss:1.8428 aux_loss:0.0109 train_time:9348158ms step_avg:1867.02ms step:5018/6250 train_loss:1.8648 aux_loss:0.0106 train_time:9351048ms step_avg:1867.22ms step:5019/6250 train_loss:1.8532 aux_loss:0.0103 train_time:9352814ms step_avg:1867.20ms step:5020/6250 train_loss:1.7802 aux_loss:0.0104 train_time:9354365ms step_avg:1867.14ms step:5021/6250 train_loss:1.8514 aux_loss:0.0104 train_time:9356252ms step_avg:1867.14ms step:5022/6250 train_loss:1.8281 aux_loss:0.0104 train_time:9357795ms step_avg:1867.08ms step:5023/6250 train_loss:1.7530 aux_loss:0.0102 train_time:9359913ms step_avg:1867.13ms step:5024/6250 train_loss:1.8673 aux_loss:0.0105 train_time:9362605ms step_avg:1867.29ms step:5025/6250 train_loss:1.8098 aux_loss:0.0101 train_time:9366396ms step_avg:1867.68ms step:5026/6250 train_loss:1.7583 aux_loss:0.0102 train_time:9371207ms step_avg:1868.26ms step:5027/6250 train_loss:1.8405 aux_loss:0.0101 train_time:9374162ms step_avg:1868.48ms step:5028/6250 train_loss:1.8354 aux_loss:0.0102 train_time:9376928ms step_avg:1868.66ms step:5029/6250 train_loss:1.7724 aux_loss:0.0100 train_time:9381262ms step_avg:1869.15ms step:5030/6250 train_loss:1.7916 aux_loss:0.0101 train_time:9384799ms step_avg:1869.48ms step:5031/6250 train_loss:1.7701 aux_loss:0.0102 train_time:9389066ms step_avg:1869.96ms step:5032/6250 train_loss:1.8443 aux_loss:0.0106 train_time:9391142ms step_avg:1870.00ms step:5033/6250 train_loss:1.7558 aux_loss:0.0101 train_time:9393359ms step_avg:1870.07ms step:5034/6250 train_loss:1.8532 aux_loss:0.0102 train_time:9396803ms step_avg:1870.38ms step:5035/6250 train_loss:1.8324 aux_loss:0.0106 train_time:9406354ms step_avg:1871.91ms step:5036/6250 train_loss:1.8461 aux_loss:0.0103 train_time:9438142ms step_avg:1877.86ms step:5037/6250 train_loss:1.8338 aux_loss:0.0100 train_time:9442332ms step_avg:1878.32ms step:5038/6250 train_loss:1.8239 aux_loss:0.0101 train_time:9444680ms step_avg:1878.42ms step:5039/6250 train_loss:1.8030 aux_loss:0.0102 train_time:9446305ms step_avg:1878.37ms step:5040/6250 train_loss:1.7615 aux_loss:0.0103 train_time:9447819ms step_avg:1878.29ms step:5041/6250 train_loss:1.7586 aux_loss:0.0100 train_time:9449194ms step_avg:1878.19ms step:5042/6250 train_loss:1.8047 aux_loss:0.0101 train_time:9450612ms step_avg:1878.10ms step:5043/6250 train_loss:1.7876 aux_loss:0.0104 train_time:9451894ms step_avg:1877.98ms step:5044/6250 train_loss:1.7625 aux_loss:0.0103 train_time:9453284ms step_avg:1877.89ms step:5045/6250 train_loss:1.7941 aux_loss:0.0099 train_time:9454801ms step_avg:1877.82ms step:5046/6250 train_loss:1.7511 aux_loss:0.0103 train_time:9456215ms step_avg:1877.72ms step:5047/6250 train_loss:1.8119 aux_loss:0.0104 train_time:9457648ms step_avg:1877.64ms step:5048/6250 train_loss:1.7584 aux_loss:0.0100 train_time:9459322ms step_avg:1877.59ms step:5049/6250 train_loss:1.7764 aux_loss:0.0103 train_time:9460707ms step_avg:1877.50ms step:5050/6250 train_loss:1.7484 aux_loss:0.0106 train_time:9463029ms step_avg:1877.59ms step:5051/6250 train_loss:1.7770 aux_loss:0.0102 train_time:9466448ms step_avg:1877.89ms step:5052/6250 train_loss:1.7791 aux_loss:0.0104 train_time:9468931ms step_avg:1878.01ms step:5053/6250 train_loss:1.8730 aux_loss:0.0105 train_time:9471869ms step_avg:1878.22ms step:5054/6250 train_loss:1.8447 aux_loss:0.0105 train_time:9473382ms step_avg:1878.15ms step:5055/6250 train_loss:1.8457 aux_loss:0.0104 train_time:9475586ms step_avg:1878.21ms step:5056/6250 train_loss:1.8591 aux_loss:0.0102 train_time:9477083ms step_avg:1878.14ms step:5057/6250 train_loss:1.7522 aux_loss:0.0102 train_time:9478718ms step_avg:1878.09ms step:5058/6250 train_loss:1.7079 aux_loss:0.0105 train_time:9480476ms step_avg:1878.07ms step:5059/6250 train_loss:1.8159 aux_loss:0.0108 train_time:9482236ms step_avg:1878.04ms step:5060/6250 train_loss:1.8016 aux_loss:0.0103 train_time:9484917ms step_avg:1878.20ms step:5061/6250 train_loss:1.7852 aux_loss:0.0101 train_time:9486599ms step_avg:1878.16ms step:5062/6250 train_loss:1.7756 aux_loss:0.0104 train_time:9489584ms step_avg:1878.38ms step:5063/6250 train_loss:1.8364 aux_loss:0.0105 train_time:9493747ms step_avg:1878.83ms step:5064/6250 train_loss:1.7385 aux_loss:0.0103 train_time:9495263ms step_avg:1878.76ms step:5065/6250 train_loss:1.8196 aux_loss:0.0104 train_time:9497502ms step_avg:1878.83ms step:5066/6250 train_loss:1.8393 aux_loss:0.0104 train_time:9502109ms step_avg:1879.37ms step:5067/6250 train_loss:1.8687 aux_loss:0.0102 train_time:9504323ms step_avg:1879.44ms step:5068/6250 train_loss:1.8105 aux_loss:0.0103 train_time:9506786ms step_avg:1879.55ms step:5069/6250 train_loss:1.8546 aux_loss:0.0100 train_time:9508907ms step_avg:1879.60ms step:5070/6250 train_loss:1.8155 aux_loss:0.0102 train_time:9511874ms step_avg:1879.82ms step:5071/6250 train_loss:1.7397 aux_loss:0.0103 train_time:9514233ms step_avg:1879.91ms step:5072/6250 train_loss:1.7920 aux_loss:0.0103 train_time:9516145ms step_avg:1879.92ms step:5073/6250 train_loss:1.8275 aux_loss:0.0105 train_time:9518272ms step_avg:1879.97ms step:5074/6250 train_loss:1.8105 aux_loss:0.0110 train_time:9521361ms step_avg:1880.21ms step:5075/6250 train_loss:1.8087 aux_loss:0.0118 train_time:9524055ms step_avg:1880.37ms step:5076/6250 train_loss:1.8319 aux_loss:0.0104 train_time:9526513ms step_avg:1880.48ms step:5077/6250 train_loss:1.7839 aux_loss:0.0103 train_time:9528129ms step_avg:1880.43ms step:5078/6250 train_loss:1.9117 aux_loss:0.0105 train_time:9529722ms step_avg:1880.37ms step:5079/6250 train_loss:1.7979 aux_loss:0.0101 train_time:9532060ms step_avg:1880.46ms step:5080/6250 train_loss:1.8624 aux_loss:0.0105 train_time:9534121ms step_avg:1880.50ms step:5081/6250 train_loss:1.7521 aux_loss:0.0100 train_time:9537161ms step_avg:1880.73ms step:5082/6250 train_loss:1.7234 aux_loss:0.0104 train_time:9539055ms step_avg:1880.73ms step:5083/6250 train_loss:1.8584 aux_loss:0.0104 train_time:9540489ms step_avg:1880.64ms step:5084/6250 train_loss:1.7881 aux_loss:0.0104 train_time:9543134ms step_avg:1880.79ms step:5085/6250 train_loss:1.7925 aux_loss:0.0104 train_time:9544583ms step_avg:1880.71ms step:5086/6250 train_loss:1.8490 aux_loss:0.0105 train_time:9546135ms step_avg:1880.64ms step:5087/6250 train_loss:1.8611 aux_loss:0.0101 train_time:9548097ms step_avg:1880.66ms step:5088/6250 train_loss:1.9413 aux_loss:0.0101 train_time:9551680ms step_avg:1880.99ms step:5089/6250 train_loss:1.7571 aux_loss:0.0096 train_time:9555725ms step_avg:1881.42ms step:5090/6250 train_loss:1.8935 aux_loss:0.0100 train_time:9557573ms step_avg:1881.41ms step:5091/6250 train_loss:1.7856 aux_loss:0.0105 train_time:9559623ms step_avg:1881.45ms step:5092/6250 train_loss:1.7683 aux_loss:0.0102 train_time:9561498ms step_avg:1881.44ms step:5093/6250 train_loss:1.8117 aux_loss:0.0101 train_time:9563639ms step_avg:1881.49ms step:5094/6250 train_loss:1.8349 aux_loss:0.0102 train_time:9565304ms step_avg:1881.45ms step:5095/6250 train_loss:1.8025 aux_loss:0.0103 train_time:9566780ms step_avg:1881.37ms step:5096/6250 train_loss:1.8755 aux_loss:0.0105 train_time:9568636ms step_avg:1881.37ms step:5097/6250 train_loss:1.7973 aux_loss:0.0102 train_time:9570725ms step_avg:1881.41ms step:5098/6250 train_loss:1.8383 aux_loss:0.0102 train_time:9572163ms step_avg:1881.32ms step:5099/6250 train_loss:1.8164 aux_loss:0.0107 train_time:9574229ms step_avg:1881.36ms step:5100/6250 train_loss:1.8878 aux_loss:0.0104 train_time:9576512ms step_avg:1881.44ms step:5101/6250 train_loss:1.8054 aux_loss:0.0100 train_time:9579552ms step_avg:1881.66ms step:5102/6250 train_loss:1.7822 aux_loss:0.0099 train_time:9581702ms step_avg:1881.72ms step:5103/6250 train_loss:1.7943 aux_loss:0.0102 train_time:9584356ms step_avg:1881.87ms step:5104/6250 train_loss:1.7729 aux_loss:0.0104 train_time:9587798ms step_avg:1882.17ms step:5105/6250 train_loss:1.7820 aux_loss:0.0106 train_time:9589480ms step_avg:1882.14ms step:5106/6250 train_loss:1.7952 aux_loss:0.0102 train_time:9590980ms step_avg:1882.06ms step:5107/6250 train_loss:1.9428 aux_loss:0.0104 train_time:9592646ms step_avg:1882.02ms step:5108/6250 train_loss:1.8158 aux_loss:0.0104 train_time:9594049ms step_avg:1881.92ms step:5109/6250 train_loss:1.8220 aux_loss:0.0102 train_time:9595453ms step_avg:1881.83ms step:5110/6250 train_loss:1.8153 aux_loss:0.0101 train_time:9596870ms step_avg:1881.74ms step:5111/6250 train_loss:1.8101 aux_loss:0.0103 train_time:9598142ms step_avg:1881.62ms step:5112/6250 train_loss:1.8306 aux_loss:0.0105 train_time:9599502ms step_avg:1881.52ms step:5113/6250 train_loss:1.8082 aux_loss:0.0102 train_time:9600965ms step_avg:1881.44ms step:5114/6250 train_loss:1.8972 aux_loss:0.0102 train_time:9602470ms step_avg:1881.36ms step:5115/6250 train_loss:1.8785 aux_loss:0.0103 train_time:9604377ms step_avg:1881.37ms step:5116/6250 train_loss:1.8452 aux_loss:0.0105 train_time:9606317ms step_avg:1881.38ms step:5117/6250 train_loss:1.8976 aux_loss:0.0106 train_time:9608156ms step_avg:1881.37ms step:5118/6250 train_loss:1.7771 aux_loss:0.0103 train_time:9609892ms step_avg:1881.34ms step:5119/6250 train_loss:1.8427 aux_loss:0.0104 train_time:9611926ms step_avg:1881.37ms step:5120/6250 train_loss:1.7854 aux_loss:0.0106 train_time:9613315ms step_avg:1881.27ms step:5121/6250 train_loss:1.7930 aux_loss:0.0105 train_time:9616512ms step_avg:1881.53ms step:5122/6250 train_loss:1.8978 aux_loss:0.0101 train_time:9619482ms step_avg:1881.75ms step:5123/6250 train_loss:1.8315 aux_loss:0.0100 train_time:9621380ms step_avg:1881.75ms step:5124/6250 train_loss:1.7303 aux_loss:0.0101 train_time:9622792ms step_avg:1881.66ms step:5125/6250 train_loss:1.7552 aux_loss:0.0103 train_time:9624252ms step_avg:1881.57ms step:5126/6250 train_loss:1.8241 aux_loss:0.0102 train_time:9625930ms step_avg:1881.53ms step:5127/6250 train_loss:1.7577 aux_loss:0.0104 train_time:9627285ms step_avg:1881.43ms step:5128/6250 train_loss:1.8077 aux_loss:0.0103 train_time:9629061ms step_avg:1881.41ms step:5129/6250 train_loss:1.8125 aux_loss:0.0099 train_time:9630690ms step_avg:1881.36ms step:5130/6250 train_loss:1.8059 aux_loss:0.0100 train_time:9632104ms step_avg:1881.27ms step:5131/6250 train_loss:1.7827 aux_loss:0.0098 train_time:9633900ms step_avg:1881.25ms step:5132/6250 train_loss:1.7724 aux_loss:0.0099 train_time:9635678ms step_avg:1881.23ms step:5133/6250 train_loss:1.8420 aux_loss:0.0100 train_time:9638292ms step_avg:1881.38ms step:5134/6250 train_loss:1.7941 aux_loss:0.0099 train_time:9640702ms step_avg:1881.48ms step:5135/6250 train_loss:1.8314 aux_loss:0.0103 train_time:9642528ms step_avg:1881.47ms step:5136/6250 train_loss:1.7526 aux_loss:0.0101 train_time:9644420ms step_avg:1881.47ms step:5137/6250 train_loss:1.7892 aux_loss:0.0100 train_time:9646835ms step_avg:1881.57ms step:5138/6250 train_loss:1.8049 aux_loss:0.0105 train_time:9649106ms step_avg:1881.65ms step:5139/6250 train_loss:1.8143 aux_loss:0.0098 train_time:9650757ms step_avg:1881.61ms step:5140/6250 train_loss:1.7461 aux_loss:0.0094 train_time:9653247ms step_avg:1881.72ms step:5141/6250 train_loss:1.8607 aux_loss:0.0100 train_time:9658253ms step_avg:1882.33ms step:5142/6250 train_loss:1.8219 aux_loss:0.0103 train_time:9660028ms step_avg:1882.31ms step:5143/6250 train_loss:1.7564 aux_loss:0.0102 train_time:9661413ms step_avg:1882.22ms step:5144/6250 train_loss:1.8322 aux_loss:0.0099 train_time:9662793ms step_avg:1882.12ms step:5145/6250 train_loss:1.7753 aux_loss:0.0099 train_time:9664147ms step_avg:1882.01ms step:5146/6250 train_loss:1.8041 aux_loss:0.0101 train_time:9665963ms step_avg:1882.00ms step:5147/6250 train_loss:1.8130 aux_loss:0.0102 train_time:9669167ms step_avg:1882.26ms step:5148/6250 train_loss:1.8751 aux_loss:0.0104 train_time:9671169ms step_avg:1882.28ms step:5149/6250 train_loss:1.7522 aux_loss:0.0101 train_time:9673547ms step_avg:1882.38ms step:5150/6250 train_loss:1.7192 aux_loss:0.0100 train_time:9675335ms step_avg:1882.36ms step:5151/6250 train_loss:1.7489 aux_loss:0.0103 train_time:9677940ms step_avg:1882.50ms step:5152/6250 train_loss:1.8189 aux_loss:0.0102 train_time:9680345ms step_avg:1882.60ms step:5153/6250 train_loss:1.8266 aux_loss:0.0101 train_time:9682145ms step_avg:1882.59ms step:5154/6250 train_loss:1.8795 aux_loss:0.0099 train_time:9683898ms step_avg:1882.56ms step:5155/6250 train_loss:1.8020 aux_loss:0.0099 train_time:9686047ms step_avg:1882.61ms step:5156/6250 train_loss:1.8397 aux_loss:0.0099 train_time:9688016ms step_avg:1882.63ms step:5157/6250 train_loss:1.8385 aux_loss:0.0102 train_time:9689402ms step_avg:1882.53ms step:5158/6250 train_loss:1.8065 aux_loss:0.0098 train_time:9691349ms step_avg:1882.55ms step:5159/6250 train_loss:1.9195 aux_loss:0.0102 train_time:9693003ms step_avg:1882.50ms step:5160/6250 train_loss:1.8277 aux_loss:0.0101 train_time:9695728ms step_avg:1882.67ms step:5161/6250 train_loss:1.7950 aux_loss:0.0100 train_time:9697265ms step_avg:1882.60ms step:5162/6250 train_loss:1.9202 aux_loss:0.0104 train_time:9698638ms step_avg:1882.50ms step:5163/6250 train_loss:1.8795 aux_loss:0.0102 train_time:9700016ms step_avg:1882.40ms step:5164/6250 train_loss:1.8949 aux_loss:0.0105 train_time:9701417ms step_avg:1882.31ms step:5165/6250 train_loss:1.8426 aux_loss:0.0101 train_time:9702632ms step_avg:1882.18ms step:5166/6250 train_loss:1.8523 aux_loss:0.0103 train_time:9704009ms step_avg:1882.08ms step:5167/6250 train_loss:1.8432 aux_loss:0.0104 train_time:9705370ms step_avg:1881.98ms step:5168/6250 train_loss:1.7159 aux_loss:0.0099 train_time:9706666ms step_avg:1881.87ms step:5169/6250 train_loss:1.7851 aux_loss:0.0102 train_time:9708055ms step_avg:1881.77ms step:5170/6250 train_loss:1.8516 aux_loss:0.0104 train_time:9710147ms step_avg:1881.81ms step:5171/6250 train_loss:1.7396 aux_loss:0.0097 train_time:9711721ms step_avg:1881.75ms step:5172/6250 train_loss:1.8440 aux_loss:0.0098 train_time:9713080ms step_avg:1881.65ms step:5173/6250 train_loss:1.8799 aux_loss:0.0103 train_time:9715761ms step_avg:1881.81ms step:5174/6250 train_loss:1.7752 aux_loss:0.0104 train_time:9717696ms step_avg:1881.82ms step:5175/6250 train_loss:1.8097 aux_loss:0.0102 train_time:9719374ms step_avg:1881.78ms step:5176/6250 train_loss:1.7461 aux_loss:0.0099 train_time:9722131ms step_avg:1881.95ms step:5177/6250 train_loss:1.7698 aux_loss:0.0103 train_time:9723438ms step_avg:1881.83ms step:5178/6250 train_loss:1.7569 aux_loss:0.0101 train_time:9724809ms step_avg:1881.74ms step:5179/6250 train_loss:1.8083 aux_loss:0.0104 train_time:9726516ms step_avg:1881.70ms step:5180/6250 train_loss:1.8157 aux_loss:0.0101 train_time:9728081ms step_avg:1881.64ms step:5181/6250 train_loss:1.8225 aux_loss:0.0101 train_time:9729396ms step_avg:1881.53ms step:5182/6250 train_loss:1.7335 aux_loss:0.0100 train_time:9730718ms step_avg:1881.42ms step:5183/6250 train_loss:1.8634 aux_loss:0.0103 train_time:9732079ms step_avg:1881.32ms step:5184/6250 train_loss:1.7155 aux_loss:0.0100 train_time:9733462ms step_avg:1881.23ms step:5185/6250 train_loss:1.7756 aux_loss:0.0099 train_time:9735212ms step_avg:1881.20ms step:5186/6250 train_loss:1.7734 aux_loss:0.0097 train_time:9737518ms step_avg:1881.28ms step:5187/6250 train_loss:1.8559 aux_loss:0.0100 train_time:9739773ms step_avg:1881.35ms step:5188/6250 train_loss:1.8640 aux_loss:0.0103 train_time:9741393ms step_avg:1881.30ms step:5189/6250 train_loss:1.7932 aux_loss:0.0100 train_time:9742684ms step_avg:1881.19ms step:5190/6250 train_loss:1.8434 aux_loss:0.0101 train_time:9744013ms step_avg:1881.08ms step:5191/6250 train_loss:1.7525 aux_loss:0.0104 train_time:9745896ms step_avg:1881.08ms step:5192/6250 train_loss:1.8307 aux_loss:0.0102 train_time:9747225ms step_avg:1880.98ms step:5193/6250 train_loss:1.8865 aux_loss:0.0102 train_time:9749672ms step_avg:1881.09ms step:5194/6250 train_loss:1.7442 aux_loss:0.0101 train_time:9751667ms step_avg:1881.11ms step:5195/6250 train_loss:1.6913 aux_loss:0.0101 train_time:9753170ms step_avg:1881.04ms step:5196/6250 train_loss:1.8051 aux_loss:0.0103 train_time:9754808ms step_avg:1880.99ms step:5197/6250 train_loss:1.8508 aux_loss:0.0104 train_time:9761585ms step_avg:1881.93ms step:5198/6250 train_loss:1.7519 aux_loss:0.0100 train_time:9765475ms step_avg:1882.32ms step:5199/6250 train_loss:1.7972 aux_loss:0.0099 train_time:9766795ms step_avg:1882.21ms step:5200/6250 train_loss:1.7981 aux_loss:0.0101 train_time:9769743ms step_avg:1882.42ms step:5201/6250 train_loss:1.8266 aux_loss:0.0101 train_time:9771132ms step_avg:1882.32ms step:5202/6250 train_loss:1.8680 aux_loss:0.0100 train_time:9772496ms step_avg:1882.22ms step:5203/6250 train_loss:1.8477 aux_loss:0.0100 train_time:9774842ms step_avg:1882.31ms step:5204/6250 train_loss:1.7711 aux_loss:0.0101 train_time:9777451ms step_avg:1882.45ms step:5205/6250 train_loss:1.8118 aux_loss:0.0105 train_time:9782748ms step_avg:1883.11ms step:5206/6250 train_loss:1.7651 aux_loss:0.0099 train_time:9786032ms step_avg:1883.38ms step:5207/6250 train_loss:1.9609 aux_loss:0.0102 train_time:9787835ms step_avg:1883.36ms step:5208/6250 train_loss:1.8858 aux_loss:0.0101 train_time:9789558ms step_avg:1883.33ms step:5209/6250 train_loss:1.9029 aux_loss:0.0100 train_time:9793160ms step_avg:1883.66ms step:5210/6250 train_loss:1.8673 aux_loss:0.0098 train_time:9800687ms step_avg:1884.75ms step:5211/6250 train_loss:1.7972 aux_loss:0.0095 train_time:9802091ms step_avg:1884.66ms step:5212/6250 train_loss:1.8113 aux_loss:0.0099 train_time:9812445ms step_avg:1886.28ms step:5213/6250 train_loss:1.7072 aux_loss:0.0100 train_time:9820657ms step_avg:1887.50ms step:5214/6250 train_loss:1.7735 aux_loss:0.0107 train_time:9822222ms step_avg:1887.44ms step:5215/6250 train_loss:1.7243 aux_loss:0.0098 train_time:9823780ms step_avg:1887.37ms step:5216/6250 train_loss:1.8142 aux_loss:0.0098 train_time:9825191ms step_avg:1887.28ms step:5217/6250 train_loss:1.7817 aux_loss:0.0099 train_time:9826581ms step_avg:1887.19ms step:5218/6250 train_loss:1.7832 aux_loss:0.0102 train_time:9828105ms step_avg:1887.12ms step:5219/6250 train_loss:1.7816 aux_loss:0.0102 train_time:9829610ms step_avg:1887.04ms step:5220/6250 train_loss:1.7801 aux_loss:0.0101 train_time:9831280ms step_avg:1887.00ms step:5221/6250 train_loss:1.8384 aux_loss:0.0102 train_time:9832592ms step_avg:1886.89ms step:5222/6250 train_loss:1.8300 aux_loss:0.0101 train_time:9834192ms step_avg:1886.84ms step:5223/6250 train_loss:1.8700 aux_loss:0.0107 train_time:9835634ms step_avg:1886.75ms step:5224/6250 train_loss:1.7974 aux_loss:0.0112 train_time:9837172ms step_avg:1886.68ms step:5225/6250 train_loss:1.7508 aux_loss:0.0099 train_time:9838559ms step_avg:1886.59ms step:5226/6250 train_loss:1.8167 aux_loss:0.0100 train_time:9840016ms step_avg:1886.51ms step:5227/6250 train_loss:1.9516 aux_loss:0.0104 train_time:9841401ms step_avg:1886.41ms step:5228/6250 train_loss:1.7674 aux_loss:0.0099 train_time:9842929ms step_avg:1886.34ms step:5229/6250 train_loss:1.8619 aux_loss:0.0096 train_time:9844337ms step_avg:1886.25ms step:5230/6250 train_loss:1.7826 aux_loss:0.0100 train_time:9846937ms step_avg:1886.39ms step:5231/6250 train_loss:1.8159 aux_loss:0.0106 train_time:9878612ms step_avg:1892.09ms step:5232/6250 train_loss:1.7916 aux_loss:0.0103 train_time:9890196ms step_avg:1893.95ms step:5233/6250 train_loss:1.7761 aux_loss:0.0101 train_time:9900208ms step_avg:1895.50ms step:5234/6250 train_loss:1.8249 aux_loss:0.0102 train_time:9921959ms step_avg:1899.30ms step:5235/6250 train_loss:1.8255 aux_loss:0.0103 train_time:9931951ms step_avg:1900.85ms step:5236/6250 train_loss:1.8547 aux_loss:0.0104 train_time:9935747ms step_avg:1901.21ms step:5237/6250 train_loss:1.7671 aux_loss:0.0098 train_time:9940809ms step_avg:1901.82ms step:5238/6250 train_loss:1.8255 aux_loss:0.0099 train_time:9942130ms step_avg:1901.71ms step:5239/6250 train_loss:1.7897 aux_loss:0.0099 train_time:9943460ms step_avg:1901.60ms step:5240/6250 train_loss:1.7521 aux_loss:0.0100 train_time:9944824ms step_avg:1901.50ms step:5241/6250 train_loss:1.7050 aux_loss:0.0101 train_time:9946369ms step_avg:1901.43ms step:5242/6250 train_loss:1.8119 aux_loss:0.0100 train_time:9947596ms step_avg:1901.30ms step:5243/6250 train_loss:1.8198 aux_loss:0.0101 train_time:9948915ms step_avg:1901.19ms step:5244/6250 train_loss:1.8363 aux_loss:0.0099 train_time:9950313ms step_avg:1901.09ms step:5245/6250 train_loss:1.7999 aux_loss:0.0102 train_time:9951804ms step_avg:1901.01ms step:5246/6250 train_loss:1.8512 aux_loss:0.0103 train_time:9953421ms step_avg:1900.96ms step:5247/6250 train_loss:1.7976 aux_loss:0.0101 train_time:9954883ms step_avg:1900.88ms step:5248/6250 train_loss:1.7742 aux_loss:0.0098 train_time:9956234ms step_avg:1900.77ms step:5249/6250 train_loss:1.7758 aux_loss:0.0102 train_time:9957689ms step_avg:1900.69ms step:5250/6250 train_loss:1.7915 aux_loss:0.0100 train_time:9959235ms step_avg:1900.62ms step:5251/6250 train_loss:1.8960 aux_loss:0.0103 train_time:9960672ms step_avg:1900.53ms step:5252/6250 train_loss:1.8215 aux_loss:0.0097 train_time:9962171ms step_avg:1900.45ms step:5253/6250 train_loss:1.7758 aux_loss:0.0097 train_time:9963719ms step_avg:1900.39ms step:5254/6250 train_loss:1.7380 aux_loss:0.0100 train_time:9967086ms step_avg:1900.66ms step:5255/6250 train_loss:1.8422 aux_loss:0.0101 train_time:9968545ms step_avg:1900.58ms step:5256/6250 train_loss:1.8007 aux_loss:0.0100 train_time:9973938ms step_avg:1901.25ms step:5257/6250 train_loss:1.7340 aux_loss:0.0098 train_time:9975456ms step_avg:1901.17ms step:5258/6250 train_loss:1.8823 aux_loss:0.0098 train_time:9977149ms step_avg:1901.13ms step:5259/6250 train_loss:1.8502 aux_loss:0.0099 train_time:9978686ms step_avg:1901.06ms step:5260/6250 train_loss:1.7923 aux_loss:0.0098 train_time:9980027ms step_avg:1900.96ms step:5261/6250 train_loss:1.7733 aux_loss:0.0096 train_time:9981512ms step_avg:1900.88ms step:5262/6250 train_loss:1.8554 aux_loss:0.0099 train_time:9983114ms step_avg:1900.82ms step:5263/6250 train_loss:1.8292 aux_loss:0.0099 train_time:9984599ms step_avg:1900.74ms step:5264/6250 train_loss:1.8142 aux_loss:0.0100 train_time:9986155ms step_avg:1900.68ms step:5265/6250 train_loss:1.8660 aux_loss:0.0098 train_time:9987584ms step_avg:1900.59ms step:5266/6250 train_loss:1.8460 aux_loss:0.0101 train_time:9989124ms step_avg:1900.52ms step:5267/6250 train_loss:1.8266 aux_loss:0.0105 train_time:9990551ms step_avg:1900.43ms step:5268/6250 train_loss:1.8929 aux_loss:0.0099 train_time:9992085ms step_avg:1900.36ms step:5269/6250 train_loss:1.8500 aux_loss:0.0094 train_time:9993588ms step_avg:1900.28ms step:5270/6250 train_loss:1.7779 aux_loss:0.0096 train_time:9994970ms step_avg:1900.18ms step:5271/6250 train_loss:1.8097 aux_loss:0.0100 train_time:9996343ms step_avg:1900.08ms step:5272/6250 train_loss:1.7774 aux_loss:0.0100 train_time:9997762ms step_avg:1899.99ms step:5273/6250 train_loss:1.8369 aux_loss:0.0103 train_time:10004233ms step_avg:1900.86ms step:5274/6250 train_loss:1.8291 aux_loss:0.0103 train_time:10005620ms step_avg:1900.76ms step:5275/6250 train_loss:1.8495 aux_loss:0.0104 train_time:10007064ms step_avg:1900.68ms step:5276/6250 train_loss:1.7662 aux_loss:0.0101 train_time:10008383ms step_avg:1900.57ms step:5277/6250 train_loss:1.7457 aux_loss:0.0098 train_time:10009942ms step_avg:1900.50ms step:5278/6250 train_loss:1.8735 aux_loss:0.0100 train_time:10011453ms step_avg:1900.43ms step:5279/6250 train_loss:1.8240 aux_loss:0.0098 train_time:10012902ms step_avg:1900.34ms step:5280/6250 train_loss:1.7544 aux_loss:0.0097 train_time:10014407ms step_avg:1900.27ms step:5281/6250 train_loss:1.8060 aux_loss:0.0098 train_time:10016045ms step_avg:1900.22ms step:5282/6250 train_loss:1.7735 aux_loss:0.0099 train_time:10017595ms step_avg:1900.15ms step:5283/6250 train_loss:1.7854 aux_loss:0.0096 train_time:10019168ms step_avg:1900.09ms step:5284/6250 train_loss:1.7801 aux_loss:0.0098 train_time:10020612ms step_avg:1900.00ms step:5285/6250 train_loss:1.7516 aux_loss:0.0098 train_time:10022179ms step_avg:1899.94ms step:5286/6250 train_loss:1.8696 aux_loss:0.0099 train_time:10023837ms step_avg:1899.89ms step:5287/6250 train_loss:1.7756 aux_loss:0.0101 train_time:10025258ms step_avg:1899.80ms step:5288/6250 train_loss:1.8593 aux_loss:0.0096 train_time:10026632ms step_avg:1899.70ms step:5289/6250 train_loss:1.9943 aux_loss:0.0100 train_time:10027906ms step_avg:1899.58ms step:5290/6250 train_loss:1.8846 aux_loss:0.0100 train_time:10029388ms step_avg:1899.51ms step:5291/6250 train_loss:1.8023 aux_loss:0.0100 train_time:10030914ms step_avg:1899.43ms step:5292/6250 train_loss:1.7287 aux_loss:0.0098 train_time:10032386ms step_avg:1899.35ms step:5293/6250 train_loss:1.8105 aux_loss:0.0100 train_time:10033855ms step_avg:1899.27ms step:5294/6250 train_loss:1.8875 aux_loss:0.0105 train_time:10035244ms step_avg:1899.18ms step:5295/6250 train_loss:1.7302 aux_loss:0.0100 train_time:10036668ms step_avg:1899.09ms step:5296/6250 train_loss:1.8317 aux_loss:0.0102 train_time:10038225ms step_avg:1899.02ms step:5297/6250 train_loss:1.8278 aux_loss:0.0101 train_time:10039746ms step_avg:1898.95ms step:5298/6250 train_loss:1.8630 aux_loss:0.0102 train_time:10041264ms step_avg:1898.88ms step:5299/6250 train_loss:1.8670 aux_loss:0.0101 train_time:10042765ms step_avg:1898.80ms step:5300/6250 train_loss:1.8142 aux_loss:0.0098 train_time:10044223ms step_avg:1898.72ms step:5301/6250 train_loss:1.7669 aux_loss:0.0100 train_time:10045686ms step_avg:1898.64ms step:5302/6250 train_loss:1.7646 aux_loss:0.0103 train_time:10047094ms step_avg:1898.54ms step:5303/6250 train_loss:1.8816 aux_loss:0.0103 train_time:10048533ms step_avg:1898.46ms step:5304/6250 train_loss:1.7892 aux_loss:0.0104 train_time:10050345ms step_avg:1898.44ms step:5305/6250 train_loss:1.8581 aux_loss:0.0104 train_time:10051964ms step_avg:1898.39ms step:5306/6250 train_loss:1.8360 aux_loss:0.0102 train_time:10053437ms step_avg:1898.31ms step:5307/6250 train_loss:1.7262 aux_loss:0.0099 train_time:10054842ms step_avg:1898.21ms step:5308/6250 train_loss:1.7328 aux_loss:0.0100 train_time:10056194ms step_avg:1898.11ms step:5309/6250 train_loss:1.7703 aux_loss:0.0097 train_time:10057579ms step_avg:1898.01ms step:5310/6250 train_loss:1.8514 aux_loss:0.0097 train_time:10059135ms step_avg:1897.95ms step:5311/6250 train_loss:1.7865 aux_loss:0.0100 train_time:10060666ms step_avg:1897.88ms step:5312/6250 train_loss:1.8377 aux_loss:0.0100 train_time:10062023ms step_avg:1897.78ms step:5313/6250 train_loss:1.8092 aux_loss:0.0097 train_time:10063435ms step_avg:1897.69ms step:5314/6250 train_loss:1.8328 aux_loss:0.0095 train_time:10064860ms step_avg:1897.60ms step:5315/6250 train_loss:1.7660 aux_loss:0.0098 train_time:10066240ms step_avg:1897.50ms step:5316/6250 train_loss:1.7895 aux_loss:0.0097 train_time:10067664ms step_avg:1897.41ms step:5317/6250 train_loss:1.7763 aux_loss:0.0099 train_time:10068997ms step_avg:1897.30ms step:5318/6250 train_loss:1.8797 aux_loss:0.0097 train_time:10070542ms step_avg:1897.24ms step:5319/6250 train_loss:1.7525 aux_loss:0.0101 train_time:10072107ms step_avg:1897.18ms step:5320/6250 train_loss:1.7754 aux_loss:0.0100 train_time:10073603ms step_avg:1897.10ms step:5321/6250 train_loss:1.8626 aux_loss:0.0105 train_time:10075383ms step_avg:1897.08ms step:5322/6250 train_loss:1.7744 aux_loss:0.0100 train_time:10077711ms step_avg:1897.16ms step:5323/6250 train_loss:1.8209 aux_loss:0.0100 train_time:10080359ms step_avg:1897.30ms step:5324/6250 train_loss:1.8470 aux_loss:0.0099 train_time:10082106ms step_avg:1897.27ms step:5325/6250 train_loss:1.7780 aux_loss:0.0097 train_time:10083701ms step_avg:1897.22ms step:5326/6250 train_loss:1.8878 aux_loss:0.0097 train_time:10085189ms step_avg:1897.14ms step:5327/6250 train_loss:1.7646 aux_loss:0.0100 train_time:10086655ms step_avg:1897.06ms step:5328/6250 train_loss:1.8044 aux_loss:0.0104 train_time:10088018ms step_avg:1896.96ms step:5329/6250 train_loss:1.8171 aux_loss:0.0100 train_time:10094428ms step_avg:1897.81ms step:5330/6250 train_loss:1.8112 aux_loss:0.0101 train_time:10111307ms step_avg:1900.62ms step:5331/6250 train_loss:1.7879 aux_loss:0.0101 train_time:10115780ms step_avg:1901.10ms step:5332/6250 train_loss:1.7875 aux_loss:0.0100 train_time:10117324ms step_avg:1901.04ms step:5333/6250 train_loss:1.8383 aux_loss:0.0099 train_time:10118666ms step_avg:1900.93ms step:5334/6250 train_loss:1.7893 aux_loss:0.0097 train_time:10120012ms step_avg:1900.83ms step:5335/6250 train_loss:1.8145 aux_loss:0.0099 train_time:10121328ms step_avg:1900.72ms step:5336/6250 train_loss:1.8226 aux_loss:0.0101 train_time:10122662ms step_avg:1900.61ms step:5337/6250 train_loss:1.7986 aux_loss:0.0098 train_time:10124206ms step_avg:1900.55ms step:5338/6250 train_loss:1.8569 aux_loss:0.0100 train_time:10125869ms step_avg:1900.50ms step:5339/6250 train_loss:1.8113 aux_loss:0.0103 train_time:10127469ms step_avg:1900.44ms step:5340/6250 train_loss:1.8878 aux_loss:0.0102 train_time:10128795ms step_avg:1900.34ms step:5341/6250 train_loss:1.8726 aux_loss:0.0100 train_time:10130271ms step_avg:1900.26ms step:5342/6250 train_loss:1.8354 aux_loss:0.0097 train_time:10131590ms step_avg:1900.15ms step:5343/6250 train_loss:1.8242 aux_loss:0.0095 train_time:10132985ms step_avg:1900.05ms step:5344/6250 train_loss:1.8352 aux_loss:0.0100 train_time:10134462ms step_avg:1899.97ms step:5345/6250 train_loss:1.8347 aux_loss:0.0105 train_time:10135915ms step_avg:1899.89ms step:5346/6250 train_loss:1.8368 aux_loss:0.0102 train_time:10137258ms step_avg:1899.79ms step:5347/6250 train_loss:1.7601 aux_loss:0.0097 train_time:10138829ms step_avg:1899.72ms step:5348/6250 train_loss:1.8409 aux_loss:0.0097 train_time:10140281ms step_avg:1899.64ms step:5349/6250 train_loss:1.7664 aux_loss:0.0100 train_time:10141809ms step_avg:1899.57ms step:5350/6250 train_loss:1.8507 aux_loss:0.0101 train_time:10143134ms step_avg:1899.46ms step:5351/6250 train_loss:1.7831 aux_loss:0.0096 train_time:10144473ms step_avg:1899.36ms step:5352/6250 train_loss:1.8508 aux_loss:0.0098 train_time:10145931ms step_avg:1899.28ms step:5353/6250 train_loss:1.7625 aux_loss:0.0097 train_time:10147643ms step_avg:1899.24ms step:5354/6250 train_loss:1.8333 aux_loss:0.0102 train_time:10149347ms step_avg:1899.20ms step:5355/6250 train_loss:1.7168 aux_loss:0.0097 train_time:10151001ms step_avg:1899.16ms step:5356/6250 train_loss:1.8054 aux_loss:0.0097 train_time:10152410ms step_avg:1899.07ms step:5357/6250 train_loss:1.8212 aux_loss:0.0096 train_time:10153880ms step_avg:1898.99ms step:5358/6250 train_loss:1.7572 aux_loss:0.0096 train_time:10155357ms step_avg:1898.91ms step:5359/6250 train_loss:1.7583 aux_loss:0.0100 train_time:10156762ms step_avg:1898.82ms step:5360/6250 train_loss:1.8154 aux_loss:0.0103 train_time:10158393ms step_avg:1898.77ms step:5361/6250 train_loss:1.7557 aux_loss:0.0098 train_time:10159828ms step_avg:1898.68ms step:5362/6250 train_loss:1.8653 aux_loss:0.0098 train_time:10161399ms step_avg:1898.62ms step:5363/6250 train_loss:1.7718 aux_loss:0.0099 train_time:10162941ms step_avg:1898.55ms step:5364/6250 train_loss:1.7485 aux_loss:0.0098 train_time:10182182ms step_avg:1901.79ms step:5365/6250 train_loss:1.8586 aux_loss:0.0101 train_time:10196434ms step_avg:1904.10ms step:5366/6250 train_loss:1.7850 aux_loss:0.0096 train_time:10197859ms step_avg:1904.01ms step:5367/6250 train_loss:1.8810 aux_loss:0.0098 train_time:10211301ms step_avg:1906.16ms step:5368/6250 train_loss:1.7878 aux_loss:0.0099 train_time:10216919ms step_avg:1906.85ms step:5369/6250 train_loss:1.7988 aux_loss:0.0098 train_time:10218205ms step_avg:1906.74ms step:5370/6250 train_loss:1.7987 aux_loss:0.0098 train_time:10219563ms step_avg:1906.63ms step:5371/6250 train_loss:1.7574 aux_loss:0.0097 train_time:10220957ms step_avg:1906.54ms step:5372/6250 train_loss:1.7866 aux_loss:0.0099 train_time:10222343ms step_avg:1906.44ms step:5373/6250 train_loss:1.7110 aux_loss:0.0096 train_time:10223699ms step_avg:1906.34ms step:5374/6250 train_loss:1.8008 aux_loss:0.0099 train_time:10225086ms step_avg:1906.24ms step:5375/6250 train_loss:1.8221 aux_loss:0.0101 train_time:10226553ms step_avg:1906.16ms step:5376/6250 train_loss:1.8016 aux_loss:0.0099 train_time:10228174ms step_avg:1906.11ms step:5377/6250 train_loss:1.8041 aux_loss:0.0100 train_time:10230506ms step_avg:1906.19ms step:5378/6250 train_loss:1.8570 aux_loss:0.0099 train_time:10232041ms step_avg:1906.12ms step:5379/6250 train_loss:1.7915 aux_loss:0.0096 train_time:10233581ms step_avg:1906.05ms step:5380/6250 train_loss:1.7667 aux_loss:0.0096 train_time:10235075ms step_avg:1905.97ms step:5381/6250 train_loss:1.7352 aux_loss:0.0097 train_time:10236568ms step_avg:1905.90ms step:5382/6250 train_loss:1.7606 aux_loss:0.0099 train_time:10238228ms step_avg:1905.85ms step:5383/6250 train_loss:1.8737 aux_loss:0.0097 train_time:10239767ms step_avg:1905.78ms step:5384/6250 train_loss:1.7809 aux_loss:0.0098 train_time:10241226ms step_avg:1905.70ms step:5385/6250 train_loss:1.8228 aux_loss:0.0098 train_time:10242809ms step_avg:1905.64ms step:5386/6250 train_loss:1.7270 aux_loss:0.0096 train_time:10244221ms step_avg:1905.55ms step:5387/6250 train_loss:1.8160 aux_loss:0.0094 train_time:10245884ms step_avg:1905.50ms step:5388/6250 train_loss:1.8648 aux_loss:0.0097 train_time:10247410ms step_avg:1905.43ms step:5389/6250 train_loss:1.8691 aux_loss:0.0099 train_time:10248886ms step_avg:1905.35ms step:5390/6250 train_loss:1.7600 aux_loss:0.0096 train_time:10250286ms step_avg:1905.26ms step:5391/6250 train_loss:1.8142 aux_loss:0.0096 train_time:10251913ms step_avg:1905.21ms step:5392/6250 train_loss:1.7966 aux_loss:0.0099 train_time:10253423ms step_avg:1905.13ms step:5393/6250 train_loss:1.7870 aux_loss:0.0100 train_time:10254998ms step_avg:1905.07ms step:5394/6250 train_loss:1.7322 aux_loss:0.0096 train_time:10256382ms step_avg:1904.97ms step:5395/6250 train_loss:1.9251 aux_loss:0.0095 train_time:10258070ms step_avg:1904.93ms step:5396/6250 train_loss:1.7944 aux_loss:0.0095 train_time:10259708ms step_avg:1904.88ms step:5397/6250 train_loss:1.7934 aux_loss:0.0096 train_time:10261324ms step_avg:1904.83ms step:5398/6250 train_loss:1.8209 aux_loss:0.0101 train_time:10262935ms step_avg:1904.78ms step:5399/6250 train_loss:1.7682 aux_loss:0.0096 train_time:10264393ms step_avg:1904.69ms step:5400/6250 train_loss:1.8759 aux_loss:0.0099 train_time:10265674ms step_avg:1904.58ms step:5401/6250 train_loss:1.7845 aux_loss:0.0097 train_time:10267136ms step_avg:1904.50ms step:5402/6250 train_loss:1.7553 aux_loss:0.0096 train_time:10268716ms step_avg:1904.44ms step:5403/6250 train_loss:1.7944 aux_loss:0.0096 train_time:10270268ms step_avg:1904.37ms step:5404/6250 train_loss:1.9195 aux_loss:0.0100 train_time:10271862ms step_avg:1904.31ms step:5405/6250 train_loss:1.7511 aux_loss:0.0096 train_time:10273346ms step_avg:1904.23ms step:5406/6250 train_loss:1.8303 aux_loss:0.0099 train_time:10282683ms step_avg:1905.61ms step:5407/6250 train_loss:1.7052 aux_loss:0.0099 train_time:10288037ms step_avg:1906.25ms step:5408/6250 train_loss:1.8102 aux_loss:0.0101 train_time:10289472ms step_avg:1906.16ms step:5409/6250 train_loss:1.9045 aux_loss:0.0099 train_time:10293729ms step_avg:1906.60ms step:5410/6250 train_loss:1.7239 aux_loss:0.0098 train_time:10295064ms step_avg:1906.49ms step:5411/6250 train_loss:1.8078 aux_loss:0.0096 train_time:10296641ms step_avg:1906.43ms step:5412/6250 train_loss:1.8456 aux_loss:0.0097 train_time:10306950ms step_avg:1907.99ms step:5413/6250 train_loss:1.7009 aux_loss:0.0098 train_time:10313082ms step_avg:1908.77ms step:5414/6250 train_loss:1.9093 aux_loss:0.0099 train_time:10315050ms step_avg:1908.78ms step:5415/6250 train_loss:1.8080 aux_loss:0.0097 train_time:10324097ms step_avg:1910.10ms step:5416/6250 train_loss:1.6829 aux_loss:0.0098 train_time:10337042ms step_avg:1912.14ms step:5417/6250 train_loss:1.8916 aux_loss:0.0100 train_time:10338436ms step_avg:1912.05ms step:5418/6250 train_loss:1.7496 aux_loss:0.0103 train_time:10340571ms step_avg:1912.09ms step:5419/6250 train_loss:1.8335 aux_loss:0.0097 train_time:10342730ms step_avg:1912.13ms step:5420/6250 train_loss:1.8054 aux_loss:0.0094 train_time:10344288ms step_avg:1912.07ms step:5421/6250 train_loss:1.8334 aux_loss:0.0098 train_time:10345647ms step_avg:1911.97ms step:5422/6250 train_loss:1.7725 aux_loss:0.0099 train_time:10347076ms step_avg:1911.88ms step:5423/6250 train_loss:1.9070 aux_loss:0.0100 train_time:10348425ms step_avg:1911.77ms step:5424/6250 train_loss:1.7982 aux_loss:0.0095 train_time:10350467ms step_avg:1911.80ms step:5425/6250 train_loss:1.8628 aux_loss:0.0098 train_time:10351961ms step_avg:1911.72ms step:5426/6250 train_loss:1.8559 aux_loss:0.0099 train_time:10353850ms step_avg:1911.72ms step:5427/6250 train_loss:1.7837 aux_loss:0.0102 train_time:10355406ms step_avg:1911.65ms step:5428/6250 train_loss:1.8286 aux_loss:0.0101 train_time:10357053ms step_avg:1911.60ms step:5429/6250 train_loss:1.8445 aux_loss:0.0102 train_time:10358472ms step_avg:1911.51ms step:5430/6250 train_loss:1.8068 aux_loss:0.0101 train_time:10360015ms step_avg:1911.44ms step:5431/6250 train_loss:1.6932 aux_loss:0.0100 train_time:10361524ms step_avg:1911.37ms step:5432/6250 train_loss:1.7880 aux_loss:0.0098 train_time:10363209ms step_avg:1911.33ms step:5433/6250 train_loss:1.7541 aux_loss:0.0095 train_time:10367486ms step_avg:1911.76ms step:5434/6250 train_loss:1.9272 aux_loss:0.0101 train_time:10368821ms step_avg:1911.66ms step:5435/6250 train_loss:1.8390 aux_loss:0.0099 train_time:10370183ms step_avg:1911.55ms step:5436/6250 train_loss:1.8025 aux_loss:0.0100 train_time:10371546ms step_avg:1911.45ms step:5437/6250 train_loss:1.8235 aux_loss:0.0097 train_time:10372764ms step_avg:1911.33ms step:5438/6250 train_loss:1.7839 aux_loss:0.0097 train_time:10374361ms step_avg:1911.27ms step:5439/6250 train_loss:1.7346 aux_loss:0.0098 train_time:10375770ms step_avg:1911.18ms step:5440/6250 train_loss:1.7680 aux_loss:0.0099 train_time:10377396ms step_avg:1911.12ms step:5441/6250 train_loss:1.8378 aux_loss:0.0099 train_time:10378960ms step_avg:1911.06ms step:5442/6250 train_loss:1.7887 aux_loss:0.0101 train_time:10380389ms step_avg:1910.97ms step:5443/6250 train_loss:1.7191 aux_loss:0.0099 train_time:10381963ms step_avg:1910.91ms step:5444/6250 train_loss:1.7976 aux_loss:0.0097 train_time:10383401ms step_avg:1910.82ms step:5445/6250 train_loss:1.7743 aux_loss:0.0097 train_time:10384889ms step_avg:1910.74ms step:5446/6250 train_loss:1.7823 aux_loss:0.0097 train_time:10386310ms step_avg:1910.65ms step:5447/6250 train_loss:1.8136 aux_loss:0.0099 train_time:10388017ms step_avg:1910.62ms step:5448/6250 train_loss:1.8057 aux_loss:0.0097 train_time:10389613ms step_avg:1910.56ms step:5449/6250 train_loss:1.8420 aux_loss:0.0100 train_time:10391442ms step_avg:1910.54ms step:5450/6250 train_loss:1.8780 aux_loss:0.0103 train_time:10392875ms step_avg:1910.46ms step:5451/6250 train_loss:1.8541 aux_loss:0.0100 train_time:10394485ms step_avg:1910.40ms step:5452/6250 train_loss:1.7826 aux_loss:0.0099 train_time:10396027ms step_avg:1910.33ms step:5453/6250 train_loss:1.8151 aux_loss:0.0102 train_time:10397386ms step_avg:1910.23ms step:5454/6250 train_loss:1.7583 aux_loss:0.0101 train_time:10398847ms step_avg:1910.15ms step:5455/6250 train_loss:1.7595 aux_loss:0.0099 train_time:10400390ms step_avg:1910.08ms step:5456/6250 train_loss:1.7677 aux_loss:0.0099 train_time:10402070ms step_avg:1910.04ms step:5457/6250 train_loss:1.8335 aux_loss:0.0098 train_time:10403619ms step_avg:1909.97ms step:5458/6250 train_loss:1.7471 aux_loss:0.0097 train_time:10405040ms step_avg:1909.88ms step:5459/6250 train_loss:1.7694 aux_loss:0.0099 train_time:10406345ms step_avg:1909.77ms step:5460/6250 train_loss:1.8103 aux_loss:0.0103 train_time:10407787ms step_avg:1909.69ms step:5461/6250 train_loss:1.7870 aux_loss:0.0103 train_time:10409108ms step_avg:1909.58ms step:5462/6250 train_loss:1.7840 aux_loss:0.0102 train_time:10410549ms step_avg:1909.49ms step:5463/6250 train_loss:1.8016 aux_loss:0.0099 train_time:10411989ms step_avg:1909.41ms step:5464/6250 train_loss:1.7779 aux_loss:0.0100 train_time:10413398ms step_avg:1909.31ms step:5465/6250 train_loss:1.7828 aux_loss:0.0100 train_time:10415008ms step_avg:1909.26ms step:5466/6250 train_loss:1.7369 aux_loss:0.0101 train_time:10416566ms step_avg:1909.19ms step:5467/6250 train_loss:1.8587 aux_loss:0.0103 train_time:10417883ms step_avg:1909.09ms step:5468/6250 train_loss:1.7104 aux_loss:0.0097 train_time:10419174ms step_avg:1908.97ms step:5469/6250 train_loss:1.7793 aux_loss:0.0100 train_time:10420468ms step_avg:1908.86ms step:5470/6250 train_loss:1.7901 aux_loss:0.0098 train_time:10421890ms step_avg:1908.77ms step:5471/6250 train_loss:1.7899 aux_loss:0.0094 train_time:10423283ms step_avg:1908.68ms step:5472/6250 train_loss:1.8002 aux_loss:0.0097 train_time:10424620ms step_avg:1908.57ms step:5473/6250 train_loss:1.8565 aux_loss:0.0099 train_time:10426025ms step_avg:1908.48ms step:5474/6250 train_loss:1.8113 aux_loss:0.0100 train_time:10427439ms step_avg:1908.39ms step:5475/6250 train_loss:1.8378 aux_loss:0.0098 train_time:10429188ms step_avg:1908.36ms step:5476/6250 train_loss:1.8151 aux_loss:0.0101 train_time:10431212ms step_avg:1908.38ms step:5477/6250 train_loss:1.7979 aux_loss:0.0101 train_time:10434027ms step_avg:1908.55ms step:5478/6250 train_loss:1.8933 aux_loss:0.0102 train_time:10435401ms step_avg:1908.45ms step:5479/6250 train_loss:1.8221 aux_loss:0.0099 train_time:10437293ms step_avg:1908.45ms step:5480/6250 train_loss:1.7788 aux_loss:0.0097 train_time:10439658ms step_avg:1908.53ms step:5481/6250 train_loss:1.7899 aux_loss:0.0101 train_time:10441828ms step_avg:1908.58ms step:5482/6250 train_loss:1.8271 aux_loss:0.0100 train_time:10444811ms step_avg:1908.77ms step:5483/6250 train_loss:1.8106 aux_loss:0.0098 train_time:10446208ms step_avg:1908.68ms step:5484/6250 train_loss:1.6535 aux_loss:0.0095 train_time:10448124ms step_avg:1908.68ms step:5485/6250 train_loss:1.7446 aux_loss:0.0099 train_time:10450711ms step_avg:1908.81ms step:5486/6250 train_loss:1.7220 aux_loss:0.0099 train_time:10452088ms step_avg:1908.71ms step:5487/6250 train_loss:1.8682 aux_loss:0.0099 train_time:10453630ms step_avg:1908.64ms step:5488/6250 train_loss:1.7594 aux_loss:0.0096 train_time:10455157ms step_avg:1908.57ms step:5489/6250 train_loss:1.7532 aux_loss:0.0099 train_time:10456463ms step_avg:1908.46ms step:5490/6250 train_loss:1.7651 aux_loss:0.0098 train_time:10457820ms step_avg:1908.36ms step:5491/6250 train_loss:1.8476 aux_loss:0.0096 train_time:10459473ms step_avg:1908.31ms step:5492/6250 train_loss:1.7792 aux_loss:0.0098 train_time:10460951ms step_avg:1908.24ms step:5493/6250 train_loss:1.8411 aux_loss:0.0101 train_time:10462440ms step_avg:1908.16ms step:5494/6250 train_loss:1.8486 aux_loss:0.0097 train_time:10463933ms step_avg:1908.08ms step:5495/6250 train_loss:1.7434 aux_loss:0.0097 train_time:10465622ms step_avg:1908.04ms step:5496/6250 train_loss:1.7765 aux_loss:0.0097 train_time:10470551ms step_avg:1908.59ms step:5497/6250 train_loss:1.7977 aux_loss:0.0104 train_time:10472622ms step_avg:1908.62ms step:5498/6250 train_loss:1.7726 aux_loss:0.0096 train_time:10474224ms step_avg:1908.57ms step:5499/6250 train_loss:1.7335 aux_loss:0.0096 train_time:10475701ms step_avg:1908.49ms step:5500/6250 train_loss:1.8813 aux_loss:0.0097 train_time:10477100ms step_avg:1908.40ms step:5501/6250 train_loss:1.8682 aux_loss:0.0098 train_time:10478666ms step_avg:1908.33ms step:5502/6250 train_loss:1.7912 aux_loss:0.0095 train_time:10480117ms step_avg:1908.25ms step:5503/6250 train_loss:1.9175 aux_loss:0.0098 train_time:10481595ms step_avg:1908.17ms step:5504/6250 train_loss:1.8126 aux_loss:0.0097 train_time:10483221ms step_avg:1908.12ms step:5505/6250 train_loss:1.7586 aux_loss:0.0097 train_time:10484835ms step_avg:1908.07ms step:5506/6250 train_loss:1.8456 aux_loss:0.0094 train_time:10486365ms step_avg:1908.00ms step:5507/6250 train_loss:1.7805 aux_loss:0.0096 train_time:10487792ms step_avg:1907.91ms step:5508/6250 train_loss:1.8279 aux_loss:0.0098 train_time:10490142ms step_avg:1907.99ms step:5509/6250 train_loss:1.8535 aux_loss:0.0096 train_time:10491604ms step_avg:1907.91ms step:5510/6250 train_loss:1.6979 aux_loss:0.0096 train_time:10493045ms step_avg:1907.83ms step:5511/6250 train_loss:1.9043 aux_loss:0.0101 train_time:10494496ms step_avg:1907.74ms step:5512/6250 train_loss:1.7858 aux_loss:0.0099 train_time:10495989ms step_avg:1907.67ms step:5513/6250 train_loss:1.6885 aux_loss:0.0093 train_time:10497535ms step_avg:1907.60ms step:5514/6250 train_loss:1.8781 aux_loss:0.0096 train_time:10499054ms step_avg:1907.53ms step:5515/6250 train_loss:1.7845 aux_loss:0.0097 train_time:10500490ms step_avg:1907.45ms step:5516/6250 train_loss:1.7819 aux_loss:0.0097 train_time:10501954ms step_avg:1907.37ms step:5517/6250 train_loss:1.8661 aux_loss:0.0097 train_time:10503487ms step_avg:1907.30ms step:5518/6250 train_loss:1.7495 aux_loss:0.0095 train_time:10505019ms step_avg:1907.23ms step:5519/6250 train_loss:1.9284 aux_loss:0.0099 train_time:10506525ms step_avg:1907.16ms step:5520/6250 train_loss:1.7894 aux_loss:0.0096 train_time:10508092ms step_avg:1907.09ms step:5521/6250 train_loss:1.7979 aux_loss:0.0095 train_time:10509679ms step_avg:1907.04ms step:5522/6250 train_loss:1.7949 aux_loss:0.0095 train_time:10511205ms step_avg:1906.97ms step:5523/6250 train_loss:1.7392 aux_loss:0.0096 train_time:10512732ms step_avg:1906.90ms step:5524/6250 train_loss:1.8545 aux_loss:0.0100 train_time:10514178ms step_avg:1906.82ms step:5525/6250 train_loss:1.8075 aux_loss:0.0097 train_time:10515736ms step_avg:1906.75ms step:5526/6250 train_loss:1.7924 aux_loss:0.0093 train_time:10517376ms step_avg:1906.70ms step:5527/6250 train_loss:1.8257 aux_loss:0.0096 train_time:10518713ms step_avg:1906.60ms step:5528/6250 train_loss:1.8491 aux_loss:0.0098 train_time:10520190ms step_avg:1906.52ms step:5529/6250 train_loss:1.7354 aux_loss:0.0095 train_time:10521644ms step_avg:1906.44ms step:5530/6250 train_loss:1.7380 aux_loss:0.0095 train_time:10523256ms step_avg:1906.39ms step:5531/6250 train_loss:1.7684 aux_loss:0.0097 train_time:10524737ms step_avg:1906.31ms step:5532/6250 train_loss:1.8150 aux_loss:0.0098 train_time:10526112ms step_avg:1906.21ms step:5533/6250 train_loss:1.7692 aux_loss:0.0098 train_time:10527463ms step_avg:1906.11ms step:5534/6250 train_loss:1.8103 aux_loss:0.0098 train_time:10528969ms step_avg:1906.04ms step:5535/6250 train_loss:1.7757 aux_loss:0.0097 train_time:10530484ms step_avg:1905.97ms step:5536/6250 train_loss:1.8172 aux_loss:0.0097 train_time:10531996ms step_avg:1905.90ms step:5537/6250 train_loss:1.8271 aux_loss:0.0098 train_time:10533456ms step_avg:1905.82ms step:5538/6250 train_loss:1.8222 aux_loss:0.0094 train_time:10534928ms step_avg:1905.74ms step:5539/6250 train_loss:1.8241 aux_loss:0.0095 train_time:10536233ms step_avg:1905.63ms step:5540/6250 train_loss:1.8862 aux_loss:0.0101 train_time:10537549ms step_avg:1905.52ms step:5541/6250 train_loss:1.7868 aux_loss:0.0098 train_time:10539243ms step_avg:1905.49ms step:5542/6250 train_loss:1.8240 aux_loss:0.0101 train_time:10542226ms step_avg:1905.68ms step:5543/6250 train_loss:1.7927 aux_loss:0.0098 train_time:10544504ms step_avg:1905.75ms step:5544/6250 train_loss:1.8072 aux_loss:0.0099 train_time:10545986ms step_avg:1905.67ms step:5545/6250 train_loss:1.8879 aux_loss:0.0097 train_time:10547430ms step_avg:1905.59ms step:5546/6250 train_loss:1.8165 aux_loss:0.0098 train_time:10548901ms step_avg:1905.51ms step:5547/6250 train_loss:1.8459 aux_loss:0.0094 train_time:10550704ms step_avg:1905.49ms step:5548/6250 train_loss:1.8075 aux_loss:0.0098 train_time:10552274ms step_avg:1905.43ms step:5549/6250 train_loss:1.8324 aux_loss:0.0098 train_time:10553819ms step_avg:1905.37ms step:5550/6250 train_loss:1.7896 aux_loss:0.0096 train_time:10555175ms step_avg:1905.27ms step:5551/6250 train_loss:1.9374 aux_loss:0.0100 train_time:10556610ms step_avg:1905.18ms step:5552/6250 train_loss:1.8004 aux_loss:0.0099 train_time:10558045ms step_avg:1905.10ms step:5553/6250 train_loss:1.7791 aux_loss:0.0097 train_time:10559307ms step_avg:1904.98ms step:5554/6250 train_loss:1.8174 aux_loss:0.0095 train_time:10560784ms step_avg:1904.90ms step:5555/6250 train_loss:1.7479 aux_loss:0.0099 train_time:10562194ms step_avg:1904.81ms step:5556/6250 train_loss:1.7902 aux_loss:0.0099 train_time:10563780ms step_avg:1904.76ms step:5557/6250 train_loss:1.9573 aux_loss:0.0113 train_time:10565207ms step_avg:1904.67ms step:5558/6250 train_loss:1.8254 aux_loss:0.0100 train_time:10566647ms step_avg:1904.59ms step:5559/6250 train_loss:1.7641 aux_loss:0.0098 train_time:10568298ms step_avg:1904.54ms step:5560/6250 train_loss:1.8475 aux_loss:0.0099 train_time:10569831ms step_avg:1904.47ms step:5561/6250 train_loss:1.8000 aux_loss:0.0097 train_time:10571217ms step_avg:1904.38ms step:5562/6250 train_loss:1.8101 aux_loss:0.0094 train_time:10572632ms step_avg:1904.29ms step:5563/6250 train_loss:1.9008 aux_loss:0.0097 train_time:10574094ms step_avg:1904.21ms step:5564/6250 train_loss:1.8599 aux_loss:0.0098 train_time:10575338ms step_avg:1904.09ms step:5565/6250 train_loss:1.8226 aux_loss:0.0098 train_time:10576793ms step_avg:1904.01ms step:5566/6250 train_loss:1.8233 aux_loss:0.0096 train_time:10578299ms step_avg:1903.94ms step:5567/6250 train_loss:1.8499 aux_loss:0.0098 train_time:10579665ms step_avg:1903.84ms step:5568/6250 train_loss:1.7958 aux_loss:0.0098 train_time:10581086ms step_avg:1903.76ms step:5569/6250 train_loss:1.7843 aux_loss:0.0099 train_time:10582464ms step_avg:1903.66ms step:5570/6250 train_loss:1.7706 aux_loss:0.0095 train_time:10584079ms step_avg:1903.61ms step:5571/6250 train_loss:1.7816 aux_loss:0.0101 train_time:10585566ms step_avg:1903.54ms step:5572/6250 train_loss:1.8214 aux_loss:0.0099 train_time:10587070ms step_avg:1903.46ms step:5573/6250 train_loss:1.8296 aux_loss:0.0100 train_time:10588595ms step_avg:1903.40ms step:5574/6250 train_loss:1.8614 aux_loss:0.0098 train_time:10590051ms step_avg:1903.32ms step:5575/6250 train_loss:1.7532 aux_loss:0.0094 train_time:10591701ms step_avg:1903.27ms step:5576/6250 train_loss:1.7652 aux_loss:0.0099 train_time:10593097ms step_avg:1903.18ms step:5577/6250 train_loss:1.8070 aux_loss:0.0099 train_time:10594626ms step_avg:1903.11ms step:5578/6250 train_loss:1.8289 aux_loss:0.0095 train_time:10596185ms step_avg:1903.05ms step:5579/6250 train_loss:1.7740 aux_loss:0.0091 train_time:10597593ms step_avg:1902.96ms step:5580/6250 train_loss:1.8868 aux_loss:0.0096 train_time:10598953ms step_avg:1902.86ms step:5581/6250 train_loss:1.8312 aux_loss:0.0100 train_time:10600535ms step_avg:1902.81ms step:5582/6250 train_loss:1.8503 aux_loss:0.0100 train_time:10602074ms step_avg:1902.74ms step:5583/6250 train_loss:1.7089 aux_loss:0.0096 train_time:10603519ms step_avg:1902.66ms step:5584/6250 train_loss:1.8750 aux_loss:0.0097 train_time:10605018ms step_avg:1902.59ms step:5585/6250 train_loss:1.7781 aux_loss:0.0096 train_time:10606427ms step_avg:1902.50ms step:5586/6250 train_loss:1.7741 aux_loss:0.0097 train_time:10608065ms step_avg:1902.45ms step:5587/6250 train_loss:1.7679 aux_loss:0.0095 train_time:10609623ms step_avg:1902.39ms step:5588/6250 train_loss:1.8111 aux_loss:0.0094 train_time:10611136ms step_avg:1902.32ms step:5589/6250 train_loss:1.7357 aux_loss:0.0093 train_time:10612438ms step_avg:1902.21ms step:5590/6250 train_loss:1.8425 aux_loss:0.0094 train_time:10614135ms step_avg:1902.17ms step:5591/6250 train_loss:1.7473 aux_loss:0.0093 train_time:10615563ms step_avg:1902.09ms step:5592/6250 train_loss:1.8282 aux_loss:0.0095 train_time:10617081ms step_avg:1902.02ms step:5593/6250 train_loss:1.8458 aux_loss:0.0097 train_time:10618386ms step_avg:1901.91ms step:5594/6250 train_loss:1.7690 aux_loss:0.0096 train_time:10619763ms step_avg:1901.82ms step:5595/6250 train_loss:1.8065 aux_loss:0.0096 train_time:10621085ms step_avg:1901.72ms step:5596/6250 train_loss:1.7902 aux_loss:0.0094 train_time:10622466ms step_avg:1901.62ms step:5597/6250 train_loss:1.8074 aux_loss:0.0097 train_time:10623759ms step_avg:1901.51ms step:5598/6250 train_loss:1.8379 aux_loss:0.0097 train_time:10625132ms step_avg:1901.42ms step:5599/6250 train_loss:1.7875 aux_loss:0.0099 train_time:10626508ms step_avg:1901.33ms step:5600/6250 train_loss:1.7530 aux_loss:0.0096 train_time:10628507ms step_avg:1901.34ms step:5601/6250 train_loss:1.7238 aux_loss:0.0093 train_time:10630937ms step_avg:1901.44ms step:5602/6250 train_loss:1.7532 aux_loss:0.0093 train_time:10632650ms step_avg:1901.40ms step:5603/6250 train_loss:1.7417 aux_loss:0.0096 train_time:10633945ms step_avg:1901.30ms step:5604/6250 train_loss:1.8499 aux_loss:0.0099 train_time:10635269ms step_avg:1901.19ms step:5605/6250 train_loss:1.7349 aux_loss:0.0096 train_time:10636602ms step_avg:1901.09ms step:5606/6250 train_loss:1.7594 aux_loss:0.0097 train_time:10638551ms step_avg:1901.10ms step:5607/6250 train_loss:1.8000 aux_loss:0.0099 train_time:10640355ms step_avg:1901.08ms step:5608/6250 train_loss:1.8507 aux_loss:0.0099 train_time:10642104ms step_avg:1901.05ms step:5609/6250 train_loss:1.8100 aux_loss:0.0096 train_time:10643489ms step_avg:1900.96ms step:5610/6250 train_loss:1.7833 aux_loss:0.0092 train_time:10644848ms step_avg:1900.87ms step:5611/6250 train_loss:1.7955 aux_loss:0.0091 train_time:10646443ms step_avg:1900.81ms step:5612/6250 train_loss:1.7881 aux_loss:0.0092 train_time:10648249ms step_avg:1900.79ms step:5613/6250 train_loss:1.7847 aux_loss:0.0096 train_time:10650964ms step_avg:1900.94ms step:5614/6250 train_loss:1.7175 aux_loss:0.0094 train_time:10653474ms step_avg:1901.05ms step:5615/6250 train_loss:1.7789 aux_loss:0.0093 train_time:10654753ms step_avg:1900.94ms step:5616/6250 train_loss:1.7434 aux_loss:0.0095 train_time:10656214ms step_avg:1900.86ms step:5617/6250 train_loss:1.8520 aux_loss:0.0096 train_time:10658290ms step_avg:1900.89ms step:5618/6250 train_loss:1.7552 aux_loss:0.0096 train_time:10660159ms step_avg:1900.88ms step:5619/6250 train_loss:1.8306 aux_loss:0.0095 train_time:10661506ms step_avg:1900.79ms step:5620/6250 train_loss:1.8205 aux_loss:0.0099 train_time:10662821ms step_avg:1900.68ms step:5621/6250 train_loss:1.8040 aux_loss:0.0101 train_time:10664276ms step_avg:1900.60ms step:5622/6250 train_loss:1.8411 aux_loss:0.0100 train_time:10665636ms step_avg:1900.51ms step:5623/6250 train_loss:1.7987 aux_loss:0.0097 train_time:10667041ms step_avg:1900.42ms step:5624/6250 train_loss:1.7410 aux_loss:0.0097 train_time:10668401ms step_avg:1900.32ms step:5625/6250 train_loss:1.7678 aux_loss:0.0098 train_time:10669840ms step_avg:1900.24ms step:5626/6250 train_loss:1.8199 aux_loss:0.0097 train_time:10672479ms step_avg:1900.37ms step:5627/6250 train_loss:1.8253 aux_loss:0.0096 train_time:10674318ms step_avg:1900.36ms step:5628/6250 train_loss:1.7980 aux_loss:0.0095 train_time:10675896ms step_avg:1900.30ms step:5629/6250 train_loss:1.8405 aux_loss:0.0095 train_time:10678383ms step_avg:1900.41ms step:5630/6250 train_loss:1.8621 aux_loss:0.0095 train_time:10680288ms step_avg:1900.41ms step:5631/6250 train_loss:1.8427 aux_loss:0.0094 train_time:10681707ms step_avg:1900.32ms step:5632/6250 train_loss:1.6830 aux_loss:0.0094 train_time:10683020ms step_avg:1900.22ms step:5633/6250 train_loss:1.7440 aux_loss:0.0096 train_time:10684831ms step_avg:1900.20ms step:5634/6250 train_loss:1.7792 aux_loss:0.0093 train_time:10686093ms step_avg:1900.09ms step:5635/6250 train_loss:1.7763 aux_loss:0.0094 train_time:10687712ms step_avg:1900.04ms step:5636/6250 train_loss:1.7693 aux_loss:0.0096 train_time:10689139ms step_avg:1899.95ms step:5637/6250 train_loss:1.7352 aux_loss:0.0096 train_time:10691264ms step_avg:1899.99ms step:5638/6250 train_loss:1.8148 aux_loss:0.0096 train_time:10693125ms step_avg:1899.99ms step:5639/6250 train_loss:1.7170 aux_loss:0.0098 train_time:10694761ms step_avg:1899.94ms step:5640/6250 train_loss:1.7570 aux_loss:0.0098 train_time:10696632ms step_avg:1899.93ms step:5641/6250 train_loss:1.9484 aux_loss:0.0100 train_time:10698524ms step_avg:1899.93ms step:5642/6250 train_loss:1.7841 aux_loss:0.0098 train_time:10701620ms step_avg:1900.15ms step:5643/6250 train_loss:1.7811 aux_loss:0.0097 train_time:10703398ms step_avg:1900.12ms step:5644/6250 train_loss:1.7777 aux_loss:0.0097 train_time:10705165ms step_avg:1900.10ms step:5645/6250 train_loss:1.8473 aux_loss:0.0097 train_time:10708877ms step_avg:1900.42ms step:5646/6250 train_loss:1.8011 aux_loss:0.0097 train_time:10711121ms step_avg:1900.48ms step:5647/6250 train_loss:1.7552 aux_loss:0.0096 train_time:10713177ms step_avg:1900.51ms step:5648/6250 train_loss:1.8309 aux_loss:0.0097 train_time:10715226ms step_avg:1900.54ms step:5649/6250 train_loss:1.7553 aux_loss:0.0106 train_time:10717864ms step_avg:1900.67ms step:5650/6250 train_loss:1.8074 aux_loss:0.0102 train_time:10720392ms step_avg:1900.78ms step:5651/6250 train_loss:1.7974 aux_loss:0.0098 train_time:10721794ms step_avg:1900.69ms step:5652/6250 train_loss:1.7246 aux_loss:0.0094 train_time:10723864ms step_avg:1900.72ms step:5653/6250 train_loss:1.8285 aux_loss:0.0102 train_time:10725250ms step_avg:1900.63ms step:5654/6250 train_loss:1.7853 aux_loss:0.0097 train_time:10727025ms step_avg:1900.61ms step:5655/6250 train_loss:1.8135 aux_loss:0.0095 train_time:10728408ms step_avg:1900.52ms step:5656/6250 train_loss:1.8430 aux_loss:0.0098 train_time:10729713ms step_avg:1900.41ms step:5657/6250 train_loss:1.7966 aux_loss:0.0097 train_time:10731174ms step_avg:1900.33ms step:5658/6250 train_loss:1.7569 aux_loss:0.0095 train_time:10732551ms step_avg:1900.24ms step:5659/6250 train_loss:1.7396 aux_loss:0.0095 train_time:10734198ms step_avg:1900.19ms step:5660/6250 train_loss:1.7506 aux_loss:0.0096 train_time:10736088ms step_avg:1900.19ms step:5661/6250 train_loss:1.7757 aux_loss:0.0094 train_time:10737361ms step_avg:1900.08ms step:5662/6250 train_loss:1.9222 aux_loss:0.0097 train_time:10738921ms step_avg:1900.02ms step:5663/6250 train_loss:1.7518 aux_loss:0.0096 train_time:10740260ms step_avg:1899.92ms step:5664/6250 train_loss:1.7962 aux_loss:0.0097 train_time:10742548ms step_avg:1899.99ms step:5665/6250 train_loss:1.8205 aux_loss:0.0093 train_time:10744027ms step_avg:1899.92ms step:5666/6250 train_loss:1.7575 aux_loss:0.0091 train_time:10745677ms step_avg:1899.87ms step:5667/6250 train_loss:1.7575 aux_loss:0.0095 train_time:10747229ms step_avg:1899.81ms step:5668/6250 train_loss:1.8457 aux_loss:0.0095 train_time:10748956ms step_avg:1899.78ms step:5669/6250 train_loss:1.8168 aux_loss:0.0094 train_time:10750571ms step_avg:1899.73ms step:5670/6250 train_loss:1.6878 aux_loss:0.0091 train_time:10752312ms step_avg:1899.70ms step:5671/6250 train_loss:1.7692 aux_loss:0.0093 train_time:10754046ms step_avg:1899.67ms step:5672/6250 train_loss:1.7935 aux_loss:0.0095 train_time:10755795ms step_avg:1899.65ms step:5673/6250 train_loss:1.8513 aux_loss:0.0094 train_time:10757155ms step_avg:1899.55ms step:5674/6250 train_loss:1.8220 aux_loss:0.0093 train_time:10758534ms step_avg:1899.46ms step:5675/6250 train_loss:1.8336 aux_loss:0.0094 train_time:10759899ms step_avg:1899.36ms step:5676/6250 train_loss:1.8320 aux_loss:0.0093 train_time:10761583ms step_avg:1899.33ms step:5677/6250 train_loss:1.7783 aux_loss:0.0095 train_time:10764952ms step_avg:1899.59ms step:5678/6250 train_loss:1.8476 aux_loss:0.0094 train_time:10766279ms step_avg:1899.48ms step:5679/6250 train_loss:1.8515 aux_loss:0.0096 train_time:10767725ms step_avg:1899.40ms step:5680/6250 train_loss:1.7924 aux_loss:0.0096 train_time:10769097ms step_avg:1899.31ms step:5681/6250 train_loss:1.8236 aux_loss:0.0096 train_time:10770464ms step_avg:1899.22ms step:5682/6250 train_loss:1.7198 aux_loss:0.0095 train_time:10772280ms step_avg:1899.20ms step:5683/6250 train_loss:1.7910 aux_loss:0.0095 train_time:10773622ms step_avg:1899.10ms step:5684/6250 train_loss:1.8020 aux_loss:0.0095 train_time:10775048ms step_avg:1899.02ms step:5685/6250 train_loss:1.8497 aux_loss:0.0098 train_time:10776573ms step_avg:1898.96ms step:5686/6250 train_loss:1.7989 aux_loss:0.0094 train_time:10778109ms step_avg:1898.89ms step:5687/6250 train_loss:1.8771 aux_loss:0.0096 train_time:10779755ms step_avg:1898.85ms step:5688/6250 train_loss:1.7991 aux_loss:0.0095 train_time:10781172ms step_avg:1898.76ms step:5689/6250 train_loss:1.6958 aux_loss:0.0093 train_time:10782620ms step_avg:1898.68ms step:5690/6250 train_loss:1.8357 aux_loss:0.0096 train_time:10784172ms step_avg:1898.62ms step:5691/6250 train_loss:1.8383 aux_loss:0.0096 train_time:10785544ms step_avg:1898.53ms step:5692/6250 train_loss:1.8208 aux_loss:0.0095 train_time:10787017ms step_avg:1898.45ms step:5693/6250 train_loss:1.8023 aux_loss:0.0096 train_time:10788531ms step_avg:1898.39ms step:5694/6250 train_loss:1.7786 aux_loss:0.0098 train_time:10789979ms step_avg:1898.31ms step:5695/6250 train_loss:1.8005 aux_loss:0.0097 train_time:10791365ms step_avg:1898.22ms step:5696/6250 train_loss:1.8003 aux_loss:0.0094 train_time:10792683ms step_avg:1898.12ms step:5697/6250 train_loss:1.8304 aux_loss:0.0093 train_time:10794016ms step_avg:1898.02ms step:5698/6250 train_loss:1.7600 aux_loss:0.0091 train_time:10795461ms step_avg:1897.94ms step:5699/6250 train_loss:1.8142 aux_loss:0.0095 train_time:10796878ms step_avg:1897.85ms step:5700/6250 train_loss:1.7899 aux_loss:0.0097 train_time:10798292ms step_avg:1897.77ms step:5701/6250 train_loss:1.8238 aux_loss:0.0095 train_time:10799795ms step_avg:1897.70ms step:5702/6250 train_loss:1.8363 aux_loss:0.0097 train_time:10801303ms step_avg:1897.63ms step:5703/6250 train_loss:1.7883 aux_loss:0.0099 train_time:10802761ms step_avg:1897.55ms step:5704/6250 train_loss:1.8600 aux_loss:0.0098 train_time:10804326ms step_avg:1897.49ms step:5705/6250 train_loss:1.7651 aux_loss:0.0097 train_time:10805739ms step_avg:1897.41ms step:5706/6250 train_loss:1.6773 aux_loss:0.0093 train_time:10807251ms step_avg:1897.34ms step:5707/6250 train_loss:1.7529 aux_loss:0.0096 train_time:10808670ms step_avg:1897.26ms step:5708/6250 train_loss:1.8461 aux_loss:0.0096 train_time:10810206ms step_avg:1897.19ms step:5709/6250 train_loss:1.8148 aux_loss:0.0090 train_time:10811560ms step_avg:1897.10ms step:5710/6250 train_loss:1.7562 aux_loss:0.0092 train_time:10813199ms step_avg:1897.05ms step:5711/6250 train_loss:1.8175 aux_loss:0.0096 train_time:10814667ms step_avg:1896.98ms step:5712/6250 train_loss:1.7369 aux_loss:0.0095 train_time:10816238ms step_avg:1896.92ms step:5713/6250 train_loss:1.9210 aux_loss:0.0099 train_time:10818048ms step_avg:1896.90ms step:5714/6250 train_loss:1.7559 aux_loss:0.0095 train_time:10819471ms step_avg:1896.82ms step:5715/6250 train_loss:1.8034 aux_loss:0.0095 train_time:10820893ms step_avg:1896.74ms step:5716/6250 train_loss:1.8323 aux_loss:0.0097 train_time:10822312ms step_avg:1896.65ms step:5717/6250 train_loss:1.7611 aux_loss:0.0092 train_time:10823827ms step_avg:1896.59ms step:5718/6250 train_loss:1.7445 aux_loss:0.0092 train_time:10825286ms step_avg:1896.51ms step:5719/6250 train_loss:1.7228 aux_loss:0.0093 train_time:10826697ms step_avg:1896.43ms step:5720/6250 train_loss:1.8648 aux_loss:0.0095 train_time:10828182ms step_avg:1896.35ms step:5721/6250 train_loss:1.7764 aux_loss:0.0096 train_time:10829610ms step_avg:1896.27ms step:5722/6250 train_loss:1.7166 aux_loss:0.0094 train_time:10831086ms step_avg:1896.20ms step:5723/6250 train_loss:1.8760 aux_loss:0.0101 train_time:10832523ms step_avg:1896.12ms step:5724/6250 train_loss:1.8157 aux_loss:0.0099 train_time:10834205ms step_avg:1896.08ms step:5725/6250 train_loss:1.7285 aux_loss:0.0099 train_time:10835229ms step_avg:1895.93ms step:5726/6250 train_loss:1.6889 aux_loss:0.0093 train_time:10836690ms step_avg:1895.85ms step:5727/6250 train_loss:1.7966 aux_loss:0.0097 train_time:10838149ms step_avg:1895.78ms step:5728/6250 train_loss:1.7033 aux_loss:0.0094 train_time:10839196ms step_avg:1895.63ms step:5729/6250 train_loss:1.7290 aux_loss:0.0098 train_time:10840575ms step_avg:1895.54ms step:5730/6250 train_loss:1.6897 aux_loss:0.0094 train_time:10842039ms step_avg:1895.46ms step:5731/6250 train_loss:1.7702 aux_loss:0.0099 train_time:10843514ms step_avg:1895.39ms step:5732/6250 train_loss:1.7199 aux_loss:0.0095 train_time:10844894ms step_avg:1895.30ms step:5733/6250 train_loss:1.7784 aux_loss:0.0094 train_time:10846245ms step_avg:1895.20ms step:5734/6250 train_loss:1.7703 aux_loss:0.0092 train_time:10847653ms step_avg:1895.12ms step:5735/6250 train_loss:1.7059 aux_loss:0.0094 train_time:10849226ms step_avg:1895.06ms step:5736/6250 train_loss:1.8214 aux_loss:0.0095 train_time:10850488ms step_avg:1894.95ms step:5737/6250 train_loss:1.8454 aux_loss:0.0097 train_time:10851948ms step_avg:1894.87ms step:5738/6250 train_loss:1.7576 aux_loss:0.0096 train_time:10853451ms step_avg:1894.81ms step:5739/6250 train_loss:1.8902 aux_loss:0.0095 train_time:10855055ms step_avg:1894.76ms step:5740/6250 train_loss:1.8369 aux_loss:0.0097 train_time:10856475ms step_avg:1894.67ms step:5741/6250 train_loss:1.7765 aux_loss:0.0094 train_time:10857856ms step_avg:1894.58ms step:5742/6250 train_loss:1.7388 aux_loss:0.0094 train_time:10859290ms step_avg:1894.50ms step:5743/6250 train_loss:1.7845 aux_loss:0.0096 train_time:10860615ms step_avg:1894.40ms step:5744/6250 train_loss:1.7556 aux_loss:0.0095 train_time:10862016ms step_avg:1894.32ms step:5745/6250 train_loss:1.7694 aux_loss:0.0094 train_time:10863386ms step_avg:1894.23ms step:5746/6250 train_loss:1.7598 aux_loss:0.0097 train_time:10864748ms step_avg:1894.13ms step:5747/6250 train_loss:1.8320 aux_loss:0.0097 train_time:10866301ms step_avg:1894.07ms step:5748/6250 train_loss:1.8255 aux_loss:0.0095 train_time:10867770ms step_avg:1894.00ms step:5749/6250 train_loss:1.8443 aux_loss:0.0095 train_time:10869155ms step_avg:1893.91ms step:5750/6250 train_loss:1.7702 aux_loss:0.0097 train_time:10870641ms step_avg:1893.84ms step:5751/6250 train_loss:1.7933 aux_loss:0.0095 train_time:10872079ms step_avg:1893.76ms step:5752/6250 train_loss:1.7967 aux_loss:0.0095 train_time:10873320ms step_avg:1893.65ms step:5753/6250 train_loss:1.7130 aux_loss:0.0095 train_time:10874797ms step_avg:1893.57ms step:5754/6250 train_loss:1.7132 aux_loss:0.0097 train_time:10876201ms step_avg:1893.49ms step:5755/6250 train_loss:1.7875 aux_loss:0.0094 train_time:10878713ms step_avg:1893.60ms step:5756/6250 train_loss:1.8076 aux_loss:0.0097 train_time:10879985ms step_avg:1893.49ms step:5757/6250 train_loss:1.7563 aux_loss:0.0096 train_time:10881303ms step_avg:1893.39ms step:5758/6250 train_loss:1.8721 aux_loss:0.0099 train_time:10882559ms step_avg:1893.28ms step:5759/6250 train_loss:1.8189 aux_loss:0.0097 train_time:10883926ms step_avg:1893.19ms step:5760/6250 train_loss:1.8212 aux_loss:0.0092 train_time:10885232ms step_avg:1893.08ms step:5761/6250 train_loss:1.8676 aux_loss:0.0095 train_time:10886727ms step_avg:1893.01ms step:5762/6250 train_loss:1.8399 aux_loss:0.0095 train_time:10889907ms step_avg:1893.24ms step:5763/6250 train_loss:1.8176 aux_loss:0.0096 train_time:10893040ms step_avg:1893.45ms step:5764/6250 train_loss:1.7340 aux_loss:0.0092 train_time:10895843ms step_avg:1893.61ms step:5765/6250 train_loss:1.7389 aux_loss:0.0089 train_time:10897345ms step_avg:1893.54ms step:5766/6250 train_loss:1.8102 aux_loss:0.0095 train_time:10899030ms step_avg:1893.51ms step:5767/6250 train_loss:1.7365 aux_loss:0.0097 train_time:10901317ms step_avg:1893.58ms step:5768/6250 train_loss:1.8621 aux_loss:0.0097 train_time:10902806ms step_avg:1893.51ms step:5769/6250 train_loss:1.7732 aux_loss:0.0096 train_time:10905109ms step_avg:1893.58ms step:5770/6250 train_loss:1.7951 aux_loss:0.0096 train_time:10907459ms step_avg:1893.66ms step:5771/6250 train_loss:1.7827 aux_loss:0.0094 train_time:10909524ms step_avg:1893.69ms step:5772/6250 train_loss:1.7681 aux_loss:0.0097 train_time:10912433ms step_avg:1893.86ms step:5773/6250 train_loss:1.9495 aux_loss:0.0096 train_time:10914570ms step_avg:1893.90ms step:5774/6250 train_loss:1.8462 aux_loss:0.0096 train_time:10916104ms step_avg:1893.84ms step:5775/6250 train_loss:1.7358 aux_loss:0.0094 train_time:10917571ms step_avg:1893.77ms step:5776/6250 train_loss:1.7095 aux_loss:0.0093 train_time:10919267ms step_avg:1893.73ms step:5777/6250 train_loss:1.8364 aux_loss:0.0098 train_time:10920979ms step_avg:1893.70ms step:5778/6250 train_loss:1.7886 aux_loss:0.0099 train_time:10922637ms step_avg:1893.66ms step:5779/6250 train_loss:1.8330 aux_loss:0.0096 train_time:10924211ms step_avg:1893.61ms step:5780/6250 train_loss:1.7516 aux_loss:0.0095 train_time:10925519ms step_avg:1893.50ms step:5781/6250 train_loss:1.7934 aux_loss:0.0093 train_time:10927044ms step_avg:1893.44ms step:5782/6250 train_loss:1.7831 aux_loss:0.0095 train_time:10928486ms step_avg:1893.36ms step:5783/6250 train_loss:1.7655 aux_loss:0.0100 train_time:10930014ms step_avg:1893.30ms step:5784/6250 train_loss:1.7127 aux_loss:0.0101 train_time:10931377ms step_avg:1893.21ms step:5785/6250 train_loss:1.7928 aux_loss:0.0094 train_time:10932755ms step_avg:1893.12ms step:5786/6250 train_loss:1.7658 aux_loss:0.0096 train_time:10934155ms step_avg:1893.03ms step:5787/6250 train_loss:1.7699 aux_loss:0.0094 train_time:10935657ms step_avg:1892.96ms step:5788/6250 train_loss:1.7307 aux_loss:0.0092 train_time:10937036ms step_avg:1892.88ms step:5789/6250 train_loss:1.8185 aux_loss:0.0092 train_time:10938464ms step_avg:1892.80ms step:5790/6250 train_loss:1.7604 aux_loss:0.0090 train_time:10939725ms step_avg:1892.69ms step:5791/6250 train_loss:1.7677 aux_loss:0.0097 train_time:10941178ms step_avg:1892.61ms step:5792/6250 train_loss:1.9286 aux_loss:0.0099 train_time:10942889ms step_avg:1892.58ms step:5793/6250 train_loss:1.8114 aux_loss:0.0100 train_time:10944327ms step_avg:1892.50ms step:5794/6250 train_loss:1.9076 aux_loss:0.0097 train_time:10945900ms step_avg:1892.44ms step:5795/6250 train_loss:1.8338 aux_loss:0.0094 train_time:10947544ms step_avg:1892.40ms step:5796/6250 train_loss:1.7533 aux_loss:0.0095 train_time:10949030ms step_avg:1892.33ms step:5797/6250 train_loss:1.8112 aux_loss:0.0100 train_time:10950622ms step_avg:1892.28ms step:5798/6250 train_loss:1.8307 aux_loss:0.0098 train_time:10952081ms step_avg:1892.20ms step:5799/6250 train_loss:1.7696 aux_loss:0.0091 train_time:10953595ms step_avg:1892.14ms step:5800/6250 train_loss:1.8453 aux_loss:0.0094 train_time:10955444ms step_avg:1892.13ms step:5801/6250 train_loss:1.7861 aux_loss:0.0094 train_time:10956816ms step_avg:1892.04ms step:5802/6250 train_loss:1.8124 aux_loss:0.0094 train_time:10958278ms step_avg:1891.97ms step:5803/6250 train_loss:1.6797 aux_loss:0.0093 train_time:10959741ms step_avg:1891.89ms step:5804/6250 train_loss:1.8413 aux_loss:0.0094 train_time:10961241ms step_avg:1891.83ms step:5805/6250 train_loss:1.7527 aux_loss:0.0094 train_time:10962966ms step_avg:1891.80ms step:5806/6250 train_loss:1.8795 aux_loss:0.0097 train_time:10964461ms step_avg:1891.73ms step:5807/6250 train_loss:1.8009 aux_loss:0.0095 train_time:10966000ms step_avg:1891.67ms step:5808/6250 train_loss:1.8531 aux_loss:0.0096 train_time:10967558ms step_avg:1891.61ms step:5809/6250 train_loss:1.7835 aux_loss:0.0096 train_time:10968987ms step_avg:1891.53ms step:5810/6250 train_loss:1.7737 aux_loss:0.0095 train_time:10971506ms step_avg:1891.64ms step:5811/6250 train_loss:1.7880 aux_loss:0.0099 train_time:10976533ms step_avg:1892.18ms step:5812/6250 train_loss:1.7599 aux_loss:0.0096 train_time:10977951ms step_avg:1892.10ms step:5813/6250 train_loss:1.8148 aux_loss:0.0095 train_time:10979257ms step_avg:1892.00ms step:5814/6250 train_loss:1.8064 aux_loss:0.0093 train_time:10980554ms step_avg:1891.89ms step:5815/6250 train_loss:1.8317 aux_loss:0.0095 train_time:10982863ms step_avg:1891.97ms step:5816/6250 train_loss:1.9166 aux_loss:0.0094 train_time:10984287ms step_avg:1891.89ms step:5817/6250 train_loss:1.7589 aux_loss:0.0091 train_time:10985823ms step_avg:1891.82ms step:5818/6250 train_loss:1.7935 aux_loss:0.0091 train_time:10987398ms step_avg:1891.77ms step:5819/6250 train_loss:1.8233 aux_loss:0.0095 train_time:10988767ms step_avg:1891.68ms step:5820/6250 train_loss:1.8413 aux_loss:0.0096 train_time:10990230ms step_avg:1891.61ms step:5821/6250 train_loss:1.7557 aux_loss:0.0092 train_time:10991672ms step_avg:1891.53ms step:5822/6250 train_loss:1.7939 aux_loss:0.0095 train_time:10993235ms step_avg:1891.47ms step:5823/6250 train_loss:1.8374 aux_loss:0.0096 train_time:10994815ms step_avg:1891.42ms step:5824/6250 train_loss:1.7357 aux_loss:0.0095 train_time:10996339ms step_avg:1891.36ms step:5825/6250 train_loss:1.7942 aux_loss:0.0094 train_time:10997896ms step_avg:1891.30ms step:5826/6250 train_loss:1.7896 aux_loss:0.0094 train_time:10999980ms step_avg:1891.33ms step:5827/6250 train_loss:1.7283 aux_loss:0.0095 train_time:11002639ms step_avg:1891.46ms step:5828/6250 train_loss:1.8233 aux_loss:0.0096 train_time:11007741ms step_avg:1892.01ms step:5829/6250 train_loss:1.7050 aux_loss:0.0092 train_time:11009032ms step_avg:1891.91ms step:5830/6250 train_loss:1.8406 aux_loss:0.0098 train_time:11010398ms step_avg:1891.82ms step:5831/6250 train_loss:1.7724 aux_loss:0.0096 train_time:11011733ms step_avg:1891.73ms step:5832/6250 train_loss:1.8317 aux_loss:0.0097 train_time:11013106ms step_avg:1891.64ms step:5833/6250 train_loss:1.8433 aux_loss:0.0096 train_time:11014464ms step_avg:1891.54ms step:5834/6250 train_loss:1.6711 aux_loss:0.0093 train_time:11016215ms step_avg:1891.52ms step:5835/6250 train_loss:1.7168 aux_loss:0.0095 train_time:11017507ms step_avg:1891.42ms step:5836/6250 train_loss:1.7678 aux_loss:0.0096 train_time:11019164ms step_avg:1891.38ms step:5837/6250 train_loss:1.7320 aux_loss:0.0094 train_time:11020912ms step_avg:1891.35ms step:5838/6250 train_loss:1.8359 aux_loss:0.0096 train_time:11022755ms step_avg:1891.34ms step:5839/6250 train_loss:1.8040 aux_loss:0.0095 train_time:11024155ms step_avg:1891.26ms step:5840/6250 train_loss:1.8726 aux_loss:0.0096 train_time:11025477ms step_avg:1891.16ms step:5841/6250 train_loss:1.7311 aux_loss:0.0093 train_time:11027236ms step_avg:1891.14ms step:5842/6250 train_loss:1.6942 aux_loss:0.0093 train_time:11028573ms step_avg:1891.04ms step:5843/6250 train_loss:1.7724 aux_loss:0.0093 train_time:11030009ms step_avg:1890.97ms step:5844/6250 train_loss:1.7394 aux_loss:0.0090 train_time:11031786ms step_avg:1890.95ms step:5845/6250 train_loss:1.7911 aux_loss:0.0093 train_time:11033456ms step_avg:1890.91ms step:5846/6250 train_loss:1.8071 aux_loss:0.0095 train_time:11034840ms step_avg:1890.82ms step:5847/6250 train_loss:1.8083 aux_loss:0.0094 train_time:11036544ms step_avg:1890.79ms step:5848/6250 train_loss:1.7015 aux_loss:0.0094 train_time:11037864ms step_avg:1890.69ms step:5849/6250 train_loss:1.7493 aux_loss:0.0095 train_time:11039189ms step_avg:1890.60ms step:5850/6250 train_loss:1.7309 aux_loss:0.0094 train_time:11040546ms step_avg:1890.50ms step:5851/6250 train_loss:1.7692 aux_loss:0.0097 train_time:11041900ms step_avg:1890.41ms step:5852/6250 train_loss:1.7895 aux_loss:0.0093 train_time:11043230ms step_avg:1890.32ms step:5853/6250 train_loss:1.8392 aux_loss:0.0097 train_time:11044643ms step_avg:1890.24ms step:5854/6250 train_loss:1.7545 aux_loss:0.0095 train_time:11046058ms step_avg:1890.15ms step:5855/6250 train_loss:1.8244 aux_loss:0.0096 train_time:11047474ms step_avg:1890.07ms step:5856/6250 train_loss:1.7677 aux_loss:0.0094 train_time:11048787ms step_avg:1889.97ms step:5857/6250 train_loss:1.8042 aux_loss:0.0093 train_time:11050098ms step_avg:1889.87ms step:5858/6250 train_loss:1.8626 aux_loss:0.0095 train_time:11051457ms step_avg:1889.78ms step:5859/6250 train_loss:1.8051 aux_loss:0.0094 train_time:11052781ms step_avg:1889.69ms step:5860/6250 train_loss:1.8489 aux_loss:0.0097 train_time:11054501ms step_avg:1889.66ms step:5861/6250 train_loss:1.7012 aux_loss:0.0094 train_time:11055882ms step_avg:1889.57ms step:5862/6250 train_loss:1.7265 aux_loss:0.0090 train_time:11057337ms step_avg:1889.50ms step:5863/6250 train_loss:1.9023 aux_loss:0.0091 train_time:11058639ms step_avg:1889.40ms step:5864/6250 train_loss:1.7986 aux_loss:0.0094 train_time:11059969ms step_avg:1889.30ms step:5865/6250 train_loss:1.7918 aux_loss:0.0093 train_time:11061220ms step_avg:1889.19ms step:5866/6250 train_loss:1.8467 aux_loss:0.0094 train_time:11062610ms step_avg:1889.11ms step:5867/6250 train_loss:1.7115 aux_loss:0.0089 train_time:11064177ms step_avg:1889.05ms step:5868/6250 train_loss:1.7449 aux_loss:0.0095 train_time:11065721ms step_avg:1888.99ms step:5869/6250 train_loss:1.8021 aux_loss:0.0095 train_time:11067461ms step_avg:1888.97ms step:5870/6250 train_loss:1.7881 aux_loss:0.0092 train_time:11068815ms step_avg:1888.88ms step:5871/6250 train_loss:1.7883 aux_loss:0.0091 train_time:11070133ms step_avg:1888.78ms step:5872/6250 train_loss:1.7878 aux_loss:0.0092 train_time:11071436ms step_avg:1888.68ms step:5873/6250 train_loss:1.6634 aux_loss:0.0095 train_time:11072796ms step_avg:1888.59ms step:5874/6250 train_loss:1.7893 aux_loss:0.0095 train_time:11074160ms step_avg:1888.50ms step:5875/6250 train_loss:1.7644 aux_loss:0.0093 train_time:11075487ms step_avg:1888.40ms step:5876/6250 train_loss:1.8472 aux_loss:0.0093 train_time:11077009ms step_avg:1888.34ms step:5877/6250 train_loss:1.8404 aux_loss:0.0094 train_time:11078414ms step_avg:1888.26ms step:5878/6250 train_loss:1.7330 aux_loss:0.0093 train_time:11079762ms step_avg:1888.17ms step:5879/6250 train_loss:1.7354 aux_loss:0.0094 train_time:11081110ms step_avg:1888.07ms step:5880/6250 train_loss:1.7581 aux_loss:0.0100 train_time:11082370ms step_avg:1887.97ms step:5881/6250 train_loss:1.7462 aux_loss:0.0096 train_time:11083753ms step_avg:1887.88ms step:5882/6250 train_loss:1.8469 aux_loss:0.0096 train_time:11085106ms step_avg:1887.79ms step:5883/6250 train_loss:1.7713 aux_loss:0.0095 train_time:11086711ms step_avg:1887.74ms step:5884/6250 train_loss:1.7076 aux_loss:0.0096 train_time:11088145ms step_avg:1887.67ms step:5885/6250 train_loss:1.8365 aux_loss:0.0097 train_time:11090205ms step_avg:1887.69ms step:5886/6250 train_loss:1.7973 aux_loss:0.0095 train_time:11091594ms step_avg:1887.61ms step:5887/6250 train_loss:1.7516 aux_loss:0.0091 train_time:11093277ms step_avg:1887.57ms step:5888/6250 train_loss:1.8045 aux_loss:0.0092 train_time:11094660ms step_avg:1887.49ms step:5889/6250 train_loss:1.7770 aux_loss:0.0094 train_time:11097368ms step_avg:1887.63ms step:5890/6250 train_loss:1.7203 aux_loss:0.0097 train_time:11100121ms step_avg:1887.78ms step:5891/6250 train_loss:1.7717 aux_loss:0.0094 train_time:11102022ms step_avg:1887.78ms step:5892/6250 train_loss:1.8120 aux_loss:0.0092 train_time:11103851ms step_avg:1887.77ms step:5893/6250 train_loss:1.7580 aux_loss:0.0091 train_time:11105715ms step_avg:1887.76ms step:5894/6250 train_loss:1.8846 aux_loss:0.0095 train_time:11107362ms step_avg:1887.72ms step:5895/6250 train_loss:1.7969 aux_loss:0.0095 train_time:11109093ms step_avg:1887.70ms step:5896/6250 train_loss:1.8861 aux_loss:0.0094 train_time:11111155ms step_avg:1887.73ms step:5897/6250 train_loss:1.7936 aux_loss:0.0094 train_time:11113299ms step_avg:1887.77ms step:5898/6250 train_loss:1.7505 aux_loss:0.0093 train_time:11115400ms step_avg:1887.81ms step:5899/6250 train_loss:1.8208 aux_loss:0.0096 train_time:11120446ms step_avg:1888.34ms step:5900/6250 train_loss:1.7301 aux_loss:0.0093 train_time:11124441ms step_avg:1888.70ms step:5901/6250 train_loss:1.7185 aux_loss:0.0093 train_time:11126641ms step_avg:1888.75ms step:5902/6250 train_loss:1.7757 aux_loss:0.0095 train_time:11129029ms step_avg:1888.84ms step:5903/6250 train_loss:1.8001 aux_loss:0.0096 train_time:11131153ms step_avg:1888.88ms step:5904/6250 train_loss:1.7839 aux_loss:0.0095 train_time:11133564ms step_avg:1888.97ms step:5905/6250 train_loss:1.7200 aux_loss:0.0096 train_time:11137918ms step_avg:1889.38ms step:5906/6250 train_loss:1.7497 aux_loss:0.0096 train_time:11139903ms step_avg:1889.40ms step:5907/6250 train_loss:1.9499 aux_loss:0.0096 train_time:11143710ms step_avg:1889.73ms step:5908/6250 train_loss:1.7591 aux_loss:0.0092 train_time:11153391ms step_avg:1891.05ms step:5909/6250 train_loss:1.8164 aux_loss:0.0092 train_time:11154752ms step_avg:1890.96ms step:5910/6250 train_loss:1.7598 aux_loss:0.0093 train_time:11156121ms step_avg:1890.87ms step:5911/6250 train_loss:1.7361 aux_loss:0.0094 train_time:11170932ms step_avg:1893.06ms step:5912/6250 train_loss:1.8276 aux_loss:0.0091 train_time:11172159ms step_avg:1892.94ms step:5913/6250 train_loss:1.7364 aux_loss:0.0093 train_time:11173443ms step_avg:1892.84ms step:5914/6250 train_loss:1.7723 aux_loss:0.0096 train_time:11175570ms step_avg:1892.88ms step:5915/6250 train_loss:1.8395 aux_loss:0.0097 train_time:11177251ms step_avg:1892.85ms step:5916/6250 train_loss:1.7749 aux_loss:0.0099 train_time:11180748ms step_avg:1893.12ms step:5917/6250 train_loss:1.8519 aux_loss:0.0092 train_time:11182126ms step_avg:1893.03ms step:5918/6250 train_loss:1.8379 aux_loss:0.0105 train_time:11184563ms step_avg:1893.12ms step:5919/6250 train_loss:1.8891 aux_loss:0.0097 train_time:11186411ms step_avg:1893.11ms step:5920/6250 train_loss:1.7666 aux_loss:0.0096 train_time:11188097ms step_avg:1893.08ms step:5921/6250 train_loss:1.7423 aux_loss:0.0095 train_time:11190675ms step_avg:1893.19ms step:5922/6250 train_loss:1.7456 aux_loss:0.0094 train_time:11192440ms step_avg:1893.17ms step:5923/6250 train_loss:1.8109 aux_loss:0.0096 train_time:11195082ms step_avg:1893.30ms step:5924/6250 train_loss:1.7352 aux_loss:0.0098 train_time:11196969ms step_avg:1893.30ms step:5925/6250 train_loss:1.7624 aux_loss:0.0095 train_time:11198311ms step_avg:1893.21ms step:5926/6250 train_loss:1.6823 aux_loss:0.0094 train_time:11200353ms step_avg:1893.23ms step:5927/6250 train_loss:1.7786 aux_loss:0.0093 train_time:11201677ms step_avg:1893.13ms step:5928/6250 train_loss:1.8575 aux_loss:0.0096 train_time:11203122ms step_avg:1893.06ms step:5929/6250 train_loss:1.8405 aux_loss:0.0094 train_time:11204491ms step_avg:1892.97ms step:5930/6250 train_loss:1.8172 aux_loss:0.0095 train_time:11207039ms step_avg:1893.08ms step:5931/6250 train_loss:1.8099 aux_loss:0.0098 train_time:11208522ms step_avg:1893.01ms step:5932/6250 train_loss:1.8038 aux_loss:0.0095 train_time:11209840ms step_avg:1892.91ms step:5933/6250 train_loss:1.7341 aux_loss:0.0095 train_time:11211179ms step_avg:1892.82ms step:5934/6250 train_loss:1.8609 aux_loss:0.0095 train_time:11212679ms step_avg:1892.75ms step:5935/6250 train_loss:1.7612 aux_loss:0.0097 train_time:11214552ms step_avg:1892.75ms step:5936/6250 train_loss:1.6822 aux_loss:0.0093 train_time:11216799ms step_avg:1892.81ms step:5937/6250 train_loss:1.7793 aux_loss:0.0089 train_time:11218110ms step_avg:1892.71ms step:5938/6250 train_loss:1.7709 aux_loss:0.0089 train_time:11219717ms step_avg:1892.66ms step:5939/6250 train_loss:1.7959 aux_loss:0.0092 train_time:11221752ms step_avg:1892.69ms step:5940/6250 train_loss:1.8732 aux_loss:0.0096 train_time:11223007ms step_avg:1892.58ms step:5941/6250 train_loss:1.8011 aux_loss:0.0095 train_time:11224239ms step_avg:1892.47ms step:5942/6250 train_loss:1.8800 aux_loss:0.0093 train_time:11225563ms step_avg:1892.37ms step:5943/6250 train_loss:1.7718 aux_loss:0.0093 train_time:11227064ms step_avg:1892.31ms step:5944/6250 train_loss:1.9032 aux_loss:0.0093 train_time:11228421ms step_avg:1892.22ms step:5945/6250 train_loss:1.7325 aux_loss:0.0092 train_time:11230247ms step_avg:1892.21ms step:5946/6250 train_loss:1.8906 aux_loss:0.0092 train_time:11231837ms step_avg:1892.16ms step:5947/6250 train_loss:1.7776 aux_loss:0.0089 train_time:11233683ms step_avg:1892.15ms step:5948/6250 train_loss:1.7909 aux_loss:0.0093 train_time:11235046ms step_avg:1892.06ms step:5949/6250 train_loss:1.7777 aux_loss:0.0095 train_time:11236361ms step_avg:1891.96ms step:5950/6250 train_loss:1.7176 aux_loss:0.0093 train_time:11238182ms step_avg:1891.95ms step:5951/6250 train_loss:1.8525 aux_loss:0.0096 train_time:11240176ms step_avg:1891.97ms step:5952/6250 train_loss:1.8454 aux_loss:0.0096 train_time:11241858ms step_avg:1891.93ms step:5953/6250 train_loss:1.7667 aux_loss:0.0093 train_time:11243591ms step_avg:1891.90ms step:5954/6250 train_loss:1.7801 aux_loss:0.0093 train_time:11245565ms step_avg:1891.92ms step:5955/6250 train_loss:1.7317 aux_loss:0.0092 train_time:11247533ms step_avg:1891.93ms step:5956/6250 train_loss:1.7941 aux_loss:0.0095 train_time:11249334ms step_avg:1891.92ms step:5957/6250 train_loss:1.8538 aux_loss:0.0094 train_time:11251662ms step_avg:1891.99ms step:5958/6250 train_loss:1.8374 aux_loss:0.0095 train_time:11252996ms step_avg:1891.90ms step:5959/6250 train_loss:1.7404 aux_loss:0.0094 train_time:11255559ms step_avg:1892.01ms step:5960/6250 train_loss:1.7616 aux_loss:0.0095 train_time:11257765ms step_avg:1892.06ms step:5961/6250 train_loss:1.8156 aux_loss:0.0094 train_time:11259674ms step_avg:1892.06ms step:5962/6250 train_loss:1.8558 aux_loss:0.0098 train_time:11261303ms step_avg:1892.02ms step:5963/6250 train_loss:1.8016 aux_loss:0.0096 train_time:11262822ms step_avg:1891.96ms step:5964/6250 train_loss:1.8286 aux_loss:0.0094 train_time:11264155ms step_avg:1891.86ms step:5965/6250 train_loss:1.7898 aux_loss:0.0093 train_time:11265421ms step_avg:1891.76ms step:5966/6250 train_loss:1.7902 aux_loss:0.0093 train_time:11266842ms step_avg:1891.68ms step:5967/6250 train_loss:1.7636 aux_loss:0.0093 train_time:11268230ms step_avg:1891.59ms step:5968/6250 train_loss:1.7931 aux_loss:0.0090 train_time:11269903ms step_avg:1891.56ms step:5969/6250 train_loss:1.7946 aux_loss:0.0090 train_time:11271483ms step_avg:1891.51ms step:5970/6250 train_loss:1.8925 aux_loss:0.0096 train_time:11273380ms step_avg:1891.51ms step:5971/6250 train_loss:1.7437 aux_loss:0.0093 train_time:11275770ms step_avg:1891.59ms step:5972/6250 train_loss:1.8394 aux_loss:0.0094 train_time:11277256ms step_avg:1891.52ms step:5973/6250 train_loss:1.7975 aux_loss:0.0095 train_time:11279463ms step_avg:1891.58ms step:5974/6250 train_loss:1.7792 aux_loss:0.0098 train_time:11282592ms step_avg:1891.78ms step:5975/6250 train_loss:1.8849 aux_loss:0.0094 train_time:11284852ms step_avg:1891.84ms step:5976/6250 train_loss:1.8267 aux_loss:0.0093 train_time:11286698ms step_avg:1891.84ms step:5977/6250 train_loss:1.8698 aux_loss:0.0096 train_time:11288420ms step_avg:1891.81ms step:5978/6250 train_loss:1.7808 aux_loss:0.0094 train_time:11291986ms step_avg:1892.09ms step:5979/6250 train_loss:1.8515 aux_loss:0.0094 train_time:11294576ms step_avg:1892.21ms step:5980/6250 train_loss:1.8152 aux_loss:0.0095 train_time:11297012ms step_avg:1892.30ms step:5981/6250 train_loss:1.8477 aux_loss:0.0095 train_time:11304526ms step_avg:1893.24ms step:5982/6250 train_loss:1.7824 aux_loss:0.0094 train_time:11306915ms step_avg:1893.32ms step:5983/6250 train_loss:1.7098 aux_loss:0.0093 train_time:11309235ms step_avg:1893.39ms step:5984/6250 train_loss:1.8061 aux_loss:0.0092 train_time:11310706ms step_avg:1893.32ms step:5985/6250 train_loss:1.7425 aux_loss:0.0096 train_time:11312073ms step_avg:1893.23ms step:5986/6250 train_loss:1.7272 aux_loss:0.0095 train_time:11313441ms step_avg:1893.15ms step:5987/6250 train_loss:1.7653 aux_loss:0.0095 train_time:11314794ms step_avg:1893.06ms step:5988/6250 train_loss:1.7574 aux_loss:0.0095 train_time:11316081ms step_avg:1892.95ms step:5989/6250 train_loss:1.7659 aux_loss:0.0100 train_time:11317799ms step_avg:1892.93ms step:5990/6250 train_loss:1.7689 aux_loss:0.0096 train_time:11319211ms step_avg:1892.84ms step:5991/6250 train_loss:1.8936 aux_loss:0.0097 train_time:11320463ms step_avg:1892.74ms step:5992/6250 train_loss:1.7286 aux_loss:0.0095 train_time:11321886ms step_avg:1892.66ms step:5993/6250 train_loss:1.8435 aux_loss:0.0094 train_time:11323705ms step_avg:1892.65ms step:5994/6250 train_loss:1.8191 aux_loss:0.0094 train_time:11325085ms step_avg:1892.56ms step:5995/6250 train_loss:1.7940 aux_loss:0.0090 train_time:11326406ms step_avg:1892.47ms step:5996/6250 train_loss:1.8455 aux_loss:0.0090 train_time:11327736ms step_avg:1892.37ms step:5997/6250 train_loss:1.8340 aux_loss:0.0094 train_time:11329027ms step_avg:1892.27ms step:5998/6250 train_loss:1.8273 aux_loss:0.0092 train_time:11330637ms step_avg:1892.22ms step:5999/6250 train_loss:1.6580 aux_loss:0.0089 train_time:11332646ms step_avg:1892.24ms step:6000/6250 train_loss:1.8356 aux_loss:0.0092 train_time:11336348ms step_avg:1892.55ms step:6000/6250 val_loss:1.7838 val_aux_loss:0.0094 train_time:11336349ms step_avg:1892.55ms step:6001/6250 train_loss:1.7712 aux_loss:0.0095 train_time:11339650ms step_avg:1892.78ms step:6002/6250 train_loss:1.7857 aux_loss:0.0092 train_time:11344080ms step_avg:1893.20ms step:6003/6250 train_loss:1.7922 aux_loss:0.0092 train_time:11346211ms step_avg:1893.24ms step:6004/6250 train_loss:1.7546 aux_loss:0.0093 train_time:11347705ms step_avg:1893.18ms step:6005/6250 train_loss:1.8618 aux_loss:0.0095 train_time:11349082ms step_avg:1893.09ms step:6006/6250 train_loss:1.7514 aux_loss:0.0090 train_time:11350495ms step_avg:1893.01ms step:6007/6250 train_loss:1.7469 aux_loss:0.0092 train_time:11361375ms step_avg:1894.51ms step:6008/6250 train_loss:1.8070 aux_loss:0.0092 train_time:11426734ms step_avg:1905.09ms step:6009/6250 train_loss:1.7909 aux_loss:0.0092 train_time:11428151ms step_avg:1905.01ms step:6010/6250 train_loss:1.8289 aux_loss:0.0095 train_time:11429699ms step_avg:1904.95ms step:6011/6250 train_loss:1.7153 aux_loss:0.0098 train_time:11431094ms step_avg:1904.86ms step:6012/6250 train_loss:1.7889 aux_loss:0.0096 train_time:11432405ms step_avg:1904.77ms step:6013/6250 train_loss:1.8065 aux_loss:0.0093 train_time:11433792ms step_avg:1904.68ms step:6014/6250 train_loss:1.8443 aux_loss:0.0096 train_time:11435243ms step_avg:1904.60ms step:6015/6250 train_loss:1.7822 aux_loss:0.0095 train_time:11436580ms step_avg:1904.51ms step:6016/6250 train_loss:1.8006 aux_loss:0.0099 train_time:11438002ms step_avg:1904.43ms step:6017/6250 train_loss:1.8142 aux_loss:0.0095 train_time:11439280ms step_avg:1904.32ms step:6018/6250 train_loss:1.7799 aux_loss:0.0098 train_time:11444682ms step_avg:1904.91ms step:6019/6250 train_loss:1.8213 aux_loss:0.0095 train_time:11446196ms step_avg:1904.84ms step:6020/6250 train_loss:1.8381 aux_loss:0.0092 train_time:11447624ms step_avg:1904.76ms step:6021/6250 train_loss:1.7473 aux_loss:0.0092 train_time:11448904ms step_avg:1904.66ms step:6022/6250 train_loss:1.8413 aux_loss:0.0092 train_time:11450201ms step_avg:1904.56ms step:6023/6250 train_loss:1.7710 aux_loss:0.0092 train_time:11515360ms step_avg:1915.08ms step:6024/6250 train_loss:1.7672 aux_loss:0.0092 train_time:11525539ms step_avg:1916.45ms step:6025/6250 train_loss:1.7760 aux_loss:0.0091 train_time:11526826ms step_avg:1916.35ms step:6026/6250 train_loss:1.8379 aux_loss:0.0091 train_time:11528211ms step_avg:1916.26ms step:6027/6250 train_loss:1.7474 aux_loss:0.0095 train_time:11529556ms step_avg:1916.16ms step:6028/6250 train_loss:1.7912 aux_loss:0.0094 train_time:11530966ms step_avg:1916.08ms step:6029/6250 train_loss:1.7879 aux_loss:0.0092 train_time:11533030ms step_avg:1916.10ms step:6030/6250 train_loss:1.7287 aux_loss:0.0091 train_time:11535438ms step_avg:1916.19ms step:6031/6250 train_loss:1.6704 aux_loss:0.0092 train_time:11536835ms step_avg:1916.10ms step:6032/6250 train_loss:1.8463 aux_loss:0.0094 train_time:11541138ms step_avg:1916.50ms step:6033/6250 train_loss:1.8149 aux_loss:0.0093 train_time:11549058ms step_avg:1917.49ms step:6034/6250 train_loss:1.7841 aux_loss:0.0091 train_time:11550599ms step_avg:1917.43ms step:6035/6250 train_loss:1.8069 aux_loss:0.0090 train_time:11552609ms step_avg:1917.45ms step:6036/6250 train_loss:1.8363 aux_loss:0.0093 train_time:11553973ms step_avg:1917.35ms step:6037/6250 train_loss:1.7536 aux_loss:0.0096 train_time:11557156ms step_avg:1917.56ms step:6038/6250 train_loss:1.8265 aux_loss:0.0099 train_time:11564089ms step_avg:1918.40ms step:6039/6250 train_loss:1.8239 aux_loss:0.0095 train_time:11569781ms step_avg:1919.02ms step:6040/6250 train_loss:1.8012 aux_loss:0.0096 train_time:11571186ms step_avg:1918.94ms step:6041/6250 train_loss:1.8373 aux_loss:0.0100 train_time:11574293ms step_avg:1919.13ms step:6042/6250 train_loss:1.8510 aux_loss:0.0097 train_time:11605381ms step_avg:1923.97ms step:6043/6250 train_loss:1.8190 aux_loss:0.0097 train_time:11613631ms step_avg:1925.02ms step:6044/6250 train_loss:1.7431 aux_loss:0.0093 train_time:11627766ms step_avg:1927.04ms step:6045/6250 train_loss:1.6603 aux_loss:0.0091 train_time:11635824ms step_avg:1928.06ms step:6046/6250 train_loss:1.7476 aux_loss:0.0092 train_time:11637197ms step_avg:1927.97ms step:6047/6250 train_loss:1.7790 aux_loss:0.0092 train_time:11638674ms step_avg:1927.89ms step:6048/6250 train_loss:1.7361 aux_loss:0.0093 train_time:11639980ms step_avg:1927.79ms step:6049/6250 train_loss:1.7349 aux_loss:0.0093 train_time:11641371ms step_avg:1927.70ms step:6050/6250 train_loss:1.8161 aux_loss:0.0092 train_time:11642759ms step_avg:1927.61ms step:6051/6250 train_loss:1.7292 aux_loss:0.0094 train_time:11644064ms step_avg:1927.51ms step:6052/6250 train_loss:1.6627 aux_loss:0.0095 train_time:11645386ms step_avg:1927.41ms step:6053/6250 train_loss:1.7675 aux_loss:0.0094 train_time:11647030ms step_avg:1927.36ms step:6054/6250 train_loss:1.7385 aux_loss:0.0091 train_time:11648337ms step_avg:1927.26ms step:6055/6250 train_loss:1.7368 aux_loss:0.0093 train_time:11649689ms step_avg:1927.16ms step:6056/6250 train_loss:1.7595 aux_loss:0.0093 train_time:11651063ms step_avg:1927.07ms step:6057/6250 train_loss:1.8826 aux_loss:0.0097 train_time:11652378ms step_avg:1926.97ms step:6058/6250 train_loss:1.6804 aux_loss:0.0095 train_time:11653749ms step_avg:1926.88ms step:6059/6250 train_loss:1.8005 aux_loss:0.0093 train_time:11655042ms step_avg:1926.77ms step:6060/6250 train_loss:1.7451 aux_loss:0.0096 train_time:11657683ms step_avg:1926.89ms step:6061/6250 train_loss:1.8085 aux_loss:0.0094 train_time:11661331ms step_avg:1927.17ms step:6062/6250 train_loss:1.8230 aux_loss:0.0093 train_time:11667239ms step_avg:1927.83ms step:6063/6250 train_loss:1.8051 aux_loss:0.0093 train_time:11670661ms step_avg:1928.08ms step:6064/6250 train_loss:1.7236 aux_loss:0.0091 train_time:11672044ms step_avg:1927.99ms step:6065/6250 train_loss:1.7353 aux_loss:0.0093 train_time:11673788ms step_avg:1927.96ms step:6066/6250 train_loss:1.7225 aux_loss:0.0093 train_time:11679784ms step_avg:1928.63ms step:6067/6250 train_loss:1.8246 aux_loss:0.0095 train_time:11681189ms step_avg:1928.54ms step:6068/6250 train_loss:1.6881 aux_loss:0.0090 train_time:11682496ms step_avg:1928.44ms step:6069/6250 train_loss:1.8487 aux_loss:0.0094 train_time:11683818ms step_avg:1928.34ms step:6070/6250 train_loss:1.7160 aux_loss:0.0093 train_time:11685701ms step_avg:1928.33ms step:6071/6250 train_loss:1.7895 aux_loss:0.0092 train_time:11687589ms step_avg:1928.33ms step:6072/6250 train_loss:1.7790 aux_loss:0.0092 train_time:11689010ms step_avg:1928.24ms step:6073/6250 train_loss:1.7514 aux_loss:0.0093 train_time:11690797ms step_avg:1928.22ms step:6074/6250 train_loss:1.8295 aux_loss:0.0093 train_time:11692160ms step_avg:1928.13ms step:6075/6250 train_loss:1.8187 aux_loss:0.0091 train_time:11694965ms step_avg:1928.27ms step:6076/6250 train_loss:1.7311 aux_loss:0.0093 train_time:11698836ms step_avg:1928.59ms step:6077/6250 train_loss:1.8202 aux_loss:0.0094 train_time:11700617ms step_avg:1928.57ms step:6078/6250 train_loss:1.7424 aux_loss:0.0095 train_time:11701983ms step_avg:1928.47ms step:6079/6250 train_loss:1.8318 aux_loss:0.0094 train_time:11704467ms step_avg:1928.57ms step:6080/6250 train_loss:1.7607 aux_loss:0.0095 train_time:11705983ms step_avg:1928.50ms step:6081/6250 train_loss:1.7830 aux_loss:0.0093 train_time:11707356ms step_avg:1928.41ms step:6082/6250 train_loss:1.7282 aux_loss:0.0093 train_time:11708776ms step_avg:1928.32ms step:6083/6250 train_loss:1.7572 aux_loss:0.0093 train_time:11710303ms step_avg:1928.26ms step:6084/6250 train_loss:1.7312 aux_loss:0.0095 train_time:11713181ms step_avg:1928.41ms step:6085/6250 train_loss:1.6971 aux_loss:0.0107 train_time:11715625ms step_avg:1928.50ms step:6086/6250 train_loss:1.8879 aux_loss:0.0096 train_time:11717968ms step_avg:1928.57ms step:6087/6250 train_loss:1.8506 aux_loss:0.0094 train_time:11719386ms step_avg:1928.48ms step:6088/6250 train_loss:1.7740 aux_loss:0.0093 train_time:11720719ms step_avg:1928.38ms step:6089/6250 train_loss:1.7819 aux_loss:0.0095 train_time:11722213ms step_avg:1928.31ms step:6090/6250 train_loss:1.6895 aux_loss:0.0093 train_time:11723649ms step_avg:1928.23ms step:6091/6250 train_loss:1.7441 aux_loss:0.0092 train_time:11724964ms step_avg:1928.13ms step:6092/6250 train_loss:1.7337 aux_loss:0.0091 train_time:11726305ms step_avg:1928.03ms step:6093/6250 train_loss:1.7135 aux_loss:0.0092 train_time:11743664ms step_avg:1930.57ms step:6094/6250 train_loss:1.7908 aux_loss:0.0094 train_time:11756905ms step_avg:1932.43ms step:6095/6250 train_loss:1.7246 aux_loss:0.0091 train_time:11769603ms step_avg:1934.20ms step:6096/6250 train_loss:1.8363 aux_loss:0.0092 train_time:11777915ms step_avg:1935.25ms step:6097/6250 train_loss:1.7122 aux_loss:0.0091 train_time:11779258ms step_avg:1935.15ms step:6098/6250 train_loss:1.7029 aux_loss:0.0092 train_time:11782494ms step_avg:1935.36ms step:6099/6250 train_loss:1.7530 aux_loss:0.0093 train_time:11799411ms step_avg:1937.82ms step:6100/6250 train_loss:1.7750 aux_loss:0.0095 train_time:11809322ms step_avg:1939.13ms step:6101/6250 train_loss:1.7295 aux_loss:0.0096 train_time:11810692ms step_avg:1939.04ms step:6102/6250 train_loss:1.7549 aux_loss:0.0097 train_time:11812329ms step_avg:1938.99ms step:6103/6250 train_loss:1.6757 aux_loss:0.0095 train_time:11814086ms step_avg:1938.96ms step:6104/6250 train_loss:1.7699 aux_loss:0.0094 train_time:11815893ms step_avg:1938.94ms step:6105/6250 train_loss:1.7588 aux_loss:0.0093 train_time:11818866ms step_avg:1939.11ms step:6106/6250 train_loss:1.8456 aux_loss:0.0098 train_time:11820559ms step_avg:1939.07ms step:6107/6250 train_loss:1.8479 aux_loss:0.0098 train_time:11824141ms step_avg:1939.34ms step:6108/6250 train_loss:1.8001 aux_loss:0.0094 train_time:11826668ms step_avg:1939.43ms step:6109/6250 train_loss:1.7970 aux_loss:0.0095 train_time:11829484ms step_avg:1939.58ms step:6110/6250 train_loss:1.7974 aux_loss:0.0094 train_time:11831840ms step_avg:1939.65ms step:6111/6250 train_loss:1.7392 aux_loss:0.0094 train_time:11834697ms step_avg:1939.80ms step:6112/6250 train_loss:1.7904 aux_loss:0.0095 train_time:11837932ms step_avg:1940.01ms step:6113/6250 train_loss:1.8046 aux_loss:0.0093 train_time:11840639ms step_avg:1940.13ms step:6114/6250 train_loss:1.7411 aux_loss:0.0093 train_time:11843808ms step_avg:1940.34ms step:6115/6250 train_loss:1.8300 aux_loss:0.0096 train_time:11846871ms step_avg:1940.52ms step:6116/6250 train_loss:1.6974 aux_loss:0.0094 train_time:11850102ms step_avg:1940.73ms step:6117/6250 train_loss:1.8576 aux_loss:0.0096 train_time:11852419ms step_avg:1940.79ms step:6118/6250 train_loss:1.7293 aux_loss:0.0095 train_time:11854223ms step_avg:1940.77ms step:6119/6250 train_loss:1.7252 aux_loss:0.0094 train_time:11856175ms step_avg:1940.77ms step:6120/6250 train_loss:1.7890 aux_loss:0.0093 train_time:11858095ms step_avg:1940.77ms step:6121/6250 train_loss:1.7460 aux_loss:0.0093 train_time:11859803ms step_avg:1940.73ms step:6122/6250 train_loss:1.7878 aux_loss:0.0096 train_time:11861604ms step_avg:1940.71ms step:6123/6250 train_loss:1.7358 aux_loss:0.0092 train_time:11864002ms step_avg:1940.78ms step:6124/6250 train_loss:1.7512 aux_loss:0.0093 train_time:11865972ms step_avg:1940.79ms step:6125/6250 train_loss:1.6424 aux_loss:0.0090 train_time:11869508ms step_avg:1941.05ms step:6126/6250 train_loss:1.7382 aux_loss:0.0093 train_time:11872242ms step_avg:1941.18ms step:6127/6250 train_loss:1.6770 aux_loss:0.0092 train_time:11874881ms step_avg:1941.29ms step:6128/6250 train_loss:1.7599 aux_loss:0.0091 train_time:11877462ms step_avg:1941.40ms step:6129/6250 train_loss:1.7222 aux_loss:0.0093 train_time:11889037ms step_avg:1942.97ms step:6130/6250 train_loss:1.8216 aux_loss:0.0093 train_time:11890922ms step_avg:1942.96ms step:6131/6250 train_loss:1.7427 aux_loss:0.0095 train_time:11897378ms step_avg:1943.70ms step:6132/6250 train_loss:1.7380 aux_loss:0.0094 train_time:11907274ms step_avg:1945.00ms step:6133/6250 train_loss:1.7543 aux_loss:0.0095 train_time:11909400ms step_avg:1945.03ms step:6134/6250 train_loss:1.6481 aux_loss:0.0094 train_time:11910932ms step_avg:1944.96ms step:6135/6250 train_loss:1.6569 aux_loss:0.0093 train_time:11912585ms step_avg:1944.91ms step:6136/6250 train_loss:1.7483 aux_loss:0.0092 train_time:11917277ms step_avg:1945.36ms step:6137/6250 train_loss:1.7629 aux_loss:0.0091 train_time:11919144ms step_avg:1945.35ms step:6138/6250 train_loss:1.6993 aux_loss:0.0095 train_time:11920398ms step_avg:1945.23ms step:6139/6250 train_loss:1.8211 aux_loss:0.0094 train_time:11928985ms step_avg:1946.32ms step:6140/6250 train_loss:1.7383 aux_loss:0.0096 train_time:11930811ms step_avg:1946.30ms step:6141/6250 train_loss:1.6836 aux_loss:0.0098 train_time:11932499ms step_avg:1946.26ms step:6142/6250 train_loss:1.7730 aux_loss:0.0095 train_time:11934133ms step_avg:1946.21ms step:6143/6250 train_loss:1.7788 aux_loss:0.0097 train_time:11936569ms step_avg:1946.29ms step:6144/6250 train_loss:1.7524 aux_loss:0.0096 train_time:11939197ms step_avg:1946.40ms step:6145/6250 train_loss:1.7384 aux_loss:0.0095 train_time:11940634ms step_avg:1946.31ms step:6146/6250 train_loss:1.7003 aux_loss:0.0095 train_time:11942338ms step_avg:1946.27ms step:6147/6250 train_loss:1.6866 aux_loss:0.0093 train_time:11943635ms step_avg:1946.17ms step:6148/6250 train_loss:1.7196 aux_loss:0.0095 train_time:11945855ms step_avg:1946.21ms step:6149/6250 train_loss:1.6003 aux_loss:0.0091 train_time:11947443ms step_avg:1946.15ms step:6150/6250 train_loss:1.7584 aux_loss:0.0094 train_time:11951073ms step_avg:1946.43ms step:6151/6250 train_loss:1.7239 aux_loss:0.0094 train_time:11954396ms step_avg:1946.65ms step:6152/6250 train_loss:1.6052 aux_loss:0.0093 train_time:11958175ms step_avg:1946.95ms step:6153/6250 train_loss:1.6865 aux_loss:0.0093 train_time:11961176ms step_avg:1947.12ms step:6154/6250 train_loss:1.7130 aux_loss:0.0095 train_time:11963198ms step_avg:1947.14ms step:6155/6250 train_loss:1.8086 aux_loss:0.0095 train_time:11964635ms step_avg:1947.05ms step:6156/6250 train_loss:1.6978 aux_loss:0.0092 train_time:11966201ms step_avg:1946.99ms step:6157/6250 train_loss:1.7264 aux_loss:0.0091 train_time:11967697ms step_avg:1946.92ms step:6158/6250 train_loss:1.8012 aux_loss:0.0094 train_time:11969151ms step_avg:1946.84ms step:6159/6250 train_loss:1.7849 aux_loss:0.0096 train_time:11970675ms step_avg:1946.77ms step:6160/6250 train_loss:1.7214 aux_loss:0.0097 train_time:11972101ms step_avg:1946.68ms step:6161/6250 train_loss:1.7196 aux_loss:0.0094 train_time:11973704ms step_avg:1946.63ms step:6162/6250 train_loss:1.7100 aux_loss:0.0099 train_time:11975148ms step_avg:1946.55ms step:6163/6250 train_loss:1.7650 aux_loss:0.0093 train_time:11976513ms step_avg:1946.45ms step:6164/6250 train_loss:1.7597 aux_loss:0.0097 train_time:11977882ms step_avg:1946.36ms step:6165/6250 train_loss:1.6793 aux_loss:0.0094 train_time:11979223ms step_avg:1946.26ms step:6166/6250 train_loss:1.6832 aux_loss:0.0095 train_time:11980778ms step_avg:1946.20ms step:6167/6250 train_loss:1.6807 aux_loss:0.0094 train_time:11982193ms step_avg:1946.11ms step:6168/6250 train_loss:1.7914 aux_loss:0.0096 train_time:11983805ms step_avg:1946.05ms step:6169/6250 train_loss:1.6909 aux_loss:0.0095 train_time:11985202ms step_avg:1945.97ms step:6170/6250 train_loss:1.6663 aux_loss:0.0092 train_time:11986710ms step_avg:1945.89ms step:6171/6250 train_loss:1.8530 aux_loss:0.0095 train_time:11988274ms step_avg:1945.83ms step:6172/6250 train_loss:1.7070 aux_loss:0.0094 train_time:11989992ms step_avg:1945.80ms step:6173/6250 train_loss:1.7122 aux_loss:0.0097 train_time:11997356ms step_avg:1946.67ms step:6174/6250 train_loss:1.6771 aux_loss:0.0092 train_time:11999331ms step_avg:1946.68ms step:6175/6250 train_loss:1.6940 aux_loss:0.0093 train_time:12000934ms step_avg:1946.62ms step:6176/6250 train_loss:1.6800 aux_loss:0.0092 train_time:12002360ms step_avg:1946.54ms step:6177/6250 train_loss:1.7163 aux_loss:0.0093 train_time:12003678ms step_avg:1946.44ms step:6178/6250 train_loss:1.7031 aux_loss:0.0093 train_time:12004943ms step_avg:1946.33ms step:6179/6250 train_loss:1.6656 aux_loss:0.0093 train_time:12006402ms step_avg:1946.25ms step:6180/6250 train_loss:1.7950 aux_loss:0.0095 train_time:12007878ms step_avg:1946.17ms step:6181/6250 train_loss:1.7621 aux_loss:0.0094 train_time:12009462ms step_avg:1946.11ms step:6182/6250 train_loss:1.7591 aux_loss:0.0094 train_time:12011036ms step_avg:1946.05ms step:6183/6250 train_loss:1.7659 aux_loss:0.0096 train_time:12012334ms step_avg:1945.95ms step:6184/6250 train_loss:1.6850 aux_loss:0.0095 train_time:12013735ms step_avg:1945.86ms step:6185/6250 train_loss:1.6458 aux_loss:0.0094 train_time:12015139ms step_avg:1945.77ms step:6186/6250 train_loss:1.6637 aux_loss:0.0093 train_time:12016472ms step_avg:1945.67ms step:6187/6250 train_loss:1.6791 aux_loss:0.0092 train_time:12017776ms step_avg:1945.57ms step:6188/6250 train_loss:1.6684 aux_loss:0.0091 train_time:12019271ms step_avg:1945.50ms step:6189/6250 train_loss:1.6685 aux_loss:0.0093 train_time:12020709ms step_avg:1945.41ms step:6190/6250 train_loss:1.8232 aux_loss:0.0095 train_time:12022102ms step_avg:1945.32ms step:6191/6250 train_loss:1.7763 aux_loss:0.0096 train_time:12023646ms step_avg:1945.26ms step:6192/6250 train_loss:1.6963 aux_loss:0.0094 train_time:12025901ms step_avg:1945.31ms step:6193/6250 train_loss:1.7387 aux_loss:0.0096 train_time:12027439ms step_avg:1945.24ms step:6194/6250 train_loss:1.7575 aux_loss:0.0095 train_time:12029819ms step_avg:1945.31ms step:6195/6250 train_loss:1.6908 aux_loss:0.0095 train_time:12031364ms step_avg:1945.25ms step:6196/6250 train_loss:1.7502 aux_loss:0.0097 train_time:12032786ms step_avg:1945.16ms step:6197/6250 train_loss:1.7203 aux_loss:0.0092 train_time:12034107ms step_avg:1945.06ms step:6198/6250 train_loss:1.6947 aux_loss:0.0093 train_time:12035430ms step_avg:1944.96ms step:6199/6250 train_loss:1.7879 aux_loss:0.0093 train_time:12036780ms step_avg:1944.87ms step:6200/6250 train_loss:1.7046 aux_loss:0.0094 train_time:12038086ms step_avg:1944.76ms step:6201/6250 train_loss:1.7384 aux_loss:0.0094 train_time:12039395ms step_avg:1944.66ms step:6202/6250 train_loss:1.7708 aux_loss:0.0095 train_time:12040622ms step_avg:1944.54ms step:6203/6250 train_loss:1.7447 aux_loss:0.0095 train_time:12041811ms step_avg:1944.42ms step:6204/6250 train_loss:1.7077 aux_loss:0.0095 train_time:12043117ms step_avg:1944.32ms step:6205/6250 train_loss:1.6142 aux_loss:0.0094 train_time:12044493ms step_avg:1944.23ms step:6206/6250 train_loss:1.7755 aux_loss:0.0095 train_time:12045703ms step_avg:1944.11ms step:6207/6250 train_loss:1.7315 aux_loss:0.0096 train_time:12046978ms step_avg:1944.00ms step:6208/6250 train_loss:1.7376 aux_loss:0.0094 train_time:12048190ms step_avg:1943.88ms step:6209/6250 train_loss:1.6740 aux_loss:0.0093 train_time:12049404ms step_avg:1943.77ms step:6210/6250 train_loss:1.7348 aux_loss:0.0094 train_time:12050779ms step_avg:1943.67ms step:6211/6250 train_loss:1.7677 aux_loss:0.0095 train_time:12052117ms step_avg:1943.58ms step:6212/6250 train_loss:1.6673 aux_loss:0.0096 train_time:12053376ms step_avg:1943.47ms step:6213/6250 train_loss:1.7221 aux_loss:0.0094 train_time:12054957ms step_avg:1943.41ms step:6214/6250 train_loss:1.6954 aux_loss:0.0092 train_time:12056243ms step_avg:1943.30ms step:6215/6250 train_loss:1.7021 aux_loss:0.0094 train_time:12057775ms step_avg:1943.24ms step:6216/6250 train_loss:1.7050 aux_loss:0.0093 train_time:12059000ms step_avg:1943.12ms step:6217/6250 train_loss:1.7904 aux_loss:0.0094 train_time:12060329ms step_avg:1943.02ms step:6218/6250 train_loss:1.6737 aux_loss:0.0093 train_time:12061636ms step_avg:1942.92ms step:6219/6250 train_loss:1.7060 aux_loss:0.0095 train_time:12062913ms step_avg:1942.81ms step:6220/6250 train_loss:1.7173 aux_loss:0.0095 train_time:12064207ms step_avg:1942.71ms step:6221/6250 train_loss:1.7376 aux_loss:0.0095 train_time:12065547ms step_avg:1942.61ms step:6222/6250 train_loss:1.7553 aux_loss:0.0093 train_time:12066846ms step_avg:1942.51ms step:6223/6250 train_loss:1.6512 aux_loss:0.0093 train_time:12068156ms step_avg:1942.40ms step:6224/6250 train_loss:1.7400 aux_loss:0.0096 train_time:12069510ms step_avg:1942.31ms step:6225/6250 train_loss:1.6896 aux_loss:0.0094 train_time:12070910ms step_avg:1942.22ms step:6226/6250 train_loss:1.7017 aux_loss:0.0097 train_time:12072181ms step_avg:1942.11ms step:6227/6250 train_loss:1.7233 aux_loss:0.0097 train_time:12073716ms step_avg:1942.05ms step:6228/6250 train_loss:1.6318 aux_loss:0.0093 train_time:12075147ms step_avg:1941.97ms step:6229/6250 train_loss:1.6811 aux_loss:0.0094 train_time:12076625ms step_avg:1941.89ms step:6230/6250 train_loss:1.6052 aux_loss:0.0095 train_time:12077909ms step_avg:1941.79ms step:6231/6250 train_loss:1.6465 aux_loss:0.0092 train_time:12079274ms step_avg:1941.69ms step:6232/6250 train_loss:1.7663 aux_loss:0.0095 train_time:12080700ms step_avg:1941.61ms step:6233/6250 train_loss:1.7312 aux_loss:0.0095 train_time:12082117ms step_avg:1941.53ms step:6234/6250 train_loss:1.7779 aux_loss:0.0095 train_time:12083517ms step_avg:1941.44ms step:6235/6250 train_loss:1.6388 aux_loss:0.0094 train_time:12085361ms step_avg:1941.42ms step:6236/6250 train_loss:1.7913 aux_loss:0.0094 train_time:12086893ms step_avg:1941.36ms step:6237/6250 train_loss:1.7409 aux_loss:0.0093 train_time:12088064ms step_avg:1941.23ms step:6238/6250 train_loss:1.7523 aux_loss:0.0095 train_time:12089325ms step_avg:1941.12ms step:6239/6250 train_loss:1.7318 aux_loss:0.0095 train_time:12090518ms step_avg:1941.00ms step:6240/6250 train_loss:1.6108 aux_loss:0.0099 train_time:12091815ms step_avg:1940.90ms step:6241/6250 train_loss:1.7163 aux_loss:0.0094 train_time:12093147ms step_avg:1940.80ms step:6242/6250 train_loss:1.6811 aux_loss:0.0095 train_time:12095259ms step_avg:1940.83ms step:6243/6250 train_loss:1.7311 aux_loss:0.0095 train_time:12097009ms step_avg:1940.80ms step:6244/6250 train_loss:1.6817 aux_loss:0.0094 train_time:12098532ms step_avg:1940.73ms step:6245/6250 train_loss:1.6756 aux_loss:0.0092 train_time:12100352ms step_avg:1940.71ms step:6246/6250 train_loss:1.6915 aux_loss:0.0094 train_time:12102190ms step_avg:1940.70ms step:6247/6250 train_loss:1.6557 aux_loss:0.0092 train_time:12103485ms step_avg:1940.59ms step:6248/6250 train_loss:1.6887 aux_loss:0.0094 train_time:12105464ms step_avg:1940.60ms step:6249/6250 train_loss:1.8126 aux_loss:0.0094 train_time:12107043ms step_avg:1940.54ms step:6250/6250 train_loss:1.6961 aux_loss:0.0094 train_time:12109837ms step_avg:1940.68ms step:6250/6250 val_loss:1.7004 val_aux_loss:0.0094 train_time:12109837ms step_avg:1940.68ms