==================================================================================================== #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-rmsXrms-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":256,"dim_head":32,"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":4} 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 19:37:48 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 | | 30% 49C P2 48W / 350W | 2168MiB / 24564MiB | 2% 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 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 55224 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:92ms step_avg:nanms step:1/6250 train_loss:10.8258 aux_loss:0.0421 train_time:775ms step_avg:nanms step:2/6250 train_loss:10.8241 aux_loss:0.0421 train_time:1307ms step_avg:nanms step:3/6250 train_loss:10.8132 aux_loss:0.0421 train_time:1838ms step_avg:nanms step:4/6250 train_loss:10.7907 aux_loss:0.0421 train_time:2371ms step_avg:nanms step:5/6250 train_loss:10.7370 aux_loss:0.0419 train_time:2903ms step_avg:nanms step:6/6250 train_loss:10.6673 aux_loss:0.0418 train_time:3435ms step_avg:nanms step:7/6250 train_loss:10.5625 aux_loss:0.0415 train_time:3966ms step_avg:nanms step:8/6250 train_loss:10.4403 aux_loss:0.0413 train_time:4496ms step_avg:nanms step:9/6250 train_loss:10.2771 aux_loss:0.0409 train_time:5024ms step_avg:nanms step:10/6250 train_loss:10.0976 aux_loss:0.0406 train_time:5556ms step_avg:nanms step:11/6250 train_loss:9.9159 aux_loss:0.0403 train_time:528ms step_avg:nanms step:12/6250 train_loss:9.7027 aux_loss:0.0399 train_time:1056ms step_avg:nanms step:13/6250 train_loss:9.4653 aux_loss:0.0395 train_time:1585ms step_avg:528.41ms step:14/6250 train_loss:9.2061 aux_loss:0.0391 train_time:2138ms step_avg:534.59ms step:15/6250 train_loss:8.9507 aux_loss:0.0387 train_time:2688ms step_avg:537.67ms step:16/6250 train_loss:8.6819 aux_loss:0.0384 train_time:3217ms step_avg:536.23ms step:17/6250 train_loss:8.3864 aux_loss:0.0379 train_time:3745ms step_avg:535.06ms step:18/6250 train_loss:8.1263 aux_loss:0.0378 train_time:4275ms step_avg:534.34ms step:19/6250 train_loss:7.8693 aux_loss:0.0376 train_time:4802ms step_avg:533.54ms step:20/6250 train_loss:7.6015 aux_loss:0.0375 train_time:5331ms step_avg:533.11ms step:21/6250 train_loss:7.3720 aux_loss:0.0375 train_time:5858ms step_avg:532.54ms step:22/6250 train_loss:7.1101 aux_loss:0.0374 train_time:6385ms step_avg:532.08ms step:23/6250 train_loss:6.9991 aux_loss:0.0380 train_time:6911ms step_avg:531.59ms step:24/6250 train_loss:6.7689 aux_loss:0.0388 train_time:7439ms step_avg:531.35ms step:25/6250 train_loss:6.6179 aux_loss:0.0385 train_time:7966ms step_avg:531.10ms step:26/6250 train_loss:6.5123 aux_loss:0.0400 train_time:8495ms step_avg:530.96ms step:27/6250 train_loss:6.4161 aux_loss:0.0411 train_time:9024ms step_avg:530.81ms step:28/6250 train_loss:6.3956 aux_loss:0.0421 train_time:9552ms step_avg:530.64ms step:29/6250 train_loss:6.2874 aux_loss:0.0425 train_time:10080ms step_avg:530.50ms step:30/6250 train_loss:6.3039 aux_loss:0.0435 train_time:10613ms step_avg:530.67ms step:31/6250 train_loss:6.1520 aux_loss:0.0423 train_time:11158ms step_avg:531.32ms step:32/6250 train_loss:6.1367 aux_loss:0.0439 train_time:11693ms step_avg:531.51ms step:33/6250 train_loss:6.1574 aux_loss:0.0444 train_time:12223ms step_avg:531.45ms step:34/6250 train_loss:6.0139 aux_loss:0.0433 train_time:12750ms step_avg:531.27ms step:35/6250 train_loss:6.0666 aux_loss:0.0430 train_time:13284ms step_avg:531.35ms step:36/6250 train_loss:6.0406 aux_loss:0.0428 train_time:13868ms step_avg:533.37ms step:37/6250 train_loss:6.1393 aux_loss:0.0403 train_time:14450ms step_avg:535.18ms step:38/6250 train_loss:6.0011 aux_loss:0.0425 train_time:15032ms step_avg:536.86ms step:39/6250 train_loss:5.9542 aux_loss:0.0428 train_time:15614ms step_avg:538.41ms step:40/6250 train_loss:6.0456 aux_loss:0.0438 train_time:16196ms step_avg:539.86ms step:41/6250 train_loss:6.0844 aux_loss:0.0428 train_time:16779ms step_avg:541.25ms step:42/6250 train_loss:6.0021 aux_loss:0.0406 train_time:17359ms step_avg:542.47ms step:43/6250 train_loss:6.0098 aux_loss:0.0412 train_time:17940ms step_avg:543.65ms step:44/6250 train_loss:6.0678 aux_loss:0.0407 train_time:18522ms step_avg:544.77ms step:45/6250 train_loss:6.0059 aux_loss:0.0416 train_time:19118ms step_avg:546.23ms step:46/6250 train_loss:5.9608 aux_loss:0.0425 train_time:19703ms step_avg:547.29ms step:47/6250 train_loss:5.9013 aux_loss:0.0432 train_time:20255ms step_avg:547.43ms step:48/6250 train_loss:5.9818 aux_loss:0.0446 train_time:20801ms step_avg:547.39ms step:49/6250 train_loss:5.9629 aux_loss:0.0437 train_time:21341ms step_avg:547.21ms step:50/6250 train_loss:6.0225 aux_loss:0.0446 train_time:21878ms step_avg:546.96ms step:51/6250 train_loss:5.9754 aux_loss:0.0431 train_time:22417ms step_avg:546.75ms step:52/6250 train_loss:5.9533 aux_loss:0.0419 train_time:22952ms step_avg:546.47ms step:53/6250 train_loss:5.9621 aux_loss:0.0412 train_time:23489ms step_avg:546.26ms step:54/6250 train_loss:5.9345 aux_loss:0.0404 train_time:24024ms step_avg:546.00ms step:55/6250 train_loss:5.9239 aux_loss:0.0398 train_time:24553ms step_avg:545.63ms step:56/6250 train_loss:5.9687 aux_loss:0.0413 train_time:25081ms step_avg:545.24ms step:57/6250 train_loss:5.9434 aux_loss:0.0413 train_time:25614ms step_avg:544.97ms step:58/6250 train_loss:5.9189 aux_loss:0.0398 train_time:26145ms step_avg:544.69ms step:59/6250 train_loss:5.9064 aux_loss:0.0398 train_time:26677ms step_avg:544.42ms step:60/6250 train_loss:5.8661 aux_loss:0.0383 train_time:27206ms step_avg:544.11ms step:61/6250 train_loss:5.9122 aux_loss:0.0442 train_time:27735ms step_avg:543.82ms step:62/6250 train_loss:5.9727 aux_loss:0.0421 train_time:28266ms step_avg:543.57ms step:63/6250 train_loss:5.9153 aux_loss:0.0396 train_time:28795ms step_avg:543.31ms step:64/6250 train_loss:5.8220 aux_loss:0.0412 train_time:29324ms step_avg:543.04ms step:65/6250 train_loss:5.9358 aux_loss:0.0386 train_time:29853ms step_avg:542.77ms step:66/6250 train_loss:5.8950 aux_loss:0.0379 train_time:30383ms step_avg:542.54ms step:67/6250 train_loss:5.9269 aux_loss:0.0377 train_time:30911ms step_avg:542.30ms step:68/6250 train_loss:5.8761 aux_loss:0.0378 train_time:31440ms step_avg:542.06ms step:69/6250 train_loss:5.8927 aux_loss:0.0380 train_time:31969ms step_avg:541.85ms step:70/6250 train_loss:5.8540 aux_loss:0.0386 train_time:32502ms step_avg:541.70ms step:71/6250 train_loss:5.8502 aux_loss:0.0390 train_time:33033ms step_avg:541.52ms step:72/6250 train_loss:5.8509 aux_loss:0.0388 train_time:33565ms step_avg:541.36ms step:73/6250 train_loss:5.8176 aux_loss:0.0383 train_time:34093ms step_avg:541.16ms step:74/6250 train_loss:5.8654 aux_loss:0.0378 train_time:34622ms step_avg:540.97ms step:75/6250 train_loss:5.8448 aux_loss:0.0380 train_time:35151ms step_avg:540.79ms step:76/6250 train_loss:5.8723 aux_loss:0.0388 train_time:35679ms step_avg:540.58ms step:77/6250 train_loss:5.7425 aux_loss:0.0390 train_time:36208ms step_avg:540.41ms step:78/6250 train_loss:5.7729 aux_loss:0.0385 train_time:36741ms step_avg:540.31ms step:79/6250 train_loss:5.7334 aux_loss:0.0382 train_time:37274ms step_avg:540.20ms step:80/6250 train_loss:5.6645 aux_loss:0.0385 train_time:37804ms step_avg:540.06ms step:81/6250 train_loss:5.6680 aux_loss:0.0383 train_time:38337ms step_avg:539.95ms step:82/6250 train_loss:5.6441 aux_loss:0.0384 train_time:38870ms step_avg:539.87ms step:83/6250 train_loss:5.5586 aux_loss:0.0392 train_time:39404ms step_avg:539.79ms step:84/6250 train_loss:5.5453 aux_loss:0.0397 train_time:39935ms step_avg:539.66ms step:85/6250 train_loss:5.5955 aux_loss:0.0400 train_time:40467ms step_avg:539.56ms step:86/6250 train_loss:5.4562 aux_loss:0.0401 train_time:40998ms step_avg:539.44ms step:87/6250 train_loss:5.4990 aux_loss:0.0396 train_time:41530ms step_avg:539.35ms step:88/6250 train_loss:5.4725 aux_loss:0.0407 train_time:42062ms step_avg:539.26ms step:89/6250 train_loss:5.4424 aux_loss:0.0406 train_time:42599ms step_avg:539.23ms step:90/6250 train_loss:5.3445 aux_loss:0.0407 train_time:43131ms step_avg:539.14ms step:91/6250 train_loss:5.3011 aux_loss:0.0412 train_time:43665ms step_avg:539.07ms step:92/6250 train_loss:5.3081 aux_loss:0.0413 train_time:44198ms step_avg:539.00ms step:93/6250 train_loss:5.3110 aux_loss:0.0410 train_time:44729ms step_avg:538.91ms step:94/6250 train_loss:5.2679 aux_loss:0.0416 train_time:45258ms step_avg:538.79ms step:95/6250 train_loss:5.1208 aux_loss:0.0426 train_time:45787ms step_avg:538.67ms step:96/6250 train_loss:5.1904 aux_loss:0.0429 train_time:46321ms step_avg:538.61ms step:97/6250 train_loss:5.2524 aux_loss:0.0424 train_time:46853ms step_avg:538.54ms step:98/6250 train_loss:5.1652 aux_loss:0.0424 train_time:47385ms step_avg:538.47ms step:99/6250 train_loss:5.1452 aux_loss:0.0428 train_time:47917ms step_avg:538.40ms step:100/6250 train_loss:5.1611 aux_loss:0.0432 train_time:48451ms step_avg:538.34ms step:101/6250 train_loss:5.1199 aux_loss:0.0432 train_time:48984ms step_avg:538.28ms step:102/6250 train_loss:5.1247 aux_loss:0.0434 train_time:49521ms step_avg:538.27ms step:103/6250 train_loss:5.0963 aux_loss:0.0437 train_time:50053ms step_avg:538.21ms step:104/6250 train_loss:5.1129 aux_loss:0.0438 train_time:50588ms step_avg:538.17ms step:105/6250 train_loss:5.0469 aux_loss:0.0435 train_time:51121ms step_avg:538.12ms step:106/6250 train_loss:5.1132 aux_loss:0.0443 train_time:51652ms step_avg:538.05ms step:107/6250 train_loss:5.0328 aux_loss:0.0444 train_time:52183ms step_avg:537.97ms step:108/6250 train_loss:4.9767 aux_loss:0.0442 train_time:52714ms step_avg:537.89ms step:109/6250 train_loss:4.9583 aux_loss:0.0446 train_time:53244ms step_avg:537.81ms step:110/6250 train_loss:4.8694 aux_loss:0.0444 train_time:53772ms step_avg:537.72ms step:111/6250 train_loss:4.9254 aux_loss:0.0445 train_time:54300ms step_avg:537.62ms step:112/6250 train_loss:4.9412 aux_loss:0.0449 train_time:54828ms step_avg:537.53ms step:113/6250 train_loss:4.8548 aux_loss:0.0446 train_time:55377ms step_avg:537.64ms step:114/6250 train_loss:4.7797 aux_loss:0.0450 train_time:55914ms step_avg:537.64ms step:115/6250 train_loss:4.8475 aux_loss:0.0450 train_time:56444ms step_avg:537.56ms step:116/6250 train_loss:4.7940 aux_loss:0.0447 train_time:56985ms step_avg:537.59ms step:117/6250 train_loss:4.7543 aux_loss:0.0448 train_time:57528ms step_avg:537.64ms step:118/6250 train_loss:4.7118 aux_loss:0.0448 train_time:58074ms step_avg:537.72ms step:119/6250 train_loss:4.8069 aux_loss:0.0451 train_time:58624ms step_avg:537.83ms step:120/6250 train_loss:4.7385 aux_loss:0.0449 train_time:59161ms step_avg:537.83ms step:121/6250 train_loss:4.7044 aux_loss:0.0450 train_time:59690ms step_avg:537.75ms step:122/6250 train_loss:4.7024 aux_loss:0.0450 train_time:60229ms step_avg:537.76ms step:123/6250 train_loss:4.6602 aux_loss:0.0452 train_time:60771ms step_avg:537.80ms step:124/6250 train_loss:4.6470 aux_loss:0.0449 train_time:61332ms step_avg:538.00ms step:125/6250 train_loss:4.6448 aux_loss:0.0459 train_time:61864ms step_avg:537.94ms step:126/6250 train_loss:4.7824 aux_loss:0.0453 train_time:62392ms step_avg:537.86ms step:127/6250 train_loss:4.6054 aux_loss:0.0454 train_time:62920ms step_avg:537.78ms step:128/6250 train_loss:4.6453 aux_loss:0.0458 train_time:63448ms step_avg:537.70ms step:129/6250 train_loss:4.5806 aux_loss:0.0457 train_time:63977ms step_avg:537.62ms step:130/6250 train_loss:4.5750 aux_loss:0.0463 train_time:64505ms step_avg:537.54ms step:131/6250 train_loss:4.5830 aux_loss:0.0463 train_time:65032ms step_avg:537.46ms step:132/6250 train_loss:4.6007 aux_loss:0.0465 train_time:65563ms step_avg:537.40ms step:133/6250 train_loss:4.5585 aux_loss:0.0468 train_time:66091ms step_avg:537.32ms step:134/6250 train_loss:4.6570 aux_loss:0.0463 train_time:66656ms step_avg:537.55ms step:135/6250 train_loss:4.4327 aux_loss:0.0464 train_time:67216ms step_avg:537.73ms step:136/6250 train_loss:4.4562 aux_loss:0.0462 train_time:67754ms step_avg:537.73ms step:137/6250 train_loss:4.4944 aux_loss:0.0464 train_time:68294ms step_avg:537.75ms step:138/6250 train_loss:4.4317 aux_loss:0.0466 train_time:68840ms step_avg:537.81ms step:139/6250 train_loss:4.4614 aux_loss:0.0461 train_time:69386ms step_avg:537.88ms step:140/6250 train_loss:4.4611 aux_loss:0.0462 train_time:69926ms step_avg:537.89ms step:141/6250 train_loss:4.3896 aux_loss:0.0465 train_time:70455ms step_avg:537.82ms step:142/6250 train_loss:4.4022 aux_loss:0.0465 train_time:70984ms step_avg:537.76ms step:143/6250 train_loss:4.3834 aux_loss:0.0460 train_time:71513ms step_avg:537.69ms step:144/6250 train_loss:4.3649 aux_loss:0.0470 train_time:72043ms step_avg:537.63ms step:145/6250 train_loss:4.4590 aux_loss:0.0471 train_time:72602ms step_avg:537.79ms step:146/6250 train_loss:4.3717 aux_loss:0.0468 train_time:73151ms step_avg:537.87ms step:147/6250 train_loss:4.3585 aux_loss:0.0466 train_time:73678ms step_avg:537.80ms step:148/6250 train_loss:4.3611 aux_loss:0.0469 train_time:74207ms step_avg:537.73ms step:149/6250 train_loss:4.3248 aux_loss:0.0466 train_time:74752ms step_avg:537.79ms step:150/6250 train_loss:4.2726 aux_loss:0.0469 train_time:75333ms step_avg:538.10ms step:151/6250 train_loss:4.3704 aux_loss:0.0471 train_time:75894ms step_avg:538.26ms step:152/6250 train_loss:4.3127 aux_loss:0.0470 train_time:76447ms step_avg:538.36ms step:153/6250 train_loss:4.3392 aux_loss:0.0472 train_time:76978ms step_avg:538.31ms step:154/6250 train_loss:4.3506 aux_loss:0.0468 train_time:77512ms step_avg:538.28ms step:155/6250 train_loss:4.2862 aux_loss:0.0473 train_time:78043ms step_avg:538.23ms step:156/6250 train_loss:4.2292 aux_loss:0.0477 train_time:78577ms step_avg:538.20ms step:157/6250 train_loss:4.2822 aux_loss:0.0476 train_time:79110ms step_avg:538.16ms step:158/6250 train_loss:4.2918 aux_loss:0.0473 train_time:79642ms step_avg:538.12ms step:159/6250 train_loss:4.2464 aux_loss:0.0475 train_time:80177ms step_avg:538.10ms step:160/6250 train_loss:4.2774 aux_loss:0.0478 train_time:80712ms step_avg:538.08ms step:161/6250 train_loss:4.3210 aux_loss:0.0479 train_time:81245ms step_avg:538.05ms step:162/6250 train_loss:4.1855 aux_loss:0.0477 train_time:81780ms step_avg:538.02ms step:163/6250 train_loss:4.1907 aux_loss:0.0476 train_time:82315ms step_avg:538.01ms step:164/6250 train_loss:4.2518 aux_loss:0.0481 train_time:82850ms step_avg:537.98ms step:165/6250 train_loss:4.1465 aux_loss:0.0483 train_time:83383ms step_avg:537.95ms step:166/6250 train_loss:4.1758 aux_loss:0.0483 train_time:83929ms step_avg:538.01ms step:167/6250 train_loss:4.2330 aux_loss:0.0484 train_time:84486ms step_avg:538.13ms step:168/6250 train_loss:4.3551 aux_loss:0.0484 train_time:85032ms step_avg:538.18ms step:169/6250 train_loss:4.2309 aux_loss:0.0486 train_time:85587ms step_avg:538.28ms step:170/6250 train_loss:4.1920 aux_loss:0.0486 train_time:86133ms step_avg:538.33ms step:171/6250 train_loss:4.1363 aux_loss:0.0485 train_time:86665ms step_avg:538.29ms step:172/6250 train_loss:4.1768 aux_loss:0.0489 train_time:87199ms step_avg:538.26ms step:173/6250 train_loss:4.1577 aux_loss:0.0486 train_time:87729ms step_avg:538.22ms step:174/6250 train_loss:4.2026 aux_loss:0.0484 train_time:88263ms step_avg:538.19ms step:175/6250 train_loss:4.1090 aux_loss:0.0486 train_time:88795ms step_avg:538.15ms step:176/6250 train_loss:4.1316 aux_loss:0.0495 train_time:89327ms step_avg:538.12ms step:177/6250 train_loss:4.0885 aux_loss:0.0491 train_time:89857ms step_avg:538.07ms step:178/6250 train_loss:4.1123 aux_loss:0.0492 train_time:90388ms step_avg:538.03ms step:179/6250 train_loss:4.0563 aux_loss:0.0493 train_time:90919ms step_avg:537.98ms step:180/6250 train_loss:4.0387 aux_loss:0.0500 train_time:91449ms step_avg:537.93ms step:181/6250 train_loss:4.0516 aux_loss:0.0498 train_time:91979ms step_avg:537.89ms step:182/6250 train_loss:4.0145 aux_loss:0.0493 train_time:92510ms step_avg:537.85ms step:183/6250 train_loss:4.0282 aux_loss:0.0496 train_time:93042ms step_avg:537.82ms step:184/6250 train_loss:4.0805 aux_loss:0.0501 train_time:93576ms step_avg:537.79ms step:185/6250 train_loss:4.0051 aux_loss:0.0497 train_time:94106ms step_avg:537.75ms step:186/6250 train_loss:4.0227 aux_loss:0.0501 train_time:94636ms step_avg:537.70ms step:187/6250 train_loss:3.9796 aux_loss:0.0494 train_time:95166ms step_avg:537.66ms step:188/6250 train_loss:4.0486 aux_loss:0.0500 train_time:95694ms step_avg:537.61ms step:189/6250 train_loss:3.9815 aux_loss:0.0498 train_time:96224ms step_avg:537.57ms step:190/6250 train_loss:3.9938 aux_loss:0.0498 train_time:96754ms step_avg:537.52ms step:191/6250 train_loss:4.0173 aux_loss:0.0502 train_time:97289ms step_avg:537.51ms step:192/6250 train_loss:3.9743 aux_loss:0.0506 train_time:97818ms step_avg:537.46ms step:193/6250 train_loss:4.0433 aux_loss:0.0497 train_time:98348ms step_avg:537.42ms step:194/6250 train_loss:3.9491 aux_loss:0.0499 train_time:98875ms step_avg:537.36ms step:195/6250 train_loss:3.9874 aux_loss:0.0507 train_time:99405ms step_avg:537.33ms step:196/6250 train_loss:3.9933 aux_loss:0.0498 train_time:99939ms step_avg:537.31ms step:197/6250 train_loss:3.8789 aux_loss:0.0496 train_time:100472ms step_avg:537.28ms step:198/6250 train_loss:3.9287 aux_loss:0.0501 train_time:101001ms step_avg:537.24ms step:199/6250 train_loss:3.9162 aux_loss:0.0503 train_time:101532ms step_avg:537.20ms step:200/6250 train_loss:3.9492 aux_loss:0.0498 train_time:102063ms step_avg:537.17ms step:201/6250 train_loss:3.9410 aux_loss:0.0504 train_time:102594ms step_avg:537.14ms step:202/6250 train_loss:3.9229 aux_loss:0.0498 train_time:103124ms step_avg:537.10ms step:203/6250 train_loss:3.9319 aux_loss:0.0507 train_time:103656ms step_avg:537.08ms step:204/6250 train_loss:4.0056 aux_loss:0.0503 train_time:104191ms step_avg:537.07ms step:205/6250 train_loss:3.9629 aux_loss:0.0503 train_time:104724ms step_avg:537.05ms step:206/6250 train_loss:4.0151 aux_loss:0.0505 train_time:105257ms step_avg:537.03ms step:207/6250 train_loss:3.8458 aux_loss:0.0501 train_time:105789ms step_avg:537.00ms step:208/6250 train_loss:3.8892 aux_loss:0.0502 train_time:106320ms step_avg:536.97ms step:209/6250 train_loss:3.9403 aux_loss:0.0503 train_time:106848ms step_avg:536.92ms step:210/6250 train_loss:3.9092 aux_loss:0.0499 train_time:107378ms step_avg:536.89ms step:211/6250 train_loss:3.8860 aux_loss:0.0503 train_time:107907ms step_avg:536.85ms step:212/6250 train_loss:3.8833 aux_loss:0.0499 train_time:108436ms step_avg:536.81ms step:213/6250 train_loss:3.8765 aux_loss:0.0512 train_time:108966ms step_avg:536.78ms step:214/6250 train_loss:3.9288 aux_loss:0.0511 train_time:109496ms step_avg:536.75ms step:215/6250 train_loss:3.8311 aux_loss:0.0501 train_time:110029ms step_avg:536.73ms step:216/6250 train_loss:3.7840 aux_loss:0.0505 train_time:110581ms step_avg:536.80ms step:217/6250 train_loss:3.8794 aux_loss:0.0509 train_time:111127ms step_avg:536.84ms step:218/6250 train_loss:3.8842 aux_loss:0.0501 train_time:111656ms step_avg:536.81ms step:219/6250 train_loss:3.8348 aux_loss:0.0505 train_time:112183ms step_avg:536.76ms step:220/6250 train_loss:3.8476 aux_loss:0.0510 train_time:112721ms step_avg:536.76ms step:221/6250 train_loss:3.8573 aux_loss:0.0501 train_time:113250ms step_avg:536.73ms step:222/6250 train_loss:3.8823 aux_loss:0.0502 train_time:113779ms step_avg:536.69ms step:223/6250 train_loss:3.8203 aux_loss:0.0510 train_time:114308ms step_avg:536.66ms step:224/6250 train_loss:3.8531 aux_loss:0.0509 train_time:114838ms step_avg:536.62ms step:225/6250 train_loss:3.8627 aux_loss:0.0504 train_time:115368ms step_avg:536.59ms step:226/6250 train_loss:3.8492 aux_loss:0.0506 train_time:115917ms step_avg:536.65ms step:227/6250 train_loss:3.8368 aux_loss:0.0507 train_time:116457ms step_avg:536.67ms step:228/6250 train_loss:3.8066 aux_loss:0.0500 train_time:117002ms step_avg:536.71ms step:229/6250 train_loss:3.7848 aux_loss:0.0499 train_time:117549ms step_avg:536.76ms step:230/6250 train_loss:3.7732 aux_loss:0.0508 train_time:118095ms step_avg:536.79ms step:231/6250 train_loss:3.8589 aux_loss:0.0502 train_time:118644ms step_avg:536.85ms step:232/6250 train_loss:3.7703 aux_loss:0.0505 train_time:119214ms step_avg:537.00ms step:233/6250 train_loss:3.8599 aux_loss:0.0507 train_time:119775ms step_avg:537.11ms step:234/6250 train_loss:3.8577 aux_loss:0.0510 train_time:120333ms step_avg:537.20ms step:235/6250 train_loss:3.6604 aux_loss:0.0513 train_time:120880ms step_avg:537.24ms step:236/6250 train_loss:3.7885 aux_loss:0.0505 train_time:121409ms step_avg:537.21ms step:237/6250 train_loss:3.8064 aux_loss:0.0503 train_time:121945ms step_avg:537.20ms step:238/6250 train_loss:3.7284 aux_loss:0.0506 train_time:122491ms step_avg:537.24ms step:239/6250 train_loss:3.6850 aux_loss:0.0513 train_time:123032ms step_avg:537.26ms step:240/6250 train_loss:3.8659 aux_loss:0.0506 train_time:123561ms step_avg:537.22ms step:241/6250 train_loss:3.8405 aux_loss:0.0493 train_time:124102ms step_avg:537.24ms step:242/6250 train_loss:3.7587 aux_loss:0.0504 train_time:124629ms step_avg:537.20ms step:243/6250 train_loss:3.7298 aux_loss:0.0507 train_time:125171ms step_avg:537.21ms step:244/6250 train_loss:3.7772 aux_loss:0.0504 train_time:125703ms step_avg:537.19ms step:245/6250 train_loss:3.7579 aux_loss:0.0497 train_time:126260ms step_avg:537.28ms step:246/6250 train_loss:3.7698 aux_loss:0.0501 train_time:126818ms step_avg:537.37ms step:247/6250 train_loss:3.6289 aux_loss:0.0508 train_time:127358ms step_avg:537.37ms step:248/6250 train_loss:3.7117 aux_loss:0.0502 train_time:127896ms step_avg:537.38ms step:249/6250 train_loss:3.7180 aux_loss:0.0503 train_time:128438ms step_avg:537.40ms step:250/6250 train_loss:3.6581 aux_loss:0.0495 train_time:128998ms step_avg:537.49ms step:251/6250 train_loss:3.7327 aux_loss:0.0498 train_time:129536ms step_avg:537.49ms step:252/6250 train_loss:3.6929 aux_loss:0.0505 train_time:130065ms step_avg:537.46ms step:253/6250 train_loss:3.7317 aux_loss:0.0503 train_time:130597ms step_avg:537.43ms step:254/6250 train_loss:3.7368 aux_loss:0.0500 train_time:131127ms step_avg:537.40ms step:255/6250 train_loss:3.7021 aux_loss:0.0500 train_time:131667ms step_avg:537.42ms step:256/6250 train_loss:3.6460 aux_loss:0.0498 train_time:132215ms step_avg:537.46ms step:257/6250 train_loss:3.6943 aux_loss:0.0505 train_time:132764ms step_avg:537.51ms step:258/6250 train_loss:3.7244 aux_loss:0.0501 train_time:133294ms step_avg:537.48ms step:259/6250 train_loss:3.7448 aux_loss:0.0491 train_time:133823ms step_avg:537.44ms step:260/6250 train_loss:3.7098 aux_loss:0.0503 train_time:134372ms step_avg:537.49ms step:261/6250 train_loss:3.7281 aux_loss:0.0506 train_time:134911ms step_avg:537.49ms step:262/6250 train_loss:3.6949 aux_loss:0.0502 train_time:135441ms step_avg:537.46ms step:263/6250 train_loss:3.5432 aux_loss:0.0497 train_time:135972ms step_avg:537.44ms step:264/6250 train_loss:3.7026 aux_loss:0.0501 train_time:136501ms step_avg:537.41ms step:265/6250 train_loss:3.6516 aux_loss:0.0499 train_time:137031ms step_avg:537.38ms step:266/6250 train_loss:3.6903 aux_loss:0.0508 train_time:137565ms step_avg:537.36ms step:267/6250 train_loss:3.5994 aux_loss:0.0502 train_time:138094ms step_avg:537.33ms step:268/6250 train_loss:3.6704 aux_loss:0.0494 train_time:138624ms step_avg:537.30ms step:269/6250 train_loss:3.5863 aux_loss:0.0499 train_time:139154ms step_avg:537.27ms step:270/6250 train_loss:3.5982 aux_loss:0.0500 train_time:139685ms step_avg:537.25ms step:271/6250 train_loss:3.6147 aux_loss:0.0501 train_time:140214ms step_avg:537.22ms step:272/6250 train_loss:3.6277 aux_loss:0.0497 train_time:140744ms step_avg:537.19ms step:273/6250 train_loss:3.5680 aux_loss:0.0499 train_time:141272ms step_avg:537.16ms step:274/6250 train_loss:3.6416 aux_loss:0.0496 train_time:141801ms step_avg:537.12ms step:275/6250 train_loss:3.5619 aux_loss:0.0497 train_time:142330ms step_avg:537.09ms step:276/6250 train_loss:3.6144 aux_loss:0.0498 train_time:142858ms step_avg:537.06ms step:277/6250 train_loss:3.6568 aux_loss:0.0493 train_time:143387ms step_avg:537.03ms step:278/6250 train_loss:3.6321 aux_loss:0.0500 train_time:143917ms step_avg:537.00ms step:279/6250 train_loss:3.5078 aux_loss:0.0499 train_time:144446ms step_avg:536.97ms step:280/6250 train_loss:3.4556 aux_loss:0.0503 train_time:144975ms step_avg:536.94ms step:281/6250 train_loss:3.6125 aux_loss:0.0504 train_time:145503ms step_avg:536.91ms step:282/6250 train_loss:3.4757 aux_loss:0.0502 train_time:146033ms step_avg:536.89ms step:283/6250 train_loss:3.5248 aux_loss:0.0500 train_time:146564ms step_avg:536.86ms step:284/6250 train_loss:3.5935 aux_loss:0.0497 train_time:147093ms step_avg:536.83ms step:285/6250 train_loss:3.5878 aux_loss:0.0498 train_time:147622ms step_avg:536.81ms step:286/6250 train_loss:3.5400 aux_loss:0.0500 train_time:148151ms step_avg:536.78ms step:287/6250 train_loss:3.5847 aux_loss:0.0497 train_time:148680ms step_avg:536.75ms step:288/6250 train_loss:3.6211 aux_loss:0.0501 train_time:149208ms step_avg:536.72ms step:289/6250 train_loss:3.5371 aux_loss:0.0497 train_time:149736ms step_avg:536.69ms step:290/6250 train_loss:3.5693 aux_loss:0.0495 train_time:150265ms step_avg:536.66ms step:291/6250 train_loss:3.6087 aux_loss:0.0500 train_time:150793ms step_avg:536.63ms step:292/6250 train_loss:3.4977 aux_loss:0.0500 train_time:151323ms step_avg:536.61ms step:293/6250 train_loss:3.5193 aux_loss:0.0496 train_time:151850ms step_avg:536.57ms step:294/6250 train_loss:3.4778 aux_loss:0.0497 train_time:152382ms step_avg:536.55ms step:295/6250 train_loss:3.5543 aux_loss:0.0495 train_time:152912ms step_avg:536.53ms step:296/6250 train_loss:3.6064 aux_loss:0.0496 train_time:153442ms step_avg:536.51ms step:297/6250 train_loss:3.4628 aux_loss:0.0498 train_time:153969ms step_avg:536.48ms step:298/6250 train_loss:3.5354 aux_loss:0.0497 train_time:154500ms step_avg:536.46ms step:299/6250 train_loss:3.5411 aux_loss:0.0501 train_time:155028ms step_avg:536.43ms step:300/6250 train_loss:3.4013 aux_loss:0.0498 train_time:155558ms step_avg:536.41ms step:301/6250 train_loss:3.5360 aux_loss:0.0493 train_time:156087ms step_avg:536.38ms step:302/6250 train_loss:3.5410 aux_loss:0.0496 train_time:156613ms step_avg:536.34ms step:303/6250 train_loss:3.5638 aux_loss:0.0495 train_time:157142ms step_avg:536.32ms step:304/6250 train_loss:3.4709 aux_loss:0.0495 train_time:157670ms step_avg:536.29ms step:305/6250 train_loss:3.3694 aux_loss:0.0494 train_time:158198ms step_avg:536.26ms step:306/6250 train_loss:3.4771 aux_loss:0.0495 train_time:158727ms step_avg:536.24ms step:307/6250 train_loss:3.4924 aux_loss:0.0494 train_time:159259ms step_avg:536.22ms step:308/6250 train_loss:3.5816 aux_loss:0.0490 train_time:159788ms step_avg:536.20ms step:309/6250 train_loss:3.3749 aux_loss:0.0496 train_time:160315ms step_avg:536.17ms step:310/6250 train_loss:3.4573 aux_loss:0.0493 train_time:160843ms step_avg:536.14ms step:311/6250 train_loss:3.4236 aux_loss:0.0493 train_time:161399ms step_avg:536.21ms step:312/6250 train_loss:3.4751 aux_loss:0.0496 train_time:161944ms step_avg:536.24ms step:313/6250 train_loss:3.6212 aux_loss:0.0492 train_time:162485ms step_avg:536.25ms step:314/6250 train_loss:3.4292 aux_loss:0.0494 train_time:163026ms step_avg:536.27ms step:315/6250 train_loss:3.5654 aux_loss:0.0496 train_time:163560ms step_avg:536.26ms step:316/6250 train_loss:3.3772 aux_loss:0.0496 train_time:164120ms step_avg:536.34ms step:317/6250 train_loss:3.4805 aux_loss:0.0494 train_time:164651ms step_avg:536.32ms step:318/6250 train_loss:3.4219 aux_loss:0.0491 train_time:165179ms step_avg:536.29ms step:319/6250 train_loss:3.4179 aux_loss:0.0491 train_time:165724ms step_avg:536.32ms step:320/6250 train_loss:3.4491 aux_loss:0.0492 train_time:166265ms step_avg:536.34ms step:321/6250 train_loss:3.3783 aux_loss:0.0491 train_time:166848ms step_avg:536.49ms step:322/6250 train_loss:3.4308 aux_loss:0.0492 train_time:167444ms step_avg:536.68ms step:323/6250 train_loss:3.3949 aux_loss:0.0495 train_time:167998ms step_avg:536.73ms step:324/6250 train_loss:3.4879 aux_loss:0.0496 train_time:168541ms step_avg:536.75ms step:325/6250 train_loss:3.4606 aux_loss:0.0493 train_time:169085ms step_avg:536.78ms step:326/6250 train_loss:3.4862 aux_loss:0.0490 train_time:169650ms step_avg:536.87ms step:327/6250 train_loss:3.3868 aux_loss:0.0491 train_time:170191ms step_avg:536.88ms step:328/6250 train_loss:3.4838 aux_loss:0.0488 train_time:170731ms step_avg:536.89ms step:329/6250 train_loss:3.3891 aux_loss:0.0490 train_time:171270ms step_avg:536.90ms step:330/6250 train_loss:3.4410 aux_loss:0.0489 train_time:171812ms step_avg:536.91ms step:331/6250 train_loss:3.3283 aux_loss:0.0497 train_time:172350ms step_avg:536.92ms step:332/6250 train_loss:3.4803 aux_loss:0.0494 train_time:172891ms step_avg:536.93ms step:333/6250 train_loss:3.4542 aux_loss:0.0487 train_time:173438ms step_avg:536.96ms step:334/6250 train_loss:3.4319 aux_loss:0.0490 train_time:173979ms step_avg:536.97ms step:335/6250 train_loss:3.3312 aux_loss:0.0494 train_time:174515ms step_avg:536.97ms step:336/6250 train_loss:3.3920 aux_loss:0.0491 train_time:175055ms step_avg:536.98ms step:337/6250 train_loss:3.4648 aux_loss:0.0488 train_time:175587ms step_avg:536.96ms step:338/6250 train_loss:3.2807 aux_loss:0.0489 train_time:176128ms step_avg:536.98ms step:339/6250 train_loss:3.4724 aux_loss:0.0489 train_time:176674ms step_avg:537.00ms step:340/6250 train_loss:3.3477 aux_loss:0.0490 train_time:177214ms step_avg:537.01ms step:341/6250 train_loss:3.3952 aux_loss:0.0491 train_time:177746ms step_avg:537.00ms step:342/6250 train_loss:3.2866 aux_loss:0.0492 train_time:178279ms step_avg:536.99ms step:343/6250 train_loss:3.3802 aux_loss:0.0492 train_time:178824ms step_avg:537.01ms step:344/6250 train_loss:3.4775 aux_loss:0.0493 train_time:179367ms step_avg:537.03ms step:345/6250 train_loss:3.3824 aux_loss:0.0492 train_time:179912ms step_avg:537.05ms step:346/6250 train_loss:3.3078 aux_loss:0.0491 train_time:180477ms step_avg:537.14ms step:347/6250 train_loss:3.3358 aux_loss:0.0491 train_time:181023ms step_avg:537.16ms step:348/6250 train_loss:3.3780 aux_loss:0.0494 train_time:181595ms step_avg:537.26ms step:349/6250 train_loss:3.2556 aux_loss:0.0492 train_time:182153ms step_avg:537.32ms step:350/6250 train_loss:3.4159 aux_loss:0.0490 train_time:182716ms step_avg:537.40ms step:351/6250 train_loss:3.2826 aux_loss:0.0490 train_time:183245ms step_avg:537.38ms step:352/6250 train_loss:3.3270 aux_loss:0.0491 train_time:183784ms step_avg:537.38ms step:353/6250 train_loss:3.3415 aux_loss:0.0486 train_time:184343ms step_avg:537.44ms step:354/6250 train_loss:3.1968 aux_loss:0.0493 train_time:184875ms step_avg:537.43ms step:355/6250 train_loss:3.4464 aux_loss:0.0488 train_time:185411ms step_avg:537.42ms step:356/6250 train_loss:3.3089 aux_loss:0.0488 train_time:185940ms step_avg:537.40ms step:357/6250 train_loss:3.3796 aux_loss:0.0487 train_time:186469ms step_avg:537.37ms step:358/6250 train_loss:3.3482 aux_loss:0.0491 train_time:186998ms step_avg:537.35ms step:359/6250 train_loss:3.2517 aux_loss:0.0489 train_time:187556ms step_avg:537.41ms step:360/6250 train_loss:3.3823 aux_loss:0.0488 train_time:188091ms step_avg:537.40ms step:361/6250 train_loss:3.2387 aux_loss:0.0489 train_time:188663ms step_avg:537.50ms step:362/6250 train_loss:3.3663 aux_loss:0.0486 train_time:189210ms step_avg:537.53ms step:363/6250 train_loss:3.3276 aux_loss:0.0487 train_time:189752ms step_avg:537.54ms step:364/6250 train_loss:3.2605 aux_loss:0.0492 train_time:190300ms step_avg:537.57ms step:365/6250 train_loss:3.2125 aux_loss:0.0491 train_time:190828ms step_avg:537.55ms step:366/6250 train_loss:3.2076 aux_loss:0.0490 train_time:191358ms step_avg:537.52ms step:367/6250 train_loss:3.3381 aux_loss:0.0490 train_time:191886ms step_avg:537.50ms step:368/6250 train_loss:3.2876 aux_loss:0.0488 train_time:192416ms step_avg:537.48ms step:369/6250 train_loss:3.2705 aux_loss:0.0484 train_time:192946ms step_avg:537.45ms step:370/6250 train_loss:3.2676 aux_loss:0.0487 train_time:193484ms step_avg:537.45ms step:371/6250 train_loss:3.2400 aux_loss:0.0491 train_time:194064ms step_avg:537.57ms step:372/6250 train_loss:3.2985 aux_loss:0.0488 train_time:194619ms step_avg:537.62ms step:373/6250 train_loss:3.2571 aux_loss:0.0489 train_time:195157ms step_avg:537.62ms step:374/6250 train_loss:3.1788 aux_loss:0.0494 train_time:195732ms step_avg:537.73ms step:375/6250 train_loss:3.2649 aux_loss:0.0491 train_time:196267ms step_avg:537.72ms step:376/6250 train_loss:3.2251 aux_loss:0.0489 train_time:196814ms step_avg:537.74ms step:377/6250 train_loss:3.2265 aux_loss:0.0487 train_time:197356ms step_avg:537.75ms step:378/6250 train_loss:3.3357 aux_loss:0.0486 train_time:197883ms step_avg:537.73ms step:379/6250 train_loss:3.3075 aux_loss:0.0486 train_time:198411ms step_avg:537.70ms step:380/6250 train_loss:3.3379 aux_loss:0.0486 train_time:198939ms step_avg:537.67ms step:381/6250 train_loss:3.2658 aux_loss:0.0488 train_time:199469ms step_avg:537.65ms step:382/6250 train_loss:3.3564 aux_loss:0.0485 train_time:199997ms step_avg:537.63ms step:383/6250 train_loss:3.2488 aux_loss:0.0488 train_time:200526ms step_avg:537.60ms step:384/6250 train_loss:3.2947 aux_loss:0.0488 train_time:201063ms step_avg:537.60ms step:385/6250 train_loss:3.3398 aux_loss:0.0487 train_time:201599ms step_avg:537.60ms step:386/6250 train_loss:3.2418 aux_loss:0.0488 train_time:202140ms step_avg:537.61ms step:387/6250 train_loss:3.1312 aux_loss:0.0490 train_time:202675ms step_avg:537.60ms step:388/6250 train_loss:3.2389 aux_loss:0.0485 train_time:203206ms step_avg:537.58ms step:389/6250 train_loss:3.2046 aux_loss:0.0485 train_time:203751ms step_avg:537.60ms step:390/6250 train_loss:3.2041 aux_loss:0.0486 train_time:204305ms step_avg:537.64ms step:391/6250 train_loss:3.2496 aux_loss:0.0485 train_time:204853ms step_avg:537.67ms step:392/6250 train_loss:3.2251 aux_loss:0.0487 train_time:205400ms step_avg:537.70ms step:393/6250 train_loss:3.2532 aux_loss:0.0486 train_time:205940ms step_avg:537.70ms step:394/6250 train_loss:3.2173 aux_loss:0.0487 train_time:206483ms step_avg:537.72ms step:395/6250 train_loss:3.2320 aux_loss:0.0489 train_time:207022ms step_avg:537.72ms step:396/6250 train_loss:3.1201 aux_loss:0.0490 train_time:207553ms step_avg:537.70ms step:397/6250 train_loss:3.3135 aux_loss:0.0486 train_time:208083ms step_avg:537.68ms step:398/6250 train_loss:3.1852 aux_loss:0.0486 train_time:208613ms step_avg:537.66ms step:399/6250 train_loss:3.1909 aux_loss:0.0487 train_time:209144ms step_avg:537.64ms step:400/6250 train_loss:3.1754 aux_loss:0.0487 train_time:209676ms step_avg:537.63ms step:401/6250 train_loss:3.2083 aux_loss:0.0484 train_time:210206ms step_avg:537.61ms step:402/6250 train_loss:3.2347 aux_loss:0.0485 train_time:210736ms step_avg:537.59ms step:403/6250 train_loss:3.1396 aux_loss:0.0489 train_time:211267ms step_avg:537.57ms step:404/6250 train_loss:3.3236 aux_loss:0.0486 train_time:211794ms step_avg:537.55ms step:405/6250 train_loss:3.1701 aux_loss:0.0488 train_time:212326ms step_avg:537.54ms step:406/6250 train_loss:3.1828 aux_loss:0.0489 train_time:212867ms step_avg:537.54ms step:407/6250 train_loss:3.1904 aux_loss:0.0484 train_time:213398ms step_avg:537.53ms step:408/6250 train_loss:3.1413 aux_loss:0.0486 train_time:213927ms step_avg:537.50ms step:409/6250 train_loss:3.1601 aux_loss:0.0486 train_time:214458ms step_avg:537.49ms step:410/6250 train_loss:3.1472 aux_loss:0.0488 train_time:215000ms step_avg:537.50ms step:411/6250 train_loss:3.2464 aux_loss:0.0485 train_time:215529ms step_avg:537.48ms step:412/6250 train_loss:3.2669 aux_loss:0.0484 train_time:216092ms step_avg:537.54ms step:413/6250 train_loss:3.2278 aux_loss:0.0485 train_time:216675ms step_avg:537.66ms step:414/6250 train_loss:3.1560 aux_loss:0.0486 train_time:217262ms step_avg:537.78ms step:415/6250 train_loss:3.2473 aux_loss:0.0483 train_time:217846ms step_avg:537.89ms step:416/6250 train_loss:3.1082 aux_loss:0.0489 train_time:218441ms step_avg:538.03ms step:417/6250 train_loss:3.1882 aux_loss:0.0485 train_time:219015ms step_avg:538.12ms step:418/6250 train_loss:3.2251 aux_loss:0.0483 train_time:219561ms step_avg:538.14ms step:419/6250 train_loss:3.2308 aux_loss:0.0481 train_time:220102ms step_avg:538.15ms step:420/6250 train_loss:3.1742 aux_loss:0.0485 train_time:220637ms step_avg:538.14ms step:421/6250 train_loss:3.0820 aux_loss:0.0487 train_time:221177ms step_avg:538.14ms step:422/6250 train_loss:3.0973 aux_loss:0.0486 train_time:221718ms step_avg:538.15ms step:423/6250 train_loss:3.2149 aux_loss:0.0484 train_time:222248ms step_avg:538.13ms step:424/6250 train_loss:3.1815 aux_loss:0.0483 train_time:222780ms step_avg:538.12ms step:425/6250 train_loss:3.1795 aux_loss:0.0483 train_time:223313ms step_avg:538.10ms step:426/6250 train_loss:3.2280 aux_loss:0.0483 train_time:223846ms step_avg:538.09ms step:427/6250 train_loss:3.1997 aux_loss:0.0484 train_time:224377ms step_avg:538.07ms step:428/6250 train_loss:3.2066 aux_loss:0.0483 train_time:224922ms step_avg:538.09ms step:429/6250 train_loss:3.1085 aux_loss:0.0491 train_time:225464ms step_avg:538.10ms step:430/6250 train_loss:3.1244 aux_loss:0.0485 train_time:225993ms step_avg:538.08ms step:431/6250 train_loss:3.1061 aux_loss:0.0484 train_time:226524ms step_avg:538.06ms step:432/6250 train_loss:3.0705 aux_loss:0.0485 train_time:227068ms step_avg:538.08ms step:433/6250 train_loss:3.1668 aux_loss:0.0484 train_time:227612ms step_avg:538.09ms step:434/6250 train_loss:3.1510 aux_loss:0.0484 train_time:228150ms step_avg:538.09ms step:435/6250 train_loss:3.1729 aux_loss:0.0481 train_time:228680ms step_avg:538.07ms step:436/6250 train_loss:3.0868 aux_loss:0.0485 train_time:229210ms step_avg:538.05ms step:437/6250 train_loss:3.2043 aux_loss:0.0482 train_time:229739ms step_avg:538.03ms step:438/6250 train_loss:3.1398 aux_loss:0.0484 train_time:230270ms step_avg:538.01ms step:439/6250 train_loss:3.1645 aux_loss:0.0483 train_time:230800ms step_avg:538.00ms step:440/6250 train_loss:3.1016 aux_loss:0.0484 train_time:231338ms step_avg:538.00ms step:441/6250 train_loss:3.1216 aux_loss:0.0485 train_time:231870ms step_avg:537.98ms step:442/6250 train_loss:3.2667 aux_loss:0.0485 train_time:232403ms step_avg:537.97ms step:443/6250 train_loss:3.1638 aux_loss:0.0482 train_time:232931ms step_avg:537.95ms step:444/6250 train_loss:3.1888 aux_loss:0.0483 train_time:233462ms step_avg:537.93ms step:445/6250 train_loss:3.0185 aux_loss:0.0487 train_time:233990ms step_avg:537.91ms step:446/6250 train_loss:3.1810 aux_loss:0.0482 train_time:234518ms step_avg:537.89ms step:447/6250 train_loss:3.1446 aux_loss:0.0480 train_time:235060ms step_avg:537.90ms step:448/6250 train_loss:3.1192 aux_loss:0.0483 train_time:235606ms step_avg:537.91ms step:449/6250 train_loss:3.1716 aux_loss:0.0484 train_time:236153ms step_avg:537.93ms step:450/6250 train_loss:3.1194 aux_loss:0.0487 train_time:236701ms step_avg:537.96ms step:451/6250 train_loss:3.0389 aux_loss:0.0486 train_time:237245ms step_avg:537.97ms step:452/6250 train_loss:3.0576 aux_loss:0.0487 train_time:237786ms step_avg:537.98ms step:453/6250 train_loss:3.1192 aux_loss:0.0486 train_time:238327ms step_avg:537.98ms step:454/6250 train_loss:3.0522 aux_loss:0.0484 train_time:238863ms step_avg:537.98ms step:455/6250 train_loss:3.1696 aux_loss:0.0480 train_time:239398ms step_avg:537.97ms step:456/6250 train_loss:3.1716 aux_loss:0.0481 train_time:239937ms step_avg:537.97ms step:457/6250 train_loss:3.0128 aux_loss:0.0487 train_time:240472ms step_avg:537.97ms step:458/6250 train_loss:3.1101 aux_loss:0.0485 train_time:241006ms step_avg:537.96ms step:459/6250 train_loss:3.0647 aux_loss:0.0487 train_time:241540ms step_avg:537.95ms step:460/6250 train_loss:3.1569 aux_loss:0.0485 train_time:242074ms step_avg:537.94ms step:461/6250 train_loss:3.0770 aux_loss:0.0486 train_time:242609ms step_avg:537.93ms step:462/6250 train_loss:3.0961 aux_loss:0.0486 train_time:243144ms step_avg:537.93ms step:463/6250 train_loss:3.1352 aux_loss:0.0484 train_time:243676ms step_avg:537.92ms step:464/6250 train_loss:3.0898 aux_loss:0.0482 train_time:244215ms step_avg:537.92ms step:465/6250 train_loss:3.1203 aux_loss:0.0488 train_time:244747ms step_avg:537.90ms step:466/6250 train_loss:2.9863 aux_loss:0.0486 train_time:245284ms step_avg:537.90ms step:467/6250 train_loss:2.9883 aux_loss:0.0485 train_time:245819ms step_avg:537.90ms step:468/6250 train_loss:3.0735 aux_loss:0.0482 train_time:246356ms step_avg:537.89ms step:469/6250 train_loss:3.0656 aux_loss:0.0484 train_time:246892ms step_avg:537.89ms step:470/6250 train_loss:3.1285 aux_loss:0.0481 train_time:247428ms step_avg:537.89ms step:471/6250 train_loss:3.0796 aux_loss:0.0483 train_time:247972ms step_avg:537.90ms step:472/6250 train_loss:3.0050 aux_loss:0.0484 train_time:248508ms step_avg:537.90ms step:473/6250 train_loss:3.0756 aux_loss:0.0485 train_time:249039ms step_avg:537.88ms step:474/6250 train_loss:3.0837 aux_loss:0.0484 train_time:249574ms step_avg:537.87ms step:475/6250 train_loss:3.0847 aux_loss:0.0482 train_time:250118ms step_avg:537.89ms step:476/6250 train_loss:3.0489 aux_loss:0.0479 train_time:250667ms step_avg:537.91ms step:477/6250 train_loss:3.0499 aux_loss:0.0481 train_time:251214ms step_avg:537.93ms step:478/6250 train_loss:2.9947 aux_loss:0.0484 train_time:251758ms step_avg:537.94ms step:479/6250 train_loss:3.0086 aux_loss:0.0484 train_time:252289ms step_avg:537.93ms step:480/6250 train_loss:3.0657 aux_loss:0.0483 train_time:252823ms step_avg:537.92ms step:481/6250 train_loss:3.0310 aux_loss:0.0487 train_time:253371ms step_avg:537.94ms step:482/6250 train_loss:3.0561 aux_loss:0.0482 train_time:253907ms step_avg:537.94ms step:483/6250 train_loss:3.1473 aux_loss:0.0482 train_time:254437ms step_avg:537.92ms step:484/6250 train_loss:3.1272 aux_loss:0.0483 train_time:254968ms step_avg:537.91ms step:485/6250 train_loss:2.9860 aux_loss:0.0480 train_time:255495ms step_avg:537.88ms step:486/6250 train_loss:3.1446 aux_loss:0.0476 train_time:256027ms step_avg:537.87ms step:487/6250 train_loss:3.0563 aux_loss:0.0484 train_time:256570ms step_avg:537.88ms step:488/6250 train_loss:3.0726 aux_loss:0.0483 train_time:257117ms step_avg:537.90ms step:489/6250 train_loss:3.0469 aux_loss:0.0487 train_time:257663ms step_avg:537.92ms step:490/6250 train_loss:3.0456 aux_loss:0.0484 train_time:258207ms step_avg:537.93ms step:491/6250 train_loss:2.9996 aux_loss:0.0484 train_time:258744ms step_avg:537.93ms step:492/6250 train_loss:3.0740 aux_loss:0.0481 train_time:259291ms step_avg:537.95ms step:493/6250 train_loss:3.0577 aux_loss:0.0483 train_time:259837ms step_avg:537.96ms step:494/6250 train_loss:3.0031 aux_loss:0.0483 train_time:260369ms step_avg:537.95ms step:495/6250 train_loss:3.0348 aux_loss:0.0480 train_time:260929ms step_avg:538.00ms step:496/6250 train_loss:2.9671 aux_loss:0.0481 train_time:261483ms step_avg:538.03ms step:497/6250 train_loss:3.0278 aux_loss:0.0484 train_time:262012ms step_avg:538.01ms step:498/6250 train_loss:2.9475 aux_loss:0.0488 train_time:262543ms step_avg:538.00ms step:499/6250 train_loss:3.0620 aux_loss:0.0484 train_time:263072ms step_avg:537.98ms step:500/6250 train_loss:3.0162 aux_loss:0.0481 train_time:263602ms step_avg:537.96ms step:501/6250 train_loss:2.9699 aux_loss:0.0480 train_time:264131ms step_avg:537.94ms step:502/6250 train_loss:3.1418 aux_loss:0.0483 train_time:264663ms step_avg:537.93ms step:503/6250 train_loss:3.0489 aux_loss:0.0483 train_time:265193ms step_avg:537.92ms step:504/6250 train_loss:3.0799 aux_loss:0.0481 train_time:265751ms step_avg:537.96ms step:505/6250 train_loss:3.0490 aux_loss:0.0483 train_time:266318ms step_avg:538.02ms step:506/6250 train_loss:3.0669 aux_loss:0.0485 train_time:266855ms step_avg:538.01ms step:507/6250 train_loss:3.0224 aux_loss:0.0482 train_time:267385ms step_avg:538.00ms step:508/6250 train_loss:3.0601 aux_loss:0.0480 train_time:267914ms step_avg:537.98ms step:509/6250 train_loss:3.1260 aux_loss:0.0481 train_time:268443ms step_avg:537.96ms step:510/6250 train_loss:3.1139 aux_loss:0.0477 train_time:268972ms step_avg:537.94ms step:511/6250 train_loss:2.9963 aux_loss:0.0482 train_time:269500ms step_avg:537.92ms step:512/6250 train_loss:3.1113 aux_loss:0.0478 train_time:270029ms step_avg:537.91ms step:513/6250 train_loss:3.0096 aux_loss:0.0482 train_time:270557ms step_avg:537.89ms step:514/6250 train_loss:3.0598 aux_loss:0.0485 train_time:271085ms step_avg:537.87ms step:515/6250 train_loss:3.0756 aux_loss:0.0482 train_time:271614ms step_avg:537.85ms step:516/6250 train_loss:3.0667 aux_loss:0.0481 train_time:272141ms step_avg:537.83ms step:517/6250 train_loss:2.9467 aux_loss:0.0486 train_time:272673ms step_avg:537.82ms step:518/6250 train_loss:3.0460 aux_loss:0.0482 train_time:273204ms step_avg:537.80ms step:519/6250 train_loss:3.0472 aux_loss:0.0481 train_time:273733ms step_avg:537.79ms step:520/6250 train_loss:3.0457 aux_loss:0.0480 train_time:274263ms step_avg:537.77ms step:521/6250 train_loss:3.0319 aux_loss:0.0481 train_time:274791ms step_avg:537.75ms step:522/6250 train_loss:2.9733 aux_loss:0.0477 train_time:275321ms step_avg:537.74ms step:523/6250 train_loss:3.0054 aux_loss:0.0478 train_time:275849ms step_avg:537.72ms step:524/6250 train_loss:3.0326 aux_loss:0.0481 train_time:276377ms step_avg:537.70ms step:525/6250 train_loss:2.9689 aux_loss:0.0480 train_time:276905ms step_avg:537.68ms step:526/6250 train_loss:2.9677 aux_loss:0.0481 train_time:277436ms step_avg:537.67ms step:527/6250 train_loss:2.9809 aux_loss:0.0483 train_time:277963ms step_avg:537.65ms step:528/6250 train_loss:3.0239 aux_loss:0.0480 train_time:278494ms step_avg:537.63ms step:529/6250 train_loss:3.0298 aux_loss:0.0480 train_time:279022ms step_avg:537.61ms step:530/6250 train_loss:3.0105 aux_loss:0.0484 train_time:279554ms step_avg:537.60ms step:531/6250 train_loss:2.9799 aux_loss:0.0486 train_time:280082ms step_avg:537.59ms step:532/6250 train_loss:2.9099 aux_loss:0.0487 train_time:280610ms step_avg:537.57ms step:533/6250 train_loss:2.9411 aux_loss:0.0483 train_time:281156ms step_avg:537.58ms step:534/6250 train_loss:2.9516 aux_loss:0.0482 train_time:281704ms step_avg:537.60ms step:535/6250 train_loss:3.0178 aux_loss:0.0481 train_time:282255ms step_avg:537.63ms step:536/6250 train_loss:2.9315 aux_loss:0.0482 train_time:282804ms step_avg:537.65ms step:537/6250 train_loss:2.9367 aux_loss:0.0484 train_time:283353ms step_avg:537.67ms step:538/6250 train_loss:2.9759 aux_loss:0.0479 train_time:283902ms step_avg:537.69ms step:539/6250 train_loss:2.9966 aux_loss:0.0483 train_time:284452ms step_avg:537.72ms step:540/6250 train_loss:2.9943 aux_loss:0.0480 train_time:285000ms step_avg:537.74ms step:541/6250 train_loss:2.9651 aux_loss:0.0480 train_time:285534ms step_avg:537.73ms step:542/6250 train_loss:2.9481 aux_loss:0.0482 train_time:286066ms step_avg:537.72ms step:543/6250 train_loss:3.0534 aux_loss:0.0484 train_time:286617ms step_avg:537.74ms step:544/6250 train_loss:2.9378 aux_loss:0.0485 train_time:287166ms step_avg:537.76ms step:545/6250 train_loss:3.0401 aux_loss:0.0480 train_time:287715ms step_avg:537.78ms step:546/6250 train_loss:2.8406 aux_loss:0.0483 train_time:288263ms step_avg:537.80ms step:547/6250 train_loss:2.9292 aux_loss:0.0482 train_time:288811ms step_avg:537.82ms step:548/6250 train_loss:3.0156 aux_loss:0.0478 train_time:289358ms step_avg:537.84ms step:549/6250 train_loss:3.0308 aux_loss:0.0478 train_time:289907ms step_avg:537.86ms step:550/6250 train_loss:2.9964 aux_loss:0.0480 train_time:290451ms step_avg:537.87ms step:551/6250 train_loss:3.0331 aux_loss:0.0480 train_time:290981ms step_avg:537.86ms step:552/6250 train_loss:2.9239 aux_loss:0.0485 train_time:291509ms step_avg:537.84ms step:553/6250 train_loss:2.9737 aux_loss:0.0483 train_time:292039ms step_avg:537.83ms step:554/6250 train_loss:3.0443 aux_loss:0.0479 train_time:292568ms step_avg:537.81ms step:555/6250 train_loss:2.9501 aux_loss:0.0480 train_time:293097ms step_avg:537.79ms step:556/6250 train_loss:2.9372 aux_loss:0.0482 train_time:293626ms step_avg:537.78ms step:557/6250 train_loss:3.0299 aux_loss:0.0480 train_time:294153ms step_avg:537.76ms step:558/6250 train_loss:2.8897 aux_loss:0.0479 train_time:294698ms step_avg:537.77ms step:559/6250 train_loss:2.9434 aux_loss:0.0483 train_time:295250ms step_avg:537.80ms step:560/6250 train_loss:2.9897 aux_loss:0.0481 train_time:295799ms step_avg:537.82ms step:561/6250 train_loss:3.0212 aux_loss:0.0477 train_time:296349ms step_avg:537.84ms step:562/6250 train_loss:3.0193 aux_loss:0.0481 train_time:296890ms step_avg:537.84ms step:563/6250 train_loss:2.9305 aux_loss:0.0484 train_time:297430ms step_avg:537.85ms step:564/6250 train_loss:2.9277 aux_loss:0.0483 train_time:297959ms step_avg:537.83ms step:565/6250 train_loss:2.9455 aux_loss:0.0481 train_time:298501ms step_avg:537.84ms step:566/6250 train_loss:2.8660 aux_loss:0.0479 train_time:299045ms step_avg:537.85ms step:567/6250 train_loss:2.9779 aux_loss:0.0480 train_time:299591ms step_avg:537.86ms step:568/6250 train_loss:2.9514 aux_loss:0.0481 train_time:300147ms step_avg:537.90ms step:569/6250 train_loss:2.9525 aux_loss:0.0480 train_time:300693ms step_avg:537.91ms step:570/6250 train_loss:2.9590 aux_loss:0.0481 train_time:301237ms step_avg:537.92ms step:571/6250 train_loss:2.9778 aux_loss:0.0479 train_time:301769ms step_avg:537.91ms step:572/6250 train_loss:2.9533 aux_loss:0.0477 train_time:302313ms step_avg:537.92ms step:573/6250 train_loss:2.8713 aux_loss:0.0482 train_time:302843ms step_avg:537.91ms step:574/6250 train_loss:2.9159 aux_loss:0.0481 train_time:303375ms step_avg:537.90ms step:575/6250 train_loss:2.9380 aux_loss:0.0479 train_time:303908ms step_avg:537.89ms step:576/6250 train_loss:2.8878 aux_loss:0.0480 train_time:304438ms step_avg:537.88ms step:577/6250 train_loss:2.9003 aux_loss:0.0482 train_time:304968ms step_avg:537.86ms step:578/6250 train_loss:2.8652 aux_loss:0.0484 train_time:305498ms step_avg:537.85ms step:579/6250 train_loss:2.8424 aux_loss:0.0484 train_time:306028ms step_avg:537.83ms step:580/6250 train_loss:2.8659 aux_loss:0.0483 train_time:306556ms step_avg:537.82ms step:581/6250 train_loss:2.8827 aux_loss:0.0483 train_time:307098ms step_avg:537.83ms step:582/6250 train_loss:2.9755 aux_loss:0.0484 train_time:307641ms step_avg:537.83ms step:583/6250 train_loss:2.9852 aux_loss:0.0480 train_time:308180ms step_avg:537.84ms step:584/6250 train_loss:2.9816 aux_loss:0.0482 train_time:308732ms step_avg:537.86ms step:585/6250 train_loss:2.8445 aux_loss:0.0482 train_time:309280ms step_avg:537.88ms step:586/6250 train_loss:2.9666 aux_loss:0.0481 train_time:309820ms step_avg:537.88ms step:587/6250 train_loss:3.0114 aux_loss:0.0480 train_time:310359ms step_avg:537.88ms step:588/6250 train_loss:2.9791 aux_loss:0.0474 train_time:310905ms step_avg:537.90ms step:589/6250 train_loss:2.9332 aux_loss:0.0481 train_time:311441ms step_avg:537.89ms step:590/6250 train_loss:2.9321 aux_loss:0.0481 train_time:311974ms step_avg:537.89ms step:591/6250 train_loss:2.8988 aux_loss:0.0481 train_time:312523ms step_avg:537.91ms step:592/6250 train_loss:2.9807 aux_loss:0.0481 train_time:313056ms step_avg:537.90ms step:593/6250 train_loss:2.9170 aux_loss:0.0481 train_time:313588ms step_avg:537.89ms step:594/6250 train_loss:2.8551 aux_loss:0.0481 train_time:314116ms step_avg:537.87ms step:595/6250 train_loss:2.9152 aux_loss:0.0477 train_time:314645ms step_avg:537.85ms step:596/6250 train_loss:2.9998 aux_loss:0.0475 train_time:315173ms step_avg:537.84ms step:597/6250 train_loss:3.0126 aux_loss:0.0479 train_time:315710ms step_avg:537.84ms step:598/6250 train_loss:2.9021 aux_loss:0.0479 train_time:316272ms step_avg:537.88ms step:599/6250 train_loss:2.8160 aux_loss:0.0480 train_time:316830ms step_avg:537.91ms step:600/6250 train_loss:2.7831 aux_loss:0.0483 train_time:317365ms step_avg:537.91ms step:601/6250 train_loss:2.8986 aux_loss:0.0480 train_time:317893ms step_avg:537.89ms step:602/6250 train_loss:2.9119 aux_loss:0.0480 train_time:318424ms step_avg:537.88ms step:603/6250 train_loss:2.9156 aux_loss:0.0481 train_time:318958ms step_avg:537.87ms step:604/6250 train_loss:2.9300 aux_loss:0.0479 train_time:319497ms step_avg:537.87ms step:605/6250 train_loss:2.8270 aux_loss:0.0483 train_time:320041ms step_avg:537.88ms step:606/6250 train_loss:2.9515 aux_loss:0.0482 train_time:320580ms step_avg:537.89ms step:607/6250 train_loss:2.9153 aux_loss:0.0480 train_time:321113ms step_avg:537.88ms step:608/6250 train_loss:2.8162 aux_loss:0.0483 train_time:321653ms step_avg:537.88ms step:609/6250 train_loss:2.9191 aux_loss:0.0478 train_time:322211ms step_avg:537.91ms step:610/6250 train_loss:2.8077 aux_loss:0.0481 train_time:322750ms step_avg:537.92ms step:611/6250 train_loss:2.8858 aux_loss:0.0482 train_time:323286ms step_avg:537.91ms step:612/6250 train_loss:2.9217 aux_loss:0.0482 train_time:323851ms step_avg:537.96ms step:613/6250 train_loss:2.9473 aux_loss:0.0481 train_time:324405ms step_avg:537.98ms step:614/6250 train_loss:2.8744 aux_loss:0.0478 train_time:324946ms step_avg:537.99ms step:615/6250 train_loss:2.8922 aux_loss:0.0478 train_time:325499ms step_avg:538.01ms step:616/6250 train_loss:2.8610 aux_loss:0.0479 train_time:326029ms step_avg:538.00ms step:617/6250 train_loss:2.9381 aux_loss:0.0475 train_time:326577ms step_avg:538.02ms step:618/6250 train_loss:2.8058 aux_loss:0.0476 train_time:327120ms step_avg:538.03ms step:619/6250 train_loss:2.8271 aux_loss:0.0479 train_time:327668ms step_avg:538.04ms step:620/6250 train_loss:2.9052 aux_loss:0.0479 train_time:328211ms step_avg:538.05ms step:621/6250 train_loss:2.9738 aux_loss:0.0476 train_time:328761ms step_avg:538.07ms step:622/6250 train_loss:2.8036 aux_loss:0.0481 train_time:329307ms step_avg:538.08ms step:623/6250 train_loss:2.9094 aux_loss:0.0479 train_time:329848ms step_avg:538.09ms step:624/6250 train_loss:2.9165 aux_loss:0.0479 train_time:330382ms step_avg:538.08ms step:625/6250 train_loss:2.8578 aux_loss:0.0480 train_time:330924ms step_avg:538.09ms step:626/6250 train_loss:2.9029 aux_loss:0.0480 train_time:331461ms step_avg:538.09ms step:627/6250 train_loss:2.8845 aux_loss:0.0481 train_time:331992ms step_avg:538.07ms step:628/6250 train_loss:2.8701 aux_loss:0.0478 train_time:332526ms step_avg:538.07ms step:629/6250 train_loss:2.9094 aux_loss:0.0478 train_time:333060ms step_avg:538.06ms step:630/6250 train_loss:2.9167 aux_loss:0.0477 train_time:333601ms step_avg:538.07ms step:631/6250 train_loss:2.9037 aux_loss:0.0475 train_time:334138ms step_avg:538.06ms step:632/6250 train_loss:2.9161 aux_loss:0.0478 train_time:334677ms step_avg:538.07ms step:633/6250 train_loss:2.9904 aux_loss:0.0476 train_time:335212ms step_avg:538.06ms step:634/6250 train_loss:2.8920 aux_loss:0.0478 train_time:335745ms step_avg:538.05ms step:635/6250 train_loss:2.8609 aux_loss:0.0480 train_time:336281ms step_avg:538.05ms step:636/6250 train_loss:2.9084 aux_loss:0.0478 train_time:336816ms step_avg:538.04ms step:637/6250 train_loss:2.7400 aux_loss:0.0477 train_time:337350ms step_avg:538.04ms step:638/6250 train_loss:2.8929 aux_loss:0.0475 train_time:337883ms step_avg:538.03ms step:639/6250 train_loss:2.8438 aux_loss:0.0477 train_time:338418ms step_avg:538.03ms step:640/6250 train_loss:2.7297 aux_loss:0.0480 train_time:338952ms step_avg:538.02ms step:641/6250 train_loss:2.9326 aux_loss:0.0479 train_time:339490ms step_avg:538.02ms step:642/6250 train_loss:2.8644 aux_loss:0.0480 train_time:340024ms step_avg:538.01ms step:643/6250 train_loss:2.8790 aux_loss:0.0481 train_time:340564ms step_avg:538.02ms step:644/6250 train_loss:2.8058 aux_loss:0.0481 train_time:341105ms step_avg:538.02ms step:645/6250 train_loss:2.8386 aux_loss:0.0477 train_time:341649ms step_avg:538.03ms step:646/6250 train_loss:2.9044 aux_loss:0.0479 train_time:342194ms step_avg:538.04ms step:647/6250 train_loss:2.7359 aux_loss:0.0481 train_time:342728ms step_avg:538.03ms step:648/6250 train_loss:2.8355 aux_loss:0.0477 train_time:343264ms step_avg:538.03ms step:649/6250 train_loss:2.7705 aux_loss:0.0476 train_time:343826ms step_avg:538.07ms step:650/6250 train_loss:2.7658 aux_loss:0.0477 train_time:344354ms step_avg:538.05ms step:651/6250 train_loss:2.8846 aux_loss:0.0478 train_time:344895ms step_avg:538.06ms step:652/6250 train_loss:2.9203 aux_loss:0.0478 train_time:345423ms step_avg:538.04ms step:653/6250 train_loss:2.8054 aux_loss:0.0479 train_time:345979ms step_avg:538.07ms step:654/6250 train_loss:2.7623 aux_loss:0.0480 train_time:346553ms step_avg:538.13ms step:655/6250 train_loss:2.8549 aux_loss:0.0478 train_time:347094ms step_avg:538.13ms step:656/6250 train_loss:2.8772 aux_loss:0.0474 train_time:347658ms step_avg:538.17ms step:657/6250 train_loss:2.7955 aux_loss:0.0478 train_time:348191ms step_avg:538.16ms step:658/6250 train_loss:2.8667 aux_loss:0.0478 train_time:348725ms step_avg:538.16ms step:659/6250 train_loss:2.8610 aux_loss:0.0478 train_time:349260ms step_avg:538.15ms step:660/6250 train_loss:2.7866 aux_loss:0.0479 train_time:349789ms step_avg:538.14ms step:661/6250 train_loss:2.8999 aux_loss:0.0478 train_time:350341ms step_avg:538.16ms step:662/6250 train_loss:2.7887 aux_loss:0.0474 train_time:350881ms step_avg:538.16ms step:663/6250 train_loss:2.8154 aux_loss:0.0474 train_time:351425ms step_avg:538.17ms step:664/6250 train_loss:2.8119 aux_loss:0.0477 train_time:351955ms step_avg:538.16ms step:665/6250 train_loss:2.9252 aux_loss:0.0475 train_time:352487ms step_avg:538.15ms step:666/6250 train_loss:2.8803 aux_loss:0.0475 train_time:353019ms step_avg:538.14ms step:667/6250 train_loss:2.9306 aux_loss:0.0475 train_time:353549ms step_avg:538.13ms step:668/6250 train_loss:2.8724 aux_loss:0.0478 train_time:354080ms step_avg:538.11ms step:669/6250 train_loss:2.8060 aux_loss:0.0479 train_time:354610ms step_avg:538.10ms step:670/6250 train_loss:2.8556 aux_loss:0.0478 train_time:355140ms step_avg:538.09ms step:671/6250 train_loss:2.7338 aux_loss:0.0479 train_time:355671ms step_avg:538.08ms step:672/6250 train_loss:2.7850 aux_loss:0.0478 train_time:356200ms step_avg:538.07ms step:673/6250 train_loss:2.7891 aux_loss:0.0479 train_time:356730ms step_avg:538.05ms step:674/6250 train_loss:2.8728 aux_loss:0.0477 train_time:357261ms step_avg:538.04ms step:675/6250 train_loss:2.7985 aux_loss:0.0478 train_time:357792ms step_avg:538.03ms step:676/6250 train_loss:2.8589 aux_loss:0.0477 train_time:358323ms step_avg:538.02ms step:677/6250 train_loss:2.8374 aux_loss:0.0477 train_time:358867ms step_avg:538.03ms step:678/6250 train_loss:2.8465 aux_loss:0.0477 train_time:359407ms step_avg:538.03ms step:679/6250 train_loss:2.8604 aux_loss:0.0474 train_time:359949ms step_avg:538.04ms step:680/6250 train_loss:2.9198 aux_loss:0.0474 train_time:360501ms step_avg:538.06ms step:681/6250 train_loss:2.8162 aux_loss:0.0476 train_time:361037ms step_avg:538.06ms step:682/6250 train_loss:2.8240 aux_loss:0.0476 train_time:361580ms step_avg:538.06ms step:683/6250 train_loss:2.9086 aux_loss:0.0476 train_time:362125ms step_avg:538.08ms step:684/6250 train_loss:2.8857 aux_loss:0.0478 train_time:362665ms step_avg:538.08ms step:685/6250 train_loss:2.8103 aux_loss:0.0477 train_time:363195ms step_avg:538.07ms step:686/6250 train_loss:2.8575 aux_loss:0.0477 train_time:363727ms step_avg:538.06ms step:687/6250 train_loss:2.8541 aux_loss:0.0475 train_time:364257ms step_avg:538.05ms step:688/6250 train_loss:2.8955 aux_loss:0.0472 train_time:364790ms step_avg:538.04ms step:689/6250 train_loss:2.9303 aux_loss:0.0475 train_time:365321ms step_avg:538.03ms step:690/6250 train_loss:2.8513 aux_loss:0.0474 train_time:365850ms step_avg:538.02ms step:691/6250 train_loss:2.8206 aux_loss:0.0472 train_time:366381ms step_avg:538.00ms step:692/6250 train_loss:2.9051 aux_loss:0.0476 train_time:366910ms step_avg:537.99ms step:693/6250 train_loss:2.8983 aux_loss:0.0475 train_time:367441ms step_avg:537.98ms step:694/6250 train_loss:2.8233 aux_loss:0.0474 train_time:367990ms step_avg:538.00ms step:695/6250 train_loss:2.7799 aux_loss:0.0478 train_time:368522ms step_avg:537.99ms step:696/6250 train_loss:2.8871 aux_loss:0.0474 train_time:369066ms step_avg:538.00ms step:697/6250 train_loss:2.8915 aux_loss:0.0473 train_time:369629ms step_avg:538.03ms step:698/6250 train_loss:2.9097 aux_loss:0.0471 train_time:370189ms step_avg:538.06ms step:699/6250 train_loss:2.8191 aux_loss:0.0473 train_time:370723ms step_avg:538.06ms step:700/6250 train_loss:2.8825 aux_loss:0.0473 train_time:371275ms step_avg:538.08ms step:701/6250 train_loss:2.9173 aux_loss:0.0472 train_time:371809ms step_avg:538.07ms step:702/6250 train_loss:2.8023 aux_loss:0.0475 train_time:372358ms step_avg:538.09ms step:703/6250 train_loss:2.7867 aux_loss:0.0474 train_time:372895ms step_avg:538.09ms step:704/6250 train_loss:2.8841 aux_loss:0.0471 train_time:373435ms step_avg:538.09ms step:705/6250 train_loss:2.7848 aux_loss:0.0478 train_time:373974ms step_avg:538.09ms step:706/6250 train_loss:2.7403 aux_loss:0.0475 train_time:374514ms step_avg:538.10ms step:707/6250 train_loss:2.7979 aux_loss:0.0475 train_time:375048ms step_avg:538.09ms step:708/6250 train_loss:2.8207 aux_loss:0.0477 train_time:375578ms step_avg:538.08ms step:709/6250 train_loss:2.8023 aux_loss:0.0474 train_time:376108ms step_avg:538.07ms step:710/6250 train_loss:2.7774 aux_loss:0.0476 train_time:376637ms step_avg:538.05ms step:711/6250 train_loss:2.7804 aux_loss:0.0476 train_time:377166ms step_avg:538.04ms step:712/6250 train_loss:2.7906 aux_loss:0.0472 train_time:377694ms step_avg:538.03ms step:713/6250 train_loss:2.7486 aux_loss:0.0474 train_time:378223ms step_avg:538.01ms step:714/6250 train_loss:2.8415 aux_loss:0.0473 train_time:378752ms step_avg:538.00ms step:715/6250 train_loss:2.8416 aux_loss:0.0473 train_time:379282ms step_avg:537.99ms step:716/6250 train_loss:2.7850 aux_loss:0.0477 train_time:379812ms step_avg:537.98ms step:717/6250 train_loss:2.7426 aux_loss:0.0475 train_time:380342ms step_avg:537.97ms step:718/6250 train_loss:2.8694 aux_loss:0.0476 train_time:380871ms step_avg:537.95ms step:719/6250 train_loss:2.8141 aux_loss:0.0473 train_time:381400ms step_avg:537.94ms step:720/6250 train_loss:2.8550 aux_loss:0.0474 train_time:381929ms step_avg:537.93ms step:721/6250 train_loss:2.7303 aux_loss:0.0475 train_time:382459ms step_avg:537.92ms step:722/6250 train_loss:2.7818 aux_loss:0.0473 train_time:382987ms step_avg:537.90ms step:723/6250 train_loss:2.7709 aux_loss:0.0470 train_time:383517ms step_avg:537.89ms step:724/6250 train_loss:2.7903 aux_loss:0.0471 train_time:384047ms step_avg:537.88ms step:725/6250 train_loss:2.8219 aux_loss:0.0473 train_time:384577ms step_avg:537.87ms step:726/6250 train_loss:2.6983 aux_loss:0.0472 train_time:385105ms step_avg:537.86ms step:727/6250 train_loss:2.8512 aux_loss:0.0472 train_time:385636ms step_avg:537.85ms step:728/6250 train_loss:2.6445 aux_loss:0.0475 train_time:386164ms step_avg:537.83ms step:729/6250 train_loss:2.7641 aux_loss:0.0471 train_time:386693ms step_avg:537.82ms step:730/6250 train_loss:2.8326 aux_loss:0.0472 train_time:387222ms step_avg:537.81ms step:731/6250 train_loss:2.8798 aux_loss:0.0472 train_time:387768ms step_avg:537.82ms step:732/6250 train_loss:2.8192 aux_loss:0.0472 train_time:388332ms step_avg:537.86ms step:733/6250 train_loss:2.7988 aux_loss:0.0476 train_time:388871ms step_avg:537.86ms step:734/6250 train_loss:2.7187 aux_loss:0.0476 train_time:389402ms step_avg:537.85ms step:735/6250 train_loss:2.7939 aux_loss:0.0472 train_time:389932ms step_avg:537.84ms step:736/6250 train_loss:2.8225 aux_loss:0.0474 train_time:390462ms step_avg:537.83ms step:737/6250 train_loss:2.8130 aux_loss:0.0477 train_time:390993ms step_avg:537.82ms step:738/6250 train_loss:2.7845 aux_loss:0.0474 train_time:391525ms step_avg:537.81ms step:739/6250 train_loss:2.7586 aux_loss:0.0476 train_time:392056ms step_avg:537.80ms step:740/6250 train_loss:2.8290 aux_loss:0.0475 train_time:392592ms step_avg:537.80ms step:741/6250 train_loss:2.7287 aux_loss:0.0475 train_time:393140ms step_avg:537.81ms step:742/6250 train_loss:2.8066 aux_loss:0.0475 train_time:393676ms step_avg:537.81ms step:743/6250 train_loss:2.7085 aux_loss:0.0475 train_time:394208ms step_avg:537.80ms step:744/6250 train_loss:2.7320 aux_loss:0.0472 train_time:394754ms step_avg:537.81ms step:745/6250 train_loss:2.7870 aux_loss:0.0470 train_time:395295ms step_avg:537.82ms step:746/6250 train_loss:2.7113 aux_loss:0.0472 train_time:395888ms step_avg:537.89ms step:747/6250 train_loss:2.8470 aux_loss:0.0473 train_time:396472ms step_avg:537.95ms step:748/6250 train_loss:2.8704 aux_loss:0.0476 train_time:397054ms step_avg:538.01ms step:749/6250 train_loss:2.6813 aux_loss:0.0475 train_time:397638ms step_avg:538.08ms step:750/6250 train_loss:2.7136 aux_loss:0.0474 train_time:398223ms step_avg:538.14ms step:751/6250 train_loss:2.7256 aux_loss:0.0474 train_time:398786ms step_avg:538.17ms step:752/6250 train_loss:2.7650 aux_loss:0.0471 train_time:399319ms step_avg:538.17ms step:753/6250 train_loss:2.8831 aux_loss:0.0469 train_time:399849ms step_avg:538.16ms step:754/6250 train_loss:2.7522 aux_loss:0.0474 train_time:400398ms step_avg:538.17ms step:755/6250 train_loss:2.7887 aux_loss:0.0473 train_time:400943ms step_avg:538.18ms step:756/6250 train_loss:2.7446 aux_loss:0.0475 train_time:401483ms step_avg:538.18ms step:757/6250 train_loss:2.7801 aux_loss:0.0474 train_time:402014ms step_avg:538.17ms step:758/6250 train_loss:2.6770 aux_loss:0.0473 train_time:402598ms step_avg:538.23ms step:759/6250 train_loss:2.7366 aux_loss:0.0470 train_time:403183ms step_avg:538.30ms step:760/6250 train_loss:2.7752 aux_loss:0.0470 train_time:403766ms step_avg:538.36ms step:761/6250 train_loss:2.7870 aux_loss:0.0472 train_time:404351ms step_avg:538.42ms step:762/6250 train_loss:2.7774 aux_loss:0.0470 train_time:404933ms step_avg:538.48ms step:763/6250 train_loss:2.8830 aux_loss:0.0469 train_time:405522ms step_avg:538.54ms step:764/6250 train_loss:2.6563 aux_loss:0.0473 train_time:406089ms step_avg:538.58ms step:765/6250 train_loss:2.8337 aux_loss:0.0470 train_time:406626ms step_avg:538.58ms step:766/6250 train_loss:2.7991 aux_loss:0.0471 train_time:407163ms step_avg:538.58ms step:767/6250 train_loss:2.7299 aux_loss:0.0470 train_time:407693ms step_avg:538.56ms step:768/6250 train_loss:2.6727 aux_loss:0.0473 train_time:408228ms step_avg:538.56ms step:769/6250 train_loss:2.7483 aux_loss:0.0470 train_time:408758ms step_avg:538.55ms step:770/6250 train_loss:2.6889 aux_loss:0.0471 train_time:409300ms step_avg:538.55ms step:771/6250 train_loss:2.8336 aux_loss:0.0467 train_time:409837ms step_avg:538.55ms step:772/6250 train_loss:2.8068 aux_loss:0.0472 train_time:410384ms step_avg:538.56ms step:773/6250 train_loss:2.8210 aux_loss:0.0471 train_time:410929ms step_avg:538.57ms step:774/6250 train_loss:2.7017 aux_loss:0.0475 train_time:411469ms step_avg:538.57ms step:775/6250 train_loss:2.8162 aux_loss:0.0472 train_time:411999ms step_avg:538.56ms step:776/6250 train_loss:2.6912 aux_loss:0.0474 train_time:412529ms step_avg:538.55ms step:777/6250 train_loss:2.7867 aux_loss:0.0471 train_time:413058ms step_avg:538.54ms step:778/6250 train_loss:2.7580 aux_loss:0.0469 train_time:413587ms step_avg:538.53ms step:779/6250 train_loss:2.7483 aux_loss:0.0471 train_time:414117ms step_avg:538.51ms step:780/6250 train_loss:2.7867 aux_loss:0.0471 train_time:414651ms step_avg:538.51ms step:781/6250 train_loss:2.8199 aux_loss:0.0468 train_time:415180ms step_avg:538.50ms step:782/6250 train_loss:2.7966 aux_loss:0.0472 train_time:415742ms step_avg:538.53ms step:783/6250 train_loss:2.7764 aux_loss:0.0469 train_time:416292ms step_avg:538.54ms step:784/6250 train_loss:2.7100 aux_loss:0.0469 train_time:416821ms step_avg:538.53ms step:785/6250 train_loss:2.7923 aux_loss:0.0471 train_time:417353ms step_avg:538.52ms step:786/6250 train_loss:2.7674 aux_loss:0.0468 train_time:417900ms step_avg:538.53ms step:787/6250 train_loss:2.7859 aux_loss:0.0468 train_time:418430ms step_avg:538.52ms step:788/6250 train_loss:2.7281 aux_loss:0.0469 train_time:418961ms step_avg:538.51ms step:789/6250 train_loss:2.6376 aux_loss:0.0471 train_time:419491ms step_avg:538.50ms step:790/6250 train_loss:2.7263 aux_loss:0.0470 train_time:420018ms step_avg:538.49ms step:791/6250 train_loss:2.7462 aux_loss:0.0471 train_time:420601ms step_avg:538.54ms step:792/6250 train_loss:2.7690 aux_loss:0.0466 train_time:421183ms step_avg:538.60ms step:793/6250 train_loss:2.7295 aux_loss:0.0469 train_time:421768ms step_avg:538.66ms step:794/6250 train_loss:2.7526 aux_loss:0.0468 train_time:422353ms step_avg:538.72ms step:795/6250 train_loss:2.7644 aux_loss:0.0469 train_time:422940ms step_avg:538.78ms step:796/6250 train_loss:2.6950 aux_loss:0.0470 train_time:423523ms step_avg:538.83ms step:797/6250 train_loss:2.7090 aux_loss:0.0469 train_time:424108ms step_avg:538.89ms step:798/6250 train_loss:2.7841 aux_loss:0.0468 train_time:424690ms step_avg:538.95ms step:799/6250 train_loss:2.6784 aux_loss:0.0469 train_time:425274ms step_avg:539.00ms step:800/6250 train_loss:2.7490 aux_loss:0.0465 train_time:425855ms step_avg:539.06ms step:801/6250 train_loss:2.7893 aux_loss:0.0468 train_time:426440ms step_avg:539.11ms step:802/6250 train_loss:2.7076 aux_loss:0.0469 train_time:427024ms step_avg:539.17ms step:803/6250 train_loss:2.7811 aux_loss:0.0469 train_time:427604ms step_avg:539.22ms step:804/6250 train_loss:2.7636 aux_loss:0.0469 train_time:428188ms step_avg:539.28ms step:805/6250 train_loss:2.7369 aux_loss:0.0469 train_time:428770ms step_avg:539.33ms step:806/6250 train_loss:2.6993 aux_loss:0.0469 train_time:429352ms step_avg:539.39ms step:807/6250 train_loss:2.8114 aux_loss:0.0466 train_time:429949ms step_avg:539.46ms step:808/6250 train_loss:2.7438 aux_loss:0.0465 train_time:430488ms step_avg:539.46ms step:809/6250 train_loss:2.7577 aux_loss:0.0468 train_time:431019ms step_avg:539.45ms step:810/6250 train_loss:2.7923 aux_loss:0.0467 train_time:431548ms step_avg:539.43ms step:811/6250 train_loss:2.7358 aux_loss:0.0467 train_time:432079ms step_avg:539.42ms step:812/6250 train_loss:2.8112 aux_loss:0.0466 train_time:432625ms step_avg:539.43ms step:813/6250 train_loss:2.8130 aux_loss:0.0464 train_time:433163ms step_avg:539.43ms step:814/6250 train_loss:2.7104 aux_loss:0.0468 train_time:433708ms step_avg:539.44ms step:815/6250 train_loss:2.7965 aux_loss:0.0469 train_time:434246ms step_avg:539.44ms step:816/6250 train_loss:2.7673 aux_loss:0.0466 train_time:434775ms step_avg:539.42ms step:817/6250 train_loss:2.6136 aux_loss:0.0470 train_time:435304ms step_avg:539.41ms step:818/6250 train_loss:2.7043 aux_loss:0.0466 train_time:435841ms step_avg:539.41ms step:819/6250 train_loss:2.7397 aux_loss:0.0469 train_time:436388ms step_avg:539.42ms step:820/6250 train_loss:2.6124 aux_loss:0.0471 train_time:436925ms step_avg:539.41ms step:821/6250 train_loss:2.8179 aux_loss:0.0469 train_time:437470ms step_avg:539.42ms step:822/6250 train_loss:2.6837 aux_loss:0.0468 train_time:438009ms step_avg:539.42ms step:823/6250 train_loss:2.7099 aux_loss:0.0469 train_time:438548ms step_avg:539.42ms step:824/6250 train_loss:2.6766 aux_loss:0.0469 train_time:439083ms step_avg:539.41ms step:825/6250 train_loss:2.6831 aux_loss:0.0467 train_time:439619ms step_avg:539.41ms step:826/6250 train_loss:2.7234 aux_loss:0.0464 train_time:440157ms step_avg:539.41ms step:827/6250 train_loss:2.7166 aux_loss:0.0467 train_time:440695ms step_avg:539.41ms step:828/6250 train_loss:2.7234 aux_loss:0.0465 train_time:441232ms step_avg:539.40ms step:829/6250 train_loss:2.6573 aux_loss:0.0466 train_time:441772ms step_avg:539.40ms step:830/6250 train_loss:2.6614 aux_loss:0.0467 train_time:442308ms step_avg:539.40ms step:831/6250 train_loss:2.7505 aux_loss:0.0469 train_time:442845ms step_avg:539.40ms step:832/6250 train_loss:2.6653 aux_loss:0.0469 train_time:443389ms step_avg:539.40ms step:833/6250 train_loss:2.6902 aux_loss:0.0469 train_time:443927ms step_avg:539.40ms step:834/6250 train_loss:2.7315 aux_loss:0.0467 train_time:444466ms step_avg:539.40ms step:835/6250 train_loss:2.8245 aux_loss:0.0465 train_time:445001ms step_avg:539.40ms step:836/6250 train_loss:2.7434 aux_loss:0.0466 train_time:445537ms step_avg:539.39ms step:837/6250 train_loss:2.6027 aux_loss:0.0470 train_time:446073ms step_avg:539.39ms step:838/6250 train_loss:2.6954 aux_loss:0.0466 train_time:446608ms step_avg:539.38ms step:839/6250 train_loss:2.6341 aux_loss:0.0468 train_time:447148ms step_avg:539.38ms step:840/6250 train_loss:2.6975 aux_loss:0.0466 train_time:447685ms step_avg:539.38ms step:841/6250 train_loss:2.7858 aux_loss:0.0466 train_time:448225ms step_avg:539.38ms step:842/6250 train_loss:2.6970 aux_loss:0.0469 train_time:448763ms step_avg:539.38ms step:843/6250 train_loss:2.7653 aux_loss:0.0466 train_time:449300ms step_avg:539.38ms step:844/6250 train_loss:2.6499 aux_loss:0.0467 train_time:449838ms step_avg:539.37ms step:845/6250 train_loss:2.6757 aux_loss:0.0467 train_time:450380ms step_avg:539.38ms step:846/6250 train_loss:2.6892 aux_loss:0.0467 train_time:450917ms step_avg:539.37ms step:847/6250 train_loss:2.6698 aux_loss:0.0464 train_time:451456ms step_avg:539.37ms step:848/6250 train_loss:2.7598 aux_loss:0.0463 train_time:451992ms step_avg:539.37ms step:849/6250 train_loss:2.7683 aux_loss:0.0463 train_time:452528ms step_avg:539.37ms step:850/6250 train_loss:2.7755 aux_loss:0.0465 train_time:453066ms step_avg:539.36ms step:851/6250 train_loss:2.7169 aux_loss:0.0466 train_time:453603ms step_avg:539.36ms step:852/6250 train_loss:2.7354 aux_loss:0.0462 train_time:454143ms step_avg:539.36ms step:853/6250 train_loss:2.7211 aux_loss:0.0465 train_time:454681ms step_avg:539.36ms step:854/6250 train_loss:2.8248 aux_loss:0.0463 train_time:455219ms step_avg:539.36ms step:855/6250 train_loss:2.7674 aux_loss:0.0467 train_time:455756ms step_avg:539.36ms step:856/6250 train_loss:2.6514 aux_loss:0.0466 train_time:456290ms step_avg:539.35ms step:857/6250 train_loss:2.6462 aux_loss:0.0464 train_time:456826ms step_avg:539.35ms step:858/6250 train_loss:2.6996 aux_loss:0.0466 train_time:457363ms step_avg:539.34ms step:859/6250 train_loss:2.7508 aux_loss:0.0467 train_time:457898ms step_avg:539.34ms step:860/6250 train_loss:2.7965 aux_loss:0.0465 train_time:458436ms step_avg:539.34ms step:861/6250 train_loss:2.7512 aux_loss:0.0466 train_time:458973ms step_avg:539.33ms step:862/6250 train_loss:2.7420 aux_loss:0.0466 train_time:459515ms step_avg:539.34ms step:863/6250 train_loss:2.6933 aux_loss:0.0467 train_time:460055ms step_avg:539.34ms step:864/6250 train_loss:2.7234 aux_loss:0.0466 train_time:460597ms step_avg:539.34ms step:865/6250 train_loss:2.7558 aux_loss:0.0463 train_time:461147ms step_avg:539.35ms step:866/6250 train_loss:2.7127 aux_loss:0.0462 train_time:461683ms step_avg:539.35ms step:867/6250 train_loss:2.6862 aux_loss:0.0461 train_time:462216ms step_avg:539.34ms step:868/6250 train_loss:2.6722 aux_loss:0.0464 train_time:462753ms step_avg:539.34ms step:869/6250 train_loss:2.7256 aux_loss:0.0465 train_time:463296ms step_avg:539.34ms step:870/6250 train_loss:2.7355 aux_loss:0.0463 train_time:463890ms step_avg:539.41ms step:871/6250 train_loss:2.6721 aux_loss:0.0464 train_time:464485ms step_avg:539.47ms step:872/6250 train_loss:2.7321 aux_loss:0.0463 train_time:465082ms step_avg:539.54ms step:873/6250 train_loss:2.6750 aux_loss:0.0466 train_time:465681ms step_avg:539.61ms step:874/6250 train_loss:2.6499 aux_loss:0.0467 train_time:466281ms step_avg:539.68ms step:875/6250 train_loss:2.6231 aux_loss:0.0467 train_time:466879ms step_avg:539.74ms step:876/6250 train_loss:2.7766 aux_loss:0.0462 train_time:467476ms step_avg:539.81ms step:877/6250 train_loss:2.7380 aux_loss:0.0462 train_time:468072ms step_avg:539.88ms step:878/6250 train_loss:2.6648 aux_loss:0.0462 train_time:468676ms step_avg:539.95ms step:879/6250 train_loss:2.7734 aux_loss:0.0460 train_time:469249ms step_avg:539.99ms step:880/6250 train_loss:2.7455 aux_loss:0.0461 train_time:469797ms step_avg:540.00ms step:881/6250 train_loss:2.6631 aux_loss:0.0464 train_time:470341ms step_avg:540.00ms step:882/6250 train_loss:2.6627 aux_loss:0.0466 train_time:470879ms step_avg:540.00ms step:883/6250 train_loss:2.7192 aux_loss:0.0464 train_time:471418ms step_avg:540.00ms step:884/6250 train_loss:2.7247 aux_loss:0.0460 train_time:471951ms step_avg:539.99ms step:885/6250 train_loss:2.6160 aux_loss:0.0465 train_time:472488ms step_avg:539.99ms step:886/6250 train_loss:2.6200 aux_loss:0.0464 train_time:473024ms step_avg:539.98ms step:887/6250 train_loss:2.5980 aux_loss:0.0464 train_time:473567ms step_avg:539.98ms step:888/6250 train_loss:2.6032 aux_loss:0.0466 train_time:474109ms step_avg:539.99ms step:889/6250 train_loss:2.6715 aux_loss:0.0465 train_time:474660ms step_avg:540.00ms step:890/6250 train_loss:2.7555 aux_loss:0.0462 train_time:475209ms step_avg:540.01ms step:891/6250 train_loss:2.6471 aux_loss:0.0465 train_time:475750ms step_avg:540.01ms step:892/6250 train_loss:2.5790 aux_loss:0.0464 train_time:476288ms step_avg:540.01ms step:893/6250 train_loss:2.7668 aux_loss:0.0459 train_time:476825ms step_avg:540.01ms step:894/6250 train_loss:2.6965 aux_loss:0.0460 train_time:477364ms step_avg:540.00ms step:895/6250 train_loss:2.7677 aux_loss:0.0461 train_time:477899ms step_avg:540.00ms step:896/6250 train_loss:2.7170 aux_loss:0.0463 train_time:478438ms step_avg:540.00ms step:897/6250 train_loss:2.7003 aux_loss:0.0463 train_time:478978ms step_avg:540.00ms step:898/6250 train_loss:2.6339 aux_loss:0.0463 train_time:479525ms step_avg:540.01ms step:899/6250 train_loss:2.6030 aux_loss:0.0462 train_time:480070ms step_avg:540.01ms step:900/6250 train_loss:2.6216 aux_loss:0.0460 train_time:480607ms step_avg:540.01ms step:901/6250 train_loss:2.7087 aux_loss:0.0461 train_time:481143ms step_avg:540.00ms step:902/6250 train_loss:2.7418 aux_loss:0.0460 train_time:481683ms step_avg:540.00ms step:903/6250 train_loss:2.6348 aux_loss:0.0461 train_time:482230ms step_avg:540.01ms step:904/6250 train_loss:2.6393 aux_loss:0.0461 train_time:482770ms step_avg:540.01ms step:905/6250 train_loss:2.6688 aux_loss:0.0460 train_time:483309ms step_avg:540.01ms step:906/6250 train_loss:2.6598 aux_loss:0.0459 train_time:483851ms step_avg:540.01ms step:907/6250 train_loss:2.7957 aux_loss:0.0458 train_time:484399ms step_avg:540.02ms step:908/6250 train_loss:2.6158 aux_loss:0.0460 train_time:484950ms step_avg:540.03ms step:909/6250 train_loss:2.6988 aux_loss:0.0460 train_time:485497ms step_avg:540.04ms step:910/6250 train_loss:2.6838 aux_loss:0.0460 train_time:486045ms step_avg:540.05ms step:911/6250 train_loss:2.6483 aux_loss:0.0460 train_time:486595ms step_avg:540.06ms step:912/6250 train_loss:2.6490 aux_loss:0.0460 train_time:487142ms step_avg:540.07ms step:913/6250 train_loss:2.7261 aux_loss:0.0460 train_time:487690ms step_avg:540.08ms step:914/6250 train_loss:2.7344 aux_loss:0.0457 train_time:488237ms step_avg:540.08ms step:915/6250 train_loss:2.6543 aux_loss:0.0459 train_time:488773ms step_avg:540.08ms step:916/6250 train_loss:2.6289 aux_loss:0.0463 train_time:489310ms step_avg:540.08ms step:917/6250 train_loss:2.6081 aux_loss:0.0462 train_time:489842ms step_avg:540.07ms step:918/6250 train_loss:2.7376 aux_loss:0.0461 train_time:490390ms step_avg:540.08ms step:919/6250 train_loss:2.5982 aux_loss:0.0462 train_time:490926ms step_avg:540.07ms step:920/6250 train_loss:2.6554 aux_loss:0.0462 train_time:491481ms step_avg:540.09ms step:921/6250 train_loss:2.6228 aux_loss:0.0460 train_time:492036ms step_avg:540.11ms step:922/6250 train_loss:2.6835 aux_loss:0.0459 train_time:492575ms step_avg:540.10ms step:923/6250 train_loss:2.6541 aux_loss:0.0460 train_time:493123ms step_avg:540.11ms step:924/6250 train_loss:2.6635 aux_loss:0.0458 train_time:493666ms step_avg:540.12ms step:925/6250 train_loss:2.7240 aux_loss:0.0460 train_time:494202ms step_avg:540.11ms step:926/6250 train_loss:2.5968 aux_loss:0.0464 train_time:494731ms step_avg:540.10ms step:927/6250 train_loss:2.5681 aux_loss:0.0462 train_time:495264ms step_avg:540.09ms step:928/6250 train_loss:2.7739 aux_loss:0.0458 train_time:495803ms step_avg:540.09ms step:929/6250 train_loss:2.6812 aux_loss:0.0459 train_time:496354ms step_avg:540.10ms step:930/6250 train_loss:2.6741 aux_loss:0.0458 train_time:496896ms step_avg:540.10ms step:931/6250 train_loss:2.7255 aux_loss:0.0456 train_time:497450ms step_avg:540.12ms step:932/6250 train_loss:2.7241 aux_loss:0.0455 train_time:498013ms step_avg:540.14ms step:933/6250 train_loss:2.6633 aux_loss:0.0459 train_time:498560ms step_avg:540.15ms step:934/6250 train_loss:2.6149 aux_loss:0.0463 train_time:499104ms step_avg:540.16ms step:935/6250 train_loss:2.7205 aux_loss:0.0462 train_time:499655ms step_avg:540.17ms step:936/6250 train_loss:2.7355 aux_loss:0.0460 train_time:500185ms step_avg:540.16ms step:937/6250 train_loss:2.6015 aux_loss:0.0463 train_time:500725ms step_avg:540.16ms step:938/6250 train_loss:2.6921 aux_loss:0.0459 train_time:501274ms step_avg:540.17ms step:939/6250 train_loss:2.6622 aux_loss:0.0458 train_time:501815ms step_avg:540.17ms step:940/6250 train_loss:2.6735 aux_loss:0.0460 train_time:502350ms step_avg:540.16ms step:941/6250 train_loss:2.6457 aux_loss:0.0459 train_time:502882ms step_avg:540.15ms step:942/6250 train_loss:2.6302 aux_loss:0.0460 train_time:503420ms step_avg:540.15ms step:943/6250 train_loss:2.6244 aux_loss:0.0461 train_time:503985ms step_avg:540.18ms step:944/6250 train_loss:2.6537 aux_loss:0.0461 train_time:504543ms step_avg:540.20ms step:945/6250 train_loss:2.6603 aux_loss:0.0459 train_time:505094ms step_avg:540.21ms step:946/6250 train_loss:2.5759 aux_loss:0.0457 train_time:505626ms step_avg:540.20ms step:947/6250 train_loss:2.5888 aux_loss:0.0460 train_time:506166ms step_avg:540.20ms step:948/6250 train_loss:2.5808 aux_loss:0.0459 train_time:506718ms step_avg:540.21ms step:949/6250 train_loss:2.8169 aux_loss:0.0457 train_time:507265ms step_avg:540.22ms step:950/6250 train_loss:2.6664 aux_loss:0.0458 train_time:507801ms step_avg:540.21ms step:951/6250 train_loss:2.6630 aux_loss:0.0459 train_time:508332ms step_avg:540.20ms step:952/6250 train_loss:2.6280 aux_loss:0.0460 train_time:508862ms step_avg:540.19ms step:953/6250 train_loss:2.6327 aux_loss:0.0461 train_time:509394ms step_avg:540.18ms step:954/6250 train_loss:2.6616 aux_loss:0.0458 train_time:509924ms step_avg:540.17ms step:955/6250 train_loss:2.6243 aux_loss:0.0461 train_time:510454ms step_avg:540.16ms step:956/6250 train_loss:2.8022 aux_loss:0.0455 train_time:510984ms step_avg:540.15ms step:957/6250 train_loss:2.6210 aux_loss:0.0461 train_time:511513ms step_avg:540.14ms step:958/6250 train_loss:2.7344 aux_loss:0.0457 train_time:512041ms step_avg:540.13ms step:959/6250 train_loss:2.6396 aux_loss:0.0458 train_time:512576ms step_avg:540.12ms step:960/6250 train_loss:2.7277 aux_loss:0.0456 train_time:513107ms step_avg:540.11ms step:961/6250 train_loss:2.6358 aux_loss:0.0458 train_time:513639ms step_avg:540.10ms step:962/6250 train_loss:2.6670 aux_loss:0.0458 train_time:514168ms step_avg:540.09ms step:963/6250 train_loss:2.6511 aux_loss:0.0457 train_time:514698ms step_avg:540.08ms step:964/6250 train_loss:2.7184 aux_loss:0.0455 train_time:515227ms step_avg:540.07ms step:965/6250 train_loss:2.5561 aux_loss:0.0459 train_time:515755ms step_avg:540.06ms step:966/6250 train_loss:2.5888 aux_loss:0.0459 train_time:516285ms step_avg:540.05ms step:967/6250 train_loss:2.6229 aux_loss:0.0456 train_time:516816ms step_avg:540.04ms step:968/6250 train_loss:2.6679 aux_loss:0.0456 train_time:517344ms step_avg:540.02ms step:969/6250 train_loss:2.6584 aux_loss:0.0456 train_time:517872ms step_avg:540.01ms step:970/6250 train_loss:2.6677 aux_loss:0.0457 train_time:518401ms step_avg:540.00ms step:971/6250 train_loss:2.6030 aux_loss:0.0458 train_time:518933ms step_avg:539.99ms step:972/6250 train_loss:2.6804 aux_loss:0.0456 train_time:519463ms step_avg:539.98ms step:973/6250 train_loss:2.6823 aux_loss:0.0458 train_time:520009ms step_avg:539.99ms step:974/6250 train_loss:2.5766 aux_loss:0.0460 train_time:520573ms step_avg:540.01ms step:975/6250 train_loss:2.6703 aux_loss:0.0455 train_time:521113ms step_avg:540.01ms step:976/6250 train_loss:2.7474 aux_loss:0.0452 train_time:521650ms step_avg:540.01ms step:977/6250 train_loss:2.6752 aux_loss:0.0455 train_time:522185ms step_avg:540.01ms step:978/6250 train_loss:2.5937 aux_loss:0.0456 train_time:522720ms step_avg:540.00ms step:979/6250 train_loss:2.6245 aux_loss:0.0458 train_time:523255ms step_avg:539.99ms step:980/6250 train_loss:2.6512 aux_loss:0.0457 train_time:523792ms step_avg:539.99ms step:981/6250 train_loss:2.6486 aux_loss:0.0454 train_time:524340ms step_avg:540.00ms step:982/6250 train_loss:2.6735 aux_loss:0.0454 train_time:524887ms step_avg:540.01ms step:983/6250 train_loss:2.5764 aux_loss:0.0455 train_time:525435ms step_avg:540.02ms step:984/6250 train_loss:2.7232 aux_loss:0.0453 train_time:525983ms step_avg:540.02ms step:985/6250 train_loss:2.7176 aux_loss:0.0453 train_time:526532ms step_avg:540.03ms step:986/6250 train_loss:2.6338 aux_loss:0.0456 train_time:527086ms step_avg:540.05ms step:987/6250 train_loss:2.8091 aux_loss:0.0456 train_time:527635ms step_avg:540.06ms step:988/6250 train_loss:2.5887 aux_loss:0.0460 train_time:528185ms step_avg:540.07ms step:989/6250 train_loss:2.7223 aux_loss:0.0458 train_time:528734ms step_avg:540.08ms step:990/6250 train_loss:2.7221 aux_loss:0.0459 train_time:529282ms step_avg:540.08ms step:991/6250 train_loss:2.6382 aux_loss:0.0456 train_time:529828ms step_avg:540.09ms step:992/6250 train_loss:2.6416 aux_loss:0.0456 train_time:530366ms step_avg:540.09ms step:993/6250 train_loss:2.6398 aux_loss:0.0455 train_time:530905ms step_avg:540.09ms step:994/6250 train_loss:2.5825 aux_loss:0.0456 train_time:531447ms step_avg:540.09ms step:995/6250 train_loss:2.5656 aux_loss:0.0457 train_time:531986ms step_avg:540.09ms step:996/6250 train_loss:2.7176 aux_loss:0.0453 train_time:532528ms step_avg:540.09ms step:997/6250 train_loss:2.5375 aux_loss:0.0457 train_time:533098ms step_avg:540.12ms step:998/6250 train_loss:2.7174 aux_loss:0.0453 train_time:533658ms step_avg:540.14ms step:999/6250 train_loss:2.6082 aux_loss:0.0455 train_time:534194ms step_avg:540.14ms step:1000/6250 train_loss:2.5937 aux_loss:0.0455 train_time:534728ms step_avg:540.13ms step:1001/6250 train_loss:2.6115 aux_loss:0.0458 train_time:535268ms step_avg:540.13ms step:1002/6250 train_loss:2.6237 aux_loss:0.0457 train_time:535814ms step_avg:540.14ms step:1003/6250 train_loss:2.6792 aux_loss:0.0457 train_time:536345ms step_avg:540.13ms step:1004/6250 train_loss:2.6514 aux_loss:0.0457 train_time:536891ms step_avg:540.13ms step:1005/6250 train_loss:2.6932 aux_loss:0.0456 train_time:537443ms step_avg:540.14ms step:1006/6250 train_loss:2.6094 aux_loss:0.0458 train_time:537994ms step_avg:540.15ms step:1007/6250 train_loss:2.5959 aux_loss:0.0456 train_time:538542ms step_avg:540.16ms step:1008/6250 train_loss:2.6458 aux_loss:0.0452 train_time:539090ms step_avg:540.17ms step:1009/6250 train_loss:2.5250 aux_loss:0.0455 train_time:539640ms step_avg:540.18ms step:1010/6250 train_loss:2.6621 aux_loss:0.0455 train_time:540189ms step_avg:540.19ms step:1011/6250 train_loss:2.5459 aux_loss:0.0460 train_time:540719ms step_avg:540.18ms step:1012/6250 train_loss:2.6768 aux_loss:0.0457 train_time:541247ms step_avg:540.17ms step:1013/6250 train_loss:2.5540 aux_loss:0.0456 train_time:541775ms step_avg:540.15ms step:1014/6250 train_loss:2.6572 aux_loss:0.0454 train_time:542303ms step_avg:540.14ms step:1015/6250 train_loss:2.5719 aux_loss:0.0455 train_time:542829ms step_avg:540.13ms step:1016/6250 train_loss:2.6041 aux_loss:0.0455 train_time:543358ms step_avg:540.12ms step:1017/6250 train_loss:2.6741 aux_loss:0.0453 train_time:543896ms step_avg:540.12ms step:1018/6250 train_loss:2.7180 aux_loss:0.0454 train_time:544426ms step_avg:540.11ms step:1019/6250 train_loss:2.5430 aux_loss:0.0457 train_time:544958ms step_avg:540.10ms step:1020/6250 train_loss:2.5752 aux_loss:0.0454 train_time:545487ms step_avg:540.09ms step:1021/6250 train_loss:2.6899 aux_loss:0.0450 train_time:546015ms step_avg:540.07ms step:1022/6250 train_loss:2.7360 aux_loss:0.0451 train_time:546542ms step_avg:540.06ms step:1023/6250 train_loss:2.6375 aux_loss:0.0455 train_time:547072ms step_avg:540.05ms step:1024/6250 train_loss:2.5762 aux_loss:0.0455 train_time:547598ms step_avg:540.04ms step:1025/6250 train_loss:2.5948 aux_loss:0.0453 train_time:548129ms step_avg:540.03ms step:1026/6250 train_loss:2.5807 aux_loss:0.0454 train_time:548674ms step_avg:540.03ms step:1027/6250 train_loss:2.5490 aux_loss:0.0456 train_time:549212ms step_avg:540.03ms step:1028/6250 train_loss:2.6489 aux_loss:0.0455 train_time:549751ms step_avg:540.03ms step:1029/6250 train_loss:2.6413 aux_loss:0.0453 train_time:550285ms step_avg:540.02ms step:1030/6250 train_loss:2.6975 aux_loss:0.0452 train_time:550822ms step_avg:540.02ms step:1031/6250 train_loss:2.5440 aux_loss:0.0453 train_time:551365ms step_avg:540.02ms step:1032/6250 train_loss:2.5591 aux_loss:0.0453 train_time:551899ms step_avg:540.02ms step:1033/6250 train_loss:2.5401 aux_loss:0.0454 train_time:552437ms step_avg:540.02ms step:1034/6250 train_loss:2.6228 aux_loss:0.0451 train_time:552973ms step_avg:540.01ms step:1035/6250 train_loss:2.6104 aux_loss:0.0449 train_time:553515ms step_avg:540.01ms step:1036/6250 train_loss:2.6032 aux_loss:0.0450 train_time:554057ms step_avg:540.02ms step:1037/6250 train_loss:2.6052 aux_loss:0.0453 train_time:554587ms step_avg:540.01ms step:1038/6250 train_loss:2.6190 aux_loss:0.0452 train_time:555122ms step_avg:540.00ms step:1039/6250 train_loss:2.6174 aux_loss:0.0454 train_time:555691ms step_avg:540.03ms step:1040/6250 train_loss:2.5713 aux_loss:0.0451 train_time:556254ms step_avg:540.05ms step:1041/6250 train_loss:2.6833 aux_loss:0.0451 train_time:556819ms step_avg:540.08ms step:1042/6250 train_loss:2.6599 aux_loss:0.0453 train_time:557358ms step_avg:540.08ms step:1043/6250 train_loss:2.5791 aux_loss:0.0452 train_time:557886ms step_avg:540.06ms step:1044/6250 train_loss:2.6725 aux_loss:0.0448 train_time:558418ms step_avg:540.06ms step:1045/6250 train_loss:2.5762 aux_loss:0.0450 train_time:558965ms step_avg:540.06ms step:1046/6250 train_loss:2.5923 aux_loss:0.0454 train_time:559500ms step_avg:540.06ms step:1047/6250 train_loss:2.5912 aux_loss:0.0454 train_time:560039ms step_avg:540.06ms step:1048/6250 train_loss:2.5957 aux_loss:0.0451 train_time:560572ms step_avg:540.05ms step:1049/6250 train_loss:2.6162 aux_loss:0.0453 train_time:561104ms step_avg:540.04ms step:1050/6250 train_loss:2.5243 aux_loss:0.0452 train_time:561644ms step_avg:540.04ms step:1051/6250 train_loss:2.6498 aux_loss:0.0453 train_time:562185ms step_avg:540.04ms step:1052/6250 train_loss:2.5914 aux_loss:0.0454 train_time:562715ms step_avg:540.03ms step:1053/6250 train_loss:2.5400 aux_loss:0.0453 train_time:563245ms step_avg:540.02ms step:1054/6250 train_loss:2.6503 aux_loss:0.0451 train_time:563782ms step_avg:540.02ms step:1055/6250 train_loss:2.6285 aux_loss:0.0450 train_time:564314ms step_avg:540.01ms step:1056/6250 train_loss:2.6064 aux_loss:0.0451 train_time:564847ms step_avg:540.01ms step:1057/6250 train_loss:2.6839 aux_loss:0.0456 train_time:565380ms step_avg:540.00ms step:1058/6250 train_loss:2.5990 aux_loss:0.0451 train_time:565912ms step_avg:539.99ms step:1059/6250 train_loss:2.6636 aux_loss:0.0451 train_time:566448ms step_avg:539.99ms step:1060/6250 train_loss:2.6261 aux_loss:0.0451 train_time:566986ms step_avg:539.99ms step:1061/6250 train_loss:2.6562 aux_loss:0.0451 train_time:567537ms step_avg:540.00ms step:1062/6250 train_loss:2.5595 aux_loss:0.0452 train_time:568100ms step_avg:540.02ms step:1063/6250 train_loss:2.6200 aux_loss:0.0450 train_time:568635ms step_avg:540.01ms step:1064/6250 train_loss:2.6136 aux_loss:0.0450 train_time:569166ms step_avg:540.01ms step:1065/6250 train_loss:2.5727 aux_loss:0.0451 train_time:569699ms step_avg:540.00ms step:1066/6250 train_loss:2.6831 aux_loss:0.0450 train_time:570231ms step_avg:539.99ms step:1067/6250 train_loss:2.5859 aux_loss:0.0450 train_time:570763ms step_avg:539.98ms step:1068/6250 train_loss:2.6126 aux_loss:0.0452 train_time:571294ms step_avg:539.98ms step:1069/6250 train_loss:2.7002 aux_loss:0.0452 train_time:571825ms step_avg:539.97ms step:1070/6250 train_loss:2.5990 aux_loss:0.0450 train_time:572357ms step_avg:539.96ms step:1071/6250 train_loss:2.6036 aux_loss:0.0448 train_time:572889ms step_avg:539.95ms step:1072/6250 train_loss:2.6059 aux_loss:0.0447 train_time:573425ms step_avg:539.95ms step:1073/6250 train_loss:2.5260 aux_loss:0.0449 train_time:573966ms step_avg:539.95ms step:1074/6250 train_loss:2.5199 aux_loss:0.0449 train_time:574513ms step_avg:539.96ms step:1075/6250 train_loss:2.5974 aux_loss:0.0450 train_time:575044ms step_avg:539.95ms step:1076/6250 train_loss:2.6373 aux_loss:0.0450 train_time:575576ms step_avg:539.94ms step:1077/6250 train_loss:2.5904 aux_loss:0.0448 train_time:576110ms step_avg:539.93ms step:1078/6250 train_loss:2.5607 aux_loss:0.0447 train_time:576645ms step_avg:539.93ms step:1079/6250 train_loss:2.6201 aux_loss:0.0445 train_time:577175ms step_avg:539.92ms step:1080/6250 train_loss:2.6112 aux_loss:0.0447 train_time:577712ms step_avg:539.92ms step:1081/6250 train_loss:2.5475 aux_loss:0.0450 train_time:578247ms step_avg:539.91ms step:1082/6250 train_loss:2.6554 aux_loss:0.0448 train_time:578810ms step_avg:539.93ms step:1083/6250 train_loss:2.5873 aux_loss:0.0449 train_time:579360ms step_avg:539.94ms step:1084/6250 train_loss:2.5690 aux_loss:0.0448 train_time:579890ms step_avg:539.94ms step:1085/6250 train_loss:2.6325 aux_loss:0.0448 train_time:580421ms step_avg:539.93ms step:1086/6250 train_loss:2.5786 aux_loss:0.0449 train_time:580951ms step_avg:539.92ms step:1087/6250 train_loss:2.5805 aux_loss:0.0450 train_time:581485ms step_avg:539.91ms step:1088/6250 train_loss:2.6698 aux_loss:0.0446 train_time:582020ms step_avg:539.91ms step:1089/6250 train_loss:2.6711 aux_loss:0.0446 train_time:582558ms step_avg:539.91ms step:1090/6250 train_loss:2.5041 aux_loss:0.0448 train_time:583125ms step_avg:539.93ms step:1091/6250 train_loss:2.5998 aux_loss:0.0447 train_time:583681ms step_avg:539.94ms step:1092/6250 train_loss:2.5493 aux_loss:0.0450 train_time:584213ms step_avg:539.94ms step:1093/6250 train_loss:2.6224 aux_loss:0.0449 train_time:584743ms step_avg:539.93ms step:1094/6250 train_loss:2.5798 aux_loss:0.0450 train_time:585273ms step_avg:539.92ms step:1095/6250 train_loss:2.6385 aux_loss:0.0448 train_time:585802ms step_avg:539.91ms step:1096/6250 train_loss:2.6111 aux_loss:0.0446 train_time:586332ms step_avg:539.90ms step:1097/6250 train_loss:2.6475 aux_loss:0.0446 train_time:586895ms step_avg:539.92ms step:1098/6250 train_loss:2.6222 aux_loss:0.0448 train_time:587494ms step_avg:539.98ms step:1099/6250 train_loss:2.6579 aux_loss:0.0449 train_time:588080ms step_avg:540.02ms step:1100/6250 train_loss:2.6100 aux_loss:0.0446 train_time:588663ms step_avg:540.06ms step:1101/6250 train_loss:2.5376 aux_loss:0.0447 train_time:589246ms step_avg:540.10ms step:1102/6250 train_loss:2.5693 aux_loss:0.0447 train_time:589829ms step_avg:540.14ms step:1103/6250 train_loss:2.5787 aux_loss:0.0450 train_time:590414ms step_avg:540.18ms step:1104/6250 train_loss:2.6993 aux_loss:0.0447 train_time:590997ms step_avg:540.22ms step:1105/6250 train_loss:2.6793 aux_loss:0.0446 train_time:591580ms step_avg:540.26ms step:1106/6250 train_loss:2.5756 aux_loss:0.0446 train_time:592164ms step_avg:540.30ms step:1107/6250 train_loss:2.6339 aux_loss:0.0445 train_time:592744ms step_avg:540.33ms step:1108/6250 train_loss:2.5502 aux_loss:0.0447 train_time:593328ms step_avg:540.37ms step:1109/6250 train_loss:2.5956 aux_loss:0.0446 train_time:593911ms step_avg:540.41ms step:1110/6250 train_loss:2.7102 aux_loss:0.0445 train_time:594496ms step_avg:540.45ms step:1111/6250 train_loss:2.5792 aux_loss:0.0444 train_time:595080ms step_avg:540.49ms step:1112/6250 train_loss:2.5863 aux_loss:0.0447 train_time:595665ms step_avg:540.53ms step:1113/6250 train_loss:2.5643 aux_loss:0.0450 train_time:596249ms step_avg:540.57ms step:1114/6250 train_loss:2.4361 aux_loss:0.0452 train_time:596837ms step_avg:540.61ms step:1115/6250 train_loss:2.5988 aux_loss:0.0450 train_time:597402ms step_avg:540.64ms step:1116/6250 train_loss:2.5355 aux_loss:0.0449 train_time:597961ms step_avg:540.65ms step:1117/6250 train_loss:2.5216 aux_loss:0.0447 train_time:598500ms step_avg:540.65ms step:1118/6250 train_loss:2.5276 aux_loss:0.0445 train_time:599052ms step_avg:540.66ms step:1119/6250 train_loss:2.6021 aux_loss:0.0444 train_time:599599ms step_avg:540.67ms step:1120/6250 train_loss:2.5555 aux_loss:0.0447 train_time:600147ms step_avg:540.67ms step:1121/6250 train_loss:2.4229 aux_loss:0.0448 train_time:600693ms step_avg:540.68ms step:1122/6250 train_loss:2.5697 aux_loss:0.0448 train_time:601242ms step_avg:540.69ms step:1123/6250 train_loss:2.6344 aux_loss:0.0445 train_time:601788ms step_avg:540.69ms step:1124/6250 train_loss:2.5438 aux_loss:0.0447 train_time:602338ms step_avg:540.70ms step:1125/6250 train_loss:2.6119 aux_loss:0.0449 train_time:602888ms step_avg:540.71ms step:1126/6250 train_loss:2.5417 aux_loss:0.0448 train_time:603437ms step_avg:540.71ms step:1127/6250 train_loss:2.6642 aux_loss:0.0445 train_time:603988ms step_avg:540.72ms step:1128/6250 train_loss:2.5611 aux_loss:0.0445 train_time:604537ms step_avg:540.73ms step:1129/6250 train_loss:2.5767 aux_loss:0.0447 train_time:605090ms step_avg:540.74ms step:1130/6250 train_loss:2.6461 aux_loss:0.0443 train_time:605639ms step_avg:540.75ms step:1131/6250 train_loss:2.5810 aux_loss:0.0445 train_time:606191ms step_avg:540.76ms step:1132/6250 train_loss:2.5555 aux_loss:0.0444 train_time:606745ms step_avg:540.77ms step:1133/6250 train_loss:2.6501 aux_loss:0.0444 train_time:607296ms step_avg:540.78ms step:1134/6250 train_loss:2.5517 aux_loss:0.0449 train_time:607848ms step_avg:540.79ms step:1135/6250 train_loss:2.6098 aux_loss:0.0449 train_time:608400ms step_avg:540.80ms step:1136/6250 train_loss:2.6217 aux_loss:0.0452 train_time:608947ms step_avg:540.81ms step:1137/6250 train_loss:2.6147 aux_loss:0.0447 train_time:609497ms step_avg:540.81ms step:1138/6250 train_loss:2.5986 aux_loss:0.0445 train_time:610049ms step_avg:540.82ms step:1139/6250 train_loss:2.5814 aux_loss:0.0445 train_time:610594ms step_avg:540.83ms step:1140/6250 train_loss:2.5236 aux_loss:0.0444 train_time:611125ms step_avg:540.82ms step:1141/6250 train_loss:2.6133 aux_loss:0.0445 train_time:611658ms step_avg:540.81ms step:1142/6250 train_loss:2.5912 aux_loss:0.0444 train_time:612201ms step_avg:540.81ms step:1143/6250 train_loss:2.4800 aux_loss:0.0446 train_time:612746ms step_avg:540.82ms step:1144/6250 train_loss:2.6356 aux_loss:0.0445 train_time:613296ms step_avg:540.83ms step:1145/6250 train_loss:2.5676 aux_loss:0.0447 train_time:613847ms step_avg:540.83ms step:1146/6250 train_loss:2.5898 aux_loss:0.0446 train_time:614396ms step_avg:540.84ms step:1147/6250 train_loss:2.5253 aux_loss:0.0444 train_time:614951ms step_avg:540.85ms step:1148/6250 train_loss:2.6251 aux_loss:0.0444 train_time:615501ms step_avg:540.86ms step:1149/6250 train_loss:2.5287 aux_loss:0.0445 train_time:616052ms step_avg:540.87ms step:1150/6250 train_loss:2.5365 aux_loss:0.0443 train_time:616603ms step_avg:540.88ms step:1151/6250 train_loss:2.6142 aux_loss:0.0443 train_time:617154ms step_avg:540.89ms step:1152/6250 train_loss:2.6234 aux_loss:0.0446 train_time:617705ms step_avg:540.90ms step:1153/6250 train_loss:2.5484 aux_loss:0.0446 train_time:618253ms step_avg:540.90ms step:1154/6250 train_loss:2.5246 aux_loss:0.0443 train_time:618803ms step_avg:540.91ms step:1155/6250 train_loss:2.6414 aux_loss:0.0443 train_time:619352ms step_avg:540.92ms step:1156/6250 train_loss:2.5415 aux_loss:0.0444 train_time:619908ms step_avg:540.93ms step:1157/6250 train_loss:2.5962 aux_loss:0.0440 train_time:620462ms step_avg:540.94ms step:1158/6250 train_loss:2.6665 aux_loss:0.0440 train_time:621016ms step_avg:540.95ms step:1159/6250 train_loss:2.5738 aux_loss:0.0443 train_time:621566ms step_avg:540.96ms step:1160/6250 train_loss:2.5172 aux_loss:0.0442 train_time:622120ms step_avg:540.97ms step:1161/6250 train_loss:2.5672 aux_loss:0.0442 train_time:622676ms step_avg:540.99ms step:1162/6250 train_loss:2.4360 aux_loss:0.0443 train_time:623244ms step_avg:541.01ms step:1163/6250 train_loss:2.5715 aux_loss:0.0443 train_time:623801ms step_avg:541.02ms step:1164/6250 train_loss:2.5541 aux_loss:0.0446 train_time:624344ms step_avg:541.03ms step:1165/6250 train_loss:2.5707 aux_loss:0.0446 train_time:624873ms step_avg:541.02ms step:1166/6250 train_loss:2.5615 aux_loss:0.0444 train_time:625426ms step_avg:541.03ms step:1167/6250 train_loss:2.6247 aux_loss:0.0440 train_time:625988ms step_avg:541.04ms step:1168/6250 train_loss:2.5146 aux_loss:0.0444 train_time:626543ms step_avg:541.06ms step:1169/6250 train_loss:2.5550 aux_loss:0.0445 train_time:627106ms step_avg:541.08ms step:1170/6250 train_loss:2.5560 aux_loss:0.0443 train_time:627655ms step_avg:541.08ms step:1171/6250 train_loss:2.5261 aux_loss:0.0438 train_time:628208ms step_avg:541.09ms step:1172/6250 train_loss:2.5823 aux_loss:0.0439 train_time:628758ms step_avg:541.10ms step:1173/6250 train_loss:2.5606 aux_loss:0.0440 train_time:629309ms step_avg:541.11ms step:1174/6250 train_loss:2.5545 aux_loss:0.0441 train_time:629864ms step_avg:541.12ms step:1175/6250 train_loss:2.6653 aux_loss:0.0441 train_time:630424ms step_avg:541.14ms step:1176/6250 train_loss:2.5215 aux_loss:0.0443 train_time:630976ms step_avg:541.15ms step:1177/6250 train_loss:2.5386 aux_loss:0.0444 train_time:631529ms step_avg:541.16ms step:1178/6250 train_loss:2.5511 aux_loss:0.0443 train_time:632082ms step_avg:541.17ms step:1179/6250 train_loss:2.5661 aux_loss:0.0440 train_time:632625ms step_avg:541.17ms step:1180/6250 train_loss:2.5873 aux_loss:0.0439 train_time:633160ms step_avg:541.16ms step:1181/6250 train_loss:2.5774 aux_loss:0.0442 train_time:633696ms step_avg:541.16ms step:1182/6250 train_loss:2.5109 aux_loss:0.0443 train_time:634228ms step_avg:541.15ms step:1183/6250 train_loss:2.6168 aux_loss:0.0442 train_time:634758ms step_avg:541.14ms step:1184/6250 train_loss:2.6308 aux_loss:0.0439 train_time:635293ms step_avg:541.14ms step:1185/6250 train_loss:2.5345 aux_loss:0.0440 train_time:635822ms step_avg:541.13ms step:1186/6250 train_loss:2.4159 aux_loss:0.0441 train_time:636354ms step_avg:541.12ms step:1187/6250 train_loss:2.5290 aux_loss:0.0437 train_time:636884ms step_avg:541.11ms step:1188/6250 train_loss:2.6327 aux_loss:0.0440 train_time:637412ms step_avg:541.10ms step:1189/6250 train_loss:2.5579 aux_loss:0.0442 train_time:637942ms step_avg:541.09ms step:1190/6250 train_loss:2.5922 aux_loss:0.0443 train_time:638474ms step_avg:541.08ms step:1191/6250 train_loss:2.5859 aux_loss:0.0442 train_time:639007ms step_avg:541.07ms step:1192/6250 train_loss:2.4983 aux_loss:0.0441 train_time:639538ms step_avg:541.06ms step:1193/6250 train_loss:2.6547 aux_loss:0.0440 train_time:640067ms step_avg:541.05ms step:1194/6250 train_loss:2.6521 aux_loss:0.0441 train_time:640603ms step_avg:541.05ms step:1195/6250 train_loss:2.4963 aux_loss:0.0443 train_time:641136ms step_avg:541.04ms step:1196/6250 train_loss:2.5964 aux_loss:0.0440 train_time:641670ms step_avg:541.04ms step:1197/6250 train_loss:2.5659 aux_loss:0.0441 train_time:642204ms step_avg:541.03ms step:1198/6250 train_loss:2.5032 aux_loss:0.0443 train_time:642734ms step_avg:541.02ms step:1199/6250 train_loss:2.5944 aux_loss:0.0444 train_time:643263ms step_avg:541.01ms step:1200/6250 train_loss:2.4867 aux_loss:0.0445 train_time:643792ms step_avg:541.00ms step:1201/6250 train_loss:2.6523 aux_loss:0.0443 train_time:644321ms step_avg:540.99ms step:1202/6250 train_loss:2.4902 aux_loss:0.0443 train_time:644851ms step_avg:540.98ms step:1203/6250 train_loss:2.5484 aux_loss:0.0442 train_time:645381ms step_avg:540.97ms step:1204/6250 train_loss:2.5235 aux_loss:0.0444 train_time:645917ms step_avg:540.97ms step:1205/6250 train_loss:2.5407 aux_loss:0.0442 train_time:646450ms step_avg:540.96ms step:1206/6250 train_loss:2.5713 aux_loss:0.0442 train_time:646987ms step_avg:540.96ms step:1207/6250 train_loss:2.5307 aux_loss:0.0441 train_time:647520ms step_avg:540.95ms step:1208/6250 train_loss:2.5370 aux_loss:0.0441 train_time:648055ms step_avg:540.95ms step:1209/6250 train_loss:2.4722 aux_loss:0.0444 train_time:648589ms step_avg:540.94ms step:1210/6250 train_loss:2.5212 aux_loss:0.0441 train_time:649119ms step_avg:540.93ms step:1211/6250 train_loss:2.5355 aux_loss:0.0442 train_time:649652ms step_avg:540.93ms step:1212/6250 train_loss:2.5826 aux_loss:0.0443 train_time:650181ms step_avg:540.92ms step:1213/6250 train_loss:2.4648 aux_loss:0.0443 train_time:650711ms step_avg:540.91ms step:1214/6250 train_loss:2.6086 aux_loss:0.0439 train_time:651241ms step_avg:540.90ms step:1215/6250 train_loss:2.5449 aux_loss:0.0437 train_time:651772ms step_avg:540.89ms step:1216/6250 train_loss:2.5532 aux_loss:0.0437 train_time:652309ms step_avg:540.89ms step:1217/6250 train_loss:2.5932 aux_loss:0.0439 train_time:652840ms step_avg:540.88ms step:1218/6250 train_loss:2.5612 aux_loss:0.0439 train_time:653373ms step_avg:540.87ms step:1219/6250 train_loss:2.5529 aux_loss:0.0443 train_time:653907ms step_avg:540.87ms step:1220/6250 train_loss:2.5670 aux_loss:0.0442 train_time:654438ms step_avg:540.86ms step:1221/6250 train_loss:2.5653 aux_loss:0.0439 train_time:654967ms step_avg:540.85ms step:1222/6250 train_loss:2.5348 aux_loss:0.0440 train_time:655497ms step_avg:540.84ms step:1223/6250 train_loss:2.4600 aux_loss:0.0441 train_time:656027ms step_avg:540.83ms step:1224/6250 train_loss:2.4931 aux_loss:0.0443 train_time:656560ms step_avg:540.82ms step:1225/6250 train_loss:2.5643 aux_loss:0.0441 train_time:657093ms step_avg:540.82ms step:1226/6250 train_loss:2.5675 aux_loss:0.0442 train_time:657621ms step_avg:540.81ms step:1227/6250 train_loss:2.5982 aux_loss:0.0440 train_time:658151ms step_avg:540.80ms step:1228/6250 train_loss:2.4986 aux_loss:0.0439 train_time:658682ms step_avg:540.79ms step:1229/6250 train_loss:2.5490 aux_loss:0.0438 train_time:659212ms step_avg:540.78ms step:1230/6250 train_loss:2.5608 aux_loss:0.0440 train_time:659742ms step_avg:540.77ms step:1231/6250 train_loss:2.5160 aux_loss:0.0438 train_time:660273ms step_avg:540.76ms step:1232/6250 train_loss:2.5241 aux_loss:0.0437 train_time:660802ms step_avg:540.75ms step:1233/6250 train_loss:2.6473 aux_loss:0.0439 train_time:661333ms step_avg:540.75ms step:1234/6250 train_loss:2.5793 aux_loss:0.0437 train_time:661863ms step_avg:540.74ms step:1235/6250 train_loss:2.5574 aux_loss:0.0436 train_time:662394ms step_avg:540.73ms step:1236/6250 train_loss:2.4869 aux_loss:0.0439 train_time:662925ms step_avg:540.72ms step:1237/6250 train_loss:2.5410 aux_loss:0.0439 train_time:663454ms step_avg:540.71ms step:1238/6250 train_loss:2.5241 aux_loss:0.0440 train_time:663982ms step_avg:540.70ms step:1239/6250 train_loss:2.6082 aux_loss:0.0437 train_time:664511ms step_avg:540.69ms step:1240/6250 train_loss:2.5704 aux_loss:0.0438 train_time:665039ms step_avg:540.68ms step:1241/6250 train_loss:2.6129 aux_loss:0.0435 train_time:665569ms step_avg:540.67ms step:1242/6250 train_loss:2.5505 aux_loss:0.0434 train_time:666098ms step_avg:540.66ms step:1243/6250 train_loss:2.5719 aux_loss:0.0435 train_time:666627ms step_avg:540.65ms step:1244/6250 train_loss:2.5846 aux_loss:0.0436 train_time:667158ms step_avg:540.65ms step:1245/6250 train_loss:2.5440 aux_loss:0.0437 train_time:667688ms step_avg:540.64ms step:1246/6250 train_loss:2.5302 aux_loss:0.0437 train_time:668217ms step_avg:540.63ms step:1247/6250 train_loss:2.5105 aux_loss:0.0437 train_time:668747ms step_avg:540.62ms step:1248/6250 train_loss:2.5826 aux_loss:0.0438 train_time:669277ms step_avg:540.61ms step:1249/6250 train_loss:2.4806 aux_loss:0.0438 train_time:669804ms step_avg:540.60ms step:1250/6250 train_loss:2.4756 aux_loss:0.0437 train_time:670335ms step_avg:540.59ms step:1251/6250 train_loss:2.4915 aux_loss:0.0439 train_time:670865ms step_avg:540.58ms step:1252/6250 train_loss:2.5644 aux_loss:0.0437 train_time:671397ms step_avg:540.58ms step:1253/6250 train_loss:2.5808 aux_loss:0.0437 train_time:671926ms step_avg:540.57ms step:1254/6250 train_loss:2.5153 aux_loss:0.0438 train_time:672459ms step_avg:540.56ms step:1255/6250 train_loss:2.4995 aux_loss:0.0437 train_time:672987ms step_avg:540.55ms step:1256/6250 train_loss:2.5317 aux_loss:0.0436 train_time:673520ms step_avg:540.55ms step:1257/6250 train_loss:2.4837 aux_loss:0.0437 train_time:674049ms step_avg:540.54ms step:1258/6250 train_loss:2.5754 aux_loss:0.0438 train_time:674580ms step_avg:540.53ms step:1259/6250 train_loss:2.5224 aux_loss:0.0440 train_time:675114ms step_avg:540.52ms step:1260/6250 train_loss:2.5402 aux_loss:0.0438 train_time:675646ms step_avg:540.52ms step:1261/6250 train_loss:2.5121 aux_loss:0.0437 train_time:676180ms step_avg:540.51ms step:1262/6250 train_loss:2.5607 aux_loss:0.0436 train_time:676710ms step_avg:540.50ms step:1263/6250 train_loss:2.5554 aux_loss:0.0434 train_time:677239ms step_avg:540.49ms step:1264/6250 train_loss:2.4461 aux_loss:0.0433 train_time:677767ms step_avg:540.48ms step:1265/6250 train_loss:2.5641 aux_loss:0.0434 train_time:678300ms step_avg:540.48ms step:1266/6250 train_loss:2.5720 aux_loss:0.0436 train_time:678829ms step_avg:540.47ms step:1267/6250 train_loss:2.5470 aux_loss:0.0435 train_time:679360ms step_avg:540.46ms step:1268/6250 train_loss:2.4487 aux_loss:0.0436 train_time:679900ms step_avg:540.46ms step:1269/6250 train_loss:2.5208 aux_loss:0.0435 train_time:680433ms step_avg:540.46ms step:1270/6250 train_loss:2.6274 aux_loss:0.0435 train_time:680962ms step_avg:540.45ms step:1271/6250 train_loss:2.5228 aux_loss:0.0437 train_time:681493ms step_avg:540.44ms step:1272/6250 train_loss:2.5261 aux_loss:0.0438 train_time:682026ms step_avg:540.43ms step:1273/6250 train_loss:2.4779 aux_loss:0.0436 train_time:682558ms step_avg:540.43ms step:1274/6250 train_loss:2.6141 aux_loss:0.0432 train_time:683091ms step_avg:540.42ms step:1275/6250 train_loss:2.5352 aux_loss:0.0436 train_time:683625ms step_avg:540.42ms step:1276/6250 train_loss:2.5603 aux_loss:0.0432 train_time:684155ms step_avg:540.41ms step:1277/6250 train_loss:2.5604 aux_loss:0.0430 train_time:684688ms step_avg:540.40ms step:1278/6250 train_loss:2.5577 aux_loss:0.0435 train_time:685219ms step_avg:540.39ms step:1279/6250 train_loss:2.6387 aux_loss:0.0436 train_time:685756ms step_avg:540.39ms step:1280/6250 train_loss:2.5012 aux_loss:0.0439 train_time:686287ms step_avg:540.38ms step:1281/6250 train_loss:2.4851 aux_loss:0.0436 train_time:686817ms step_avg:540.38ms step:1282/6250 train_loss:2.5052 aux_loss:0.0436 train_time:687352ms step_avg:540.37ms step:1283/6250 train_loss:2.5981 aux_loss:0.0434 train_time:687882ms step_avg:540.36ms step:1284/6250 train_loss:2.5568 aux_loss:0.0433 train_time:688414ms step_avg:540.36ms step:1285/6250 train_loss:2.4787 aux_loss:0.0433 train_time:688942ms step_avg:540.35ms step:1286/6250 train_loss:2.5525 aux_loss:0.0431 train_time:689470ms step_avg:540.34ms step:1287/6250 train_loss:2.5462 aux_loss:0.0434 train_time:690009ms step_avg:540.34ms step:1288/6250 train_loss:2.4495 aux_loss:0.0436 train_time:690540ms step_avg:540.33ms step:1289/6250 train_loss:2.5043 aux_loss:0.0435 train_time:691069ms step_avg:540.32ms step:1290/6250 train_loss:2.5331 aux_loss:0.0434 train_time:691603ms step_avg:540.31ms step:1291/6250 train_loss:2.4345 aux_loss:0.0434 train_time:692133ms step_avg:540.31ms step:1292/6250 train_loss:2.4476 aux_loss:0.0436 train_time:692667ms step_avg:540.30ms step:1293/6250 train_loss:2.5210 aux_loss:0.0434 train_time:693197ms step_avg:540.29ms step:1294/6250 train_loss:2.4753 aux_loss:0.0432 train_time:693728ms step_avg:540.29ms step:1295/6250 train_loss:2.5898 aux_loss:0.0436 train_time:694259ms step_avg:540.28ms step:1296/6250 train_loss:2.5550 aux_loss:0.0438 train_time:694786ms step_avg:540.27ms step:1297/6250 train_loss:2.5234 aux_loss:0.0435 train_time:695317ms step_avg:540.26ms step:1298/6250 train_loss:2.5771 aux_loss:0.0431 train_time:695850ms step_avg:540.26ms step:1299/6250 train_loss:2.5063 aux_loss:0.0432 train_time:696380ms step_avg:540.25ms step:1300/6250 train_loss:2.4890 aux_loss:0.0432 train_time:696909ms step_avg:540.24ms step:1301/6250 train_loss:2.5022 aux_loss:0.0433 train_time:697439ms step_avg:540.23ms step:1302/6250 train_loss:2.3806 aux_loss:0.0431 train_time:697972ms step_avg:540.23ms step:1303/6250 train_loss:2.5267 aux_loss:0.0431 train_time:698509ms step_avg:540.22ms step:1304/6250 train_loss:2.5006 aux_loss:0.0433 train_time:699038ms step_avg:540.21ms step:1305/6250 train_loss:2.5098 aux_loss:0.0436 train_time:699570ms step_avg:540.21ms step:1306/6250 train_loss:2.5367 aux_loss:0.0434 train_time:700101ms step_avg:540.20ms step:1307/6250 train_loss:2.5534 aux_loss:0.0432 train_time:700633ms step_avg:540.19ms step:1308/6250 train_loss:2.5389 aux_loss:0.0431 train_time:701162ms step_avg:540.19ms step:1309/6250 train_loss:2.4717 aux_loss:0.0433 train_time:701746ms step_avg:540.22ms step:1310/6250 train_loss:2.6102 aux_loss:0.0434 train_time:702303ms step_avg:540.23ms step:1311/6250 train_loss:2.5935 aux_loss:0.0434 train_time:702865ms step_avg:540.25ms step:1312/6250 train_loss:2.5994 aux_loss:0.0431 train_time:703395ms step_avg:540.24ms step:1313/6250 train_loss:2.5348 aux_loss:0.0432 train_time:703925ms step_avg:540.23ms step:1314/6250 train_loss:2.5428 aux_loss:0.0432 train_time:704456ms step_avg:540.23ms step:1315/6250 train_loss:2.6190 aux_loss:0.0429 train_time:704984ms step_avg:540.22ms step:1316/6250 train_loss:2.5314 aux_loss:0.0429 train_time:705512ms step_avg:540.21ms step:1317/6250 train_loss:2.5861 aux_loss:0.0429 train_time:706041ms step_avg:540.20ms step:1318/6250 train_loss:2.5945 aux_loss:0.0431 train_time:706570ms step_avg:540.19ms step:1319/6250 train_loss:2.5688 aux_loss:0.0431 train_time:707100ms step_avg:540.18ms step:1320/6250 train_loss:2.4057 aux_loss:0.0433 train_time:707630ms step_avg:540.18ms step:1321/6250 train_loss:2.6400 aux_loss:0.0433 train_time:708162ms step_avg:540.17ms step:1322/6250 train_loss:2.5065 aux_loss:0.0432 train_time:708696ms step_avg:540.16ms step:1323/6250 train_loss:2.4979 aux_loss:0.0432 train_time:709228ms step_avg:540.16ms step:1324/6250 train_loss:2.5666 aux_loss:0.0431 train_time:709763ms step_avg:540.15ms step:1325/6250 train_loss:2.5844 aux_loss:0.0430 train_time:710300ms step_avg:540.15ms step:1326/6250 train_loss:2.6645 aux_loss:0.0426 train_time:710836ms step_avg:540.15ms step:1327/6250 train_loss:2.4724 aux_loss:0.0429 train_time:711370ms step_avg:540.14ms step:1328/6250 train_loss:2.5630 aux_loss:0.0429 train_time:711905ms step_avg:540.14ms step:1329/6250 train_loss:2.6099 aux_loss:0.0430 train_time:712439ms step_avg:540.14ms step:1330/6250 train_loss:2.5026 aux_loss:0.0432 train_time:712975ms step_avg:540.13ms step:1331/6250 train_loss:2.5575 aux_loss:0.0428 train_time:713507ms step_avg:540.13ms step:1332/6250 train_loss:2.5760 aux_loss:0.0429 train_time:714048ms step_avg:540.13ms step:1333/6250 train_loss:2.4299 aux_loss:0.0433 train_time:714576ms step_avg:540.12ms step:1334/6250 train_loss:2.5831 aux_loss:0.0432 train_time:715110ms step_avg:540.11ms step:1335/6250 train_loss:2.4984 aux_loss:0.0429 train_time:715645ms step_avg:540.11ms step:1336/6250 train_loss:2.4971 aux_loss:0.0430 train_time:716180ms step_avg:540.11ms step:1337/6250 train_loss:2.4946 aux_loss:0.0429 train_time:716717ms step_avg:540.10ms step:1338/6250 train_loss:2.4844 aux_loss:0.0428 train_time:717253ms step_avg:540.10ms step:1339/6250 train_loss:2.6078 aux_loss:0.0430 train_time:717789ms step_avg:540.10ms step:1340/6250 train_loss:2.5680 aux_loss:0.0428 train_time:718327ms step_avg:540.10ms step:1341/6250 train_loss:2.5352 aux_loss:0.0428 train_time:718860ms step_avg:540.09ms step:1342/6250 train_loss:2.5447 aux_loss:0.0428 train_time:719395ms step_avg:540.09ms step:1343/6250 train_loss:2.5821 aux_loss:0.0431 train_time:719953ms step_avg:540.10ms step:1344/6250 train_loss:2.4564 aux_loss:0.0430 train_time:720485ms step_avg:540.09ms step:1345/6250 train_loss:2.5608 aux_loss:0.0427 train_time:721017ms step_avg:540.09ms step:1346/6250 train_loss:2.4699 aux_loss:0.0426 train_time:721547ms step_avg:540.08ms step:1347/6250 train_loss:2.4578 aux_loss:0.0428 train_time:722077ms step_avg:540.07ms step:1348/6250 train_loss:2.5064 aux_loss:0.0430 train_time:722610ms step_avg:540.07ms step:1349/6250 train_loss:2.5789 aux_loss:0.0429 train_time:723139ms step_avg:540.06ms step:1350/6250 train_loss:2.4565 aux_loss:0.0431 train_time:723669ms step_avg:540.05ms step:1351/6250 train_loss:2.5680 aux_loss:0.0431 train_time:724199ms step_avg:540.04ms step:1352/6250 train_loss:2.5616 aux_loss:0.0430 train_time:724728ms step_avg:540.04ms step:1353/6250 train_loss:2.5606 aux_loss:0.0427 train_time:725258ms step_avg:540.03ms step:1354/6250 train_loss:2.4743 aux_loss:0.0426 train_time:725787ms step_avg:540.02ms step:1355/6250 train_loss:2.5528 aux_loss:0.0426 train_time:726319ms step_avg:540.01ms step:1356/6250 train_loss:2.5134 aux_loss:0.0431 train_time:726855ms step_avg:540.01ms step:1357/6250 train_loss:2.5249 aux_loss:0.0431 train_time:727385ms step_avg:540.00ms step:1358/6250 train_loss:2.4234 aux_loss:0.0432 train_time:727914ms step_avg:540.00ms step:1359/6250 train_loss:2.4654 aux_loss:0.0426 train_time:728444ms step_avg:539.99ms step:1360/6250 train_loss:2.4516 aux_loss:0.0426 train_time:728973ms step_avg:539.98ms step:1361/6250 train_loss:2.5406 aux_loss:0.0429 train_time:729507ms step_avg:539.98ms step:1362/6250 train_loss:2.4280 aux_loss:0.0431 train_time:730037ms step_avg:539.97ms step:1363/6250 train_loss:2.5466 aux_loss:0.0429 train_time:730573ms step_avg:539.96ms step:1364/6250 train_loss:2.6448 aux_loss:0.0426 train_time:731108ms step_avg:539.96ms step:1365/6250 train_loss:2.5016 aux_loss:0.0426 train_time:731638ms step_avg:539.95ms step:1366/6250 train_loss:2.5630 aux_loss:0.0427 train_time:732169ms step_avg:539.95ms step:1367/6250 train_loss:2.5461 aux_loss:0.0426 train_time:732701ms step_avg:539.94ms step:1368/6250 train_loss:2.5469 aux_loss:0.0425 train_time:733234ms step_avg:539.94ms step:1369/6250 train_loss:2.4619 aux_loss:0.0427 train_time:733766ms step_avg:539.93ms step:1370/6250 train_loss:2.4539 aux_loss:0.0428 train_time:734300ms step_avg:539.93ms step:1371/6250 train_loss:2.5312 aux_loss:0.0425 train_time:734838ms step_avg:539.92ms step:1372/6250 train_loss:2.5198 aux_loss:0.0430 train_time:735370ms step_avg:539.92ms step:1373/6250 train_loss:2.5152 aux_loss:0.0427 train_time:735901ms step_avg:539.91ms step:1374/6250 train_loss:2.4853 aux_loss:0.0427 train_time:736431ms step_avg:539.91ms step:1375/6250 train_loss:2.4808 aux_loss:0.0426 train_time:736962ms step_avg:539.90ms step:1376/6250 train_loss:2.5391 aux_loss:0.0424 train_time:737492ms step_avg:539.89ms step:1377/6250 train_loss:2.5177 aux_loss:0.0425 train_time:738022ms step_avg:539.88ms step:1378/6250 train_loss:2.4756 aux_loss:0.0425 train_time:738551ms step_avg:539.88ms step:1379/6250 train_loss:2.4754 aux_loss:0.0426 train_time:739081ms step_avg:539.87ms step:1380/6250 train_loss:2.3682 aux_loss:0.0428 train_time:739611ms step_avg:539.86ms step:1381/6250 train_loss:2.4619 aux_loss:0.0428 train_time:740141ms step_avg:539.85ms step:1382/6250 train_loss:2.5110 aux_loss:0.0428 train_time:740672ms step_avg:539.85ms step:1383/6250 train_loss:2.5243 aux_loss:0.0428 train_time:741201ms step_avg:539.84ms step:1384/6250 train_loss:2.5709 aux_loss:0.0423 train_time:741730ms step_avg:539.83ms step:1385/6250 train_loss:2.5008 aux_loss:0.0423 train_time:742261ms step_avg:539.83ms step:1386/6250 train_loss:2.5370 aux_loss:0.0425 train_time:742787ms step_avg:539.82ms step:1387/6250 train_loss:2.4524 aux_loss:0.0428 train_time:743317ms step_avg:539.81ms step:1388/6250 train_loss:2.5476 aux_loss:0.0426 train_time:743847ms step_avg:539.80ms step:1389/6250 train_loss:2.4724 aux_loss:0.0424 train_time:744378ms step_avg:539.80ms step:1390/6250 train_loss:2.4913 aux_loss:0.0424 train_time:744907ms step_avg:539.79ms step:1391/6250 train_loss:2.5391 aux_loss:0.0424 train_time:745437ms step_avg:539.78ms step:1392/6250 train_loss:2.5118 aux_loss:0.0424 train_time:745968ms step_avg:539.77ms step:1393/6250 train_loss:2.5178 aux_loss:0.0426 train_time:746498ms step_avg:539.77ms step:1394/6250 train_loss:2.4895 aux_loss:0.0426 train_time:747028ms step_avg:539.76ms step:1395/6250 train_loss:2.5208 aux_loss:0.0428 train_time:747560ms step_avg:539.75ms step:1396/6250 train_loss:2.5452 aux_loss:0.0427 train_time:748089ms step_avg:539.75ms step:1397/6250 train_loss:2.3849 aux_loss:0.0428 train_time:748617ms step_avg:539.74ms step:1398/6250 train_loss:2.5154 aux_loss:0.0426 train_time:749148ms step_avg:539.73ms step:1399/6250 train_loss:2.5218 aux_loss:0.0426 train_time:749679ms step_avg:539.73ms step:1400/6250 train_loss:2.4760 aux_loss:0.0427 train_time:750216ms step_avg:539.72ms step:1401/6250 train_loss:2.4575 aux_loss:0.0429 train_time:750757ms step_avg:539.72ms step:1402/6250 train_loss:2.5693 aux_loss:0.0425 train_time:751303ms step_avg:539.73ms step:1403/6250 train_loss:2.4931 aux_loss:0.0425 train_time:751832ms step_avg:539.72ms step:1404/6250 train_loss:2.4847 aux_loss:0.0426 train_time:752374ms step_avg:539.72ms step:1405/6250 train_loss:2.4800 aux_loss:0.0426 train_time:752909ms step_avg:539.72ms step:1406/6250 train_loss:2.4474 aux_loss:0.0426 train_time:753455ms step_avg:539.72ms step:1407/6250 train_loss:2.4701 aux_loss:0.0426 train_time:754015ms step_avg:539.74ms step:1408/6250 train_loss:2.4983 aux_loss:0.0427 train_time:754551ms step_avg:539.74ms step:1409/6250 train_loss:2.5690 aux_loss:0.0424 train_time:755079ms step_avg:539.73ms step:1410/6250 train_loss:2.4982 aux_loss:0.0424 train_time:755610ms step_avg:539.72ms step:1411/6250 train_loss:2.5259 aux_loss:0.0422 train_time:756138ms step_avg:539.71ms step:1412/6250 train_loss:2.4031 aux_loss:0.0424 train_time:756675ms step_avg:539.71ms step:1413/6250 train_loss:2.4860 aux_loss:0.0424 train_time:757218ms step_avg:539.71ms step:1414/6250 train_loss:2.4942 aux_loss:0.0422 train_time:757776ms step_avg:539.73ms step:1415/6250 train_loss:2.4666 aux_loss:0.0421 train_time:758318ms step_avg:539.73ms step:1416/6250 train_loss:2.6011 aux_loss:0.0420 train_time:758857ms step_avg:539.73ms step:1417/6250 train_loss:2.5419 aux_loss:0.0422 train_time:759396ms step_avg:539.73ms step:1418/6250 train_loss:2.5375 aux_loss:0.0424 train_time:759943ms step_avg:539.73ms step:1419/6250 train_loss:2.4696 aux_loss:0.0423 train_time:760494ms step_avg:539.74ms step:1420/6250 train_loss:2.5355 aux_loss:0.0422 train_time:761099ms step_avg:539.79ms step:1421/6250 train_loss:2.5876 aux_loss:0.0421 train_time:761674ms step_avg:539.81ms step:1422/6250 train_loss:2.5184 aux_loss:0.0423 train_time:762214ms step_avg:539.81ms step:1423/6250 train_loss:2.4921 aux_loss:0.0425 train_time:762754ms step_avg:539.81ms step:1424/6250 train_loss:2.4590 aux_loss:0.0425 train_time:763294ms step_avg:539.81ms step:1425/6250 train_loss:2.5705 aux_loss:0.0424 train_time:763827ms step_avg:539.81ms step:1426/6250 train_loss:2.4106 aux_loss:0.0425 train_time:764366ms step_avg:539.81ms step:1427/6250 train_loss:2.4494 aux_loss:0.0423 train_time:764898ms step_avg:539.80ms step:1428/6250 train_loss:2.5352 aux_loss:0.0421 train_time:765437ms step_avg:539.80ms step:1429/6250 train_loss:2.4584 aux_loss:0.0422 train_time:765973ms step_avg:539.80ms step:1430/6250 train_loss:2.5518 aux_loss:0.0420 train_time:766517ms step_avg:539.80ms step:1431/6250 train_loss:2.5180 aux_loss:0.0424 train_time:767051ms step_avg:539.80ms step:1432/6250 train_loss:2.4667 aux_loss:0.0426 train_time:767594ms step_avg:539.80ms step:1433/6250 train_loss:2.4737 aux_loss:0.0426 train_time:768131ms step_avg:539.80ms step:1434/6250 train_loss:2.4635 aux_loss:0.0422 train_time:768667ms step_avg:539.79ms step:1435/6250 train_loss:2.4703 aux_loss:0.0422 train_time:769201ms step_avg:539.79ms step:1436/6250 train_loss:2.4510 aux_loss:0.0424 train_time:769751ms step_avg:539.80ms step:1437/6250 train_loss:2.4953 aux_loss:0.0424 train_time:770301ms step_avg:539.80ms step:1438/6250 train_loss:2.4557 aux_loss:0.0424 train_time:770842ms step_avg:539.81ms step:1439/6250 train_loss:2.4045 aux_loss:0.0422 train_time:771376ms step_avg:539.80ms step:1440/6250 train_loss:2.4834 aux_loss:0.0420 train_time:771922ms step_avg:539.81ms step:1441/6250 train_loss:2.5695 aux_loss:0.0422 train_time:772463ms step_avg:539.81ms step:1442/6250 train_loss:2.5662 aux_loss:0.0422 train_time:773015ms step_avg:539.82ms step:1443/6250 train_loss:2.4220 aux_loss:0.0422 train_time:773555ms step_avg:539.81ms step:1444/6250 train_loss:2.4474 aux_loss:0.0421 train_time:774090ms step_avg:539.81ms step:1445/6250 train_loss:2.5484 aux_loss:0.0421 train_time:774642ms step_avg:539.82ms step:1446/6250 train_loss:2.5287 aux_loss:0.0422 train_time:775186ms step_avg:539.82ms step:1447/6250 train_loss:2.4245 aux_loss:0.0424 train_time:775729ms step_avg:539.83ms step:1448/6250 train_loss:2.5716 aux_loss:0.0421 train_time:776264ms step_avg:539.82ms step:1449/6250 train_loss:2.5472 aux_loss:0.0420 train_time:776803ms step_avg:539.82ms step:1450/6250 train_loss:2.5797 aux_loss:0.0423 train_time:777340ms step_avg:539.82ms step:1451/6250 train_loss:2.5000 aux_loss:0.0424 train_time:777874ms step_avg:539.82ms step:1452/6250 train_loss:2.4873 aux_loss:0.0423 train_time:778409ms step_avg:539.81ms step:1453/6250 train_loss:2.4817 aux_loss:0.0419 train_time:778949ms step_avg:539.81ms step:1454/6250 train_loss:2.4814 aux_loss:0.0421 train_time:779484ms step_avg:539.81ms step:1455/6250 train_loss:2.5377 aux_loss:0.0421 train_time:780028ms step_avg:539.81ms step:1456/6250 train_loss:2.5031 aux_loss:0.0426 train_time:780563ms step_avg:539.81ms step:1457/6250 train_loss:2.4992 aux_loss:0.0425 train_time:781102ms step_avg:539.81ms step:1458/6250 train_loss:2.6047 aux_loss:0.0421 train_time:781651ms step_avg:539.81ms step:1459/6250 train_loss:2.4543 aux_loss:0.0421 train_time:782189ms step_avg:539.81ms step:1460/6250 train_loss:2.4939 aux_loss:0.0422 train_time:782732ms step_avg:539.82ms step:1461/6250 train_loss:2.5146 aux_loss:0.0422 train_time:783274ms step_avg:539.82ms step:1462/6250 train_loss:2.5113 aux_loss:0.0421 train_time:783837ms step_avg:539.83ms step:1463/6250 train_loss:2.4615 aux_loss:0.0420 train_time:784408ms step_avg:539.85ms step:1464/6250 train_loss:2.5365 aux_loss:0.0420 train_time:784952ms step_avg:539.86ms step:1465/6250 train_loss:2.4695 aux_loss:0.0423 train_time:785516ms step_avg:539.87ms step:1466/6250 train_loss:2.4374 aux_loss:0.0421 train_time:786056ms step_avg:539.87ms step:1467/6250 train_loss:2.4406 aux_loss:0.0420 train_time:786597ms step_avg:539.87ms step:1468/6250 train_loss:2.4576 aux_loss:0.0421 train_time:787151ms step_avg:539.88ms step:1469/6250 train_loss:2.5239 aux_loss:0.0422 train_time:787692ms step_avg:539.89ms step:1470/6250 train_loss:2.4691 aux_loss:0.0422 train_time:788227ms step_avg:539.88ms step:1471/6250 train_loss:2.4727 aux_loss:0.0419 train_time:788762ms step_avg:539.88ms step:1472/6250 train_loss:2.5157 aux_loss:0.0419 train_time:789316ms step_avg:539.89ms step:1473/6250 train_loss:2.4637 aux_loss:0.0421 train_time:789856ms step_avg:539.89ms step:1474/6250 train_loss:2.4171 aux_loss:0.0421 train_time:790414ms step_avg:539.90ms step:1475/6250 train_loss:2.4016 aux_loss:0.0421 train_time:790960ms step_avg:539.90ms step:1476/6250 train_loss:2.4852 aux_loss:0.0421 train_time:791502ms step_avg:539.91ms step:1477/6250 train_loss:2.4669 aux_loss:0.0421 train_time:792034ms step_avg:539.90ms step:1478/6250 train_loss:2.4777 aux_loss:0.0417 train_time:792565ms step_avg:539.89ms step:1479/6250 train_loss:2.4236 aux_loss:0.0420 train_time:793096ms step_avg:539.89ms step:1480/6250 train_loss:2.3674 aux_loss:0.0418 train_time:793640ms step_avg:539.89ms step:1481/6250 train_loss:2.5940 aux_loss:0.0416 train_time:794169ms step_avg:539.88ms step:1482/6250 train_loss:2.4023 aux_loss:0.0416 train_time:794696ms step_avg:539.88ms step:1483/6250 train_loss:2.5813 aux_loss:0.0419 train_time:795239ms step_avg:539.88ms step:1484/6250 train_loss:2.3936 aux_loss:0.0421 train_time:795780ms step_avg:539.88ms step:1485/6250 train_loss:2.5129 aux_loss:0.0418 train_time:796330ms step_avg:539.88ms step:1486/6250 train_loss:2.5485 aux_loss:0.0416 train_time:796892ms step_avg:539.90ms step:1487/6250 train_loss:2.4153 aux_loss:0.0420 train_time:797439ms step_avg:539.90ms step:1488/6250 train_loss:2.4675 aux_loss:0.0419 train_time:797974ms step_avg:539.90ms step:1489/6250 train_loss:2.4408 aux_loss:0.0416 train_time:798533ms step_avg:539.91ms step:1490/6250 train_loss:2.4880 aux_loss:0.0414 train_time:799083ms step_avg:539.92ms step:1491/6250 train_loss:2.4118 aux_loss:0.0418 train_time:799637ms step_avg:539.93ms step:1492/6250 train_loss:2.5147 aux_loss:0.0421 train_time:800192ms step_avg:539.94ms step:1493/6250 train_loss:2.5107 aux_loss:0.0419 train_time:800738ms step_avg:539.94ms step:1494/6250 train_loss:2.4103 aux_loss:0.0415 train_time:801286ms step_avg:539.95ms step:1495/6250 train_loss:2.4989 aux_loss:0.0416 train_time:801844ms step_avg:539.96ms step:1496/6250 train_loss:2.4345 aux_loss:0.0420 train_time:802388ms step_avg:539.97ms step:1497/6250 train_loss:2.5157 aux_loss:0.0421 train_time:802951ms step_avg:539.98ms step:1498/6250 train_loss:2.4411 aux_loss:0.0420 train_time:803513ms step_avg:540.00ms step:1499/6250 train_loss:2.4525 aux_loss:0.0421 train_time:804074ms step_avg:540.01ms step:1500/6250 train_loss:2.5507 aux_loss:0.0421 train_time:804638ms step_avg:540.03ms step:1501/6250 train_loss:2.4328 aux_loss:0.0419 train_time:805193ms step_avg:540.04ms step:1502/6250 train_loss:2.4201 aux_loss:0.0418 train_time:805746ms step_avg:540.04ms step:1503/6250 train_loss:2.4608 aux_loss:0.0417 train_time:806300ms step_avg:540.05ms step:1504/6250 train_loss:2.4501 aux_loss:0.0416 train_time:806838ms step_avg:540.05ms step:1505/6250 train_loss:2.4847 aux_loss:0.0416 train_time:807374ms step_avg:540.05ms step:1506/6250 train_loss:2.5535 aux_loss:0.0415 train_time:807911ms step_avg:540.05ms step:1507/6250 train_loss:2.5242 aux_loss:0.0417 train_time:808448ms step_avg:540.05ms step:1508/6250 train_loss:2.4727 aux_loss:0.0418 train_time:808984ms step_avg:540.04ms step:1509/6250 train_loss:2.5201 aux_loss:0.0417 train_time:809522ms step_avg:540.04ms step:1510/6250 train_loss:2.5508 aux_loss:0.0416 train_time:810060ms step_avg:540.04ms step:1511/6250 train_loss:2.5193 aux_loss:0.0415 train_time:810598ms step_avg:540.04ms step:1512/6250 train_loss:2.5436 aux_loss:0.0417 train_time:811143ms step_avg:540.04ms step:1513/6250 train_loss:2.4689 aux_loss:0.0419 train_time:811692ms step_avg:540.05ms step:1514/6250 train_loss:2.5487 aux_loss:0.0422 train_time:812253ms step_avg:540.06ms step:1515/6250 train_loss:2.4118 aux_loss:0.0418 train_time:812812ms step_avg:540.07ms step:1516/6250 train_loss:2.5606 aux_loss:0.0417 train_time:813370ms step_avg:540.09ms step:1517/6250 train_loss:2.5039 aux_loss:0.0415 train_time:813935ms step_avg:540.10ms step:1518/6250 train_loss:2.5008 aux_loss:0.0414 train_time:814486ms step_avg:540.11ms step:1519/6250 train_loss:2.3861 aux_loss:0.0418 train_time:815030ms step_avg:540.11ms step:1520/6250 train_loss:2.4618 aux_loss:0.0416 train_time:815586ms step_avg:540.12ms step:1521/6250 train_loss:2.4113 aux_loss:0.0418 train_time:816135ms step_avg:540.13ms step:1522/6250 train_loss:2.4276 aux_loss:0.0416 train_time:816679ms step_avg:540.13ms step:1523/6250 train_loss:2.4019 aux_loss:0.0413 train_time:817215ms step_avg:540.13ms step:1524/6250 train_loss:2.3854 aux_loss:0.0413 train_time:817758ms step_avg:540.13ms step:1525/6250 train_loss:2.5330 aux_loss:0.0414 train_time:818313ms step_avg:540.14ms step:1526/6250 train_loss:2.4852 aux_loss:0.0415 train_time:818953ms step_avg:540.21ms step:1527/6250 train_loss:2.3993 aux_loss:0.0414 train_time:819491ms step_avg:540.20ms step:1528/6250 train_loss:2.4354 aux_loss:0.0414 train_time:820026ms step_avg:540.20ms step:1529/6250 train_loss:2.3381 aux_loss:0.0413 train_time:820564ms step_avg:540.20ms step:1530/6250 train_loss:2.4467 aux_loss:0.0416 train_time:821099ms step_avg:540.20ms step:1531/6250 train_loss:2.4585 aux_loss:0.0417 train_time:821651ms step_avg:540.20ms step:1532/6250 train_loss:2.4369 aux_loss:0.0413 train_time:822188ms step_avg:540.20ms step:1533/6250 train_loss:2.4429 aux_loss:0.0413 train_time:822732ms step_avg:540.20ms step:1534/6250 train_loss:2.4858 aux_loss:0.0417 train_time:823286ms step_avg:540.21ms step:1535/6250 train_loss:2.4255 aux_loss:0.0418 train_time:823829ms step_avg:540.22ms step:1536/6250 train_loss:2.4271 aux_loss:0.0414 train_time:824370ms step_avg:540.22ms step:1537/6250 train_loss:2.4984 aux_loss:0.0415 train_time:824907ms step_avg:540.21ms step:1538/6250 train_loss:2.3711 aux_loss:0.0415 train_time:825443ms step_avg:540.21ms step:1539/6250 train_loss:2.5781 aux_loss:0.0412 train_time:825981ms step_avg:540.21ms step:1540/6250 train_loss:2.4500 aux_loss:0.0415 train_time:826518ms step_avg:540.21ms step:1541/6250 train_loss:2.4927 aux_loss:0.0414 train_time:827054ms step_avg:540.20ms step:1542/6250 train_loss:2.5076 aux_loss:0.0412 train_time:827590ms step_avg:540.20ms step:1543/6250 train_loss:2.4524 aux_loss:0.0413 train_time:828126ms step_avg:540.20ms step:1544/6250 train_loss:2.4255 aux_loss:0.0416 train_time:828662ms step_avg:540.20ms step:1545/6250 train_loss:2.4825 aux_loss:0.0415 train_time:829198ms step_avg:540.19ms step:1546/6250 train_loss:2.4813 aux_loss:0.0413 train_time:829734ms step_avg:540.19ms step:1547/6250 train_loss:2.4409 aux_loss:0.0414 train_time:830271ms step_avg:540.19ms step:1548/6250 train_loss:2.4691 aux_loss:0.0417 train_time:830807ms step_avg:540.19ms step:1549/6250 train_loss:2.5361 aux_loss:0.0414 train_time:831343ms step_avg:540.18ms step:1550/6250 train_loss:2.4823 aux_loss:0.0412 train_time:831881ms step_avg:540.18ms step:1551/6250 train_loss:2.4365 aux_loss:0.0414 train_time:832416ms step_avg:540.18ms step:1552/6250 train_loss:2.5931 aux_loss:0.0414 train_time:832952ms step_avg:540.18ms step:1553/6250 train_loss:2.4780 aux_loss:0.0415 train_time:833489ms step_avg:540.17ms step:1554/6250 train_loss:2.4305 aux_loss:0.0413 train_time:834026ms step_avg:540.17ms step:1555/6250 train_loss:2.4906 aux_loss:0.0412 train_time:834582ms step_avg:540.18ms step:1556/6250 train_loss:2.5597 aux_loss:0.0413 train_time:835130ms step_avg:540.19ms step:1557/6250 train_loss:2.4373 aux_loss:0.0415 train_time:835676ms step_avg:540.19ms step:1558/6250 train_loss:2.4674 aux_loss:0.0414 train_time:836218ms step_avg:540.19ms step:1559/6250 train_loss:2.3821 aux_loss:0.0418 train_time:836757ms step_avg:540.19ms step:1560/6250 train_loss:2.5652 aux_loss:0.0417 train_time:837298ms step_avg:540.19ms step:1561/6250 train_loss:2.3736 aux_loss:0.0415 train_time:837835ms step_avg:540.19ms step:1562/6250 train_loss:2.4143 aux_loss:0.0413 train_time:838374ms step_avg:540.19ms step:1563/6250 train_loss:2.3939 aux_loss:0.0412 train_time:838918ms step_avg:540.19ms step:1564/6250 train_loss:2.4120 aux_loss:0.0414 train_time:839453ms step_avg:540.19ms step:1565/6250 train_loss:2.4821 aux_loss:0.0413 train_time:839988ms step_avg:540.19ms step:1566/6250 train_loss:2.4218 aux_loss:0.0410 train_time:840524ms step_avg:540.18ms step:1567/6250 train_loss:2.4321 aux_loss:0.0412 train_time:841061ms step_avg:540.18ms step:1568/6250 train_loss:2.4155 aux_loss:0.0413 train_time:841596ms step_avg:540.18ms step:1569/6250 train_loss:2.4369 aux_loss:0.0411 train_time:842133ms step_avg:540.18ms step:1570/6250 train_loss:2.3509 aux_loss:0.0411 train_time:842669ms step_avg:540.17ms step:1571/6250 train_loss:2.4323 aux_loss:0.0412 train_time:843205ms step_avg:540.17ms step:1572/6250 train_loss:2.4556 aux_loss:0.0412 train_time:843741ms step_avg:540.17ms step:1573/6250 train_loss:2.4683 aux_loss:0.0412 train_time:844278ms step_avg:540.16ms step:1574/6250 train_loss:2.4545 aux_loss:0.0413 train_time:844813ms step_avg:540.16ms step:1575/6250 train_loss:2.3640 aux_loss:0.0416 train_time:845351ms step_avg:540.16ms step:1576/6250 train_loss:2.3501 aux_loss:0.0418 train_time:845887ms step_avg:540.16ms step:1577/6250 train_loss:2.6029 aux_loss:0.0414 train_time:846423ms step_avg:540.16ms step:1578/6250 train_loss:2.3233 aux_loss:0.0413 train_time:846959ms step_avg:540.15ms step:1579/6250 train_loss:2.4309 aux_loss:0.0412 train_time:847495ms step_avg:540.15ms step:1580/6250 train_loss:2.5018 aux_loss:0.0411 train_time:848029ms step_avg:540.15ms step:1581/6250 train_loss:2.4485 aux_loss:0.0411 train_time:848565ms step_avg:540.14ms step:1582/6250 train_loss:2.5195 aux_loss:0.0411 train_time:849101ms step_avg:540.14ms step:1583/6250 train_loss:2.4787 aux_loss:0.0410 train_time:849636ms step_avg:540.14ms step:1584/6250 train_loss:2.4595 aux_loss:0.0412 train_time:850171ms step_avg:540.13ms step:1585/6250 train_loss:2.3339 aux_loss:0.0412 train_time:850707ms step_avg:540.13ms step:1586/6250 train_loss:2.5942 aux_loss:0.0408 train_time:851241ms step_avg:540.13ms step:1587/6250 train_loss:2.4701 aux_loss:0.0411 train_time:851777ms step_avg:540.12ms step:1588/6250 train_loss:2.4049 aux_loss:0.0412 train_time:852311ms step_avg:540.12ms step:1589/6250 train_loss:2.3538 aux_loss:0.0413 train_time:852845ms step_avg:540.12ms step:1590/6250 train_loss:2.5047 aux_loss:0.0407 train_time:853381ms step_avg:540.11ms step:1591/6250 train_loss:2.4801 aux_loss:0.0410 train_time:853916ms step_avg:540.11ms step:1592/6250 train_loss:2.4362 aux_loss:0.0413 train_time:854450ms step_avg:540.11ms step:1593/6250 train_loss:2.4285 aux_loss:0.0414 train_time:854985ms step_avg:540.10ms step:1594/6250 train_loss:2.4538 aux_loss:0.0410 train_time:855521ms step_avg:540.10ms step:1595/6250 train_loss:2.4648 aux_loss:0.0407 train_time:856057ms step_avg:540.10ms step:1596/6250 train_loss:2.5065 aux_loss:0.0410 train_time:856592ms step_avg:540.10ms step:1597/6250 train_loss:2.5788 aux_loss:0.0410 train_time:857128ms step_avg:540.09ms step:1598/6250 train_loss:2.3567 aux_loss:0.0410 train_time:857662ms step_avg:540.09ms step:1599/6250 train_loss:2.4262 aux_loss:0.0409 train_time:858198ms step_avg:540.09ms step:1600/6250 train_loss:2.4340 aux_loss:0.0408 train_time:858734ms step_avg:540.08ms step:1601/6250 train_loss:2.5171 aux_loss:0.0410 train_time:859271ms step_avg:540.08ms step:1602/6250 train_loss:2.4815 aux_loss:0.0412 train_time:859805ms step_avg:540.08ms step:1603/6250 train_loss:2.5243 aux_loss:0.0413 train_time:860341ms step_avg:540.08ms step:1604/6250 train_loss:2.4520 aux_loss:0.0412 train_time:860877ms step_avg:540.07ms step:1605/6250 train_loss:2.5038 aux_loss:0.0410 train_time:861413ms step_avg:540.07ms step:1606/6250 train_loss:2.3872 aux_loss:0.0410 train_time:861948ms step_avg:540.07ms step:1607/6250 train_loss:2.4852 aux_loss:0.0410 train_time:862485ms step_avg:540.07ms step:1608/6250 train_loss:2.4240 aux_loss:0.0410 train_time:863019ms step_avg:540.06ms step:1609/6250 train_loss:2.5079 aux_loss:0.0408 train_time:863555ms step_avg:540.06ms step:1610/6250 train_loss:2.4876 aux_loss:0.0409 train_time:864090ms step_avg:540.06ms step:1611/6250 train_loss:2.4543 aux_loss:0.0409 train_time:864625ms step_avg:540.05ms step:1612/6250 train_loss:2.4650 aux_loss:0.0408 train_time:865159ms step_avg:540.05ms step:1613/6250 train_loss:2.4618 aux_loss:0.0407 train_time:865695ms step_avg:540.05ms step:1614/6250 train_loss:2.4468 aux_loss:0.0407 train_time:866229ms step_avg:540.04ms step:1615/6250 train_loss:2.3648 aux_loss:0.0409 train_time:866764ms step_avg:540.04ms step:1616/6250 train_loss:2.3670 aux_loss:0.0409 train_time:867302ms step_avg:540.04ms step:1617/6250 train_loss:2.4384 aux_loss:0.0409 train_time:867838ms step_avg:540.04ms step:1618/6250 train_loss:2.4436 aux_loss:0.0408 train_time:868376ms step_avg:540.03ms step:1619/6250 train_loss:2.4307 aux_loss:0.0405 train_time:868913ms step_avg:540.03ms step:1620/6250 train_loss:2.4557 aux_loss:0.0408 train_time:869457ms step_avg:540.04ms step:1621/6250 train_loss:2.3167 aux_loss:0.0410 train_time:869992ms step_avg:540.03ms step:1622/6250 train_loss:2.5694 aux_loss:0.0409 train_time:870530ms step_avg:540.03ms step:1623/6250 train_loss:2.4845 aux_loss:0.0408 train_time:871066ms step_avg:540.03ms step:1624/6250 train_loss:2.4649 aux_loss:0.0407 train_time:871604ms step_avg:540.03ms step:1625/6250 train_loss:2.5214 aux_loss:0.0409 train_time:872140ms step_avg:540.02ms step:1626/6250 train_loss:2.4519 aux_loss:0.0409 train_time:872676ms step_avg:540.02ms step:1627/6250 train_loss:2.4688 aux_loss:0.0407 train_time:873214ms step_avg:540.02ms step:1628/6250 train_loss:2.5738 aux_loss:0.0407 train_time:873752ms step_avg:540.02ms step:1629/6250 train_loss:2.4327 aux_loss:0.0406 train_time:874289ms step_avg:540.02ms step:1630/6250 train_loss:2.4186 aux_loss:0.0407 train_time:874825ms step_avg:540.02ms step:1631/6250 train_loss:2.4228 aux_loss:0.0408 train_time:875363ms step_avg:540.01ms step:1632/6250 train_loss:2.5014 aux_loss:0.0412 train_time:875899ms step_avg:540.01ms step:1633/6250 train_loss:2.3896 aux_loss:0.0411 train_time:876437ms step_avg:540.01ms step:1634/6250 train_loss:2.4848 aux_loss:0.0407 train_time:876973ms step_avg:540.01ms step:1635/6250 train_loss:2.3864 aux_loss:0.0408 train_time:877510ms step_avg:540.01ms step:1636/6250 train_loss:2.4098 aux_loss:0.0409 train_time:878050ms step_avg:540.01ms step:1637/6250 train_loss:2.5165 aux_loss:0.0412 train_time:878596ms step_avg:540.01ms step:1638/6250 train_loss:2.4972 aux_loss:0.0410 train_time:879135ms step_avg:540.01ms step:1639/6250 train_loss:2.3925 aux_loss:0.0409 train_time:879674ms step_avg:540.01ms step:1640/6250 train_loss:2.4284 aux_loss:0.0410 train_time:880213ms step_avg:540.01ms step:1641/6250 train_loss:2.4464 aux_loss:0.0409 train_time:880752ms step_avg:540.01ms step:1642/6250 train_loss:2.4718 aux_loss:0.0409 train_time:881295ms step_avg:540.01ms step:1643/6250 train_loss:2.4702 aux_loss:0.0410 train_time:881832ms step_avg:540.01ms step:1644/6250 train_loss:2.5031 aux_loss:0.0410 train_time:882367ms step_avg:540.00ms step:1645/6250 train_loss:2.4383 aux_loss:0.0407 train_time:882902ms step_avg:540.00ms step:1646/6250 train_loss:2.5063 aux_loss:0.0405 train_time:883439ms step_avg:540.00ms step:1647/6250 train_loss:2.3513 aux_loss:0.0404 train_time:883974ms step_avg:540.00ms step:1648/6250 train_loss:2.4709 aux_loss:0.0403 train_time:884511ms step_avg:539.99ms step:1649/6250 train_loss:2.4838 aux_loss:0.0406 train_time:885046ms step_avg:539.99ms step:1650/6250 train_loss:2.4857 aux_loss:0.0409 train_time:885583ms step_avg:539.99ms step:1651/6250 train_loss:2.4615 aux_loss:0.0407 train_time:886119ms step_avg:539.99ms step:1652/6250 train_loss:2.4240 aux_loss:0.0405 train_time:886654ms step_avg:539.98ms step:1653/6250 train_loss:2.3948 aux_loss:0.0407 train_time:887190ms step_avg:539.98ms step:1654/6250 train_loss:2.3399 aux_loss:0.0408 train_time:887727ms step_avg:539.98ms step:1655/6250 train_loss:2.5163 aux_loss:0.0406 train_time:888263ms step_avg:539.98ms step:1656/6250 train_loss:2.5816 aux_loss:0.0404 train_time:888797ms step_avg:539.97ms step:1657/6250 train_loss:2.4629 aux_loss:0.0406 train_time:889335ms step_avg:539.97ms step:1658/6250 train_loss:2.3446 aux_loss:0.0407 train_time:889871ms step_avg:539.97ms step:1659/6250 train_loss:2.5343 aux_loss:0.0405 train_time:890407ms step_avg:539.97ms step:1660/6250 train_loss:2.3893 aux_loss:0.0406 train_time:890942ms step_avg:539.97ms step:1661/6250 train_loss:2.4685 aux_loss:0.0410 train_time:891478ms step_avg:539.96ms step:1662/6250 train_loss:2.3828 aux_loss:0.0411 train_time:892013ms step_avg:539.96ms step:1663/6250 train_loss:2.3398 aux_loss:0.0408 train_time:892551ms step_avg:539.96ms step:1664/6250 train_loss:2.5226 aux_loss:0.0405 train_time:893087ms step_avg:539.96ms step:1665/6250 train_loss:2.5272 aux_loss:0.0407 train_time:893623ms step_avg:539.95ms step:1666/6250 train_loss:2.4667 aux_loss:0.0404 train_time:894160ms step_avg:539.95ms step:1667/6250 train_loss:2.3826 aux_loss:0.0405 train_time:894695ms step_avg:539.95ms step:1668/6250 train_loss:2.4961 aux_loss:0.0404 train_time:895230ms step_avg:539.95ms step:1669/6250 train_loss:2.5204 aux_loss:0.0404 train_time:895765ms step_avg:539.94ms step:1670/6250 train_loss:2.4235 aux_loss:0.0402 train_time:896304ms step_avg:539.94ms step:1671/6250 train_loss:2.4198 aux_loss:0.0401 train_time:896840ms step_avg:539.94ms step:1672/6250 train_loss:2.4746 aux_loss:0.0402 train_time:897377ms step_avg:539.94ms step:1673/6250 train_loss:2.3899 aux_loss:0.0404 train_time:897919ms step_avg:539.94ms step:1674/6250 train_loss:2.3823 aux_loss:0.0404 train_time:898474ms step_avg:539.95ms step:1675/6250 train_loss:2.4039 aux_loss:0.0401 train_time:899025ms step_avg:539.95ms step:1676/6250 train_loss:2.5108 aux_loss:0.0406 train_time:899568ms step_avg:539.96ms step:1677/6250 train_loss:2.4615 aux_loss:0.0406 train_time:900112ms step_avg:539.96ms step:1678/6250 train_loss:2.4649 aux_loss:0.0407 train_time:900657ms step_avg:539.96ms step:1679/6250 train_loss:2.2871 aux_loss:0.0400 train_time:901208ms step_avg:539.97ms step:1680/6250 train_loss:2.4832 aux_loss:0.0401 train_time:901754ms step_avg:539.97ms step:1681/6250 train_loss:2.4005 aux_loss:0.0403 train_time:902303ms step_avg:539.98ms step:1682/6250 train_loss:2.3903 aux_loss:0.0403 train_time:902851ms step_avg:539.98ms step:1683/6250 train_loss:2.4410 aux_loss:0.0403 train_time:903399ms step_avg:539.99ms step:1684/6250 train_loss:2.3843 aux_loss:0.0404 train_time:903943ms step_avg:539.99ms step:1685/6250 train_loss:2.5189 aux_loss:0.0406 train_time:904479ms step_avg:539.99ms step:1686/6250 train_loss:2.3671 aux_loss:0.0403 train_time:905015ms step_avg:539.99ms step:1687/6250 train_loss:2.4280 aux_loss:0.0403 train_time:905550ms step_avg:539.98ms step:1688/6250 train_loss:2.3605 aux_loss:0.0400 train_time:906098ms step_avg:539.99ms step:1689/6250 train_loss:2.4005 aux_loss:0.0400 train_time:906683ms step_avg:540.01ms step:1690/6250 train_loss:2.4815 aux_loss:0.0402 train_time:907240ms step_avg:540.02ms step:1691/6250 train_loss:2.4664 aux_loss:0.0406 train_time:907804ms step_avg:540.04ms step:1692/6250 train_loss:2.4736 aux_loss:0.0404 train_time:908367ms step_avg:540.05ms step:1693/6250 train_loss:2.4254 aux_loss:0.0401 train_time:908941ms step_avg:540.07ms step:1694/6250 train_loss:2.4975 aux_loss:0.0403 train_time:909492ms step_avg:540.08ms step:1695/6250 train_loss:2.3926 aux_loss:0.0403 train_time:910027ms step_avg:540.08ms step:1696/6250 train_loss:2.4253 aux_loss:0.0402 train_time:910570ms step_avg:540.08ms step:1697/6250 train_loss:2.3571 aux_loss:0.0402 train_time:911105ms step_avg:540.07ms step:1698/6250 train_loss:2.5252 aux_loss:0.0405 train_time:911642ms step_avg:540.07ms step:1699/6250 train_loss:2.4012 aux_loss:0.0405 train_time:912179ms step_avg:540.07ms step:1700/6250 train_loss:2.3975 aux_loss:0.0401 train_time:912715ms step_avg:540.07ms step:1701/6250 train_loss:2.4198 aux_loss:0.0398 train_time:913251ms step_avg:540.07ms step:1702/6250 train_loss:2.3845 aux_loss:0.0401 train_time:913785ms step_avg:540.06ms step:1703/6250 train_loss:2.3937 aux_loss:0.0404 train_time:914320ms step_avg:540.06ms step:1704/6250 train_loss:2.4250 aux_loss:0.0403 train_time:914893ms step_avg:540.08ms step:1705/6250 train_loss:2.3629 aux_loss:0.0405 train_time:915440ms step_avg:540.08ms step:1706/6250 train_loss:2.3669 aux_loss:0.0403 train_time:915977ms step_avg:540.08ms step:1707/6250 train_loss:2.4366 aux_loss:0.0401 train_time:916514ms step_avg:540.08ms step:1708/6250 train_loss:2.3949 aux_loss:0.0400 train_time:917051ms step_avg:540.08ms step:1709/6250 train_loss:2.4488 aux_loss:0.0402 train_time:917586ms step_avg:540.07ms step:1710/6250 train_loss:2.4982 aux_loss:0.0404 train_time:918129ms step_avg:540.08ms step:1711/6250 train_loss:2.4269 aux_loss:0.0400 train_time:918666ms step_avg:540.07ms step:1712/6250 train_loss:2.3848 aux_loss:0.0399 train_time:919202ms step_avg:540.07ms step:1713/6250 train_loss:2.3727 aux_loss:0.0400 train_time:919737ms step_avg:540.07ms step:1714/6250 train_loss:2.4153 aux_loss:0.0401 train_time:920274ms step_avg:540.07ms step:1715/6250 train_loss:2.4681 aux_loss:0.0403 train_time:920810ms step_avg:540.06ms step:1716/6250 train_loss:2.4587 aux_loss:0.0403 train_time:921347ms step_avg:540.06ms step:1717/6250 train_loss:2.4385 aux_loss:0.0403 train_time:921882ms step_avg:540.06ms step:1718/6250 train_loss:2.4982 aux_loss:0.0403 train_time:922418ms step_avg:540.06ms step:1719/6250 train_loss:2.4360 aux_loss:0.0404 train_time:922963ms step_avg:540.06ms step:1720/6250 train_loss:2.3947 aux_loss:0.0403 train_time:923509ms step_avg:540.06ms step:1721/6250 train_loss:2.4564 aux_loss:0.0403 train_time:924046ms step_avg:540.06ms step:1722/6250 train_loss:2.3767 aux_loss:0.0401 train_time:924583ms step_avg:540.06ms step:1723/6250 train_loss:2.4341 aux_loss:0.0400 train_time:925141ms step_avg:540.07ms step:1724/6250 train_loss:2.3870 aux_loss:0.0401 train_time:925685ms step_avg:540.07ms step:1725/6250 train_loss:2.4183 aux_loss:0.0401 train_time:926230ms step_avg:540.08ms step:1726/6250 train_loss:2.4625 aux_loss:0.0400 train_time:926766ms step_avg:540.07ms step:1727/6250 train_loss:2.4652 aux_loss:0.0399 train_time:927302ms step_avg:540.07ms step:1728/6250 train_loss:2.4719 aux_loss:0.0398 train_time:927853ms step_avg:540.08ms step:1729/6250 train_loss:2.4904 aux_loss:0.0399 train_time:928405ms step_avg:540.08ms step:1730/6250 train_loss:2.3952 aux_loss:0.0398 train_time:928949ms step_avg:540.09ms step:1731/6250 train_loss:2.5113 aux_loss:0.0400 train_time:929488ms step_avg:540.09ms step:1732/6250 train_loss:2.3964 aux_loss:0.0397 train_time:930025ms step_avg:540.08ms step:1733/6250 train_loss:2.4133 aux_loss:0.0395 train_time:930562ms step_avg:540.08ms step:1734/6250 train_loss:2.3684 aux_loss:0.0401 train_time:931100ms step_avg:540.08ms step:1735/6250 train_loss:2.3294 aux_loss:0.0403 train_time:931642ms step_avg:540.08ms step:1736/6250 train_loss:2.3947 aux_loss:0.0399 train_time:932178ms step_avg:540.08ms step:1737/6250 train_loss:2.4365 aux_loss:0.0397 train_time:932715ms step_avg:540.08ms step:1738/6250 train_loss:2.4825 aux_loss:0.0396 train_time:933250ms step_avg:540.08ms step:1739/6250 train_loss:2.4929 aux_loss:0.0399 train_time:933785ms step_avg:540.07ms step:1740/6250 train_loss:2.3535 aux_loss:0.0401 train_time:934323ms step_avg:540.07ms step:1741/6250 train_loss:2.3672 aux_loss:0.0403 train_time:934858ms step_avg:540.07ms step:1742/6250 train_loss:2.4385 aux_loss:0.0403 train_time:935394ms step_avg:540.07ms step:1743/6250 train_loss:2.4367 aux_loss:0.0401 train_time:935929ms step_avg:540.06ms step:1744/6250 train_loss:2.4269 aux_loss:0.0402 train_time:936472ms step_avg:540.06ms step:1745/6250 train_loss:2.3790 aux_loss:0.0401 train_time:937027ms step_avg:540.07ms step:1746/6250 train_loss:2.4620 aux_loss:0.0399 train_time:937568ms step_avg:540.07ms step:1747/6250 train_loss:2.4352 aux_loss:0.0398 train_time:938111ms step_avg:540.08ms step:1748/6250 train_loss:2.4198 aux_loss:0.0398 train_time:938647ms step_avg:540.07ms step:1749/6250 train_loss:2.3566 aux_loss:0.0397 train_time:939189ms step_avg:540.07ms step:1750/6250 train_loss:2.4528 aux_loss:0.0395 train_time:939725ms step_avg:540.07ms step:1751/6250 train_loss:2.4133 aux_loss:0.0398 train_time:940262ms step_avg:540.07ms step:1752/6250 train_loss:2.3651 aux_loss:0.0400 train_time:940797ms step_avg:540.07ms step:1753/6250 train_loss:2.3886 aux_loss:0.0401 train_time:941334ms step_avg:540.07ms step:1754/6250 train_loss:2.4217 aux_loss:0.0401 train_time:941868ms step_avg:540.06ms step:1755/6250 train_loss:2.4559 aux_loss:0.0400 train_time:942403ms step_avg:540.06ms step:1756/6250 train_loss:2.3586 aux_loss:0.0398 train_time:942939ms step_avg:540.06ms step:1757/6250 train_loss:2.4741 aux_loss:0.0394 train_time:943474ms step_avg:540.05ms step:1758/6250 train_loss:2.4719 aux_loss:0.0395 train_time:944010ms step_avg:540.05ms step:1759/6250 train_loss:2.4039 aux_loss:0.0397 train_time:944547ms step_avg:540.05ms step:1760/6250 train_loss:2.4237 aux_loss:0.0399 train_time:945082ms step_avg:540.05ms step:1761/6250 train_loss:2.4048 aux_loss:0.0394 train_time:945616ms step_avg:540.04ms step:1762/6250 train_loss:2.3331 aux_loss:0.0394 train_time:946152ms step_avg:540.04ms step:1763/6250 train_loss:2.3289 aux_loss:0.0397 train_time:946689ms step_avg:540.04ms step:1764/6250 train_loss:2.5197 aux_loss:0.0397 train_time:947225ms step_avg:540.04ms step:1765/6250 train_loss:2.3663 aux_loss:0.0397 train_time:947761ms step_avg:540.03ms step:1766/6250 train_loss:2.4425 aux_loss:0.0398 train_time:948295ms step_avg:540.03ms step:1767/6250 train_loss:2.4645 aux_loss:0.0395 train_time:948832ms step_avg:540.03ms step:1768/6250 train_loss:2.4392 aux_loss:0.0395 train_time:949368ms step_avg:540.03ms step:1769/6250 train_loss:2.4849 aux_loss:0.0398 train_time:949905ms step_avg:540.03ms step:1770/6250 train_loss:2.4243 aux_loss:0.0400 train_time:950453ms step_avg:540.03ms step:1771/6250 train_loss:2.4201 aux_loss:0.0398 train_time:951004ms step_avg:540.04ms step:1772/6250 train_loss:2.4192 aux_loss:0.0398 train_time:951570ms step_avg:540.05ms step:1773/6250 train_loss:2.4495 aux_loss:0.0399 train_time:952109ms step_avg:540.05ms step:1774/6250 train_loss:2.4161 aux_loss:0.0401 train_time:952650ms step_avg:540.05ms step:1775/6250 train_loss:2.4319 aux_loss:0.0398 train_time:953196ms step_avg:540.05ms step:1776/6250 train_loss:2.4888 aux_loss:0.0398 train_time:953726ms step_avg:540.05ms step:1777/6250 train_loss:2.4114 aux_loss:0.0398 train_time:954256ms step_avg:540.04ms step:1778/6250 train_loss:2.4002 aux_loss:0.0396 train_time:954785ms step_avg:540.04ms step:1779/6250 train_loss:2.3402 aux_loss:0.0396 train_time:955314ms step_avg:540.03ms step:1780/6250 train_loss:2.4608 aux_loss:0.0397 train_time:955843ms step_avg:540.02ms step:1781/6250 train_loss:2.4488 aux_loss:0.0396 train_time:956373ms step_avg:540.02ms step:1782/6250 train_loss:2.4224 aux_loss:0.0397 train_time:956902ms step_avg:540.01ms step:1783/6250 train_loss:2.4158 aux_loss:0.0397 train_time:957430ms step_avg:540.01ms step:1784/6250 train_loss:2.3771 aux_loss:0.0399 train_time:957958ms step_avg:540.00ms step:1785/6250 train_loss:2.3810 aux_loss:0.0397 train_time:958489ms step_avg:539.99ms step:1786/6250 train_loss:2.4010 aux_loss:0.0397 train_time:959020ms step_avg:539.99ms step:1787/6250 train_loss:2.4531 aux_loss:0.0393 train_time:959549ms step_avg:539.98ms step:1788/6250 train_loss:2.4616 aux_loss:0.0394 train_time:960079ms step_avg:539.98ms step:1789/6250 train_loss:2.2986 aux_loss:0.0398 train_time:960609ms step_avg:539.97ms step:1790/6250 train_loss:2.3674 aux_loss:0.0396 train_time:961138ms step_avg:539.97ms step:1791/6250 train_loss:2.4929 aux_loss:0.0395 train_time:961667ms step_avg:539.96ms step:1792/6250 train_loss:2.4180 aux_loss:0.0393 train_time:962197ms step_avg:539.95ms step:1793/6250 train_loss:2.3802 aux_loss:0.0393 train_time:962728ms step_avg:539.95ms step:1794/6250 train_loss:2.4542 aux_loss:0.0394 train_time:963268ms step_avg:539.95ms step:1795/6250 train_loss:2.3618 aux_loss:0.0393 train_time:963816ms step_avg:539.95ms step:1796/6250 train_loss:2.3948 aux_loss:0.0393 train_time:964346ms step_avg:539.95ms step:1797/6250 train_loss:2.4038 aux_loss:0.0396 train_time:964877ms step_avg:539.94ms step:1798/6250 train_loss:2.3765 aux_loss:0.0397 train_time:965405ms step_avg:539.94ms step:1799/6250 train_loss:2.4840 aux_loss:0.0393 train_time:965933ms step_avg:539.93ms step:1800/6250 train_loss:2.4373 aux_loss:0.0390 train_time:966461ms step_avg:539.92ms step:1801/6250 train_loss:2.4455 aux_loss:0.0394 train_time:966988ms step_avg:539.91ms step:1802/6250 train_loss:2.3925 aux_loss:0.0400 train_time:967535ms step_avg:539.92ms step:1803/6250 train_loss:2.4626 aux_loss:0.0398 train_time:968091ms step_avg:539.93ms step:1804/6250 train_loss:2.5134 aux_loss:0.0397 train_time:968635ms step_avg:539.93ms step:1805/6250 train_loss:2.4564 aux_loss:0.0394 train_time:969184ms step_avg:539.94ms step:1806/6250 train_loss:2.4129 aux_loss:0.0391 train_time:969713ms step_avg:539.93ms step:1807/6250 train_loss:2.4546 aux_loss:0.0391 train_time:970250ms step_avg:539.93ms step:1808/6250 train_loss:2.3562 aux_loss:0.0392 train_time:970789ms step_avg:539.93ms step:1809/6250 train_loss:2.3925 aux_loss:0.0394 train_time:971323ms step_avg:539.92ms step:1810/6250 train_loss:2.4285 aux_loss:0.0395 train_time:971864ms step_avg:539.92ms step:1811/6250 train_loss:2.4888 aux_loss:0.0390 train_time:972404ms step_avg:539.92ms step:1812/6250 train_loss:2.3481 aux_loss:0.0393 train_time:972937ms step_avg:539.92ms step:1813/6250 train_loss:2.4473 aux_loss:0.0394 train_time:973467ms step_avg:539.92ms step:1814/6250 train_loss:2.4121 aux_loss:0.0395 train_time:973998ms step_avg:539.91ms step:1815/6250 train_loss:2.3508 aux_loss:0.0395 train_time:974546ms step_avg:539.91ms step:1816/6250 train_loss:2.3782 aux_loss:0.0393 train_time:975094ms step_avg:539.92ms step:1817/6250 train_loss:2.4501 aux_loss:0.0392 train_time:975631ms step_avg:539.92ms step:1818/6250 train_loss:2.3896 aux_loss:0.0394 train_time:976167ms step_avg:539.92ms step:1819/6250 train_loss:2.3457 aux_loss:0.0391 train_time:976706ms step_avg:539.91ms step:1820/6250 train_loss:2.3863 aux_loss:0.0390 train_time:977238ms step_avg:539.91ms step:1821/6250 train_loss:2.4057 aux_loss:0.0396 train_time:977768ms step_avg:539.91ms step:1822/6250 train_loss:2.4499 aux_loss:0.0396 train_time:978300ms step_avg:539.90ms step:1823/6250 train_loss:2.4185 aux_loss:0.0391 train_time:978829ms step_avg:539.89ms step:1824/6250 train_loss:2.4805 aux_loss:0.0391 train_time:979369ms step_avg:539.89ms step:1825/6250 train_loss:2.4144 aux_loss:0.0395 train_time:979908ms step_avg:539.89ms step:1826/6250 train_loss:2.4694 aux_loss:0.0398 train_time:980446ms step_avg:539.89ms step:1827/6250 train_loss:2.4150 aux_loss:0.0396 train_time:980991ms step_avg:539.90ms step:1828/6250 train_loss:2.3741 aux_loss:0.0399 train_time:981545ms step_avg:539.90ms step:1829/6250 train_loss:2.5965 aux_loss:0.0396 train_time:982092ms step_avg:539.91ms step:1830/6250 train_loss:2.3734 aux_loss:0.0394 train_time:982638ms step_avg:539.91ms step:1831/6250 train_loss:2.4966 aux_loss:0.0395 train_time:983182ms step_avg:539.91ms step:1832/6250 train_loss:2.3463 aux_loss:0.0393 train_time:983719ms step_avg:539.91ms step:1833/6250 train_loss:2.5205 aux_loss:0.0393 train_time:984263ms step_avg:539.91ms step:1834/6250 train_loss:2.4466 aux_loss:0.0391 train_time:984825ms step_avg:539.93ms step:1835/6250 train_loss:2.4185 aux_loss:0.0390 train_time:985364ms step_avg:539.93ms step:1836/6250 train_loss:2.5107 aux_loss:0.0391 train_time:985894ms step_avg:539.92ms step:1837/6250 train_loss:2.4077 aux_loss:0.0393 train_time:986426ms step_avg:539.92ms step:1838/6250 train_loss:2.4360 aux_loss:0.0391 train_time:986955ms step_avg:539.91ms step:1839/6250 train_loss:2.3987 aux_loss:0.0389 train_time:987486ms step_avg:539.90ms step:1840/6250 train_loss:2.3835 aux_loss:0.0391 train_time:988016ms step_avg:539.90ms step:1841/6250 train_loss:2.4446 aux_loss:0.0394 train_time:988545ms step_avg:539.89ms step:1842/6250 train_loss:2.3878 aux_loss:0.0392 train_time:989074ms step_avg:539.89ms step:1843/6250 train_loss:2.4655 aux_loss:0.0391 train_time:989607ms step_avg:539.88ms step:1844/6250 train_loss:2.4739 aux_loss:0.0394 train_time:990136ms step_avg:539.88ms step:1845/6250 train_loss:2.3793 aux_loss:0.0392 train_time:990665ms step_avg:539.87ms step:1846/6250 train_loss:2.4433 aux_loss:0.0393 train_time:991200ms step_avg:539.87ms step:1847/6250 train_loss:2.4469 aux_loss:0.0392 train_time:991735ms step_avg:539.87ms step:1848/6250 train_loss:2.3011 aux_loss:0.0392 train_time:992271ms step_avg:539.86ms step:1849/6250 train_loss:2.4542 aux_loss:0.0395 train_time:992827ms step_avg:539.87ms step:1850/6250 train_loss:2.4609 aux_loss:0.0393 train_time:993356ms step_avg:539.87ms step:1851/6250 train_loss:2.3256 aux_loss:0.0392 train_time:993885ms step_avg:539.86ms step:1852/6250 train_loss:2.3541 aux_loss:0.0394 train_time:994428ms step_avg:539.86ms step:1853/6250 train_loss:2.4403 aux_loss:0.0396 train_time:994976ms step_avg:539.87ms step:1854/6250 train_loss:2.3820 aux_loss:0.0393 train_time:995514ms step_avg:539.87ms step:1855/6250 train_loss:2.3833 aux_loss:0.0393 train_time:996043ms step_avg:539.86ms step:1856/6250 train_loss:2.4737 aux_loss:0.0393 train_time:996572ms step_avg:539.85ms step:1857/6250 train_loss:2.4730 aux_loss:0.0392 train_time:997101ms step_avg:539.85ms step:1858/6250 train_loss:2.3750 aux_loss:0.0394 train_time:997630ms step_avg:539.84ms step:1859/6250 train_loss:2.5097 aux_loss:0.0395 train_time:998159ms step_avg:539.84ms step:1860/6250 train_loss:2.3025 aux_loss:0.0394 train_time:998688ms step_avg:539.83ms step:1861/6250 train_loss:2.5417 aux_loss:0.0388 train_time:999216ms step_avg:539.83ms step:1862/6250 train_loss:2.4374 aux_loss:0.0391 train_time:999746ms step_avg:539.82ms step:1863/6250 train_loss:2.4028 aux_loss:0.0392 train_time:1000276ms step_avg:539.81ms step:1864/6250 train_loss:2.4508 aux_loss:0.0392 train_time:1000806ms step_avg:539.81ms step:1865/6250 train_loss:2.3365 aux_loss:0.0391 train_time:1001336ms step_avg:539.80ms step:1866/6250 train_loss:2.3436 aux_loss:0.0390 train_time:1001869ms step_avg:539.80ms step:1867/6250 train_loss:2.4367 aux_loss:0.0390 train_time:1002400ms step_avg:539.80ms step:1868/6250 train_loss:2.3709 aux_loss:0.0390 train_time:1002928ms step_avg:539.79ms step:1869/6250 train_loss:2.4116 aux_loss:0.0395 train_time:1003458ms step_avg:539.78ms step:1870/6250 train_loss:2.4263 aux_loss:0.0393 train_time:1003988ms step_avg:539.78ms step:1871/6250 train_loss:2.4057 aux_loss:0.0392 train_time:1004517ms step_avg:539.77ms step:1872/6250 train_loss:2.5002 aux_loss:0.0390 train_time:1005046ms step_avg:539.77ms step:1873/6250 train_loss:2.3026 aux_loss:0.0389 train_time:1005579ms step_avg:539.76ms step:1874/6250 train_loss:2.4101 aux_loss:0.0391 train_time:1006113ms step_avg:539.76ms step:1875/6250 train_loss:2.2714 aux_loss:0.0394 train_time:1006642ms step_avg:539.75ms step:1876/6250 train_loss:2.4610 aux_loss:0.0392 train_time:1007186ms step_avg:539.76ms step:1877/6250 train_loss:2.4150 aux_loss:0.0391 train_time:1007718ms step_avg:539.75ms step:1878/6250 train_loss:2.4405 aux_loss:0.0387 train_time:1008258ms step_avg:539.75ms step:1879/6250 train_loss:2.4361 aux_loss:0.0388 train_time:1008800ms step_avg:539.75ms step:1880/6250 train_loss:2.4164 aux_loss:0.0391 train_time:1009338ms step_avg:539.75ms step:1881/6250 train_loss:2.3798 aux_loss:0.0390 train_time:1009870ms step_avg:539.75ms step:1882/6250 train_loss:2.3364 aux_loss:0.0392 train_time:1010412ms step_avg:539.75ms step:1883/6250 train_loss:2.4526 aux_loss:0.0391 train_time:1010943ms step_avg:539.75ms step:1884/6250 train_loss:2.3870 aux_loss:0.0390 train_time:1011474ms step_avg:539.74ms step:1885/6250 train_loss:2.4002 aux_loss:0.0387 train_time:1012009ms step_avg:539.74ms step:1886/6250 train_loss:2.4030 aux_loss:0.0386 train_time:1012540ms step_avg:539.73ms step:1887/6250 train_loss:2.3685 aux_loss:0.0386 train_time:1013069ms step_avg:539.73ms step:1888/6250 train_loss:2.3255 aux_loss:0.0389 train_time:1013599ms step_avg:539.72ms step:1889/6250 train_loss:2.3682 aux_loss:0.0387 train_time:1014129ms step_avg:539.72ms step:1890/6250 train_loss:2.3765 aux_loss:0.0390 train_time:1014660ms step_avg:539.71ms step:1891/6250 train_loss:2.3883 aux_loss:0.0389 train_time:1015188ms step_avg:539.71ms step:1892/6250 train_loss:2.3938 aux_loss:0.0391 train_time:1015717ms step_avg:539.70ms step:1893/6250 train_loss:2.4534 aux_loss:0.0389 train_time:1016247ms step_avg:539.70ms step:1894/6250 train_loss:2.4144 aux_loss:0.0387 train_time:1016777ms step_avg:539.69ms step:1895/6250 train_loss:2.4025 aux_loss:0.0386 train_time:1017307ms step_avg:539.69ms step:1896/6250 train_loss:2.4511 aux_loss:0.0388 train_time:1017836ms step_avg:539.68ms step:1897/6250 train_loss:2.3397 aux_loss:0.0388 train_time:1018367ms step_avg:539.68ms step:1898/6250 train_loss:2.3979 aux_loss:0.0386 train_time:1018899ms step_avg:539.67ms step:1899/6250 train_loss:2.4839 aux_loss:0.0386 train_time:1019441ms step_avg:539.67ms step:1900/6250 train_loss:2.3889 aux_loss:0.0387 train_time:1019975ms step_avg:539.67ms step:1901/6250 train_loss:2.3512 aux_loss:0.0390 train_time:1020512ms step_avg:539.67ms step:1902/6250 train_loss:2.3818 aux_loss:0.0388 train_time:1021045ms step_avg:539.66ms step:1903/6250 train_loss:2.3864 aux_loss:0.0386 train_time:1021584ms step_avg:539.66ms step:1904/6250 train_loss:2.3857 aux_loss:0.0391 train_time:1022114ms step_avg:539.66ms step:1905/6250 train_loss:2.3977 aux_loss:0.0396 train_time:1022646ms step_avg:539.66ms step:1906/6250 train_loss:2.3411 aux_loss:0.0391 train_time:1023190ms step_avg:539.66ms step:1907/6250 train_loss:2.4508 aux_loss:0.0390 train_time:1023725ms step_avg:539.65ms step:1908/6250 train_loss:2.4186 aux_loss:0.0390 train_time:1024257ms step_avg:539.65ms step:1909/6250 train_loss:2.3457 aux_loss:0.0389 train_time:1024788ms step_avg:539.65ms step:1910/6250 train_loss:2.4246 aux_loss:0.0389 train_time:1025322ms step_avg:539.64ms step:1911/6250 train_loss:2.3800 aux_loss:0.0387 train_time:1025851ms step_avg:539.64ms step:1912/6250 train_loss:2.5348 aux_loss:0.0386 train_time:1026381ms step_avg:539.63ms step:1913/6250 train_loss:2.3509 aux_loss:0.0387 train_time:1026910ms step_avg:539.63ms step:1914/6250 train_loss:2.3775 aux_loss:0.0388 train_time:1027438ms step_avg:539.62ms step:1915/6250 train_loss:2.4076 aux_loss:0.0386 train_time:1027968ms step_avg:539.62ms step:1916/6250 train_loss:2.4693 aux_loss:0.0387 train_time:1028497ms step_avg:539.61ms step:1917/6250 train_loss:2.2955 aux_loss:0.0389 train_time:1029026ms step_avg:539.60ms step:1918/6250 train_loss:2.4486 aux_loss:0.0389 train_time:1029555ms step_avg:539.60ms step:1919/6250 train_loss:2.4025 aux_loss:0.0389 train_time:1030086ms step_avg:539.59ms step:1920/6250 train_loss:2.4440 aux_loss:0.0390 train_time:1030620ms step_avg:539.59ms step:1921/6250 train_loss:2.3872 aux_loss:0.0390 train_time:1031148ms step_avg:539.59ms step:1922/6250 train_loss:2.3480 aux_loss:0.0388 train_time:1031677ms step_avg:539.58ms step:1923/6250 train_loss:2.4541 aux_loss:0.0385 train_time:1032208ms step_avg:539.58ms step:1924/6250 train_loss:2.4323 aux_loss:0.0385 train_time:1032737ms step_avg:539.57ms step:1925/6250 train_loss:2.4784 aux_loss:0.0386 train_time:1033265ms step_avg:539.56ms step:1926/6250 train_loss:2.3656 aux_loss:0.0386 train_time:1033791ms step_avg:539.56ms step:1927/6250 train_loss:2.3577 aux_loss:0.0385 train_time:1034326ms step_avg:539.55ms step:1928/6250 train_loss:2.4217 aux_loss:0.0383 train_time:1034860ms step_avg:539.55ms step:1929/6250 train_loss:2.4360 aux_loss:0.0387 train_time:1035399ms step_avg:539.55ms step:1930/6250 train_loss:2.3688 aux_loss:0.0389 train_time:1035946ms step_avg:539.56ms step:1931/6250 train_loss:2.4135 aux_loss:0.0387 train_time:1036486ms step_avg:539.56ms step:1932/6250 train_loss:2.3747 aux_loss:0.0384 train_time:1037034ms step_avg:539.56ms step:1933/6250 train_loss:2.3645 aux_loss:0.0385 train_time:1037571ms step_avg:539.56ms step:1934/6250 train_loss:2.4466 aux_loss:0.0387 train_time:1038109ms step_avg:539.56ms step:1935/6250 train_loss:2.4246 aux_loss:0.0388 train_time:1038650ms step_avg:539.56ms step:1936/6250 train_loss:2.4336 aux_loss:0.0386 train_time:1039181ms step_avg:539.55ms step:1937/6250 train_loss:2.4070 aux_loss:0.0387 train_time:1039715ms step_avg:539.55ms step:1938/6250 train_loss:2.3719 aux_loss:0.0386 train_time:1040255ms step_avg:539.55ms step:1939/6250 train_loss:2.3886 aux_loss:0.0385 train_time:1040797ms step_avg:539.55ms step:1940/6250 train_loss:2.3714 aux_loss:0.0385 train_time:1041328ms step_avg:539.55ms step:1941/6250 train_loss:2.3759 aux_loss:0.0384 train_time:1041861ms step_avg:539.54ms step:1942/6250 train_loss:2.5093 aux_loss:0.0386 train_time:1042403ms step_avg:539.55ms step:1943/6250 train_loss:2.3580 aux_loss:0.0387 train_time:1042941ms step_avg:539.55ms step:1944/6250 train_loss:2.4057 aux_loss:0.0384 train_time:1043471ms step_avg:539.54ms step:1945/6250 train_loss:2.3916 aux_loss:0.0384 train_time:1043998ms step_avg:539.53ms step:1946/6250 train_loss:2.4680 aux_loss:0.0386 train_time:1044527ms step_avg:539.53ms step:1947/6250 train_loss:2.4629 aux_loss:0.0386 train_time:1045070ms step_avg:539.53ms step:1948/6250 train_loss:2.4072 aux_loss:0.0382 train_time:1045630ms step_avg:539.54ms step:1949/6250 train_loss:2.3551 aux_loss:0.0382 train_time:1046160ms step_avg:539.54ms step:1950/6250 train_loss:2.3204 aux_loss:0.0383 train_time:1046689ms step_avg:539.53ms step:1951/6250 train_loss:2.3817 aux_loss:0.0385 train_time:1047217ms step_avg:539.52ms step:1952/6250 train_loss:2.3849 aux_loss:0.0385 train_time:1047744ms step_avg:539.52ms step:1953/6250 train_loss:2.2999 aux_loss:0.0383 train_time:1048273ms step_avg:539.51ms step:1954/6250 train_loss:2.3733 aux_loss:0.0385 train_time:1048803ms step_avg:539.51ms step:1955/6250 train_loss:2.4018 aux_loss:0.0384 train_time:1049332ms step_avg:539.50ms step:1956/6250 train_loss:2.3268 aux_loss:0.0383 train_time:1049859ms step_avg:539.50ms step:1957/6250 train_loss:2.3928 aux_loss:0.0384 train_time:1050389ms step_avg:539.49ms step:1958/6250 train_loss:2.4075 aux_loss:0.0386 train_time:1050918ms step_avg:539.49ms step:1959/6250 train_loss:2.3798 aux_loss:0.0385 train_time:1051446ms step_avg:539.48ms step:1960/6250 train_loss:2.4259 aux_loss:0.0385 train_time:1051976ms step_avg:539.48ms step:1961/6250 train_loss:2.4187 aux_loss:0.0381 train_time:1052506ms step_avg:539.47ms step:1962/6250 train_loss:2.3674 aux_loss:0.0382 train_time:1053036ms step_avg:539.47ms step:1963/6250 train_loss:2.4268 aux_loss:0.0386 train_time:1053565ms step_avg:539.46ms step:1964/6250 train_loss:2.3357 aux_loss:0.0388 train_time:1054092ms step_avg:539.45ms step:1965/6250 train_loss:2.5468 aux_loss:0.0385 train_time:1054621ms step_avg:539.45ms step:1966/6250 train_loss:2.3201 aux_loss:0.0382 train_time:1055148ms step_avg:539.44ms step:1967/6250 train_loss:2.4464 aux_loss:0.0381 train_time:1055678ms step_avg:539.44ms step:1968/6250 train_loss:2.3998 aux_loss:0.0381 train_time:1056205ms step_avg:539.43ms step:1969/6250 train_loss:2.3674 aux_loss:0.0385 train_time:1056734ms step_avg:539.43ms step:1970/6250 train_loss:2.3626 aux_loss:0.0386 train_time:1057263ms step_avg:539.42ms step:1971/6250 train_loss:2.3834 aux_loss:0.0387 train_time:1057791ms step_avg:539.41ms step:1972/6250 train_loss:2.3934 aux_loss:0.0386 train_time:1058319ms step_avg:539.41ms step:1973/6250 train_loss:2.3967 aux_loss:0.0382 train_time:1058847ms step_avg:539.40ms step:1974/6250 train_loss:2.3325 aux_loss:0.0383 train_time:1059378ms step_avg:539.40ms step:1975/6250 train_loss:2.3268 aux_loss:0.0385 train_time:1059906ms step_avg:539.39ms step:1976/6250 train_loss:2.4107 aux_loss:0.0385 train_time:1060433ms step_avg:539.39ms step:1977/6250 train_loss:2.4484 aux_loss:0.0385 train_time:1060963ms step_avg:539.38ms step:1978/6250 train_loss:2.3638 aux_loss:0.0386 train_time:1061492ms step_avg:539.38ms step:1979/6250 train_loss:2.2986 aux_loss:0.0384 train_time:1062020ms step_avg:539.37ms step:1980/6250 train_loss:2.4552 aux_loss:0.0383 train_time:1062549ms step_avg:539.36ms step:1981/6250 train_loss:2.3300 aux_loss:0.0384 train_time:1063077ms step_avg:539.36ms step:1982/6250 train_loss:2.3993 aux_loss:0.0384 train_time:1063605ms step_avg:539.35ms step:1983/6250 train_loss:2.3362 aux_loss:0.0386 train_time:1064134ms step_avg:539.35ms step:1984/6250 train_loss:2.3987 aux_loss:0.0384 train_time:1064662ms step_avg:539.34ms step:1985/6250 train_loss:2.4455 aux_loss:0.0383 train_time:1065199ms step_avg:539.34ms step:1986/6250 train_loss:2.4762 aux_loss:0.0380 train_time:1065734ms step_avg:539.34ms step:1987/6250 train_loss:2.4048 aux_loss:0.0384 train_time:1066267ms step_avg:539.34ms step:1988/6250 train_loss:2.3561 aux_loss:0.0381 train_time:1066804ms step_avg:539.33ms step:1989/6250 train_loss:2.4514 aux_loss:0.0383 train_time:1067334ms step_avg:539.33ms step:1990/6250 train_loss:2.4490 aux_loss:0.0381 train_time:1067863ms step_avg:539.32ms step:1991/6250 train_loss:2.3263 aux_loss:0.0384 train_time:1068393ms step_avg:539.32ms step:1992/6250 train_loss:2.3401 aux_loss:0.0384 train_time:1068922ms step_avg:539.31ms step:1993/6250 train_loss:2.4933 aux_loss:0.0381 train_time:1069470ms step_avg:539.32ms step:1994/6250 train_loss:2.4926 aux_loss:0.0382 train_time:1070008ms step_avg:539.32ms step:1995/6250 train_loss:2.3056 aux_loss:0.0384 train_time:1070556ms step_avg:539.32ms step:1996/6250 train_loss:2.3730 aux_loss:0.0383 train_time:1071103ms step_avg:539.33ms step:1997/6250 train_loss:2.3956 aux_loss:0.0379 train_time:1071651ms step_avg:539.33ms step:1998/6250 train_loss:2.4528 aux_loss:0.0383 train_time:1072181ms step_avg:539.33ms step:1999/6250 train_loss:2.4144 aux_loss:0.0386 train_time:1072710ms step_avg:539.32ms step:2000/6250 train_loss:2.3795 aux_loss:0.0384 train_time:1073240ms step_avg:539.32ms step:2000/6250 val_loss:2.3788 val_aux_loss:0.0381 train_time:1073240ms step_avg:539.32ms step:2001/6250 train_loss:2.4545 aux_loss:0.0382 train_time:1073768ms step_avg:539.31ms step:2002/6250 train_loss:2.3871 aux_loss:0.0383 train_time:1074297ms step_avg:539.31ms step:2003/6250 train_loss:2.3312 aux_loss:0.0383 train_time:1074826ms step_avg:539.30ms step:2004/6250 train_loss:2.4082 aux_loss:0.0384 train_time:1075355ms step_avg:539.30ms step:2005/6250 train_loss:2.3062 aux_loss:0.0386 train_time:1075885ms step_avg:539.29ms step:2006/6250 train_loss:2.3597 aux_loss:0.0386 train_time:1076415ms step_avg:539.29ms step:2007/6250 train_loss:2.4294 aux_loss:0.0386 train_time:1076945ms step_avg:539.28ms step:2008/6250 train_loss:2.3593 aux_loss:0.0385 train_time:1077475ms step_avg:539.28ms step:2009/6250 train_loss:2.3906 aux_loss:0.0382 train_time:1078004ms step_avg:539.27ms step:2010/6250 train_loss:2.3585 aux_loss:0.0381 train_time:1078534ms step_avg:539.27ms step:2011/6250 train_loss:2.4038 aux_loss:0.0383 train_time:1079063ms step_avg:539.26ms step:2012/6250 train_loss:2.4187 aux_loss:0.0387 train_time:1079592ms step_avg:539.26ms step:2013/6250 train_loss:2.3496 aux_loss:0.0383 train_time:1080121ms step_avg:539.25ms step:2014/6250 train_loss:2.3130 aux_loss:0.0380 train_time:1080650ms step_avg:539.25ms step:2015/6250 train_loss:2.4033 aux_loss:0.0377 train_time:1081178ms step_avg:539.24ms step:2016/6250 train_loss:2.4186 aux_loss:0.0381 train_time:1081707ms step_avg:539.24ms step:2017/6250 train_loss:2.4260 aux_loss:0.0385 train_time:1082236ms step_avg:539.23ms step:2018/6250 train_loss:2.3702 aux_loss:0.0384 train_time:1082766ms step_avg:539.23ms step:2019/6250 train_loss:2.3819 aux_loss:0.0386 train_time:1083293ms step_avg:539.22ms step:2020/6250 train_loss:2.3759 aux_loss:0.0387 train_time:1083820ms step_avg:539.21ms step:2021/6250 train_loss:2.3997 aux_loss:0.0385 train_time:1084351ms step_avg:539.21ms step:2022/6250 train_loss:2.3743 aux_loss:0.0383 train_time:1084879ms step_avg:539.20ms step:2023/6250 train_loss:2.3151 aux_loss:0.0383 train_time:1085409ms step_avg:539.20ms step:2024/6250 train_loss:2.4187 aux_loss:0.0388 train_time:1085937ms step_avg:539.19ms step:2025/6250 train_loss:2.3969 aux_loss:0.0386 train_time:1086470ms step_avg:539.19ms step:2026/6250 train_loss:2.3611 aux_loss:0.0383 train_time:1087001ms step_avg:539.19ms step:2027/6250 train_loss:2.3487 aux_loss:0.0382 train_time:1087539ms step_avg:539.19ms step:2028/6250 train_loss:2.3298 aux_loss:0.0385 train_time:1088071ms step_avg:539.18ms step:2029/6250 train_loss:2.4140 aux_loss:0.0386 train_time:1088602ms step_avg:539.18ms step:2030/6250 train_loss:2.3847 aux_loss:0.0383 train_time:1089131ms step_avg:539.17ms step:2031/6250 train_loss:2.4241 aux_loss:0.0383 train_time:1089663ms step_avg:539.17ms step:2032/6250 train_loss:2.3484 aux_loss:0.0381 train_time:1090193ms step_avg:539.17ms step:2033/6250 train_loss:2.3418 aux_loss:0.0379 train_time:1090722ms step_avg:539.16ms step:2034/6250 train_loss:2.4312 aux_loss:0.0379 train_time:1091250ms step_avg:539.16ms step:2035/6250 train_loss:2.3496 aux_loss:0.0379 train_time:1091781ms step_avg:539.15ms step:2036/6250 train_loss:2.3073 aux_loss:0.0383 train_time:1092309ms step_avg:539.15ms step:2037/6250 train_loss:2.4533 aux_loss:0.0381 train_time:1092840ms step_avg:539.14ms step:2038/6250 train_loss:2.3468 aux_loss:0.0379 train_time:1093371ms step_avg:539.14ms step:2039/6250 train_loss:2.3409 aux_loss:0.0379 train_time:1093909ms step_avg:539.14ms step:2040/6250 train_loss:2.3049 aux_loss:0.0379 train_time:1094448ms step_avg:539.14ms step:2041/6250 train_loss:2.3878 aux_loss:0.0382 train_time:1094984ms step_avg:539.14ms step:2042/6250 train_loss:2.4004 aux_loss:0.0378 train_time:1095524ms step_avg:539.14ms step:2043/6250 train_loss:2.3338 aux_loss:0.0377 train_time:1096061ms step_avg:539.14ms step:2044/6250 train_loss:2.2812 aux_loss:0.0378 train_time:1096604ms step_avg:539.14ms step:2045/6250 train_loss:2.3349 aux_loss:0.0382 train_time:1097139ms step_avg:539.13ms step:2046/6250 train_loss:2.4729 aux_loss:0.0382 train_time:1097676ms step_avg:539.13ms step:2047/6250 train_loss:2.4265 aux_loss:0.0380 train_time:1098208ms step_avg:539.13ms step:2048/6250 train_loss:2.3724 aux_loss:0.0379 train_time:1098737ms step_avg:539.12ms step:2049/6250 train_loss:2.3873 aux_loss:0.0380 train_time:1099271ms step_avg:539.12ms step:2050/6250 train_loss:2.2705 aux_loss:0.0382 train_time:1099804ms step_avg:539.12ms step:2051/6250 train_loss:2.3667 aux_loss:0.0379 train_time:1100332ms step_avg:539.11ms step:2052/6250 train_loss:2.3602 aux_loss:0.0375 train_time:1100866ms step_avg:539.11ms step:2053/6250 train_loss:2.2916 aux_loss:0.0378 train_time:1101397ms step_avg:539.11ms step:2054/6250 train_loss:2.3920 aux_loss:0.0381 train_time:1101924ms step_avg:539.10ms step:2055/6250 train_loss:2.2541 aux_loss:0.0380 train_time:1102454ms step_avg:539.10ms step:2056/6250 train_loss:2.4817 aux_loss:0.0377 train_time:1102983ms step_avg:539.09ms step:2057/6250 train_loss:2.3393 aux_loss:0.0376 train_time:1103515ms step_avg:539.09ms step:2058/6250 train_loss:2.3402 aux_loss:0.0380 train_time:1104048ms step_avg:539.09ms step:2059/6250 train_loss:2.4206 aux_loss:0.0382 train_time:1104583ms step_avg:539.08ms step:2060/6250 train_loss:2.3811 aux_loss:0.0380 train_time:1105124ms step_avg:539.08ms step:2061/6250 train_loss:2.2994 aux_loss:0.0377 train_time:1105658ms step_avg:539.08ms step:2062/6250 train_loss:2.4446 aux_loss:0.0378 train_time:1106201ms step_avg:539.08ms step:2063/6250 train_loss:2.3313 aux_loss:0.0381 train_time:1106735ms step_avg:539.08ms step:2064/6250 train_loss:2.3227 aux_loss:0.0385 train_time:1107267ms step_avg:539.08ms step:2065/6250 train_loss:2.4221 aux_loss:0.0384 train_time:1107817ms step_avg:539.08ms step:2066/6250 train_loss:2.3538 aux_loss:0.0382 train_time:1108396ms step_avg:539.10ms step:2067/6250 train_loss:2.3759 aux_loss:0.0379 train_time:1108954ms step_avg:539.11ms step:2068/6250 train_loss:2.2947 aux_loss:0.0378 train_time:1109492ms step_avg:539.11ms step:2069/6250 train_loss:2.3191 aux_loss:0.0380 train_time:1110062ms step_avg:539.13ms step:2070/6250 train_loss:2.4334 aux_loss:0.0380 train_time:1110628ms step_avg:539.14ms step:2071/6250 train_loss:2.3998 aux_loss:0.0379 train_time:1111160ms step_avg:539.14ms step:2072/6250 train_loss:2.4035 aux_loss:0.0380 train_time:1111699ms step_avg:539.14ms step:2073/6250 train_loss:2.3377 aux_loss:0.0378 train_time:1112273ms step_avg:539.15ms step:2074/6250 train_loss:2.3698 aux_loss:0.0377 train_time:1112856ms step_avg:539.17ms step:2075/6250 train_loss:2.4240 aux_loss:0.0379 train_time:1113440ms step_avg:539.20ms step:2076/6250 train_loss:2.3815 aux_loss:0.0377 train_time:1114016ms step_avg:539.21ms step:2077/6250 train_loss:2.3960 aux_loss:0.0378 train_time:1114552ms step_avg:539.21ms step:2078/6250 train_loss:2.4337 aux_loss:0.0379 train_time:1115085ms step_avg:539.21ms step:2079/6250 train_loss:2.4267 aux_loss:0.0377 train_time:1115615ms step_avg:539.20ms step:2080/6250 train_loss:2.3196 aux_loss:0.0376 train_time:1116143ms step_avg:539.20ms step:2081/6250 train_loss:2.4002 aux_loss:0.0377 train_time:1116673ms step_avg:539.20ms step:2082/6250 train_loss:2.3430 aux_loss:0.0379 train_time:1117202ms step_avg:539.19ms step:2083/6250 train_loss:2.4554 aux_loss:0.0378 train_time:1117733ms step_avg:539.19ms step:2084/6250 train_loss:2.4064 aux_loss:0.0373 train_time:1118263ms step_avg:539.18ms step:2085/6250 train_loss:2.2922 aux_loss:0.0375 train_time:1118791ms step_avg:539.18ms step:2086/6250 train_loss:2.2941 aux_loss:0.0377 train_time:1119321ms step_avg:539.17ms step:2087/6250 train_loss:2.4584 aux_loss:0.0379 train_time:1119853ms step_avg:539.17ms step:2088/6250 train_loss:2.4058 aux_loss:0.0376 train_time:1120386ms step_avg:539.17ms step:2089/6250 train_loss:2.4237 aux_loss:0.0373 train_time:1120914ms step_avg:539.16ms step:2090/6250 train_loss:2.3094 aux_loss:0.0375 train_time:1121445ms step_avg:539.16ms step:2091/6250 train_loss:2.4420 aux_loss:0.0377 train_time:1121974ms step_avg:539.15ms step:2092/6250 train_loss:2.3953 aux_loss:0.0377 train_time:1122508ms step_avg:539.15ms step:2093/6250 train_loss:2.3312 aux_loss:0.0374 train_time:1123037ms step_avg:539.14ms step:2094/6250 train_loss:2.2366 aux_loss:0.0376 train_time:1123569ms step_avg:539.14ms step:2095/6250 train_loss:2.3669 aux_loss:0.0375 train_time:1124100ms step_avg:539.14ms step:2096/6250 train_loss:2.3858 aux_loss:0.0377 train_time:1124629ms step_avg:539.13ms step:2097/6250 train_loss:2.3389 aux_loss:0.0376 train_time:1125160ms step_avg:539.13ms step:2098/6250 train_loss:2.4025 aux_loss:0.0374 train_time:1125696ms step_avg:539.13ms step:2099/6250 train_loss:2.3298 aux_loss:0.0374 train_time:1126225ms step_avg:539.12ms step:2100/6250 train_loss:2.3834 aux_loss:0.0377 train_time:1126757ms step_avg:539.12ms step:2101/6250 train_loss:2.2836 aux_loss:0.0377 train_time:1127288ms step_avg:539.11ms step:2102/6250 train_loss:2.4518 aux_loss:0.0374 train_time:1127818ms step_avg:539.11ms step:2103/6250 train_loss:2.3024 aux_loss:0.0374 train_time:1128361ms step_avg:539.11ms step:2104/6250 train_loss:2.4073 aux_loss:0.0374 train_time:1128897ms step_avg:539.11ms step:2105/6250 train_loss:2.3900 aux_loss:0.0375 train_time:1129431ms step_avg:539.11ms step:2106/6250 train_loss:2.3424 aux_loss:0.0377 train_time:1129967ms step_avg:539.11ms step:2107/6250 train_loss:2.3049 aux_loss:0.0376 train_time:1130501ms step_avg:539.10ms step:2108/6250 train_loss:2.4055 aux_loss:0.0375 train_time:1131038ms step_avg:539.10ms step:2109/6250 train_loss:2.3842 aux_loss:0.0375 train_time:1131576ms step_avg:539.10ms step:2110/6250 train_loss:2.4255 aux_loss:0.0376 train_time:1132108ms step_avg:539.10ms step:2111/6250 train_loss:2.3650 aux_loss:0.0375 train_time:1132649ms step_avg:539.10ms step:2112/6250 train_loss:2.3235 aux_loss:0.0374 train_time:1133183ms step_avg:539.10ms step:2113/6250 train_loss:2.4138 aux_loss:0.0376 train_time:1133711ms step_avg:539.09ms step:2114/6250 train_loss:2.3157 aux_loss:0.0376 train_time:1134240ms step_avg:539.09ms step:2115/6250 train_loss:2.3916 aux_loss:0.0378 train_time:1134769ms step_avg:539.08ms step:2116/6250 train_loss:2.2846 aux_loss:0.0377 train_time:1135298ms step_avg:539.08ms step:2117/6250 train_loss:2.4516 aux_loss:0.0377 train_time:1135827ms step_avg:539.07ms step:2118/6250 train_loss:2.3584 aux_loss:0.0381 train_time:1136357ms step_avg:539.07ms step:2119/6250 train_loss:2.4013 aux_loss:0.0380 train_time:1136887ms step_avg:539.06ms step:2120/6250 train_loss:2.3310 aux_loss:0.0377 train_time:1137417ms step_avg:539.06ms step:2121/6250 train_loss:2.4290 aux_loss:0.0374 train_time:1137945ms step_avg:539.05ms step:2122/6250 train_loss:2.3379 aux_loss:0.0376 train_time:1138475ms step_avg:539.05ms step:2123/6250 train_loss:2.4308 aux_loss:0.0377 train_time:1139003ms step_avg:539.05ms step:2124/6250 train_loss:2.4367 aux_loss:0.0375 train_time:1139534ms step_avg:539.04ms step:2125/6250 train_loss:2.3688 aux_loss:0.0373 train_time:1140062ms step_avg:539.04ms step:2126/6250 train_loss:2.3197 aux_loss:0.0374 train_time:1140592ms step_avg:539.03ms step:2127/6250 train_loss:2.4022 aux_loss:0.0376 train_time:1141121ms step_avg:539.03ms step:2128/6250 train_loss:2.3558 aux_loss:0.0377 train_time:1141651ms step_avg:539.02ms step:2129/6250 train_loss:2.3830 aux_loss:0.0378 train_time:1142180ms step_avg:539.02ms step:2130/6250 train_loss:2.3245 aux_loss:0.0378 train_time:1142710ms step_avg:539.01ms step:2131/6250 train_loss:2.3524 aux_loss:0.0375 train_time:1143239ms step_avg:539.01ms step:2132/6250 train_loss:2.3470 aux_loss:0.0376 train_time:1143768ms step_avg:539.00ms step:2133/6250 train_loss:2.4426 aux_loss:0.0375 train_time:1144297ms step_avg:539.00ms step:2134/6250 train_loss:2.2662 aux_loss:0.0374 train_time:1144825ms step_avg:538.99ms step:2135/6250 train_loss:2.2617 aux_loss:0.0375 train_time:1145354ms step_avg:538.99ms step:2136/6250 train_loss:2.2174 aux_loss:0.0375 train_time:1145883ms step_avg:538.99ms step:2137/6250 train_loss:2.3805 aux_loss:0.0374 train_time:1146414ms step_avg:538.98ms step:2138/6250 train_loss:2.2533 aux_loss:0.0374 train_time:1146945ms step_avg:538.98ms step:2139/6250 train_loss:2.3924 aux_loss:0.0376 train_time:1147474ms step_avg:538.97ms step:2140/6250 train_loss:2.3893 aux_loss:0.0377 train_time:1148002ms step_avg:538.97ms step:2141/6250 train_loss:2.3605 aux_loss:0.0375 train_time:1148532ms step_avg:538.96ms step:2142/6250 train_loss:2.3459 aux_loss:0.0373 train_time:1149061ms step_avg:538.96ms step:2143/6250 train_loss:2.3038 aux_loss:0.0375 train_time:1149591ms step_avg:538.96ms step:2144/6250 train_loss:2.3339 aux_loss:0.0375 train_time:1150119ms step_avg:538.95ms step:2145/6250 train_loss:2.4086 aux_loss:0.0377 train_time:1150649ms step_avg:538.95ms step:2146/6250 train_loss:2.4514 aux_loss:0.0375 train_time:1151176ms step_avg:538.94ms step:2147/6250 train_loss:2.3551 aux_loss:0.0375 train_time:1151706ms step_avg:538.94ms step:2148/6250 train_loss:2.4062 aux_loss:0.0373 train_time:1152235ms step_avg:538.93ms step:2149/6250 train_loss:2.3259 aux_loss:0.0374 train_time:1152764ms step_avg:538.93ms step:2150/6250 train_loss:2.3449 aux_loss:0.0375 train_time:1153293ms step_avg:538.92ms step:2151/6250 train_loss:2.4357 aux_loss:0.0377 train_time:1153824ms step_avg:538.92ms step:2152/6250 train_loss:2.3601 aux_loss:0.0379 train_time:1154354ms step_avg:538.91ms step:2153/6250 train_loss:2.3524 aux_loss:0.0376 train_time:1154885ms step_avg:538.91ms step:2154/6250 train_loss:2.3432 aux_loss:0.0377 train_time:1155415ms step_avg:538.91ms step:2155/6250 train_loss:2.2586 aux_loss:0.0375 train_time:1155943ms step_avg:538.90ms step:2156/6250 train_loss:2.3236 aux_loss:0.0376 train_time:1156474ms step_avg:538.90ms step:2157/6250 train_loss:2.2507 aux_loss:0.0374 train_time:1157002ms step_avg:538.89ms step:2158/6250 train_loss:2.3574 aux_loss:0.0373 train_time:1157532ms step_avg:538.89ms step:2159/6250 train_loss:2.3940 aux_loss:0.0370 train_time:1158060ms step_avg:538.88ms step:2160/6250 train_loss:2.3505 aux_loss:0.0372 train_time:1158590ms step_avg:538.88ms step:2161/6250 train_loss:2.2915 aux_loss:0.0371 train_time:1159119ms step_avg:538.87ms step:2162/6250 train_loss:2.3339 aux_loss:0.0371 train_time:1159648ms step_avg:538.87ms step:2163/6250 train_loss:2.4528 aux_loss:0.0368 train_time:1160178ms step_avg:538.87ms step:2164/6250 train_loss:2.2344 aux_loss:0.0371 train_time:1160708ms step_avg:538.86ms step:2165/6250 train_loss:2.3169 aux_loss:0.0374 train_time:1161236ms step_avg:538.86ms step:2166/6250 train_loss:2.3345 aux_loss:0.0372 train_time:1161766ms step_avg:538.85ms step:2167/6250 train_loss:2.3919 aux_loss:0.0371 train_time:1162293ms step_avg:538.85ms step:2168/6250 train_loss:2.3326 aux_loss:0.0372 train_time:1162824ms step_avg:538.84ms step:2169/6250 train_loss:2.3394 aux_loss:0.0374 train_time:1163363ms step_avg:538.84ms step:2170/6250 train_loss:2.3421 aux_loss:0.0373 train_time:1163896ms step_avg:538.84ms step:2171/6250 train_loss:2.4276 aux_loss:0.0370 train_time:1164437ms step_avg:538.84ms step:2172/6250 train_loss:2.4127 aux_loss:0.0369 train_time:1164987ms step_avg:538.85ms step:2173/6250 train_loss:2.3675 aux_loss:0.0372 train_time:1165532ms step_avg:538.85ms step:2174/6250 train_loss:2.3618 aux_loss:0.0371 train_time:1166068ms step_avg:538.85ms step:2175/6250 train_loss:2.2793 aux_loss:0.0368 train_time:1166612ms step_avg:538.85ms step:2176/6250 train_loss:2.4141 aux_loss:0.0369 train_time:1167147ms step_avg:538.85ms step:2177/6250 train_loss:2.3774 aux_loss:0.0372 train_time:1167679ms step_avg:538.85ms step:2178/6250 train_loss:2.3893 aux_loss:0.0374 train_time:1168211ms step_avg:538.84ms step:2179/6250 train_loss:2.3606 aux_loss:0.0371 train_time:1168750ms step_avg:538.84ms step:2180/6250 train_loss:2.4086 aux_loss:0.0373 train_time:1169296ms step_avg:538.85ms step:2181/6250 train_loss:2.2853 aux_loss:0.0369 train_time:1169849ms step_avg:538.85ms step:2182/6250 train_loss:2.3282 aux_loss:0.0372 train_time:1170395ms step_avg:538.86ms step:2183/6250 train_loss:2.3122 aux_loss:0.0372 train_time:1170952ms step_avg:538.86ms step:2184/6250 train_loss:2.3961 aux_loss:0.0375 train_time:1171507ms step_avg:538.87ms step:2185/6250 train_loss:2.3605 aux_loss:0.0374 train_time:1172036ms step_avg:538.87ms step:2186/6250 train_loss:2.4285 aux_loss:0.0371 train_time:1172575ms step_avg:538.87ms step:2187/6250 train_loss:2.3854 aux_loss:0.0374 train_time:1173123ms step_avg:538.87ms step:2188/6250 train_loss:2.3665 aux_loss:0.0372 train_time:1173672ms step_avg:538.88ms step:2189/6250 train_loss:2.4280 aux_loss:0.0374 train_time:1174206ms step_avg:538.87ms step:2190/6250 train_loss:2.4036 aux_loss:0.0374 train_time:1174735ms step_avg:538.87ms step:2191/6250 train_loss:2.3132 aux_loss:0.0373 train_time:1175264ms step_avg:538.86ms step:2192/6250 train_loss:2.2558 aux_loss:0.0373 train_time:1175793ms step_avg:538.86ms step:2193/6250 train_loss:2.4173 aux_loss:0.0372 train_time:1176322ms step_avg:538.86ms step:2194/6250 train_loss:2.4428 aux_loss:0.0375 train_time:1176851ms step_avg:538.85ms step:2195/6250 train_loss:2.2815 aux_loss:0.0374 train_time:1177380ms step_avg:538.85ms step:2196/6250 train_loss:2.4602 aux_loss:0.0376 train_time:1177909ms step_avg:538.84ms step:2197/6250 train_loss:2.3616 aux_loss:0.0374 train_time:1178453ms step_avg:538.84ms step:2198/6250 train_loss:2.4307 aux_loss:0.0373 train_time:1178998ms step_avg:538.85ms step:2199/6250 train_loss:2.3504 aux_loss:0.0372 train_time:1179529ms step_avg:538.84ms step:2200/6250 train_loss:2.3493 aux_loss:0.0370 train_time:1180065ms step_avg:538.84ms step:2201/6250 train_loss:2.3246 aux_loss:0.0369 train_time:1180601ms step_avg:538.84ms step:2202/6250 train_loss:2.3174 aux_loss:0.0372 train_time:1181129ms step_avg:538.84ms step:2203/6250 train_loss:2.3448 aux_loss:0.0371 train_time:1181658ms step_avg:538.83ms step:2204/6250 train_loss:2.3518 aux_loss:0.0369 train_time:1182186ms step_avg:538.83ms step:2205/6250 train_loss:2.4038 aux_loss:0.0373 train_time:1182714ms step_avg:538.82ms step:2206/6250 train_loss:2.3878 aux_loss:0.0371 train_time:1183242ms step_avg:538.82ms step:2207/6250 train_loss:2.4294 aux_loss:0.0368 train_time:1183772ms step_avg:538.81ms step:2208/6250 train_loss:2.2232 aux_loss:0.0365 train_time:1184300ms step_avg:538.81ms step:2209/6250 train_loss:2.3942 aux_loss:0.0368 train_time:1184830ms step_avg:538.80ms step:2210/6250 train_loss:2.3385 aux_loss:0.0365 train_time:1185359ms step_avg:538.80ms step:2211/6250 train_loss:2.3385 aux_loss:0.0368 train_time:1185889ms step_avg:538.80ms step:2212/6250 train_loss:2.2288 aux_loss:0.0369 train_time:1186420ms step_avg:538.79ms step:2213/6250 train_loss:2.2822 aux_loss:0.0370 train_time:1186959ms step_avg:538.79ms step:2214/6250 train_loss:2.2796 aux_loss:0.0365 train_time:1187491ms step_avg:538.79ms step:2215/6250 train_loss:2.3218 aux_loss:0.0370 train_time:1188020ms step_avg:538.78ms step:2216/6250 train_loss:2.3489 aux_loss:0.0371 train_time:1188561ms step_avg:538.79ms step:2217/6250 train_loss:2.3542 aux_loss:0.0371 train_time:1189089ms step_avg:538.78ms step:2218/6250 train_loss:2.3204 aux_loss:0.0369 train_time:1189620ms step_avg:538.78ms step:2219/6250 train_loss:2.3413 aux_loss:0.0369 train_time:1190148ms step_avg:538.77ms step:2220/6250 train_loss:2.2829 aux_loss:0.0365 train_time:1190679ms step_avg:538.77ms step:2221/6250 train_loss:2.4232 aux_loss:0.0367 train_time:1191206ms step_avg:538.76ms step:2222/6250 train_loss:2.4144 aux_loss:0.0373 train_time:1191735ms step_avg:538.76ms step:2223/6250 train_loss:2.2779 aux_loss:0.0372 train_time:1192264ms step_avg:538.75ms step:2224/6250 train_loss:2.3485 aux_loss:0.0367 train_time:1192793ms step_avg:538.75ms step:2225/6250 train_loss:2.3398 aux_loss:0.0367 train_time:1193321ms step_avg:538.75ms step:2226/6250 train_loss:2.3957 aux_loss:0.0371 train_time:1193850ms step_avg:538.74ms step:2227/6250 train_loss:2.4003 aux_loss:0.0372 train_time:1194380ms step_avg:538.74ms step:2228/6250 train_loss:2.3479 aux_loss:0.0370 train_time:1194908ms step_avg:538.73ms step:2229/6250 train_loss:2.3742 aux_loss:0.0370 train_time:1195438ms step_avg:538.73ms step:2230/6250 train_loss:2.3063 aux_loss:0.0369 train_time:1195968ms step_avg:538.72ms step:2231/6250 train_loss:2.5754 aux_loss:0.0372 train_time:1196498ms step_avg:538.72ms step:2232/6250 train_loss:2.4294 aux_loss:0.0373 train_time:1197026ms step_avg:538.72ms step:2233/6250 train_loss:2.3377 aux_loss:0.0367 train_time:1197556ms step_avg:538.71ms step:2234/6250 train_loss:2.3226 aux_loss:0.0369 train_time:1198083ms step_avg:538.71ms step:2235/6250 train_loss:2.3595 aux_loss:0.0369 train_time:1198612ms step_avg:538.70ms step:2236/6250 train_loss:2.4068 aux_loss:0.0369 train_time:1199142ms step_avg:538.70ms step:2237/6250 train_loss:2.3915 aux_loss:0.0367 train_time:1199671ms step_avg:538.69ms step:2238/6250 train_loss:2.3874 aux_loss:0.0370 train_time:1200201ms step_avg:538.69ms step:2239/6250 train_loss:2.2572 aux_loss:0.0370 train_time:1200730ms step_avg:538.69ms step:2240/6250 train_loss:2.2830 aux_loss:0.0370 train_time:1201258ms step_avg:538.68ms step:2241/6250 train_loss:2.2692 aux_loss:0.0372 train_time:1201787ms step_avg:538.68ms step:2242/6250 train_loss:2.3577 aux_loss:0.0369 train_time:1202318ms step_avg:538.67ms step:2243/6250 train_loss:2.3457 aux_loss:0.0370 train_time:1202846ms step_avg:538.67ms step:2244/6250 train_loss:2.4032 aux_loss:0.0370 train_time:1203375ms step_avg:538.66ms step:2245/6250 train_loss:2.3084 aux_loss:0.0369 train_time:1203902ms step_avg:538.66ms step:2246/6250 train_loss:2.3601 aux_loss:0.0370 train_time:1204433ms step_avg:538.66ms step:2247/6250 train_loss:2.3621 aux_loss:0.0369 train_time:1204973ms step_avg:538.66ms step:2248/6250 train_loss:2.3002 aux_loss:0.0371 train_time:1205505ms step_avg:538.65ms step:2249/6250 train_loss:2.4462 aux_loss:0.0368 train_time:1206033ms step_avg:538.65ms step:2250/6250 train_loss:2.2816 aux_loss:0.0365 train_time:1206567ms step_avg:538.65ms step:2251/6250 train_loss:2.3280 aux_loss:0.0370 train_time:1207110ms step_avg:538.65ms step:2252/6250 train_loss:2.3768 aux_loss:0.0374 train_time:1207654ms step_avg:538.65ms step:2253/6250 train_loss:2.4106 aux_loss:0.0373 train_time:1208183ms step_avg:538.65ms step:2254/6250 train_loss:2.3201 aux_loss:0.0370 train_time:1208718ms step_avg:538.64ms step:2255/6250 train_loss:2.3539 aux_loss:0.0369 train_time:1209259ms step_avg:538.65ms step:2256/6250 train_loss:2.3265 aux_loss:0.0370 train_time:1209789ms step_avg:538.64ms step:2257/6250 train_loss:2.3138 aux_loss:0.0371 train_time:1210334ms step_avg:538.64ms step:2258/6250 train_loss:2.2544 aux_loss:0.0367 train_time:1210864ms step_avg:538.64ms step:2259/6250 train_loss:2.3850 aux_loss:0.0367 train_time:1211397ms step_avg:538.64ms step:2260/6250 train_loss:2.2624 aux_loss:0.0370 train_time:1211930ms step_avg:538.64ms step:2261/6250 train_loss:2.2685 aux_loss:0.0370 train_time:1212460ms step_avg:538.63ms step:2262/6250 train_loss:2.3049 aux_loss:0.0366 train_time:1212987ms step_avg:538.63ms step:2263/6250 train_loss:2.2672 aux_loss:0.0365 train_time:1213517ms step_avg:538.62ms step:2264/6250 train_loss:2.3444 aux_loss:0.0366 train_time:1214046ms step_avg:538.62ms step:2265/6250 train_loss:2.3476 aux_loss:0.0366 train_time:1214576ms step_avg:538.61ms step:2266/6250 train_loss:2.2838 aux_loss:0.0365 train_time:1215104ms step_avg:538.61ms step:2267/6250 train_loss:2.3172 aux_loss:0.0367 train_time:1215634ms step_avg:538.61ms step:2268/6250 train_loss:2.3416 aux_loss:0.0366 train_time:1216163ms step_avg:538.60ms step:2269/6250 train_loss:2.5442 aux_loss:0.0366 train_time:1216692ms step_avg:538.60ms step:2270/6250 train_loss:2.3488 aux_loss:0.0366 train_time:1217221ms step_avg:538.59ms step:2271/6250 train_loss:2.3634 aux_loss:0.0371 train_time:1217750ms step_avg:538.59ms step:2272/6250 train_loss:2.2863 aux_loss:0.0369 train_time:1218278ms step_avg:538.58ms step:2273/6250 train_loss:2.3653 aux_loss:0.0369 train_time:1218808ms step_avg:538.58ms step:2274/6250 train_loss:2.3496 aux_loss:0.0367 train_time:1219337ms step_avg:538.58ms step:2275/6250 train_loss:2.3378 aux_loss:0.0370 train_time:1219866ms step_avg:538.57ms step:2276/6250 train_loss:2.4889 aux_loss:0.0370 train_time:1220397ms step_avg:538.57ms step:2277/6250 train_loss:2.3130 aux_loss:0.0368 train_time:1220926ms step_avg:538.56ms step:2278/6250 train_loss:2.3595 aux_loss:0.0366 train_time:1221456ms step_avg:538.56ms step:2279/6250 train_loss:2.2981 aux_loss:0.0368 train_time:1221986ms step_avg:538.56ms step:2280/6250 train_loss:2.3402 aux_loss:0.0368 train_time:1222515ms step_avg:538.55ms step:2281/6250 train_loss:2.3574 aux_loss:0.0364 train_time:1223044ms step_avg:538.55ms step:2282/6250 train_loss:2.3342 aux_loss:0.0365 train_time:1223574ms step_avg:538.54ms step:2283/6250 train_loss:2.2524 aux_loss:0.0365 train_time:1224102ms step_avg:538.54ms step:2284/6250 train_loss:2.5085 aux_loss:0.0367 train_time:1224632ms step_avg:538.54ms step:2285/6250 train_loss:2.3543 aux_loss:0.0366 train_time:1225160ms step_avg:538.53ms step:2286/6250 train_loss:2.3751 aux_loss:0.0367 train_time:1225690ms step_avg:538.53ms step:2287/6250 train_loss:2.3527 aux_loss:0.0367 train_time:1226218ms step_avg:538.52ms step:2288/6250 train_loss:2.3304 aux_loss:0.0368 train_time:1226747ms step_avg:538.52ms step:2289/6250 train_loss:2.2980 aux_loss:0.0366 train_time:1227276ms step_avg:538.52ms step:2290/6250 train_loss:2.4405 aux_loss:0.0364 train_time:1227804ms step_avg:538.51ms step:2291/6250 train_loss:2.3902 aux_loss:0.0369 train_time:1228333ms step_avg:538.51ms step:2292/6250 train_loss:2.3744 aux_loss:0.0368 train_time:1228862ms step_avg:538.50ms step:2293/6250 train_loss:2.3531 aux_loss:0.0366 train_time:1229391ms step_avg:538.50ms step:2294/6250 train_loss:2.2951 aux_loss:0.0362 train_time:1229920ms step_avg:538.49ms step:2295/6250 train_loss:2.3951 aux_loss:0.0364 train_time:1230450ms step_avg:538.49ms step:2296/6250 train_loss:2.4146 aux_loss:0.0369 train_time:1230980ms step_avg:538.49ms step:2297/6250 train_loss:2.4008 aux_loss:0.0368 train_time:1231511ms step_avg:538.48ms step:2298/6250 train_loss:2.3513 aux_loss:0.0365 train_time:1232039ms step_avg:538.48ms step:2299/6250 train_loss:2.3817 aux_loss:0.0364 train_time:1232569ms step_avg:538.47ms step:2300/6250 train_loss:2.4073 aux_loss:0.0364 train_time:1233097ms step_avg:538.47ms step:2301/6250 train_loss:2.3631 aux_loss:0.0366 train_time:1233627ms step_avg:538.47ms step:2302/6250 train_loss:2.2891 aux_loss:0.0369 train_time:1234155ms step_avg:538.46ms step:2303/6250 train_loss:2.3140 aux_loss:0.0364 train_time:1234685ms step_avg:538.46ms step:2304/6250 train_loss:2.2957 aux_loss:0.0365 train_time:1235213ms step_avg:538.45ms step:2305/6250 train_loss:2.3700 aux_loss:0.0367 train_time:1235741ms step_avg:538.45ms step:2306/6250 train_loss:2.2965 aux_loss:0.0370 train_time:1236269ms step_avg:538.44ms step:2307/6250 train_loss:2.3011 aux_loss:0.0364 train_time:1236799ms step_avg:538.44ms step:2308/6250 train_loss:2.3148 aux_loss:0.0364 train_time:1237329ms step_avg:538.44ms step:2309/6250 train_loss:2.3037 aux_loss:0.0365 train_time:1237857ms step_avg:538.43ms step:2310/6250 train_loss:2.2876 aux_loss:0.0368 train_time:1238386ms step_avg:538.43ms step:2311/6250 train_loss:2.3217 aux_loss:0.0369 train_time:1238914ms step_avg:538.42ms step:2312/6250 train_loss:2.3072 aux_loss:0.0366 train_time:1239443ms step_avg:538.42ms step:2313/6250 train_loss:2.3737 aux_loss:0.0365 train_time:1239972ms step_avg:538.42ms step:2314/6250 train_loss:2.2929 aux_loss:0.0365 train_time:1240503ms step_avg:538.41ms step:2315/6250 train_loss:2.4132 aux_loss:0.0368 train_time:1241031ms step_avg:538.41ms step:2316/6250 train_loss:2.3830 aux_loss:0.0364 train_time:1241559ms step_avg:538.40ms step:2317/6250 train_loss:2.4033 aux_loss:0.0364 train_time:1242087ms step_avg:538.40ms step:2318/6250 train_loss:2.3280 aux_loss:0.0363 train_time:1242617ms step_avg:538.40ms step:2319/6250 train_loss:2.3554 aux_loss:0.0362 train_time:1243146ms step_avg:538.39ms step:2320/6250 train_loss:2.2870 aux_loss:0.0360 train_time:1243697ms step_avg:538.40ms step:2321/6250 train_loss:2.3026 aux_loss:0.0359 train_time:1244258ms step_avg:538.41ms step:2322/6250 train_loss:2.3285 aux_loss:0.0363 train_time:1244795ms step_avg:538.41ms step:2323/6250 train_loss:2.3021 aux_loss:0.0364 train_time:1245324ms step_avg:538.40ms step:2324/6250 train_loss:2.3691 aux_loss:0.0363 train_time:1245852ms step_avg:538.40ms step:2325/6250 train_loss:2.3883 aux_loss:0.0364 train_time:1246381ms step_avg:538.39ms step:2326/6250 train_loss:2.3869 aux_loss:0.0367 train_time:1246909ms step_avg:538.39ms step:2327/6250 train_loss:2.3743 aux_loss:0.0370 train_time:1247447ms step_avg:538.39ms step:2328/6250 train_loss:2.3344 aux_loss:0.0364 train_time:1247990ms step_avg:538.39ms step:2329/6250 train_loss:2.3644 aux_loss:0.0360 train_time:1248523ms step_avg:538.39ms step:2330/6250 train_loss:2.3338 aux_loss:0.0363 train_time:1249056ms step_avg:538.39ms step:2331/6250 train_loss:2.2710 aux_loss:0.0367 train_time:1249591ms step_avg:538.38ms step:2332/6250 train_loss:2.4181 aux_loss:0.0366 train_time:1250129ms step_avg:538.38ms step:2333/6250 train_loss:2.3016 aux_loss:0.0364 train_time:1250670ms step_avg:538.39ms step:2334/6250 train_loss:2.2734 aux_loss:0.0361 train_time:1251214ms step_avg:538.39ms step:2335/6250 train_loss:2.3405 aux_loss:0.0363 train_time:1251758ms step_avg:538.39ms step:2336/6250 train_loss:2.3447 aux_loss:0.0364 train_time:1252317ms step_avg:538.40ms step:2337/6250 train_loss:2.3509 aux_loss:0.0366 train_time:1252853ms step_avg:538.40ms step:2338/6250 train_loss:2.3462 aux_loss:0.0363 train_time:1253384ms step_avg:538.40ms step:2339/6250 train_loss:2.3774 aux_loss:0.0364 train_time:1253916ms step_avg:538.39ms step:2340/6250 train_loss:2.3393 aux_loss:0.0367 train_time:1254476ms step_avg:538.40ms step:2341/6250 train_loss:2.3490 aux_loss:0.0366 train_time:1255060ms step_avg:538.42ms step:2342/6250 train_loss:2.3293 aux_loss:0.0365 train_time:1255644ms step_avg:538.44ms step:2343/6250 train_loss:2.3800 aux_loss:0.0361 train_time:1256229ms step_avg:538.46ms step:2344/6250 train_loss:2.3386 aux_loss:0.0362 train_time:1256814ms step_avg:538.48ms step:2345/6250 train_loss:2.3039 aux_loss:0.0366 train_time:1257397ms step_avg:538.50ms step:2346/6250 train_loss:2.4756 aux_loss:0.0365 train_time:1257984ms step_avg:538.52ms step:2347/6250 train_loss:2.2625 aux_loss:0.0363 train_time:1258568ms step_avg:538.54ms step:2348/6250 train_loss:2.4398 aux_loss:0.0365 train_time:1259153ms step_avg:538.56ms step:2349/6250 train_loss:2.4031 aux_loss:0.0366 train_time:1259738ms step_avg:538.58ms step:2350/6250 train_loss:2.2433 aux_loss:0.0367 train_time:1260313ms step_avg:538.60ms step:2351/6250 train_loss:2.2561 aux_loss:0.0364 train_time:1260850ms step_avg:538.59ms step:2352/6250 train_loss:2.3839 aux_loss:0.0366 train_time:1261400ms step_avg:538.60ms step:2353/6250 train_loss:2.3276 aux_loss:0.0363 train_time:1261949ms step_avg:538.60ms step:2354/6250 train_loss:2.3535 aux_loss:0.0362 train_time:1262491ms step_avg:538.61ms step:2355/6250 train_loss:2.3379 aux_loss:0.0366 train_time:1263021ms step_avg:538.60ms step:2356/6250 train_loss:2.3682 aux_loss:0.0364 train_time:1263572ms step_avg:538.61ms step:2357/6250 train_loss:2.4035 aux_loss:0.0362 train_time:1264119ms step_avg:538.61ms step:2358/6250 train_loss:2.2824 aux_loss:0.0360 train_time:1264663ms step_avg:538.61ms step:2359/6250 train_loss:2.3604 aux_loss:0.0363 train_time:1265220ms step_avg:538.62ms step:2360/6250 train_loss:2.2396 aux_loss:0.0363 train_time:1265759ms step_avg:538.62ms step:2361/6250 train_loss:2.4281 aux_loss:0.0362 train_time:1266289ms step_avg:538.62ms step:2362/6250 train_loss:2.3178 aux_loss:0.0361 train_time:1266832ms step_avg:538.62ms step:2363/6250 train_loss:2.4426 aux_loss:0.0362 train_time:1267377ms step_avg:538.62ms step:2364/6250 train_loss:2.3178 aux_loss:0.0363 train_time:1267929ms step_avg:538.63ms step:2365/6250 train_loss:2.2974 aux_loss:0.0360 train_time:1268470ms step_avg:538.63ms step:2366/6250 train_loss:2.4117 aux_loss:0.0361 train_time:1269010ms step_avg:538.63ms step:2367/6250 train_loss:2.3622 aux_loss:0.0360 train_time:1269542ms step_avg:538.63ms step:2368/6250 train_loss:2.3424 aux_loss:0.0360 train_time:1270074ms step_avg:538.62ms step:2369/6250 train_loss:2.3407 aux_loss:0.0364 train_time:1270604ms step_avg:538.62ms step:2370/6250 train_loss:2.2841 aux_loss:0.0364 train_time:1271134ms step_avg:538.62ms step:2371/6250 train_loss:2.2854 aux_loss:0.0364 train_time:1271668ms step_avg:538.61ms step:2372/6250 train_loss:2.3407 aux_loss:0.0359 train_time:1272201ms step_avg:538.61ms step:2373/6250 train_loss:2.3884 aux_loss:0.0362 train_time:1272735ms step_avg:538.61ms step:2374/6250 train_loss:2.2969 aux_loss:0.0366 train_time:1273267ms step_avg:538.61ms step:2375/6250 train_loss:2.2925 aux_loss:0.0364 train_time:1273801ms step_avg:538.61ms step:2376/6250 train_loss:2.2967 aux_loss:0.0360 train_time:1274333ms step_avg:538.60ms step:2377/6250 train_loss:2.3196 aux_loss:0.0363 train_time:1274871ms step_avg:538.60ms step:2378/6250 train_loss:2.2083 aux_loss:0.0363 train_time:1275407ms step_avg:538.60ms step:2379/6250 train_loss:2.3710 aux_loss:0.0365 train_time:1275939ms step_avg:538.60ms step:2380/6250 train_loss:2.3400 aux_loss:0.0361 train_time:1276474ms step_avg:538.60ms step:2381/6250 train_loss:2.3132 aux_loss:0.0360 train_time:1277006ms step_avg:538.59ms step:2382/6250 train_loss:2.3000 aux_loss:0.0361 train_time:1277540ms step_avg:538.59ms step:2383/6250 train_loss:2.3986 aux_loss:0.0363 train_time:1278070ms step_avg:538.59ms step:2384/6250 train_loss:2.3207 aux_loss:0.0360 train_time:1278601ms step_avg:538.59ms step:2385/6250 train_loss:2.3829 aux_loss:0.0361 train_time:1279134ms step_avg:538.58ms step:2386/6250 train_loss:2.4072 aux_loss:0.0363 train_time:1279665ms step_avg:538.58ms step:2387/6250 train_loss:2.3627 aux_loss:0.0367 train_time:1280207ms step_avg:538.58ms step:2388/6250 train_loss:2.3067 aux_loss:0.0366 train_time:1280739ms step_avg:538.58ms step:2389/6250 train_loss:2.3799 aux_loss:0.0363 train_time:1281269ms step_avg:538.57ms step:2390/6250 train_loss:2.3462 aux_loss:0.0358 train_time:1281800ms step_avg:538.57ms step:2391/6250 train_loss:2.2902 aux_loss:0.0360 train_time:1282335ms step_avg:538.57ms step:2392/6250 train_loss:2.3741 aux_loss:0.0361 train_time:1282866ms step_avg:538.57ms step:2393/6250 train_loss:2.2717 aux_loss:0.0359 train_time:1283399ms step_avg:538.56ms step:2394/6250 train_loss:2.3382 aux_loss:0.0355 train_time:1283929ms step_avg:538.56ms step:2395/6250 train_loss:2.3913 aux_loss:0.0356 train_time:1284461ms step_avg:538.56ms step:2396/6250 train_loss:2.4426 aux_loss:0.0360 train_time:1284992ms step_avg:538.55ms step:2397/6250 train_loss:2.2705 aux_loss:0.0358 train_time:1285523ms step_avg:538.55ms step:2398/6250 train_loss:2.3680 aux_loss:0.0357 train_time:1286055ms step_avg:538.55ms step:2399/6250 train_loss:2.4574 aux_loss:0.0356 train_time:1286587ms step_avg:538.55ms step:2400/6250 train_loss:2.3572 aux_loss:0.0357 train_time:1287118ms step_avg:538.54ms step:2401/6250 train_loss:2.3690 aux_loss:0.0362 train_time:1287647ms step_avg:538.54ms step:2402/6250 train_loss:2.3403 aux_loss:0.0362 train_time:1288179ms step_avg:538.54ms step:2403/6250 train_loss:2.3406 aux_loss:0.0359 train_time:1288710ms step_avg:538.53ms step:2404/6250 train_loss:2.3645 aux_loss:0.0357 train_time:1289242ms step_avg:538.53ms step:2405/6250 train_loss:2.3305 aux_loss:0.0359 train_time:1289775ms step_avg:538.53ms step:2406/6250 train_loss:2.3572 aux_loss:0.0360 train_time:1290305ms step_avg:538.52ms step:2407/6250 train_loss:2.3285 aux_loss:0.0358 train_time:1290836ms step_avg:538.52ms step:2408/6250 train_loss:2.3921 aux_loss:0.0357 train_time:1291366ms step_avg:538.52ms step:2409/6250 train_loss:2.2645 aux_loss:0.0354 train_time:1291894ms step_avg:538.51ms step:2410/6250 train_loss:2.4341 aux_loss:0.0357 train_time:1292423ms step_avg:538.51ms step:2411/6250 train_loss:2.3863 aux_loss:0.0358 train_time:1292953ms step_avg:538.51ms step:2412/6250 train_loss:2.3889 aux_loss:0.0360 train_time:1293485ms step_avg:538.50ms step:2413/6250 train_loss:2.4014 aux_loss:0.0358 train_time:1294016ms step_avg:538.50ms step:2414/6250 train_loss:2.2434 aux_loss:0.0359 train_time:1294548ms step_avg:538.50ms step:2415/6250 train_loss:2.3298 aux_loss:0.0361 train_time:1295077ms step_avg:538.49ms step:2416/6250 train_loss:2.3246 aux_loss:0.0359 train_time:1295610ms step_avg:538.49ms step:2417/6250 train_loss:2.4558 aux_loss:0.0360 train_time:1296143ms step_avg:538.49ms step:2418/6250 train_loss:2.3270 aux_loss:0.0359 train_time:1296673ms step_avg:538.49ms step:2419/6250 train_loss:2.3356 aux_loss:0.0358 train_time:1297202ms step_avg:538.48ms step:2420/6250 train_loss:2.2011 aux_loss:0.0361 train_time:1297732ms step_avg:538.48ms step:2421/6250 train_loss:2.3094 aux_loss:0.0360 train_time:1298261ms step_avg:538.47ms step:2422/6250 train_loss:2.3184 aux_loss:0.0359 train_time:1298791ms step_avg:538.47ms step:2423/6250 train_loss:2.2995 aux_loss:0.0359 train_time:1299321ms step_avg:538.47ms step:2424/6250 train_loss:2.3369 aux_loss:0.0357 train_time:1299851ms step_avg:538.46ms step:2425/6250 train_loss:2.3421 aux_loss:0.0360 train_time:1300382ms step_avg:538.46ms step:2426/6250 train_loss:2.3571 aux_loss:0.0359 train_time:1300913ms step_avg:538.46ms step:2427/6250 train_loss:2.3118 aux_loss:0.0361 train_time:1301443ms step_avg:538.45ms step:2428/6250 train_loss:2.3818 aux_loss:0.0359 train_time:1301973ms step_avg:538.45ms step:2429/6250 train_loss:2.2561 aux_loss:0.0360 train_time:1302504ms step_avg:538.45ms step:2430/6250 train_loss:2.2866 aux_loss:0.0359 train_time:1303034ms step_avg:538.44ms step:2431/6250 train_loss:2.3407 aux_loss:0.0357 train_time:1303566ms step_avg:538.44ms step:2432/6250 train_loss:2.3025 aux_loss:0.0357 train_time:1304096ms step_avg:538.44ms step:2433/6250 train_loss:2.3220 aux_loss:0.0358 train_time:1304625ms step_avg:538.43ms step:2434/6250 train_loss:2.2897 aux_loss:0.0356 train_time:1305156ms step_avg:538.43ms step:2435/6250 train_loss:2.4282 aux_loss:0.0356 train_time:1305684ms step_avg:538.43ms step:2436/6250 train_loss:2.3581 aux_loss:0.0359 train_time:1306214ms step_avg:538.42ms step:2437/6250 train_loss:2.3029 aux_loss:0.0360 train_time:1306746ms step_avg:538.42ms step:2438/6250 train_loss:2.3632 aux_loss:0.0359 train_time:1307279ms step_avg:538.42ms step:2439/6250 train_loss:2.3507 aux_loss:0.0358 train_time:1307812ms step_avg:538.42ms step:2440/6250 train_loss:2.2680 aux_loss:0.0358 train_time:1308341ms step_avg:538.41ms step:2441/6250 train_loss:2.2799 aux_loss:0.0357 train_time:1308871ms step_avg:538.41ms step:2442/6250 train_loss:2.2602 aux_loss:0.0359 train_time:1309403ms step_avg:538.41ms step:2443/6250 train_loss:2.3278 aux_loss:0.0357 train_time:1309935ms step_avg:538.40ms step:2444/6250 train_loss:2.4377 aux_loss:0.0359 train_time:1310466ms step_avg:538.40ms step:2445/6250 train_loss:2.2902 aux_loss:0.0359 train_time:1310995ms step_avg:538.40ms step:2446/6250 train_loss:2.4144 aux_loss:0.0357 train_time:1311528ms step_avg:538.39ms step:2447/6250 train_loss:2.3284 aux_loss:0.0357 train_time:1312060ms step_avg:538.39ms step:2448/6250 train_loss:2.3590 aux_loss:0.0358 train_time:1312592ms step_avg:538.39ms step:2449/6250 train_loss:2.3477 aux_loss:0.0356 train_time:1313123ms step_avg:538.39ms step:2450/6250 train_loss:2.2964 aux_loss:0.0357 train_time:1313654ms step_avg:538.38ms step:2451/6250 train_loss:2.3023 aux_loss:0.0356 train_time:1314184ms step_avg:538.38ms step:2452/6250 train_loss:2.2705 aux_loss:0.0357 train_time:1314713ms step_avg:538.38ms step:2453/6250 train_loss:2.2588 aux_loss:0.0356 train_time:1315242ms step_avg:538.37ms step:2454/6250 train_loss:2.3949 aux_loss:0.0358 train_time:1315772ms step_avg:538.37ms step:2455/6250 train_loss:2.3510 aux_loss:0.0358 train_time:1316302ms step_avg:538.37ms step:2456/6250 train_loss:2.3259 aux_loss:0.0357 train_time:1316833ms step_avg:538.36ms step:2457/6250 train_loss:2.2934 aux_loss:0.0361 train_time:1317363ms step_avg:538.36ms step:2458/6250 train_loss:2.3348 aux_loss:0.0360 train_time:1317892ms step_avg:538.35ms step:2459/6250 train_loss:2.2531 aux_loss:0.0358 train_time:1318423ms step_avg:538.35ms step:2460/6250 train_loss:2.3521 aux_loss:0.0359 train_time:1318953ms step_avg:538.35ms step:2461/6250 train_loss:2.3191 aux_loss:0.0360 train_time:1319484ms step_avg:538.35ms step:2462/6250 train_loss:2.3222 aux_loss:0.0361 train_time:1320016ms step_avg:538.34ms step:2463/6250 train_loss:2.3708 aux_loss:0.0357 train_time:1320548ms step_avg:538.34ms step:2464/6250 train_loss:2.3428 aux_loss:0.0356 train_time:1321079ms step_avg:538.34ms step:2465/6250 train_loss:2.3192 aux_loss:0.0355 train_time:1321610ms step_avg:538.33ms step:2466/6250 train_loss:2.3337 aux_loss:0.0352 train_time:1322140ms step_avg:538.33ms step:2467/6250 train_loss:2.2840 aux_loss:0.0354 train_time:1322671ms step_avg:538.33ms step:2468/6250 train_loss:2.3472 aux_loss:0.0355 train_time:1323201ms step_avg:538.32ms step:2469/6250 train_loss:2.3098 aux_loss:0.0353 train_time:1323732ms step_avg:538.32ms step:2470/6250 train_loss:2.2953 aux_loss:0.0358 train_time:1324261ms step_avg:538.32ms step:2471/6250 train_loss:2.3588 aux_loss:0.0356 train_time:1324792ms step_avg:538.31ms step:2472/6250 train_loss:2.4013 aux_loss:0.0353 train_time:1325323ms step_avg:538.31ms step:2473/6250 train_loss:2.3711 aux_loss:0.0355 train_time:1325854ms step_avg:538.31ms step:2474/6250 train_loss:2.3752 aux_loss:0.0358 train_time:1326385ms step_avg:538.31ms step:2475/6250 train_loss:2.2899 aux_loss:0.0357 train_time:1326917ms step_avg:538.30ms step:2476/6250 train_loss:2.3040 aux_loss:0.0354 train_time:1327448ms step_avg:538.30ms step:2477/6250 train_loss:2.3105 aux_loss:0.0357 train_time:1327980ms step_avg:538.30ms step:2478/6250 train_loss:2.3713 aux_loss:0.0359 train_time:1328511ms step_avg:538.29ms step:2479/6250 train_loss:2.3330 aux_loss:0.0362 train_time:1329040ms step_avg:538.29ms step:2480/6250 train_loss:2.3705 aux_loss:0.0359 train_time:1329572ms step_avg:538.29ms step:2481/6250 train_loss:2.3763 aux_loss:0.0357 train_time:1330103ms step_avg:538.29ms step:2482/6250 train_loss:2.3858 aux_loss:0.0356 train_time:1330633ms step_avg:538.28ms step:2483/6250 train_loss:2.2641 aux_loss:0.0359 train_time:1331164ms step_avg:538.28ms step:2484/6250 train_loss:2.3691 aux_loss:0.0360 train_time:1331693ms step_avg:538.28ms step:2485/6250 train_loss:2.3147 aux_loss:0.0359 train_time:1332224ms step_avg:538.27ms step:2486/6250 train_loss:2.3067 aux_loss:0.0356 train_time:1332755ms step_avg:538.27ms step:2487/6250 train_loss:2.2339 aux_loss:0.0357 train_time:1333285ms step_avg:538.27ms step:2488/6250 train_loss:2.3075 aux_loss:0.0357 train_time:1333815ms step_avg:538.26ms step:2489/6250 train_loss:2.2719 aux_loss:0.0358 train_time:1334347ms step_avg:538.26ms step:2490/6250 train_loss:2.3289 aux_loss:0.0356 train_time:1334881ms step_avg:538.26ms step:2491/6250 train_loss:2.3617 aux_loss:0.0356 train_time:1335415ms step_avg:538.26ms step:2492/6250 train_loss:2.2994 aux_loss:0.0357 train_time:1335945ms step_avg:538.25ms step:2493/6250 train_loss:2.4415 aux_loss:0.0356 train_time:1336475ms step_avg:538.25ms step:2494/6250 train_loss:2.2162 aux_loss:0.0357 train_time:1337007ms step_avg:538.25ms step:2495/6250 train_loss:2.3125 aux_loss:0.0359 train_time:1337537ms step_avg:538.24ms step:2496/6250 train_loss:2.2737 aux_loss:0.0359 train_time:1338066ms step_avg:538.24ms step:2497/6250 train_loss:2.2468 aux_loss:0.0358 train_time:1338596ms step_avg:538.24ms step:2498/6250 train_loss:2.2982 aux_loss:0.0356 train_time:1339126ms step_avg:538.23ms step:2499/6250 train_loss:2.3141 aux_loss:0.0355 train_time:1339655ms step_avg:538.23ms step:2500/6250 train_loss:2.3542 aux_loss:0.0354 train_time:1340185ms step_avg:538.23ms step:2501/6250 train_loss:2.3886 aux_loss:0.0358 train_time:1340714ms step_avg:538.22ms step:2502/6250 train_loss:2.2229 aux_loss:0.0360 train_time:1341246ms step_avg:538.22ms step:2503/6250 train_loss:2.2748 aux_loss:0.0357 train_time:1341777ms step_avg:538.22ms step:2504/6250 train_loss:2.3518 aux_loss:0.0354 train_time:1342308ms step_avg:538.21ms step:2505/6250 train_loss:2.2860 aux_loss:0.0353 train_time:1342838ms step_avg:538.21ms step:2506/6250 train_loss:2.3089 aux_loss:0.0358 train_time:1343369ms step_avg:538.21ms step:2507/6250 train_loss:2.2178 aux_loss:0.0359 train_time:1343902ms step_avg:538.21ms step:2508/6250 train_loss:2.2185 aux_loss:0.0356 train_time:1344438ms step_avg:538.21ms step:2509/6250 train_loss:2.3068 aux_loss:0.0350 train_time:1344968ms step_avg:538.20ms step:2510/6250 train_loss:2.3312 aux_loss:0.0353 train_time:1345498ms step_avg:538.20ms step:2511/6250 train_loss:2.3040 aux_loss:0.0357 train_time:1346029ms step_avg:538.20ms step:2512/6250 train_loss:2.4254 aux_loss:0.0355 train_time:1346559ms step_avg:538.19ms step:2513/6250 train_loss:2.2941 aux_loss:0.0353 train_time:1347092ms step_avg:538.19ms step:2514/6250 train_loss:2.3699 aux_loss:0.0354 train_time:1347644ms step_avg:538.20ms step:2515/6250 train_loss:2.3055 aux_loss:0.0354 train_time:1348174ms step_avg:538.19ms step:2516/6250 train_loss:2.3041 aux_loss:0.0350 train_time:1348703ms step_avg:538.19ms step:2517/6250 train_loss:2.3042 aux_loss:0.0354 train_time:1349232ms step_avg:538.19ms step:2518/6250 train_loss:2.2715 aux_loss:0.0356 train_time:1349762ms step_avg:538.18ms step:2519/6250 train_loss:2.2815 aux_loss:0.0356 train_time:1350293ms step_avg:538.18ms step:2520/6250 train_loss:2.2963 aux_loss:0.0354 train_time:1350823ms step_avg:538.18ms step:2521/6250 train_loss:2.2366 aux_loss:0.0353 train_time:1351353ms step_avg:538.17ms step:2522/6250 train_loss:2.3434 aux_loss:0.0354 train_time:1351884ms step_avg:538.17ms step:2523/6250 train_loss:2.3867 aux_loss:0.0356 train_time:1352415ms step_avg:538.17ms step:2524/6250 train_loss:2.3240 aux_loss:0.0359 train_time:1352944ms step_avg:538.16ms step:2525/6250 train_loss:2.3083 aux_loss:0.0357 train_time:1353476ms step_avg:538.16ms step:2526/6250 train_loss:2.2870 aux_loss:0.0353 train_time:1354006ms step_avg:538.16ms step:2527/6250 train_loss:2.3404 aux_loss:0.0355 train_time:1354538ms step_avg:538.16ms step:2528/6250 train_loss:2.3555 aux_loss:0.0354 train_time:1355067ms step_avg:538.15ms step:2529/6250 train_loss:2.2693 aux_loss:0.0358 train_time:1355598ms step_avg:538.15ms step:2530/6250 train_loss:2.4462 aux_loss:0.0356 train_time:1356126ms step_avg:538.15ms step:2531/6250 train_loss:2.3060 aux_loss:0.0356 train_time:1356656ms step_avg:538.14ms step:2532/6250 train_loss:2.3711 aux_loss:0.0356 train_time:1357186ms step_avg:538.14ms step:2533/6250 train_loss:2.4063 aux_loss:0.0358 train_time:1357716ms step_avg:538.14ms step:2534/6250 train_loss:2.2683 aux_loss:0.0351 train_time:1358245ms step_avg:538.13ms step:2535/6250 train_loss:2.4564 aux_loss:0.0354 train_time:1358776ms step_avg:538.13ms step:2536/6250 train_loss:2.3973 aux_loss:0.0355 train_time:1359306ms step_avg:538.13ms step:2537/6250 train_loss:2.2830 aux_loss:0.0350 train_time:1359837ms step_avg:538.12ms step:2538/6250 train_loss:2.2872 aux_loss:0.0350 train_time:1360375ms step_avg:538.12ms step:2539/6250 train_loss:2.2995 aux_loss:0.0355 train_time:1360912ms step_avg:538.12ms step:2540/6250 train_loss:2.3333 aux_loss:0.0358 train_time:1361444ms step_avg:538.12ms step:2541/6250 train_loss:2.2581 aux_loss:0.0357 train_time:1361982ms step_avg:538.12ms step:2542/6250 train_loss:2.2729 aux_loss:0.0351 train_time:1362513ms step_avg:538.12ms step:2543/6250 train_loss:2.2769 aux_loss:0.0350 train_time:1363043ms step_avg:538.11ms step:2544/6250 train_loss:2.2711 aux_loss:0.0355 train_time:1363573ms step_avg:538.11ms step:2545/6250 train_loss:2.2760 aux_loss:0.0356 train_time:1364103ms step_avg:538.11ms step:2546/6250 train_loss:2.3762 aux_loss:0.0355 train_time:1364633ms step_avg:538.10ms step:2547/6250 train_loss:2.2170 aux_loss:0.0353 train_time:1365163ms step_avg:538.10ms step:2548/6250 train_loss:2.3166 aux_loss:0.0354 train_time:1365693ms step_avg:538.10ms step:2549/6250 train_loss:2.3371 aux_loss:0.0353 train_time:1366222ms step_avg:538.09ms step:2550/6250 train_loss:2.3248 aux_loss:0.0353 train_time:1366754ms step_avg:538.09ms step:2551/6250 train_loss:2.3854 aux_loss:0.0353 train_time:1367283ms step_avg:538.09ms step:2552/6250 train_loss:2.3386 aux_loss:0.0354 train_time:1367814ms step_avg:538.09ms step:2553/6250 train_loss:2.2838 aux_loss:0.0356 train_time:1368345ms step_avg:538.08ms step:2554/6250 train_loss:2.2968 aux_loss:0.0352 train_time:1368874ms step_avg:538.08ms step:2555/6250 train_loss:2.3155 aux_loss:0.0355 train_time:1369405ms step_avg:538.08ms step:2556/6250 train_loss:2.3328 aux_loss:0.0353 train_time:1369935ms step_avg:538.07ms step:2557/6250 train_loss:2.3339 aux_loss:0.0352 train_time:1370466ms step_avg:538.07ms step:2558/6250 train_loss:2.3582 aux_loss:0.0358 train_time:1370995ms step_avg:538.07ms step:2559/6250 train_loss:2.2684 aux_loss:0.0359 train_time:1371526ms step_avg:538.06ms step:2560/6250 train_loss:2.2988 aux_loss:0.0353 train_time:1372055ms step_avg:538.06ms step:2561/6250 train_loss:2.2925 aux_loss:0.0354 train_time:1372584ms step_avg:538.06ms step:2562/6250 train_loss:2.2934 aux_loss:0.0357 train_time:1373114ms step_avg:538.05ms step:2563/6250 train_loss:2.2582 aux_loss:0.0353 train_time:1373647ms step_avg:538.05ms step:2564/6250 train_loss:2.2540 aux_loss:0.0351 train_time:1374176ms step_avg:538.05ms step:2565/6250 train_loss:2.3729 aux_loss:0.0355 train_time:1374707ms step_avg:538.05ms step:2566/6250 train_loss:2.3392 aux_loss:0.0356 train_time:1375236ms step_avg:538.04ms step:2567/6250 train_loss:2.3364 aux_loss:0.0351 train_time:1375766ms step_avg:538.04ms step:2568/6250 train_loss:2.2396 aux_loss:0.0350 train_time:1376296ms step_avg:538.04ms step:2569/6250 train_loss:2.3195 aux_loss:0.0353 train_time:1376826ms step_avg:538.03ms step:2570/6250 train_loss:2.3846 aux_loss:0.0353 train_time:1377355ms step_avg:538.03ms step:2571/6250 train_loss:2.3723 aux_loss:0.0353 train_time:1377885ms step_avg:538.03ms step:2572/6250 train_loss:2.2428 aux_loss:0.0350 train_time:1378416ms step_avg:538.02ms step:2573/6250 train_loss:2.3522 aux_loss:0.0354 train_time:1378950ms step_avg:538.02ms step:2574/6250 train_loss:2.2784 aux_loss:0.0354 train_time:1379480ms step_avg:538.02ms step:2575/6250 train_loss:2.3307 aux_loss:0.0353 train_time:1380008ms step_avg:538.01ms step:2576/6250 train_loss:2.3847 aux_loss:0.0354 train_time:1380538ms step_avg:538.01ms step:2577/6250 train_loss:2.3367 aux_loss:0.0356 train_time:1381068ms step_avg:538.01ms step:2578/6250 train_loss:2.4061 aux_loss:0.0352 train_time:1381599ms step_avg:538.01ms step:2579/6250 train_loss:2.3198 aux_loss:0.0352 train_time:1382130ms step_avg:538.00ms step:2580/6250 train_loss:2.3175 aux_loss:0.0352 train_time:1382677ms step_avg:538.01ms step:2581/6250 train_loss:2.3313 aux_loss:0.0352 train_time:1383212ms step_avg:538.01ms step:2582/6250 train_loss:2.2716 aux_loss:0.0351 train_time:1383743ms step_avg:538.00ms step:2583/6250 train_loss:2.3332 aux_loss:0.0353 train_time:1384277ms step_avg:538.00ms step:2584/6250 train_loss:2.3192 aux_loss:0.0354 train_time:1384809ms step_avg:538.00ms step:2585/6250 train_loss:2.3464 aux_loss:0.0352 train_time:1385339ms step_avg:538.00ms step:2586/6250 train_loss:2.2906 aux_loss:0.0351 train_time:1385900ms step_avg:538.00ms step:2587/6250 train_loss:2.3313 aux_loss:0.0350 train_time:1386460ms step_avg:538.01ms step:2588/6250 train_loss:2.2896 aux_loss:0.0348 train_time:1387022ms step_avg:538.02ms step:2589/6250 train_loss:2.4100 aux_loss:0.0349 train_time:1387569ms step_avg:538.03ms step:2590/6250 train_loss:2.2494 aux_loss:0.0350 train_time:1388109ms step_avg:538.03ms step:2591/6250 train_loss:2.3150 aux_loss:0.0349 train_time:1388640ms step_avg:538.02ms step:2592/6250 train_loss:2.3643 aux_loss:0.0347 train_time:1389175ms step_avg:538.02ms step:2593/6250 train_loss:2.3075 aux_loss:0.0350 train_time:1389707ms step_avg:538.02ms step:2594/6250 train_loss:2.3148 aux_loss:0.0354 train_time:1390242ms step_avg:538.02ms step:2595/6250 train_loss:2.3117 aux_loss:0.0354 train_time:1390777ms step_avg:538.02ms step:2596/6250 train_loss:2.3148 aux_loss:0.0351 train_time:1391310ms step_avg:538.02ms step:2597/6250 train_loss:2.4346 aux_loss:0.0355 train_time:1391842ms step_avg:538.01ms step:2598/6250 train_loss:2.2800 aux_loss:0.0356 train_time:1392372ms step_avg:538.01ms step:2599/6250 train_loss:2.2673 aux_loss:0.0354 train_time:1392902ms step_avg:538.01ms step:2600/6250 train_loss:2.2257 aux_loss:0.0356 train_time:1393436ms step_avg:538.01ms step:2601/6250 train_loss:2.2888 aux_loss:0.0352 train_time:1393966ms step_avg:538.00ms step:2602/6250 train_loss:2.2792 aux_loss:0.0354 train_time:1394500ms step_avg:538.00ms step:2603/6250 train_loss:2.3534 aux_loss:0.0349 train_time:1395036ms step_avg:538.00ms step:2604/6250 train_loss:2.3164 aux_loss:0.0348 train_time:1395571ms step_avg:538.00ms step:2605/6250 train_loss:2.4035 aux_loss:0.0354 train_time:1396102ms step_avg:538.00ms step:2606/6250 train_loss:2.3351 aux_loss:0.0353 train_time:1396633ms step_avg:537.99ms step:2607/6250 train_loss:2.4570 aux_loss:0.0353 train_time:1397164ms step_avg:537.99ms step:2608/6250 train_loss:2.2717 aux_loss:0.0351 train_time:1397694ms step_avg:537.99ms step:2609/6250 train_loss:2.3338 aux_loss:0.0352 train_time:1398225ms step_avg:537.99ms step:2610/6250 train_loss:2.3641 aux_loss:0.0352 train_time:1398757ms step_avg:537.98ms step:2611/6250 train_loss:2.3330 aux_loss:0.0352 train_time:1399287ms step_avg:537.98ms step:2612/6250 train_loss:2.3203 aux_loss:0.0353 train_time:1399837ms step_avg:537.99ms step:2613/6250 train_loss:2.2986 aux_loss:0.0351 train_time:1400398ms step_avg:537.99ms step:2614/6250 train_loss:2.3977 aux_loss:0.0349 train_time:1400929ms step_avg:537.99ms step:2615/6250 train_loss:2.2578 aux_loss:0.0346 train_time:1401479ms step_avg:538.00ms step:2616/6250 train_loss:2.3344 aux_loss:0.0349 train_time:1402011ms step_avg:537.99ms step:2617/6250 train_loss:2.3792 aux_loss:0.0351 train_time:1402551ms step_avg:537.99ms step:2618/6250 train_loss:2.3504 aux_loss:0.0352 train_time:1403082ms step_avg:537.99ms step:2619/6250 train_loss:2.2595 aux_loss:0.0352 train_time:1403612ms step_avg:537.99ms step:2620/6250 train_loss:2.2902 aux_loss:0.0355 train_time:1404143ms step_avg:537.99ms step:2621/6250 train_loss:2.2978 aux_loss:0.0349 train_time:1404673ms step_avg:537.98ms step:2622/6250 train_loss:2.3447 aux_loss:0.0349 train_time:1405202ms step_avg:537.98ms step:2623/6250 train_loss:2.1960 aux_loss:0.0348 train_time:1405733ms step_avg:537.98ms step:2624/6250 train_loss:2.3064 aux_loss:0.0349 train_time:1406264ms step_avg:537.97ms step:2625/6250 train_loss:2.3927 aux_loss:0.0353 train_time:1406794ms step_avg:537.97ms step:2626/6250 train_loss:2.3063 aux_loss:0.0354 train_time:1407324ms step_avg:537.97ms step:2627/6250 train_loss:2.3002 aux_loss:0.0350 train_time:1407854ms step_avg:537.96ms step:2628/6250 train_loss:2.3022 aux_loss:0.0350 train_time:1408384ms step_avg:537.96ms step:2629/6250 train_loss:2.3781 aux_loss:0.0351 train_time:1408915ms step_avg:537.96ms step:2630/6250 train_loss:2.3364 aux_loss:0.0350 train_time:1409445ms step_avg:537.96ms step:2631/6250 train_loss:2.2818 aux_loss:0.0347 train_time:1409974ms step_avg:537.95ms step:2632/6250 train_loss:2.2931 aux_loss:0.0353 train_time:1410505ms step_avg:537.95ms step:2633/6250 train_loss:2.2599 aux_loss:0.0354 train_time:1411037ms step_avg:537.95ms step:2634/6250 train_loss:2.3645 aux_loss:0.0348 train_time:1411567ms step_avg:537.94ms step:2635/6250 train_loss:2.3814 aux_loss:0.0345 train_time:1412097ms step_avg:537.94ms step:2636/6250 train_loss:2.2793 aux_loss:0.0348 train_time:1412627ms step_avg:537.94ms step:2637/6250 train_loss:2.3261 aux_loss:0.0347 train_time:1413156ms step_avg:537.94ms step:2638/6250 train_loss:2.2783 aux_loss:0.0348 train_time:1413686ms step_avg:537.93ms step:2639/6250 train_loss:2.3214 aux_loss:0.0350 train_time:1414214ms step_avg:537.93ms step:2640/6250 train_loss:2.3921 aux_loss:0.0351 train_time:1414744ms step_avg:537.93ms step:2641/6250 train_loss:2.2975 aux_loss:0.0348 train_time:1415273ms step_avg:537.92ms step:2642/6250 train_loss:2.2923 aux_loss:0.0350 train_time:1415803ms step_avg:537.92ms step:2643/6250 train_loss:2.3741 aux_loss:0.0350 train_time:1416333ms step_avg:537.92ms step:2644/6250 train_loss:2.2280 aux_loss:0.0349 train_time:1416863ms step_avg:537.91ms step:2645/6250 train_loss:2.2634 aux_loss:0.0350 train_time:1417396ms step_avg:537.91ms step:2646/6250 train_loss:2.2307 aux_loss:0.0349 train_time:1417931ms step_avg:537.91ms step:2647/6250 train_loss:2.4256 aux_loss:0.0349 train_time:1418462ms step_avg:537.91ms step:2648/6250 train_loss:2.3855 aux_loss:0.0349 train_time:1418990ms step_avg:537.90ms step:2649/6250 train_loss:2.3230 aux_loss:0.0348 train_time:1419523ms step_avg:537.90ms step:2650/6250 train_loss:2.3622 aux_loss:0.0351 train_time:1420080ms step_avg:537.91ms step:2651/6250 train_loss:2.3942 aux_loss:0.0349 train_time:1420610ms step_avg:537.91ms step:2652/6250 train_loss:2.4263 aux_loss:0.0352 train_time:1421138ms step_avg:537.90ms step:2653/6250 train_loss:2.3228 aux_loss:0.0352 train_time:1421668ms step_avg:537.90ms step:2654/6250 train_loss:2.3085 aux_loss:0.0348 train_time:1422198ms step_avg:537.90ms step:2655/6250 train_loss:2.3188 aux_loss:0.0350 train_time:1422749ms step_avg:537.90ms step:2656/6250 train_loss:2.3910 aux_loss:0.0351 train_time:1423279ms step_avg:537.90ms step:2657/6250 train_loss:2.3734 aux_loss:0.0351 train_time:1423807ms step_avg:537.89ms step:2658/6250 train_loss:2.3429 aux_loss:0.0348 train_time:1424339ms step_avg:537.89ms step:2659/6250 train_loss:2.1903 aux_loss:0.0348 train_time:1424870ms step_avg:537.89ms step:2660/6250 train_loss:2.3433 aux_loss:0.0350 train_time:1425399ms step_avg:537.89ms step:2661/6250 train_loss:2.2473 aux_loss:0.0349 train_time:1425928ms step_avg:537.88ms step:2662/6250 train_loss:2.3055 aux_loss:0.0346 train_time:1426459ms step_avg:537.88ms step:2663/6250 train_loss:2.3492 aux_loss:0.0347 train_time:1426988ms step_avg:537.88ms step:2664/6250 train_loss:2.2730 aux_loss:0.0348 train_time:1427519ms step_avg:537.87ms step:2665/6250 train_loss:2.2404 aux_loss:0.0347 train_time:1428048ms step_avg:537.87ms step:2666/6250 train_loss:2.3556 aux_loss:0.0346 train_time:1428588ms step_avg:537.87ms step:2667/6250 train_loss:2.4019 aux_loss:0.0351 train_time:1429132ms step_avg:537.87ms step:2668/6250 train_loss:2.3399 aux_loss:0.0346 train_time:1429678ms step_avg:537.88ms step:2669/6250 train_loss:2.2801 aux_loss:0.0342 train_time:1430227ms step_avg:537.88ms step:2670/6250 train_loss:2.3321 aux_loss:0.0348 train_time:1430775ms step_avg:537.89ms step:2671/6250 train_loss:2.2948 aux_loss:0.0350 train_time:1431312ms step_avg:537.89ms step:2672/6250 train_loss:2.2904 aux_loss:0.0350 train_time:1431842ms step_avg:537.88ms step:2673/6250 train_loss:2.2849 aux_loss:0.0352 train_time:1432374ms step_avg:537.88ms step:2674/6250 train_loss:2.4530 aux_loss:0.0350 train_time:1432918ms step_avg:537.88ms step:2675/6250 train_loss:2.3034 aux_loss:0.0348 train_time:1433489ms step_avg:537.89ms step:2676/6250 train_loss:2.2171 aux_loss:0.0346 train_time:1434056ms step_avg:537.91ms step:2677/6250 train_loss:2.3866 aux_loss:0.0348 train_time:1434604ms step_avg:537.91ms step:2678/6250 train_loss:2.3243 aux_loss:0.0349 train_time:1435133ms step_avg:537.91ms step:2679/6250 train_loss:2.2895 aux_loss:0.0348 train_time:1435663ms step_avg:537.90ms step:2680/6250 train_loss:2.3601 aux_loss:0.0348 train_time:1436196ms step_avg:537.90ms step:2681/6250 train_loss:2.3581 aux_loss:0.0349 train_time:1436731ms step_avg:537.90ms step:2682/6250 train_loss:2.2736 aux_loss:0.0349 train_time:1437275ms step_avg:537.90ms step:2683/6250 train_loss:2.2691 aux_loss:0.0345 train_time:1437818ms step_avg:537.90ms step:2684/6250 train_loss:2.2490 aux_loss:0.0346 train_time:1438350ms step_avg:537.90ms step:2685/6250 train_loss:2.2998 aux_loss:0.0346 train_time:1438880ms step_avg:537.90ms step:2686/6250 train_loss:2.3324 aux_loss:0.0348 train_time:1439415ms step_avg:537.90ms step:2687/6250 train_loss:2.3698 aux_loss:0.0345 train_time:1439959ms step_avg:537.90ms step:2688/6250 train_loss:2.2534 aux_loss:0.0348 train_time:1440497ms step_avg:537.90ms step:2689/6250 train_loss:2.2825 aux_loss:0.0348 train_time:1441026ms step_avg:537.90ms step:2690/6250 train_loss:2.2813 aux_loss:0.0344 train_time:1441565ms step_avg:537.90ms step:2691/6250 train_loss:2.3044 aux_loss:0.0348 train_time:1442103ms step_avg:537.90ms step:2692/6250 train_loss:2.3601 aux_loss:0.0347 train_time:1442647ms step_avg:537.90ms step:2693/6250 train_loss:2.3426 aux_loss:0.0347 train_time:1443190ms step_avg:537.90ms step:2694/6250 train_loss:2.2638 aux_loss:0.0346 train_time:1443724ms step_avg:537.90ms step:2695/6250 train_loss:2.3437 aux_loss:0.0347 train_time:1444253ms step_avg:537.90ms step:2696/6250 train_loss:2.3691 aux_loss:0.0347 train_time:1444790ms step_avg:537.90ms step:2697/6250 train_loss:2.2838 aux_loss:0.0348 train_time:1445323ms step_avg:537.89ms step:2698/6250 train_loss:2.2564 aux_loss:0.0347 train_time:1445855ms step_avg:537.89ms step:2699/6250 train_loss:2.2740 aux_loss:0.0342 train_time:1446387ms step_avg:537.89ms step:2700/6250 train_loss:2.3503 aux_loss:0.0342 train_time:1446922ms step_avg:537.89ms step:2701/6250 train_loss:2.3398 aux_loss:0.0345 train_time:1447454ms step_avg:537.89ms step:2702/6250 train_loss:2.2652 aux_loss:0.0348 train_time:1447983ms step_avg:537.88ms step:2703/6250 train_loss:2.2815 aux_loss:0.0347 train_time:1448513ms step_avg:537.88ms step:2704/6250 train_loss:2.1870 aux_loss:0.0345 train_time:1449043ms step_avg:537.88ms step:2705/6250 train_loss:2.2645 aux_loss:0.0348 train_time:1449572ms step_avg:537.87ms step:2706/6250 train_loss:2.2740 aux_loss:0.0349 train_time:1450100ms step_avg:537.87ms step:2707/6250 train_loss:2.2612 aux_loss:0.0346 train_time:1450630ms step_avg:537.87ms step:2708/6250 train_loss:2.3121 aux_loss:0.0347 train_time:1451159ms step_avg:537.86ms step:2709/6250 train_loss:2.2127 aux_loss:0.0350 train_time:1451689ms step_avg:537.86ms step:2710/6250 train_loss:2.3201 aux_loss:0.0349 train_time:1452218ms step_avg:537.86ms step:2711/6250 train_loss:2.3005 aux_loss:0.0345 train_time:1452748ms step_avg:537.86ms step:2712/6250 train_loss:2.3322 aux_loss:0.0341 train_time:1453296ms step_avg:537.86ms step:2713/6250 train_loss:2.3855 aux_loss:0.0343 train_time:1453834ms step_avg:537.86ms step:2714/6250 train_loss:2.3601 aux_loss:0.0348 train_time:1454372ms step_avg:537.86ms step:2715/6250 train_loss:2.2679 aux_loss:0.0347 train_time:1454917ms step_avg:537.86ms step:2716/6250 train_loss:2.1896 aux_loss:0.0341 train_time:1455454ms step_avg:537.86ms step:2717/6250 train_loss:2.4606 aux_loss:0.0347 train_time:1455993ms step_avg:537.86ms step:2718/6250 train_loss:2.2163 aux_loss:0.0353 train_time:1456528ms step_avg:537.86ms step:2719/6250 train_loss:2.3705 aux_loss:0.0349 train_time:1457061ms step_avg:537.86ms step:2720/6250 train_loss:2.3531 aux_loss:0.0343 train_time:1457596ms step_avg:537.86ms step:2721/6250 train_loss:2.3314 aux_loss:0.0347 train_time:1458129ms step_avg:537.86ms step:2722/6250 train_loss:2.2977 aux_loss:0.0352 train_time:1458662ms step_avg:537.85ms step:2723/6250 train_loss:2.2318 aux_loss:0.0349 train_time:1459197ms step_avg:537.85ms step:2724/6250 train_loss:2.3061 aux_loss:0.0345 train_time:1459730ms step_avg:537.85ms step:2725/6250 train_loss:2.2173 aux_loss:0.0343 train_time:1460264ms step_avg:537.85ms step:2726/6250 train_loss:2.3052 aux_loss:0.0349 train_time:1460803ms step_avg:537.85ms step:2727/6250 train_loss:2.3124 aux_loss:0.0348 train_time:1461340ms step_avg:537.85ms step:2728/6250 train_loss:2.2295 aux_loss:0.0346 train_time:1461875ms step_avg:537.85ms step:2729/6250 train_loss:2.4462 aux_loss:0.0350 train_time:1462408ms step_avg:537.85ms step:2730/6250 train_loss:2.3046 aux_loss:0.0352 train_time:1462940ms step_avg:537.85ms step:2731/6250 train_loss:2.2975 aux_loss:0.0348 train_time:1463476ms step_avg:537.85ms step:2732/6250 train_loss:2.2862 aux_loss:0.0348 train_time:1464011ms step_avg:537.84ms step:2733/6250 train_loss:2.3850 aux_loss:0.0345 train_time:1464542ms step_avg:537.84ms step:2734/6250 train_loss:2.3166 aux_loss:0.0348 train_time:1465074ms step_avg:537.84ms step:2735/6250 train_loss:2.2709 aux_loss:0.0348 train_time:1465609ms step_avg:537.84ms step:2736/6250 train_loss:2.3314 aux_loss:0.0345 train_time:1466143ms step_avg:537.84ms step:2737/6250 train_loss:2.2560 aux_loss:0.0347 train_time:1466676ms step_avg:537.83ms step:2738/6250 train_loss:2.2784 aux_loss:0.0347 train_time:1467207ms step_avg:537.83ms step:2739/6250 train_loss:2.2861 aux_loss:0.0345 train_time:1467740ms step_avg:537.83ms step:2740/6250 train_loss:2.3351 aux_loss:0.0345 train_time:1468275ms step_avg:537.83ms step:2741/6250 train_loss:2.2908 aux_loss:0.0347 train_time:1468816ms step_avg:537.83ms step:2742/6250 train_loss:2.3135 aux_loss:0.0351 train_time:1469369ms step_avg:537.84ms step:2743/6250 train_loss:2.2999 aux_loss:0.0348 train_time:1469913ms step_avg:537.84ms step:2744/6250 train_loss:2.3512 aux_loss:0.0350 train_time:1470462ms step_avg:537.84ms step:2745/6250 train_loss:2.3025 aux_loss:0.0344 train_time:1471001ms step_avg:537.84ms step:2746/6250 train_loss:2.2926 aux_loss:0.0342 train_time:1471547ms step_avg:537.85ms step:2747/6250 train_loss:2.3556 aux_loss:0.0344 train_time:1472096ms step_avg:537.85ms step:2748/6250 train_loss:2.2434 aux_loss:0.0346 train_time:1472641ms step_avg:537.85ms step:2749/6250 train_loss:2.3016 aux_loss:0.0347 train_time:1473177ms step_avg:537.85ms step:2750/6250 train_loss:2.3187 aux_loss:0.0345 train_time:1473729ms step_avg:537.86ms step:2751/6250 train_loss:2.3075 aux_loss:0.0346 train_time:1474275ms step_avg:537.86ms step:2752/6250 train_loss:2.3065 aux_loss:0.0343 train_time:1474810ms step_avg:537.86ms step:2753/6250 train_loss:2.2930 aux_loss:0.0340 train_time:1475353ms step_avg:537.86ms step:2754/6250 train_loss:2.2775 aux_loss:0.0342 train_time:1475894ms step_avg:537.86ms step:2755/6250 train_loss:2.2980 aux_loss:0.0343 train_time:1476446ms step_avg:537.87ms step:2756/6250 train_loss:2.2834 aux_loss:0.0345 train_time:1476988ms step_avg:537.87ms step:2757/6250 train_loss:2.2288 aux_loss:0.0343 train_time:1477532ms step_avg:537.87ms step:2758/6250 train_loss:2.2104 aux_loss:0.0342 train_time:1478071ms step_avg:537.87ms step:2759/6250 train_loss:2.2090 aux_loss:0.0344 train_time:1478613ms step_avg:537.87ms step:2760/6250 train_loss:2.2086 aux_loss:0.0346 train_time:1479153ms step_avg:537.87ms step:2761/6250 train_loss:2.2859 aux_loss:0.0344 train_time:1479715ms step_avg:537.88ms step:2762/6250 train_loss:2.3789 aux_loss:0.0341 train_time:1480263ms step_avg:537.89ms step:2763/6250 train_loss:2.2237 aux_loss:0.0345 train_time:1480802ms step_avg:537.89ms step:2764/6250 train_loss:2.2743 aux_loss:0.0347 train_time:1481343ms step_avg:537.89ms step:2765/6250 train_loss:2.2422 aux_loss:0.0346 train_time:1481873ms step_avg:537.88ms step:2766/6250 train_loss:2.2627 aux_loss:0.0345 train_time:1482409ms step_avg:537.88ms step:2767/6250 train_loss:2.3194 aux_loss:0.0346 train_time:1482953ms step_avg:537.89ms step:2768/6250 train_loss:2.3047 aux_loss:0.0347 train_time:1483493ms step_avg:537.89ms step:2769/6250 train_loss:2.3293 aux_loss:0.0350 train_time:1484045ms step_avg:537.89ms step:2770/6250 train_loss:2.4517 aux_loss:0.0351 train_time:1484600ms step_avg:537.90ms step:2771/6250 train_loss:2.3125 aux_loss:0.0346 train_time:1485130ms step_avg:537.90ms step:2772/6250 train_loss:2.3503 aux_loss:0.0346 train_time:1485665ms step_avg:537.89ms step:2773/6250 train_loss:2.2831 aux_loss:0.0348 train_time:1486211ms step_avg:537.90ms step:2774/6250 train_loss:2.2839 aux_loss:0.0348 train_time:1486747ms step_avg:537.90ms step:2775/6250 train_loss:2.3416 aux_loss:0.0345 train_time:1487277ms step_avg:537.89ms step:2776/6250 train_loss:2.2396 aux_loss:0.0346 train_time:1487809ms step_avg:537.89ms step:2777/6250 train_loss:2.4288 aux_loss:0.0352 train_time:1488342ms step_avg:537.89ms step:2778/6250 train_loss:2.2744 aux_loss:0.0352 train_time:1488877ms step_avg:537.89ms step:2779/6250 train_loss:2.2740 aux_loss:0.0348 train_time:1489461ms step_avg:537.91ms step:2780/6250 train_loss:2.4120 aux_loss:0.0347 train_time:1490013ms step_avg:537.91ms step:2781/6250 train_loss:2.2858 aux_loss:0.0349 train_time:1490550ms step_avg:537.91ms step:2782/6250 train_loss:2.3498 aux_loss:0.0345 train_time:1491107ms step_avg:537.92ms step:2783/6250 train_loss:2.2876 aux_loss:0.0341 train_time:1491665ms step_avg:537.92ms step:2784/6250 train_loss:2.1351 aux_loss:0.0340 train_time:1492220ms step_avg:537.93ms step:2785/6250 train_loss:2.1952 aux_loss:0.0344 train_time:1492772ms step_avg:537.94ms step:2786/6250 train_loss:2.3119 aux_loss:0.0348 train_time:1493325ms step_avg:537.94ms step:2787/6250 train_loss:2.3545 aux_loss:0.0347 train_time:1493878ms step_avg:537.95ms step:2788/6250 train_loss:2.3743 aux_loss:0.0346 train_time:1494427ms step_avg:537.95ms step:2789/6250 train_loss:2.3257 aux_loss:0.0341 train_time:1494983ms step_avg:537.96ms step:2790/6250 train_loss:2.3185 aux_loss:0.0340 train_time:1495528ms step_avg:537.96ms step:2791/6250 train_loss:2.3522 aux_loss:0.0341 train_time:1496082ms step_avg:537.97ms step:2792/6250 train_loss:2.2880 aux_loss:0.0345 train_time:1496633ms step_avg:537.97ms step:2793/6250 train_loss:2.2672 aux_loss:0.0345 train_time:1497185ms step_avg:537.98ms step:2794/6250 train_loss:2.1635 aux_loss:0.0340 train_time:1497738ms step_avg:537.98ms step:2795/6250 train_loss:2.3208 aux_loss:0.0341 train_time:1498296ms step_avg:537.99ms step:2796/6250 train_loss:2.2733 aux_loss:0.0344 train_time:1498853ms step_avg:537.99ms step:2797/6250 train_loss:2.3514 aux_loss:0.0343 train_time:1499406ms step_avg:538.00ms step:2798/6250 train_loss:2.2061 aux_loss:0.0342 train_time:1499966ms step_avg:538.01ms step:2799/6250 train_loss:2.2918 aux_loss:0.0343 train_time:1500515ms step_avg:538.01ms step:2800/6250 train_loss:2.3935 aux_loss:0.0343 train_time:1501060ms step_avg:538.01ms step:2801/6250 train_loss:2.2545 aux_loss:0.0342 train_time:1501595ms step_avg:538.01ms step:2802/6250 train_loss:2.2927 aux_loss:0.0344 train_time:1502143ms step_avg:538.02ms step:2803/6250 train_loss:2.3261 aux_loss:0.0343 train_time:1502692ms step_avg:538.02ms step:2804/6250 train_loss:2.2118 aux_loss:0.0340 train_time:1503242ms step_avg:538.02ms step:2805/6250 train_loss:2.2672 aux_loss:0.0340 train_time:1503794ms step_avg:538.03ms step:2806/6250 train_loss:2.3898 aux_loss:0.0341 train_time:1504348ms step_avg:538.04ms step:2807/6250 train_loss:2.3436 aux_loss:0.0342 train_time:1504897ms step_avg:538.04ms step:2808/6250 train_loss:2.2595 aux_loss:0.0339 train_time:1505432ms step_avg:538.04ms step:2809/6250 train_loss:2.1961 aux_loss:0.0339 train_time:1505976ms step_avg:538.04ms step:2810/6250 train_loss:2.2587 aux_loss:0.0344 train_time:1506521ms step_avg:538.04ms step:2811/6250 train_loss:2.2396 aux_loss:0.0344 train_time:1507058ms step_avg:538.04ms step:2812/6250 train_loss:2.3139 aux_loss:0.0344 train_time:1507601ms step_avg:538.04ms step:2813/6250 train_loss:2.4578 aux_loss:0.0346 train_time:1508152ms step_avg:538.05ms step:2814/6250 train_loss:2.3919 aux_loss:0.0344 train_time:1508719ms step_avg:538.06ms step:2815/6250 train_loss:2.3794 aux_loss:0.0346 train_time:1509259ms step_avg:538.06ms step:2816/6250 train_loss:2.2600 aux_loss:0.0343 train_time:1509789ms step_avg:538.06ms step:2817/6250 train_loss:2.2393 aux_loss:0.0337 train_time:1510320ms step_avg:538.05ms step:2818/6250 train_loss:2.3185 aux_loss:0.0342 train_time:1510858ms step_avg:538.05ms step:2819/6250 train_loss:2.3423 aux_loss:0.0343 train_time:1511420ms step_avg:538.06ms step:2820/6250 train_loss:2.2375 aux_loss:0.0342 train_time:1512007ms step_avg:538.08ms step:2821/6250 train_loss:2.2575 aux_loss:0.0340 train_time:1512554ms step_avg:538.08ms step:2822/6250 train_loss:2.3254 aux_loss:0.0343 train_time:1513086ms step_avg:538.08ms step:2823/6250 train_loss:2.4152 aux_loss:0.0344 train_time:1513626ms step_avg:538.08ms step:2824/6250 train_loss:2.2946 aux_loss:0.0345 train_time:1514169ms step_avg:538.08ms step:2825/6250 train_loss:2.3428 aux_loss:0.0344 train_time:1514722ms step_avg:538.09ms step:2826/6250 train_loss:2.3881 aux_loss:0.0343 train_time:1515257ms step_avg:538.09ms step:2827/6250 train_loss:2.3037 aux_loss:0.0348 train_time:1515807ms step_avg:538.09ms step:2828/6250 train_loss:2.2526 aux_loss:0.0341 train_time:1516344ms step_avg:538.09ms step:2829/6250 train_loss:2.3740 aux_loss:0.0343 train_time:1516888ms step_avg:538.09ms step:2830/6250 train_loss:2.2367 aux_loss:0.0342 train_time:1517422ms step_avg:538.09ms step:2831/6250 train_loss:2.3308 aux_loss:0.0346 train_time:1517954ms step_avg:538.09ms step:2832/6250 train_loss:2.2809 aux_loss:0.0343 train_time:1518509ms step_avg:538.10ms step:2833/6250 train_loss:2.2529 aux_loss:0.0339 train_time:1519042ms step_avg:538.10ms step:2834/6250 train_loss:2.2842 aux_loss:0.0341 train_time:1519573ms step_avg:538.09ms step:2835/6250 train_loss:2.3872 aux_loss:0.0342 train_time:1520104ms step_avg:538.09ms step:2836/6250 train_loss:2.3547 aux_loss:0.0341 train_time:1520635ms step_avg:538.09ms step:2837/6250 train_loss:2.3463 aux_loss:0.0341 train_time:1521174ms step_avg:538.09ms step:2838/6250 train_loss:2.2971 aux_loss:0.0340 train_time:1521705ms step_avg:538.09ms step:2839/6250 train_loss:2.1723 aux_loss:0.0342 train_time:1522237ms step_avg:538.08ms step:2840/6250 train_loss:2.3414 aux_loss:0.0343 train_time:1522769ms step_avg:538.08ms step:2841/6250 train_loss:2.2533 aux_loss:0.0340 train_time:1523308ms step_avg:538.08ms step:2842/6250 train_loss:2.3283 aux_loss:0.0338 train_time:1523836ms step_avg:538.08ms step:2843/6250 train_loss:2.3388 aux_loss:0.0339 train_time:1524367ms step_avg:538.08ms step:2844/6250 train_loss:2.2257 aux_loss:0.0340 train_time:1524899ms step_avg:538.07ms step:2845/6250 train_loss:2.2974 aux_loss:0.0341 train_time:1525431ms step_avg:538.07ms step:2846/6250 train_loss:2.2781 aux_loss:0.0344 train_time:1525978ms step_avg:538.07ms step:2847/6250 train_loss:2.2577 aux_loss:0.0342 train_time:1526529ms step_avg:538.08ms step:2848/6250 train_loss:2.3103 aux_loss:0.0335 train_time:1527078ms step_avg:538.08ms step:2849/6250 train_loss:2.2799 aux_loss:0.0339 train_time:1527615ms step_avg:538.08ms step:2850/6250 train_loss:2.2799 aux_loss:0.0340 train_time:1528165ms step_avg:538.09ms step:2851/6250 train_loss:2.3193 aux_loss:0.0337 train_time:1528698ms step_avg:538.08ms step:2852/6250 train_loss:2.3821 aux_loss:0.0338 train_time:1529230ms step_avg:538.08ms step:2853/6250 train_loss:2.2714 aux_loss:0.0338 train_time:1529761ms step_avg:538.08ms step:2854/6250 train_loss:2.3292 aux_loss:0.0344 train_time:1530293ms step_avg:538.08ms step:2855/6250 train_loss:2.3339 aux_loss:0.0341 train_time:1530823ms step_avg:538.07ms step:2856/6250 train_loss:2.2568 aux_loss:0.0338 train_time:1531354ms step_avg:538.07ms step:2857/6250 train_loss:2.2643 aux_loss:0.0338 train_time:1531932ms step_avg:538.09ms step:2858/6250 train_loss:2.2468 aux_loss:0.0344 train_time:1532511ms step_avg:538.10ms step:2859/6250 train_loss:2.2825 aux_loss:0.0342 train_time:1533048ms step_avg:538.10ms step:2860/6250 train_loss:2.3032 aux_loss:0.0340 train_time:1533581ms step_avg:538.10ms step:2861/6250 train_loss:2.2880 aux_loss:0.0339 train_time:1534119ms step_avg:538.10ms step:2862/6250 train_loss:2.2523 aux_loss:0.0337 train_time:1534667ms step_avg:538.10ms step:2863/6250 train_loss:2.2526 aux_loss:0.0340 train_time:1535218ms step_avg:538.11ms step:2864/6250 train_loss:2.4028 aux_loss:0.0349 train_time:1535749ms step_avg:538.10ms step:2865/6250 train_loss:2.2783 aux_loss:0.0349 train_time:1536280ms step_avg:538.10ms step:2866/6250 train_loss:2.3385 aux_loss:0.0339 train_time:1536812ms step_avg:538.10ms step:2867/6250 train_loss:2.2747 aux_loss:0.0339 train_time:1537343ms step_avg:538.10ms step:2868/6250 train_loss:2.2868 aux_loss:0.0343 train_time:1537874ms step_avg:538.09ms step:2869/6250 train_loss:2.2881 aux_loss:0.0342 train_time:1538443ms step_avg:538.11ms step:2870/6250 train_loss:2.3054 aux_loss:0.0345 train_time:1538987ms step_avg:538.11ms step:2871/6250 train_loss:2.3217 aux_loss:0.0344 train_time:1539527ms step_avg:538.11ms step:2872/6250 train_loss:2.2627 aux_loss:0.0346 train_time:1540056ms step_avg:538.10ms step:2873/6250 train_loss:2.3202 aux_loss:0.0343 train_time:1540598ms step_avg:538.11ms step:2874/6250 train_loss:2.2405 aux_loss:0.0341 train_time:1541136ms step_avg:538.11ms step:2875/6250 train_loss:2.3002 aux_loss:0.0343 train_time:1541687ms step_avg:538.11ms step:2876/6250 train_loss:2.2644 aux_loss:0.0343 train_time:1542218ms step_avg:538.11ms step:2877/6250 train_loss:2.2786 aux_loss:0.0340 train_time:1542751ms step_avg:538.11ms step:2878/6250 train_loss:2.3308 aux_loss:0.0341 train_time:1543283ms step_avg:538.10ms step:2879/6250 train_loss:2.3582 aux_loss:0.0340 train_time:1543813ms step_avg:538.10ms step:2880/6250 train_loss:2.2027 aux_loss:0.0339 train_time:1544346ms step_avg:538.10ms step:2881/6250 train_loss:2.2824 aux_loss:0.0340 train_time:1544878ms step_avg:538.10ms step:2882/6250 train_loss:2.2551 aux_loss:0.0339 train_time:1545409ms step_avg:538.09ms step:2883/6250 train_loss:2.2558 aux_loss:0.0339 train_time:1545939ms step_avg:538.09ms step:2884/6250 train_loss:2.2026 aux_loss:0.0339 train_time:1546472ms step_avg:538.09ms step:2885/6250 train_loss:2.3834 aux_loss:0.0338 train_time:1547003ms step_avg:538.09ms step:2886/6250 train_loss:2.2429 aux_loss:0.0337 train_time:1547535ms step_avg:538.09ms step:2887/6250 train_loss:2.3060 aux_loss:0.0342 train_time:1548065ms step_avg:538.08ms step:2888/6250 train_loss:2.2848 aux_loss:0.0343 train_time:1548597ms step_avg:538.08ms step:2889/6250 train_loss:2.3775 aux_loss:0.0342 train_time:1549132ms step_avg:538.08ms step:2890/6250 train_loss:2.2322 aux_loss:0.0339 train_time:1549669ms step_avg:538.08ms step:2891/6250 train_loss:2.2421 aux_loss:0.0336 train_time:1550201ms step_avg:538.08ms step:2892/6250 train_loss:2.2380 aux_loss:0.0335 train_time:1550733ms step_avg:538.08ms step:2893/6250 train_loss:2.2910 aux_loss:0.0339 train_time:1551270ms step_avg:538.07ms step:2894/6250 train_loss:2.3235 aux_loss:0.0338 train_time:1551812ms step_avg:538.08ms step:2895/6250 train_loss:2.3313 aux_loss:0.0336 train_time:1552372ms step_avg:538.08ms step:2896/6250 train_loss:2.2124 aux_loss:0.0339 train_time:1552917ms step_avg:538.09ms step:2897/6250 train_loss:2.3447 aux_loss:0.0342 train_time:1553457ms step_avg:538.09ms step:2898/6250 train_loss:2.3096 aux_loss:0.0339 train_time:1553990ms step_avg:538.09ms step:2899/6250 train_loss:2.2356 aux_loss:0.0337 train_time:1554554ms step_avg:538.09ms step:2900/6250 train_loss:2.3836 aux_loss:0.0338 train_time:1555086ms step_avg:538.09ms step:2901/6250 train_loss:2.2865 aux_loss:0.0338 train_time:1555619ms step_avg:538.09ms step:2902/6250 train_loss:2.3265 aux_loss:0.0338 train_time:1556151ms step_avg:538.09ms step:2903/6250 train_loss:2.2172 aux_loss:0.0339 train_time:1556683ms step_avg:538.09ms step:2904/6250 train_loss:2.2768 aux_loss:0.0339 train_time:1557215ms step_avg:538.08ms step:2905/6250 train_loss:2.1990 aux_loss:0.0339 train_time:1557747ms step_avg:538.08ms step:2906/6250 train_loss:2.2640 aux_loss:0.0338 train_time:1558283ms step_avg:538.08ms step:2907/6250 train_loss:2.2895 aux_loss:0.0338 train_time:1558818ms step_avg:538.08ms step:2908/6250 train_loss:2.2581 aux_loss:0.0337 train_time:1559349ms step_avg:538.08ms step:2909/6250 train_loss:2.2977 aux_loss:0.0335 train_time:1559899ms step_avg:538.08ms step:2910/6250 train_loss:2.2595 aux_loss:0.0336 train_time:1560435ms step_avg:538.08ms step:2911/6250 train_loss:2.2771 aux_loss:0.0338 train_time:1561001ms step_avg:538.09ms step:2912/6250 train_loss:2.3110 aux_loss:0.0337 train_time:1561533ms step_avg:538.09ms step:2913/6250 train_loss:2.2813 aux_loss:0.0342 train_time:1562064ms step_avg:538.09ms step:2914/6250 train_loss:2.2889 aux_loss:0.0338 train_time:1562597ms step_avg:538.08ms step:2915/6250 train_loss:2.3936 aux_loss:0.0338 train_time:1563128ms step_avg:538.08ms step:2916/6250 train_loss:2.2262 aux_loss:0.0337 train_time:1563660ms step_avg:538.08ms step:2917/6250 train_loss:2.1544 aux_loss:0.0337 train_time:1564190ms step_avg:538.08ms step:2918/6250 train_loss:2.2192 aux_loss:0.0338 train_time:1564733ms step_avg:538.08ms step:2919/6250 train_loss:2.1624 aux_loss:0.0338 train_time:1565297ms step_avg:538.09ms step:2920/6250 train_loss:2.2485 aux_loss:0.0337 train_time:1565841ms step_avg:538.09ms step:2921/6250 train_loss:2.2981 aux_loss:0.0337 train_time:1566373ms step_avg:538.09ms step:2922/6250 train_loss:2.3082 aux_loss:0.0343 train_time:1566904ms step_avg:538.09ms step:2923/6250 train_loss:2.1877 aux_loss:0.0341 train_time:1567437ms step_avg:538.08ms step:2924/6250 train_loss:2.3235 aux_loss:0.0348 train_time:1567968ms step_avg:538.08ms step:2925/6250 train_loss:2.2581 aux_loss:0.0337 train_time:1568501ms step_avg:538.08ms step:2926/6250 train_loss:2.3097 aux_loss:0.0340 train_time:1569032ms step_avg:538.08ms step:2927/6250 train_loss:2.2204 aux_loss:0.0338 train_time:1569564ms step_avg:538.07ms step:2928/6250 train_loss:2.2661 aux_loss:0.0336 train_time:1570093ms step_avg:538.07ms step:2929/6250 train_loss:2.2179 aux_loss:0.0337 train_time:1570626ms step_avg:538.07ms step:2930/6250 train_loss:2.3785 aux_loss:0.0336 train_time:1571158ms step_avg:538.07ms step:2931/6250 train_loss:2.2634 aux_loss:0.0331 train_time:1571691ms step_avg:538.07ms step:2932/6250 train_loss:2.3300 aux_loss:0.0337 train_time:1572250ms step_avg:538.07ms step:2933/6250 train_loss:2.2808 aux_loss:0.0340 train_time:1572807ms step_avg:538.08ms step:2934/6250 train_loss:2.4629 aux_loss:0.0339 train_time:1573341ms step_avg:538.08ms step:2935/6250 train_loss:2.3758 aux_loss:0.0337 train_time:1573877ms step_avg:538.08ms step:2936/6250 train_loss:2.1859 aux_loss:0.0339 train_time:1574409ms step_avg:538.08ms step:2937/6250 train_loss:2.3576 aux_loss:0.0341 train_time:1574957ms step_avg:538.08ms step:2938/6250 train_loss:2.3554 aux_loss:0.0337 train_time:1575522ms step_avg:538.09ms step:2939/6250 train_loss:2.3558 aux_loss:0.0335 train_time:1576056ms step_avg:538.09ms step:2940/6250 train_loss:2.1633 aux_loss:0.0338 train_time:1576588ms step_avg:538.08ms step:2941/6250 train_loss:2.2082 aux_loss:0.0340 train_time:1577119ms step_avg:538.08ms step:2942/6250 train_loss:2.2737 aux_loss:0.0338 train_time:1577650ms step_avg:538.08ms step:2943/6250 train_loss:2.2708 aux_loss:0.0340 train_time:1578181ms step_avg:538.08ms step:2944/6250 train_loss:2.2827 aux_loss:0.0337 train_time:1578726ms step_avg:538.08ms step:2945/6250 train_loss:2.3856 aux_loss:0.0338 train_time:1579258ms step_avg:538.08ms step:2946/6250 train_loss:2.1424 aux_loss:0.0340 train_time:1579796ms step_avg:538.08ms step:2947/6250 train_loss:2.3505 aux_loss:0.0341 train_time:1580326ms step_avg:538.07ms step:2948/6250 train_loss:2.1836 aux_loss:0.0336 train_time:1580867ms step_avg:538.08ms step:2949/6250 train_loss:2.3536 aux_loss:0.0337 train_time:1581414ms step_avg:538.08ms step:2950/6250 train_loss:2.2682 aux_loss:0.0341 train_time:1581950ms step_avg:538.08ms step:2951/6250 train_loss:2.3013 aux_loss:0.0337 train_time:1582488ms step_avg:538.08ms step:2952/6250 train_loss:2.2866 aux_loss:0.0341 train_time:1583031ms step_avg:538.08ms step:2953/6250 train_loss:2.3920 aux_loss:0.0340 train_time:1583565ms step_avg:538.08ms step:2954/6250 train_loss:2.2520 aux_loss:0.0335 train_time:1584100ms step_avg:538.08ms step:2955/6250 train_loss:2.2948 aux_loss:0.0336 train_time:1584633ms step_avg:538.08ms step:2956/6250 train_loss:2.2711 aux_loss:0.0340 train_time:1585164ms step_avg:538.07ms step:2957/6250 train_loss:2.2075 aux_loss:0.0340 train_time:1585696ms step_avg:538.07ms step:2958/6250 train_loss:2.1649 aux_loss:0.0338 train_time:1586240ms step_avg:538.07ms step:2959/6250 train_loss:2.2572 aux_loss:0.0341 train_time:1586773ms step_avg:538.07ms step:2960/6250 train_loss:2.3232 aux_loss:0.0339 train_time:1587305ms step_avg:538.07ms step:2961/6250 train_loss:2.2764 aux_loss:0.0338 train_time:1587853ms step_avg:538.07ms step:2962/6250 train_loss:2.2420 aux_loss:0.0336 train_time:1588390ms step_avg:538.07ms step:2963/6250 train_loss:2.2746 aux_loss:0.0333 train_time:1588954ms step_avg:538.08ms step:2964/6250 train_loss:2.2814 aux_loss:0.0333 train_time:1589501ms step_avg:538.08ms step:2965/6250 train_loss:2.3074 aux_loss:0.0334 train_time:1590045ms step_avg:538.09ms step:2966/6250 train_loss:2.2257 aux_loss:0.0335 train_time:1590599ms step_avg:538.09ms step:2967/6250 train_loss:2.2523 aux_loss:0.0336 train_time:1591130ms step_avg:538.09ms step:2968/6250 train_loss:2.2856 aux_loss:0.0334 train_time:1591662ms step_avg:538.09ms step:2969/6250 train_loss:2.2401 aux_loss:0.0335 train_time:1592213ms step_avg:538.09ms step:2970/6250 train_loss:2.2641 aux_loss:0.0341 train_time:1592778ms step_avg:538.10ms step:2971/6250 train_loss:2.1938 aux_loss:0.0339 train_time:1593312ms step_avg:538.10ms step:2972/6250 train_loss:2.1750 aux_loss:0.0334 train_time:1593849ms step_avg:538.10ms step:2973/6250 train_loss:2.3320 aux_loss:0.0337 train_time:1594380ms step_avg:538.10ms step:2974/6250 train_loss:2.3505 aux_loss:0.0338 train_time:1594911ms step_avg:538.09ms step:2975/6250 train_loss:2.2929 aux_loss:0.0333 train_time:1595446ms step_avg:538.09ms step:2976/6250 train_loss:2.2124 aux_loss:0.0335 train_time:1595984ms step_avg:538.09ms step:2977/6250 train_loss:2.2305 aux_loss:0.0336 train_time:1596516ms step_avg:538.09ms step:2978/6250 train_loss:2.2139 aux_loss:0.0333 train_time:1597049ms step_avg:538.09ms step:2979/6250 train_loss:2.3032 aux_loss:0.0334 train_time:1597581ms step_avg:538.09ms step:2980/6250 train_loss:2.3375 aux_loss:0.0337 train_time:1598121ms step_avg:538.09ms step:2981/6250 train_loss:2.2870 aux_loss:0.0338 train_time:1598653ms step_avg:538.09ms step:2982/6250 train_loss:2.3197 aux_loss:0.0337 train_time:1599184ms step_avg:538.08ms step:2983/6250 train_loss:2.3743 aux_loss:0.0340 train_time:1599717ms step_avg:538.08ms step:2984/6250 train_loss:2.2749 aux_loss:0.0336 train_time:1600249ms step_avg:538.08ms step:2985/6250 train_loss:2.2108 aux_loss:0.0336 train_time:1600794ms step_avg:538.08ms step:2986/6250 train_loss:2.4113 aux_loss:0.0336 train_time:1601325ms step_avg:538.08ms step:2987/6250 train_loss:2.3287 aux_loss:0.0336 train_time:1601889ms step_avg:538.09ms step:2988/6250 train_loss:2.2612 aux_loss:0.0331 train_time:1602449ms step_avg:538.10ms step:2989/6250 train_loss:2.2757 aux_loss:0.0337 train_time:1602980ms step_avg:538.09ms step:2990/6250 train_loss:2.2646 aux_loss:0.0333 train_time:1603511ms step_avg:538.09ms step:2991/6250 train_loss:2.3197 aux_loss:0.0337 train_time:1604064ms step_avg:538.10ms step:2992/6250 train_loss:2.3807 aux_loss:0.0335 train_time:1604595ms step_avg:538.09ms step:2993/6250 train_loss:2.2232 aux_loss:0.0331 train_time:1605126ms step_avg:538.09ms step:2994/6250 train_loss:2.3758 aux_loss:0.0335 train_time:1605657ms step_avg:538.09ms step:2995/6250 train_loss:2.3454 aux_loss:0.0340 train_time:1606190ms step_avg:538.09ms step:2996/6250 train_loss:2.3724 aux_loss:0.0340 train_time:1606734ms step_avg:538.09ms step:2997/6250 train_loss:2.2968 aux_loss:0.0334 train_time:1607281ms step_avg:538.09ms step:2998/6250 train_loss:2.2889 aux_loss:0.0337 train_time:1607831ms step_avg:538.10ms step:2999/6250 train_loss:2.3125 aux_loss:0.0337 train_time:1608376ms step_avg:538.10ms step:3000/6250 train_loss:2.2810 aux_loss:0.0334 train_time:1608926ms step_avg:538.10ms step:3001/6250 train_loss:2.3501 aux_loss:0.0332 train_time:1609463ms step_avg:538.10ms step:3002/6250 train_loss:2.2635 aux_loss:0.0332 train_time:1609995ms step_avg:538.10ms step:3003/6250 train_loss:2.2206 aux_loss:0.0336 train_time:1610528ms step_avg:538.10ms step:3004/6250 train_loss:2.2541 aux_loss:0.0335 train_time:1611071ms step_avg:538.10ms step:3005/6250 train_loss:2.1857 aux_loss:0.0332 train_time:1611603ms step_avg:538.10ms step:3006/6250 train_loss:2.1729 aux_loss:0.0331 train_time:1612151ms step_avg:538.10ms step:3007/6250 train_loss:2.3338 aux_loss:0.0338 train_time:1612709ms step_avg:538.11ms step:3008/6250 train_loss:2.2637 aux_loss:0.0333 train_time:1613251ms step_avg:538.11ms step:3009/6250 train_loss:2.2242 aux_loss:0.0331 train_time:1613783ms step_avg:538.11ms step:3010/6250 train_loss:2.2573 aux_loss:0.0332 train_time:1614313ms step_avg:538.10ms step:3011/6250 train_loss:2.2689 aux_loss:0.0336 train_time:1614844ms step_avg:538.10ms step:3012/6250 train_loss:2.3256 aux_loss:0.0338 train_time:1615375ms step_avg:538.10ms step:3013/6250 train_loss:2.1854 aux_loss:0.0331 train_time:1615925ms step_avg:538.10ms step:3014/6250 train_loss:2.3325 aux_loss:0.0333 train_time:1616456ms step_avg:538.10ms step:3015/6250 train_loss:2.3335 aux_loss:0.0335 train_time:1616987ms step_avg:538.10ms step:3016/6250 train_loss:2.3485 aux_loss:0.0335 train_time:1617518ms step_avg:538.10ms step:3017/6250 train_loss:2.2688 aux_loss:0.0331 train_time:1618049ms step_avg:538.09ms step:3018/6250 train_loss:2.3842 aux_loss:0.0332 train_time:1618580ms step_avg:538.09ms step:3019/6250 train_loss:2.2563 aux_loss:0.0334 train_time:1619122ms step_avg:538.09ms step:3020/6250 train_loss:2.2672 aux_loss:0.0337 train_time:1619662ms step_avg:538.09ms step:3021/6250 train_loss:2.3307 aux_loss:0.0337 train_time:1620212ms step_avg:538.10ms step:3022/6250 train_loss:2.2614 aux_loss:0.0332 train_time:1620781ms step_avg:538.11ms step:3023/6250 train_loss:2.3184 aux_loss:0.0335 train_time:1621333ms step_avg:538.11ms step:3024/6250 train_loss:2.2212 aux_loss:0.0336 train_time:1621875ms step_avg:538.11ms step:3025/6250 train_loss:2.3006 aux_loss:0.0333 train_time:1622420ms step_avg:538.12ms step:3026/6250 train_loss:2.2774 aux_loss:0.0332 train_time:1622971ms step_avg:538.12ms step:3027/6250 train_loss:2.3322 aux_loss:0.0330 train_time:1623518ms step_avg:538.12ms step:3028/6250 train_loss:2.2021 aux_loss:0.0331 train_time:1624060ms step_avg:538.12ms step:3029/6250 train_loss:2.2321 aux_loss:0.0333 train_time:1624604ms step_avg:538.13ms step:3030/6250 train_loss:2.2306 aux_loss:0.0333 train_time:1625150ms step_avg:538.13ms step:3031/6250 train_loss:2.3379 aux_loss:0.0331 train_time:1625693ms step_avg:538.13ms step:3032/6250 train_loss:2.2844 aux_loss:0.0332 train_time:1626241ms step_avg:538.13ms step:3033/6250 train_loss:2.1751 aux_loss:0.0333 train_time:1626784ms step_avg:538.14ms step:3034/6250 train_loss:2.2589 aux_loss:0.0333 train_time:1627325ms step_avg:538.14ms step:3035/6250 train_loss:2.2672 aux_loss:0.0333 train_time:1627868ms step_avg:538.14ms step:3036/6250 train_loss:2.2654 aux_loss:0.0336 train_time:1628417ms step_avg:538.14ms step:3037/6250 train_loss:2.2044 aux_loss:0.0336 train_time:1628982ms step_avg:538.15ms step:3038/6250 train_loss:2.4222 aux_loss:0.0336 train_time:1629527ms step_avg:538.15ms step:3039/6250 train_loss:2.2873 aux_loss:0.0331 train_time:1630106ms step_avg:538.17ms step:3040/6250 train_loss:2.2542 aux_loss:0.0332 train_time:1630646ms step_avg:538.17ms step:3041/6250 train_loss:2.2526 aux_loss:0.0335 train_time:1631206ms step_avg:538.17ms step:3042/6250 train_loss:2.3386 aux_loss:0.0332 train_time:1631771ms step_avg:538.18ms step:3043/6250 train_loss:2.2690 aux_loss:0.0333 train_time:1632322ms step_avg:538.19ms step:3044/6250 train_loss:2.3240 aux_loss:0.0336 train_time:1632886ms step_avg:538.20ms step:3045/6250 train_loss:2.2796 aux_loss:0.0336 train_time:1633437ms step_avg:538.20ms step:3046/6250 train_loss:2.2481 aux_loss:0.0333 train_time:1633982ms step_avg:538.20ms step:3047/6250 train_loss:2.3564 aux_loss:0.0334 train_time:1634533ms step_avg:538.21ms step:3048/6250 train_loss:2.2776 aux_loss:0.0334 train_time:1635084ms step_avg:538.21ms step:3049/6250 train_loss:2.3156 aux_loss:0.0333 train_time:1635624ms step_avg:538.21ms step:3050/6250 train_loss:2.2332 aux_loss:0.0331 train_time:1636167ms step_avg:538.21ms step:3051/6250 train_loss:2.3559 aux_loss:0.0330 train_time:1636703ms step_avg:538.21ms step:3052/6250 train_loss:2.2875 aux_loss:0.0334 train_time:1637368ms step_avg:538.25ms step:3053/6250 train_loss:2.2277 aux_loss:0.0335 train_time:1637917ms step_avg:538.26ms step:3054/6250 train_loss:2.1967 aux_loss:0.0332 train_time:1638464ms step_avg:538.26ms step:3055/6250 train_loss:2.2280 aux_loss:0.0327 train_time:1638994ms step_avg:538.26ms step:3056/6250 train_loss:2.1545 aux_loss:0.0330 train_time:1639525ms step_avg:538.25ms step:3057/6250 train_loss:2.3275 aux_loss:0.0334 train_time:1640057ms step_avg:538.25ms step:3058/6250 train_loss:2.2011 aux_loss:0.0332 train_time:1640589ms step_avg:538.25ms step:3059/6250 train_loss:2.2905 aux_loss:0.0333 train_time:1641120ms step_avg:538.25ms step:3060/6250 train_loss:2.2773 aux_loss:0.0330 train_time:1641650ms step_avg:538.25ms step:3061/6250 train_loss:2.2544 aux_loss:0.0330 train_time:1642192ms step_avg:538.25ms step:3062/6250 train_loss:2.2486 aux_loss:0.0329 train_time:1642723ms step_avg:538.24ms step:3063/6250 train_loss:2.2706 aux_loss:0.0330 train_time:1643255ms step_avg:538.24ms step:3064/6250 train_loss:2.2735 aux_loss:0.0334 train_time:1643785ms step_avg:538.24ms step:3065/6250 train_loss:2.2482 aux_loss:0.0334 train_time:1644316ms step_avg:538.24ms step:3066/6250 train_loss:2.2667 aux_loss:0.0335 train_time:1644886ms step_avg:538.25ms step:3067/6250 train_loss:2.2894 aux_loss:0.0331 train_time:1645436ms step_avg:538.25ms step:3068/6250 train_loss:2.2645 aux_loss:0.0335 train_time:1645993ms step_avg:538.26ms step:3069/6250 train_loss:2.2178 aux_loss:0.0336 train_time:1646559ms step_avg:538.27ms step:3070/6250 train_loss:2.3182 aux_loss:0.0332 train_time:1647097ms step_avg:538.27ms step:3071/6250 train_loss:2.2434 aux_loss:0.0332 train_time:1647628ms step_avg:538.26ms step:3072/6250 train_loss:2.2028 aux_loss:0.0336 train_time:1648167ms step_avg:538.26ms step:3073/6250 train_loss:2.3102 aux_loss:0.0334 train_time:1648701ms step_avg:538.26ms step:3074/6250 train_loss:2.2852 aux_loss:0.0335 train_time:1649232ms step_avg:538.26ms step:3075/6250 train_loss:2.2510 aux_loss:0.0335 train_time:1649763ms step_avg:538.26ms step:3076/6250 train_loss:2.2442 aux_loss:0.0333 train_time:1650301ms step_avg:538.26ms step:3077/6250 train_loss:2.2679 aux_loss:0.0331 train_time:1650832ms step_avg:538.26ms step:3078/6250 train_loss:2.1935 aux_loss:0.0332 train_time:1651370ms step_avg:538.26ms step:3079/6250 train_loss:2.2614 aux_loss:0.0328 train_time:1651902ms step_avg:538.25ms step:3080/6250 train_loss:2.2838 aux_loss:0.0327 train_time:1652449ms step_avg:538.26ms step:3081/6250 train_loss:2.1723 aux_loss:0.0332 train_time:1653007ms step_avg:538.26ms step:3082/6250 train_loss:2.3616 aux_loss:0.0336 train_time:1653549ms step_avg:538.26ms step:3083/6250 train_loss:2.2650 aux_loss:0.0327 train_time:1654085ms step_avg:538.26ms step:3084/6250 train_loss:2.3182 aux_loss:0.0330 train_time:1654643ms step_avg:538.27ms step:3085/6250 train_loss:2.2343 aux_loss:0.0334 train_time:1655193ms step_avg:538.27ms step:3086/6250 train_loss:2.2244 aux_loss:0.0336 train_time:1655737ms step_avg:538.28ms step:3087/6250 train_loss:2.2083 aux_loss:0.0332 train_time:1656269ms step_avg:538.27ms step:3088/6250 train_loss:2.2713 aux_loss:0.0334 train_time:1656800ms step_avg:538.27ms step:3089/6250 train_loss:2.2969 aux_loss:0.0333 train_time:1657332ms step_avg:538.27ms step:3090/6250 train_loss:2.2892 aux_loss:0.0332 train_time:1657863ms step_avg:538.27ms step:3091/6250 train_loss:2.2100 aux_loss:0.0329 train_time:1658403ms step_avg:538.27ms step:3092/6250 train_loss:2.2989 aux_loss:0.0331 train_time:1658944ms step_avg:538.27ms step:3093/6250 train_loss:2.2808 aux_loss:0.0334 train_time:1659499ms step_avg:538.27ms step:3094/6250 train_loss:2.2854 aux_loss:0.0333 train_time:1660039ms step_avg:538.27ms step:3095/6250 train_loss:2.2681 aux_loss:0.0329 train_time:1660571ms step_avg:538.27ms step:3096/6250 train_loss:2.2700 aux_loss:0.0333 train_time:1661109ms step_avg:538.27ms step:3097/6250 train_loss:2.2537 aux_loss:0.0331 train_time:1661648ms step_avg:538.27ms step:3098/6250 train_loss:2.1620 aux_loss:0.0329 train_time:1662198ms step_avg:538.28ms step:3099/6250 train_loss:2.2896 aux_loss:0.0330 train_time:1662730ms step_avg:538.27ms step:3100/6250 train_loss:2.2122 aux_loss:0.0331 train_time:1663272ms step_avg:538.28ms step:3101/6250 train_loss:2.2953 aux_loss:0.0332 train_time:1663829ms step_avg:538.28ms step:3102/6250 train_loss:2.2414 aux_loss:0.0329 train_time:1664369ms step_avg:538.28ms step:3103/6250 train_loss:2.1962 aux_loss:0.0329 train_time:1664924ms step_avg:538.29ms step:3104/6250 train_loss:2.2957 aux_loss:0.0334 train_time:1665479ms step_avg:538.29ms step:3105/6250 train_loss:2.2623 aux_loss:0.0333 train_time:1666010ms step_avg:538.29ms step:3106/6250 train_loss:2.2738 aux_loss:0.0327 train_time:1666554ms step_avg:538.29ms step:3107/6250 train_loss:2.3236 aux_loss:0.0329 train_time:1667096ms step_avg:538.29ms step:3108/6250 train_loss:2.2744 aux_loss:0.0331 train_time:1667641ms step_avg:538.30ms step:3109/6250 train_loss:2.2281 aux_loss:0.0330 train_time:1668185ms step_avg:538.30ms step:3110/6250 train_loss:2.2735 aux_loss:0.0329 train_time:1668732ms step_avg:538.30ms step:3111/6250 train_loss:2.2738 aux_loss:0.0333 train_time:1669278ms step_avg:538.30ms step:3112/6250 train_loss:2.2565 aux_loss:0.0335 train_time:1669822ms step_avg:538.30ms step:3113/6250 train_loss:2.2888 aux_loss:0.0329 train_time:1670353ms step_avg:538.30ms step:3114/6250 train_loss:2.1777 aux_loss:0.0329 train_time:1670886ms step_avg:538.30ms step:3115/6250 train_loss:2.1517 aux_loss:0.0333 train_time:1671425ms step_avg:538.30ms step:3116/6250 train_loss:2.2895 aux_loss:0.0334 train_time:1671956ms step_avg:538.30ms step:3117/6250 train_loss:2.3771 aux_loss:0.0330 train_time:1672499ms step_avg:538.30ms step:3118/6250 train_loss:2.1580 aux_loss:0.0327 train_time:1673058ms step_avg:538.31ms step:3119/6250 train_loss:2.2930 aux_loss:0.0333 train_time:1673610ms step_avg:538.31ms step:3120/6250 train_loss:2.3047 aux_loss:0.0329 train_time:1674159ms step_avg:538.31ms step:3121/6250 train_loss:2.2927 aux_loss:0.0328 train_time:1674697ms step_avg:538.31ms step:3122/6250 train_loss:2.2639 aux_loss:0.0332 train_time:1675230ms step_avg:538.31ms step:3123/6250 train_loss:2.2673 aux_loss:0.0329 train_time:1675763ms step_avg:538.31ms step:3124/6250 train_loss:2.2675 aux_loss:0.0333 train_time:1676296ms step_avg:538.31ms step:3125/6250 train_loss:2.3121 aux_loss:0.0333 train_time:1676828ms step_avg:538.31ms step:3126/6250 train_loss:2.2668 aux_loss:0.0332 train_time:1677359ms step_avg:538.31ms step:3127/6250 train_loss:2.2684 aux_loss:0.0330 train_time:1677891ms step_avg:538.30ms step:3128/6250 train_loss:2.2613 aux_loss:0.0334 train_time:1678423ms step_avg:538.30ms step:3129/6250 train_loss:2.2305 aux_loss:0.0332 train_time:1678955ms step_avg:538.30ms step:3130/6250 train_loss:2.2209 aux_loss:0.0329 train_time:1679501ms step_avg:538.30ms step:3131/6250 train_loss:2.2828 aux_loss:0.0331 train_time:1680034ms step_avg:538.30ms step:3132/6250 train_loss:2.3852 aux_loss:0.0333 train_time:1680575ms step_avg:538.30ms step:3133/6250 train_loss:2.3258 aux_loss:0.0329 train_time:1681107ms step_avg:538.30ms step:3134/6250 train_loss:2.1651 aux_loss:0.0333 train_time:1681647ms step_avg:538.30ms step:3135/6250 train_loss:2.2586 aux_loss:0.0330 train_time:1682185ms step_avg:538.30ms step:3136/6250 train_loss:2.2450 aux_loss:0.0324 train_time:1682733ms step_avg:538.30ms step:3137/6250 train_loss:2.2785 aux_loss:0.0325 train_time:1683270ms step_avg:538.30ms step:3138/6250 train_loss:2.2875 aux_loss:0.0332 train_time:1683807ms step_avg:538.30ms step:3139/6250 train_loss:2.3297 aux_loss:0.0330 train_time:1684356ms step_avg:538.30ms step:3140/6250 train_loss:2.2602 aux_loss:0.0324 train_time:1684902ms step_avg:538.31ms step:3141/6250 train_loss:2.2908 aux_loss:0.0327 train_time:1685452ms step_avg:538.31ms step:3142/6250 train_loss:2.2162 aux_loss:0.0327 train_time:1685998ms step_avg:538.31ms step:3143/6250 train_loss:2.2612 aux_loss:0.0327 train_time:1686532ms step_avg:538.31ms step:3144/6250 train_loss:2.2526 aux_loss:0.0326 train_time:1687064ms step_avg:538.31ms step:3145/6250 train_loss:2.2838 aux_loss:0.0324 train_time:1687595ms step_avg:538.31ms step:3146/6250 train_loss:2.3625 aux_loss:0.0328 train_time:1688132ms step_avg:538.31ms step:3147/6250 train_loss:2.2524 aux_loss:0.0328 train_time:1688681ms step_avg:538.31ms step:3148/6250 train_loss:2.2383 aux_loss:0.0328 train_time:1689216ms step_avg:538.31ms step:3149/6250 train_loss:2.3191 aux_loss:0.0328 train_time:1689764ms step_avg:538.31ms step:3150/6250 train_loss:2.2641 aux_loss:0.0329 train_time:1690300ms step_avg:538.31ms step:3151/6250 train_loss:2.3228 aux_loss:0.0333 train_time:1690837ms step_avg:538.31ms step:3152/6250 train_loss:2.2338 aux_loss:0.0331 train_time:1691370ms step_avg:538.31ms step:3153/6250 train_loss:2.2531 aux_loss:0.0329 train_time:1691905ms step_avg:538.31ms step:3154/6250 train_loss:2.3784 aux_loss:0.0325 train_time:1692453ms step_avg:538.31ms step:3155/6250 train_loss:2.2437 aux_loss:0.0325 train_time:1693029ms step_avg:538.32ms step:3156/6250 train_loss:2.2182 aux_loss:0.0329 train_time:1693594ms step_avg:538.33ms step:3157/6250 train_loss:2.2417 aux_loss:0.0325 train_time:1694124ms step_avg:538.33ms step:3158/6250 train_loss:2.3795 aux_loss:0.0325 train_time:1694658ms step_avg:538.33ms step:3159/6250 train_loss:2.2737 aux_loss:0.0327 train_time:1695195ms step_avg:538.33ms step:3160/6250 train_loss:2.2813 aux_loss:0.0333 train_time:1695732ms step_avg:538.33ms step:3161/6250 train_loss:2.2888 aux_loss:0.0332 train_time:1696271ms step_avg:538.33ms step:3162/6250 train_loss:2.3173 aux_loss:0.0325 train_time:1696823ms step_avg:538.33ms step:3163/6250 train_loss:2.2184 aux_loss:0.0325 train_time:1697355ms step_avg:538.33ms step:3164/6250 train_loss:2.2878 aux_loss:0.0330 train_time:1697888ms step_avg:538.33ms step:3165/6250 train_loss:2.2767 aux_loss:0.0329 train_time:1698421ms step_avg:538.33ms step:3166/6250 train_loss:2.2534 aux_loss:0.0328 train_time:1698953ms step_avg:538.32ms step:3167/6250 train_loss:2.2777 aux_loss:0.0324 train_time:1699486ms step_avg:538.32ms step:3168/6250 train_loss:2.3183 aux_loss:0.0325 train_time:1700016ms step_avg:538.32ms step:3169/6250 train_loss:2.3809 aux_loss:0.0327 train_time:1700549ms step_avg:538.32ms step:3170/6250 train_loss:2.3445 aux_loss:0.0328 train_time:1701080ms step_avg:538.32ms step:3171/6250 train_loss:2.2208 aux_loss:0.0328 train_time:1701614ms step_avg:538.32ms step:3172/6250 train_loss:2.2971 aux_loss:0.0328 train_time:1702146ms step_avg:538.31ms step:3173/6250 train_loss:2.2520 aux_loss:0.0330 train_time:1702678ms step_avg:538.31ms step:3174/6250 train_loss:2.2760 aux_loss:0.0332 train_time:1703210ms step_avg:538.31ms step:3175/6250 train_loss:2.3016 aux_loss:0.0331 train_time:1703752ms step_avg:538.31ms step:3176/6250 train_loss:2.3061 aux_loss:0.0332 train_time:1704283ms step_avg:538.31ms step:3177/6250 train_loss:2.1969 aux_loss:0.0328 train_time:1704815ms step_avg:538.31ms step:3178/6250 train_loss:2.2794 aux_loss:0.0330 train_time:1705345ms step_avg:538.30ms step:3179/6250 train_loss:2.2814 aux_loss:0.0331 train_time:1705877ms step_avg:538.30ms step:3180/6250 train_loss:2.2630 aux_loss:0.0330 train_time:1706409ms step_avg:538.30ms step:3181/6250 train_loss:2.2801 aux_loss:0.0328 train_time:1706940ms step_avg:538.30ms step:3182/6250 train_loss:2.1983 aux_loss:0.0324 train_time:1707472ms step_avg:538.30ms step:3183/6250 train_loss:2.1938 aux_loss:0.0327 train_time:1708003ms step_avg:538.29ms step:3184/6250 train_loss:2.2856 aux_loss:0.0331 train_time:1708538ms step_avg:538.29ms step:3185/6250 train_loss:2.2575 aux_loss:0.0331 train_time:1709069ms step_avg:538.29ms step:3186/6250 train_loss:2.2741 aux_loss:0.0328 train_time:1709601ms step_avg:538.29ms step:3187/6250 train_loss:2.3192 aux_loss:0.0331 train_time:1710138ms step_avg:538.29ms step:3188/6250 train_loss:2.2218 aux_loss:0.0326 train_time:1710670ms step_avg:538.29ms step:3189/6250 train_loss:2.3096 aux_loss:0.0327 train_time:1711204ms step_avg:538.28ms step:3190/6250 train_loss:2.3458 aux_loss:0.0335 train_time:1711738ms step_avg:538.28ms step:3191/6250 train_loss:2.2219 aux_loss:0.0332 train_time:1712271ms step_avg:538.28ms step:3192/6250 train_loss:2.2842 aux_loss:0.0325 train_time:1712817ms step_avg:538.28ms step:3193/6250 train_loss:2.2586 aux_loss:0.0326 train_time:1713376ms step_avg:538.29ms step:3194/6250 train_loss:2.2118 aux_loss:0.0326 train_time:1713916ms step_avg:538.29ms step:3195/6250 train_loss:2.2912 aux_loss:0.0326 train_time:1714447ms step_avg:538.29ms step:3196/6250 train_loss:2.2386 aux_loss:0.0326 train_time:1714978ms step_avg:538.29ms step:3197/6250 train_loss:2.3816 aux_loss:0.0329 train_time:1715511ms step_avg:538.28ms step:3198/6250 train_loss:2.2108 aux_loss:0.0325 train_time:1716045ms step_avg:538.28ms step:3199/6250 train_loss:2.2466 aux_loss:0.0325 train_time:1716578ms step_avg:538.28ms step:3200/6250 train_loss:2.1992 aux_loss:0.0328 train_time:1717110ms step_avg:538.28ms step:3201/6250 train_loss:2.2949 aux_loss:0.0329 train_time:1717643ms step_avg:538.28ms step:3202/6250 train_loss:2.1973 aux_loss:0.0323 train_time:1718174ms step_avg:538.28ms step:3203/6250 train_loss:2.2453 aux_loss:0.0324 train_time:1718706ms step_avg:538.27ms step:3204/6250 train_loss:2.2859 aux_loss:0.0328 train_time:1719238ms step_avg:538.27ms step:3205/6250 train_loss:2.4150 aux_loss:0.0330 train_time:1719770ms step_avg:538.27ms step:3206/6250 train_loss:2.1794 aux_loss:0.0325 train_time:1720301ms step_avg:538.27ms step:3207/6250 train_loss:2.2812 aux_loss:0.0329 train_time:1720833ms step_avg:538.27ms step:3208/6250 train_loss:2.2142 aux_loss:0.0330 train_time:1721364ms step_avg:538.26ms step:3209/6250 train_loss:2.2292 aux_loss:0.0329 train_time:1721897ms step_avg:538.26ms step:3210/6250 train_loss:2.3274 aux_loss:0.0328 train_time:1722432ms step_avg:538.26ms step:3211/6250 train_loss:2.2372 aux_loss:0.0324 train_time:1722965ms step_avg:538.26ms step:3212/6250 train_loss:2.3024 aux_loss:0.0328 train_time:1723498ms step_avg:538.26ms step:3213/6250 train_loss:2.2259 aux_loss:0.0327 train_time:1724033ms step_avg:538.26ms step:3214/6250 train_loss:2.2739 aux_loss:0.0325 train_time:1724567ms step_avg:538.25ms step:3215/6250 train_loss:2.3367 aux_loss:0.0326 train_time:1725099ms step_avg:538.25ms step:3216/6250 train_loss:2.2971 aux_loss:0.0328 train_time:1725631ms step_avg:538.25ms step:3217/6250 train_loss:2.2186 aux_loss:0.0327 train_time:1726163ms step_avg:538.25ms step:3218/6250 train_loss:2.4320 aux_loss:0.0327 train_time:1726695ms step_avg:538.25ms step:3219/6250 train_loss:2.2369 aux_loss:0.0327 train_time:1727227ms step_avg:538.24ms step:3220/6250 train_loss:2.2938 aux_loss:0.0328 train_time:1727759ms step_avg:538.24ms step:3221/6250 train_loss:2.2737 aux_loss:0.0328 train_time:1728290ms step_avg:538.24ms step:3222/6250 train_loss:2.2702 aux_loss:0.0326 train_time:1728822ms step_avg:538.24ms step:3223/6250 train_loss:2.2428 aux_loss:0.0327 train_time:1729353ms step_avg:538.24ms step:3224/6250 train_loss:2.3216 aux_loss:0.0328 train_time:1729884ms step_avg:538.23ms step:3225/6250 train_loss:2.3146 aux_loss:0.0325 train_time:1730416ms step_avg:538.23ms step:3226/6250 train_loss:2.3123 aux_loss:0.0326 train_time:1730947ms step_avg:538.23ms step:3227/6250 train_loss:2.1810 aux_loss:0.0329 train_time:1731478ms step_avg:538.23ms step:3228/6250 train_loss:2.2786 aux_loss:0.0330 train_time:1732011ms step_avg:538.23ms step:3229/6250 train_loss:2.2200 aux_loss:0.0324 train_time:1732544ms step_avg:538.22ms step:3230/6250 train_loss:2.3070 aux_loss:0.0328 train_time:1733100ms step_avg:538.23ms step:3231/6250 train_loss:2.1586 aux_loss:0.0326 train_time:1733659ms step_avg:538.24ms step:3232/6250 train_loss:2.1800 aux_loss:0.0322 train_time:1734199ms step_avg:538.24ms step:3233/6250 train_loss:2.2494 aux_loss:0.0324 train_time:1734735ms step_avg:538.24ms step:3234/6250 train_loss:2.2374 aux_loss:0.0328 train_time:1735270ms step_avg:538.24ms step:3235/6250 train_loss:2.2528 aux_loss:0.0326 train_time:1735805ms step_avg:538.23ms step:3236/6250 train_loss:2.2979 aux_loss:0.0327 train_time:1736335ms step_avg:538.23ms step:3237/6250 train_loss:2.2935 aux_loss:0.0324 train_time:1736867ms step_avg:538.23ms step:3238/6250 train_loss:2.2271 aux_loss:0.0322 train_time:1737397ms step_avg:538.23ms step:3239/6250 train_loss:2.2536 aux_loss:0.0323 train_time:1737929ms step_avg:538.23ms step:3240/6250 train_loss:2.2091 aux_loss:0.0322 train_time:1738460ms step_avg:538.22ms step:3241/6250 train_loss:2.2381 aux_loss:0.0328 train_time:1738991ms step_avg:538.22ms step:3242/6250 train_loss:2.2803 aux_loss:0.0327 train_time:1739524ms step_avg:538.22ms step:3243/6250 train_loss:2.2723 aux_loss:0.0325 train_time:1740055ms step_avg:538.22ms step:3244/6250 train_loss:2.2895 aux_loss:0.0325 train_time:1740589ms step_avg:538.22ms step:3245/6250 train_loss:2.2108 aux_loss:0.0324 train_time:1741159ms step_avg:538.23ms step:3246/6250 train_loss:2.3435 aux_loss:0.0326 train_time:1741707ms step_avg:538.23ms step:3247/6250 train_loss:2.3005 aux_loss:0.0326 train_time:1742249ms step_avg:538.23ms step:3248/6250 train_loss:2.1932 aux_loss:0.0323 train_time:1742783ms step_avg:538.23ms step:3249/6250 train_loss:2.1606 aux_loss:0.0323 train_time:1743316ms step_avg:538.23ms step:3250/6250 train_loss:2.2339 aux_loss:0.0326 train_time:1743848ms step_avg:538.22ms step:3251/6250 train_loss:2.3290 aux_loss:0.0326 train_time:1744380ms step_avg:538.22ms step:3252/6250 train_loss:2.2431 aux_loss:0.0325 train_time:1744912ms step_avg:538.22ms step:3253/6250 train_loss:2.2351 aux_loss:0.0326 train_time:1745442ms step_avg:538.22ms step:3254/6250 train_loss:2.1684 aux_loss:0.0325 train_time:1745975ms step_avg:538.22ms step:3255/6250 train_loss:2.2418 aux_loss:0.0323 train_time:1746507ms step_avg:538.21ms step:3256/6250 train_loss:2.2326 aux_loss:0.0322 train_time:1747039ms step_avg:538.21ms step:3257/6250 train_loss:2.2017 aux_loss:0.0322 train_time:1747583ms step_avg:538.21ms step:3258/6250 train_loss:2.2955 aux_loss:0.0322 train_time:1748138ms step_avg:538.22ms step:3259/6250 train_loss:2.2150 aux_loss:0.0324 train_time:1748698ms step_avg:538.23ms step:3260/6250 train_loss:2.3198 aux_loss:0.0325 train_time:1749227ms step_avg:538.22ms step:3261/6250 train_loss:2.3814 aux_loss:0.0323 train_time:1749760ms step_avg:538.22ms step:3262/6250 train_loss:2.2947 aux_loss:0.0323 train_time:1750291ms step_avg:538.22ms step:3263/6250 train_loss:2.2876 aux_loss:0.0327 train_time:1750825ms step_avg:538.22ms step:3264/6250 train_loss:2.2526 aux_loss:0.0325 train_time:1751358ms step_avg:538.22ms step:3265/6250 train_loss:2.3095 aux_loss:0.0324 train_time:1751891ms step_avg:538.22ms step:3266/6250 train_loss:2.2708 aux_loss:0.0323 train_time:1752422ms step_avg:538.21ms step:3267/6250 train_loss:2.2628 aux_loss:0.0326 train_time:1752975ms step_avg:538.22ms step:3268/6250 train_loss:2.2871 aux_loss:0.0327 train_time:1753542ms step_avg:538.23ms step:3269/6250 train_loss:2.2315 aux_loss:0.0331 train_time:1754085ms step_avg:538.23ms step:3270/6250 train_loss:2.2832 aux_loss:0.0327 train_time:1754619ms step_avg:538.23ms step:3271/6250 train_loss:2.2946 aux_loss:0.0324 train_time:1755152ms step_avg:538.23ms step:3272/6250 train_loss:2.2367 aux_loss:0.0323 train_time:1755685ms step_avg:538.22ms step:3273/6250 train_loss:2.2466 aux_loss:0.0325 train_time:1756216ms step_avg:538.22ms step:3274/6250 train_loss:2.2572 aux_loss:0.0329 train_time:1756753ms step_avg:538.22ms step:3275/6250 train_loss:2.2206 aux_loss:0.0327 train_time:1757293ms step_avg:538.22ms step:3276/6250 train_loss:2.3409 aux_loss:0.0330 train_time:1757826ms step_avg:538.22ms step:3277/6250 train_loss:2.1953 aux_loss:0.0324 train_time:1758357ms step_avg:538.22ms step:3278/6250 train_loss:2.2676 aux_loss:0.0325 train_time:1758892ms step_avg:538.22ms step:3279/6250 train_loss:2.2610 aux_loss:0.0319 train_time:1759425ms step_avg:538.22ms step:3280/6250 train_loss:2.4251 aux_loss:0.0324 train_time:1759955ms step_avg:538.21ms step:3281/6250 train_loss:2.3155 aux_loss:0.0325 train_time:1760486ms step_avg:538.21ms step:3282/6250 train_loss:2.1632 aux_loss:0.0321 train_time:1761029ms step_avg:538.21ms step:3283/6250 train_loss:2.2754 aux_loss:0.0320 train_time:1761586ms step_avg:538.22ms step:3284/6250 train_loss:2.1958 aux_loss:0.0324 train_time:1762117ms step_avg:538.22ms step:3285/6250 train_loss:2.3237 aux_loss:0.0324 train_time:1762648ms step_avg:538.21ms step:3286/6250 train_loss:2.3329 aux_loss:0.0323 train_time:1763178ms step_avg:538.21ms step:3287/6250 train_loss:2.2204 aux_loss:0.0321 train_time:1763710ms step_avg:538.21ms step:3288/6250 train_loss:2.2693 aux_loss:0.0323 train_time:1764240ms step_avg:538.21ms step:3289/6250 train_loss:2.1929 aux_loss:0.0320 train_time:1764789ms step_avg:538.21ms step:3290/6250 train_loss:2.2309 aux_loss:0.0323 train_time:1765338ms step_avg:538.21ms step:3291/6250 train_loss:2.2848 aux_loss:0.0326 train_time:1765874ms step_avg:538.21ms step:3292/6250 train_loss:2.3080 aux_loss:0.0319 train_time:1766406ms step_avg:538.21ms step:3293/6250 train_loss:2.2694 aux_loss:0.0324 train_time:1766936ms step_avg:538.21ms step:3294/6250 train_loss:2.1858 aux_loss:0.0324 train_time:1767467ms step_avg:538.21ms step:3295/6250 train_loss:2.2094 aux_loss:0.0319 train_time:1768010ms step_avg:538.21ms step:3296/6250 train_loss:2.1786 aux_loss:0.0325 train_time:1768556ms step_avg:538.21ms step:3297/6250 train_loss:2.2321 aux_loss:0.0327 train_time:1769094ms step_avg:538.21ms step:3298/6250 train_loss:2.2223 aux_loss:0.0324 train_time:1769661ms step_avg:538.22ms step:3299/6250 train_loss:2.2375 aux_loss:0.0321 train_time:1770193ms step_avg:538.22ms step:3300/6250 train_loss:2.1831 aux_loss:0.0323 train_time:1770741ms step_avg:538.22ms step:3301/6250 train_loss:2.2290 aux_loss:0.0323 train_time:1771284ms step_avg:538.22ms step:3302/6250 train_loss:2.1988 aux_loss:0.0321 train_time:1771835ms step_avg:538.22ms step:3303/6250 train_loss:2.3732 aux_loss:0.0323 train_time:1772366ms step_avg:538.22ms step:3304/6250 train_loss:2.3233 aux_loss:0.0323 train_time:1772901ms step_avg:538.22ms step:3305/6250 train_loss:2.1759 aux_loss:0.0324 train_time:1773462ms step_avg:538.23ms step:3306/6250 train_loss:2.2794 aux_loss:0.0325 train_time:1774023ms step_avg:538.24ms step:3307/6250 train_loss:2.2024 aux_loss:0.0321 train_time:1774564ms step_avg:538.24ms step:3308/6250 train_loss:2.3435 aux_loss:0.0323 train_time:1775094ms step_avg:538.23ms step:3309/6250 train_loss:2.2567 aux_loss:0.0326 train_time:1775627ms step_avg:538.23ms step:3310/6250 train_loss:2.2813 aux_loss:0.0324 train_time:1776180ms step_avg:538.24ms step:3311/6250 train_loss:2.2343 aux_loss:0.0323 train_time:1776712ms step_avg:538.23ms step:3312/6250 train_loss:2.2859 aux_loss:0.0324 train_time:1777243ms step_avg:538.23ms step:3313/6250 train_loss:2.2630 aux_loss:0.0324 train_time:1777773ms step_avg:538.23ms step:3314/6250 train_loss:2.2710 aux_loss:0.0321 train_time:1778304ms step_avg:538.23ms step:3315/6250 train_loss:2.2627 aux_loss:0.0323 train_time:1778834ms step_avg:538.23ms step:3316/6250 train_loss:2.3876 aux_loss:0.0323 train_time:1779367ms step_avg:538.22ms step:3317/6250 train_loss:2.2946 aux_loss:0.0325 train_time:1779911ms step_avg:538.23ms step:3318/6250 train_loss:2.1833 aux_loss:0.0324 train_time:1780446ms step_avg:538.22ms step:3319/6250 train_loss:2.2522 aux_loss:0.0326 train_time:1780977ms step_avg:538.22ms step:3320/6250 train_loss:2.2183 aux_loss:0.0322 train_time:1781509ms step_avg:538.22ms step:3321/6250 train_loss:2.2270 aux_loss:0.0322 train_time:1782040ms step_avg:538.22ms step:3322/6250 train_loss:2.2424 aux_loss:0.0327 train_time:1782572ms step_avg:538.22ms step:3323/6250 train_loss:2.1902 aux_loss:0.0326 train_time:1783102ms step_avg:538.21ms step:3324/6250 train_loss:2.2631 aux_loss:0.0326 train_time:1783634ms step_avg:538.21ms step:3325/6250 train_loss:2.1723 aux_loss:0.0322 train_time:1784164ms step_avg:538.21ms step:3326/6250 train_loss:2.2194 aux_loss:0.0322 train_time:1784696ms step_avg:538.21ms step:3327/6250 train_loss:2.2129 aux_loss:0.0319 train_time:1785227ms step_avg:538.21ms step:3328/6250 train_loss:2.2785 aux_loss:0.0323 train_time:1785758ms step_avg:538.20ms step:3329/6250 train_loss:2.2193 aux_loss:0.0322 train_time:1786289ms step_avg:538.20ms step:3330/6250 train_loss:2.2596 aux_loss:0.0321 train_time:1786819ms step_avg:538.20ms step:3331/6250 train_loss:2.2165 aux_loss:0.0321 train_time:1787351ms step_avg:538.20ms step:3332/6250 train_loss:2.2647 aux_loss:0.0324 train_time:1787882ms step_avg:538.19ms step:3333/6250 train_loss:2.2725 aux_loss:0.0320 train_time:1788417ms step_avg:538.19ms step:3334/6250 train_loss:2.2960 aux_loss:0.0322 train_time:1788957ms step_avg:538.19ms step:3335/6250 train_loss:2.1927 aux_loss:0.0323 train_time:1789489ms step_avg:538.19ms step:3336/6250 train_loss:2.1971 aux_loss:0.0324 train_time:1790036ms step_avg:538.19ms step:3337/6250 train_loss:2.2187 aux_loss:0.0324 train_time:1790567ms step_avg:538.19ms step:3338/6250 train_loss:2.1772 aux_loss:0.0321 train_time:1791099ms step_avg:538.19ms step:3339/6250 train_loss:2.2403 aux_loss:0.0320 train_time:1791631ms step_avg:538.19ms step:3340/6250 train_loss:2.2015 aux_loss:0.0321 train_time:1792161ms step_avg:538.19ms step:3341/6250 train_loss:2.2972 aux_loss:0.0323 train_time:1792693ms step_avg:538.18ms step:3342/6250 train_loss:2.3084 aux_loss:0.0324 train_time:1793240ms step_avg:538.19ms step:3343/6250 train_loss:2.2753 aux_loss:0.0318 train_time:1793801ms step_avg:538.19ms step:3344/6250 train_loss:2.3207 aux_loss:0.0321 train_time:1794340ms step_avg:538.19ms step:3345/6250 train_loss:2.2478 aux_loss:0.0324 train_time:1794872ms step_avg:538.19ms step:3346/6250 train_loss:2.1787 aux_loss:0.0321 train_time:1795404ms step_avg:538.19ms step:3347/6250 train_loss:2.2399 aux_loss:0.0318 train_time:1795935ms step_avg:538.19ms step:3348/6250 train_loss:2.3554 aux_loss:0.0323 train_time:1796466ms step_avg:538.19ms step:3349/6250 train_loss:2.2509 aux_loss:0.0325 train_time:1796997ms step_avg:538.18ms step:3350/6250 train_loss:2.2158 aux_loss:0.0320 train_time:1797529ms step_avg:538.18ms step:3351/6250 train_loss:2.2826 aux_loss:0.0323 train_time:1798061ms step_avg:538.18ms step:3352/6250 train_loss:2.3503 aux_loss:0.0324 train_time:1798593ms step_avg:538.18ms step:3353/6250 train_loss:2.2100 aux_loss:0.0320 train_time:1799126ms step_avg:538.18ms step:3354/6250 train_loss:2.1788 aux_loss:0.0319 train_time:1799656ms step_avg:538.17ms step:3355/6250 train_loss:2.2477 aux_loss:0.0323 train_time:1800187ms step_avg:538.17ms step:3356/6250 train_loss:2.2781 aux_loss:0.0321 train_time:1800717ms step_avg:538.17ms step:3357/6250 train_loss:2.2377 aux_loss:0.0321 train_time:1801249ms step_avg:538.17ms step:3358/6250 train_loss:2.2381 aux_loss:0.0327 train_time:1801779ms step_avg:538.17ms step:3359/6250 train_loss:2.1933 aux_loss:0.0331 train_time:1802309ms step_avg:538.16ms step:3360/6250 train_loss:2.1889 aux_loss:0.0328 train_time:1802840ms step_avg:538.16ms step:3361/6250 train_loss:2.3061 aux_loss:0.0322 train_time:1803371ms step_avg:538.16ms step:3362/6250 train_loss:2.2271 aux_loss:0.0322 train_time:1803901ms step_avg:538.16ms step:3363/6250 train_loss:2.2861 aux_loss:0.0322 train_time:1804433ms step_avg:538.15ms step:3364/6250 train_loss:2.2981 aux_loss:0.0323 train_time:1804963ms step_avg:538.15ms step:3365/6250 train_loss:2.2901 aux_loss:0.0329 train_time:1805495ms step_avg:538.15ms step:3366/6250 train_loss:2.1672 aux_loss:0.0323 train_time:1806026ms step_avg:538.15ms step:3367/6250 train_loss:2.2742 aux_loss:0.0318 train_time:1806559ms step_avg:538.15ms step:3368/6250 train_loss:2.3378 aux_loss:0.0320 train_time:1807087ms step_avg:538.14ms step:3369/6250 train_loss:2.2622 aux_loss:0.0324 train_time:1807620ms step_avg:538.14ms step:3370/6250 train_loss:2.2448 aux_loss:0.0322 train_time:1808161ms step_avg:538.14ms step:3371/6250 train_loss:2.2439 aux_loss:0.0321 train_time:1808693ms step_avg:538.14ms step:3372/6250 train_loss:2.3446 aux_loss:0.0323 train_time:1809224ms step_avg:538.14ms step:3373/6250 train_loss:2.2743 aux_loss:0.0319 train_time:1809756ms step_avg:538.14ms step:3374/6250 train_loss:2.2444 aux_loss:0.0320 train_time:1810287ms step_avg:538.14ms step:3375/6250 train_loss:2.2730 aux_loss:0.0324 train_time:1810828ms step_avg:538.14ms step:3376/6250 train_loss:2.2235 aux_loss:0.0317 train_time:1811369ms step_avg:538.14ms step:3377/6250 train_loss:2.2239 aux_loss:0.0317 train_time:1811903ms step_avg:538.14ms step:3378/6250 train_loss:2.2828 aux_loss:0.0318 train_time:1812442ms step_avg:538.14ms step:3379/6250 train_loss:2.3234 aux_loss:0.0322 train_time:1812982ms step_avg:538.14ms step:3380/6250 train_loss:2.3208 aux_loss:0.0321 train_time:1813538ms step_avg:538.14ms step:3381/6250 train_loss:2.3312 aux_loss:0.0317 train_time:1814106ms step_avg:538.15ms step:3382/6250 train_loss:2.2287 aux_loss:0.0315 train_time:1814642ms step_avg:538.15ms step:3383/6250 train_loss:2.2045 aux_loss:0.0319 train_time:1815174ms step_avg:538.15ms step:3384/6250 train_loss:2.1630 aux_loss:0.0315 train_time:1815709ms step_avg:538.15ms step:3385/6250 train_loss:2.1872 aux_loss:0.0317 train_time:1816242ms step_avg:538.15ms step:3386/6250 train_loss:2.2564 aux_loss:0.0317 train_time:1816776ms step_avg:538.14ms step:3387/6250 train_loss:2.2752 aux_loss:0.0312 train_time:1817317ms step_avg:538.15ms step:3388/6250 train_loss:2.2272 aux_loss:0.0319 train_time:1817851ms step_avg:538.14ms step:3389/6250 train_loss:2.3488 aux_loss:0.0322 train_time:1818382ms step_avg:538.14ms step:3390/6250 train_loss:2.2686 aux_loss:0.0319 train_time:1818913ms step_avg:538.14ms step:3391/6250 train_loss:2.3473 aux_loss:0.0319 train_time:1819447ms step_avg:538.14ms step:3392/6250 train_loss:2.1879 aux_loss:0.0321 train_time:1819977ms step_avg:538.14ms step:3393/6250 train_loss:2.2055 aux_loss:0.0316 train_time:1820512ms step_avg:538.14ms step:3394/6250 train_loss:2.2857 aux_loss:0.0318 train_time:1821043ms step_avg:538.13ms step:3395/6250 train_loss:2.2414 aux_loss:0.0320 train_time:1821575ms step_avg:538.13ms step:3396/6250 train_loss:2.2826 aux_loss:0.0320 train_time:1822106ms step_avg:538.13ms step:3397/6250 train_loss:2.2382 aux_loss:0.0317 train_time:1822639ms step_avg:538.13ms step:3398/6250 train_loss:2.2801 aux_loss:0.0317 train_time:1823174ms step_avg:538.13ms step:3399/6250 train_loss:2.2562 aux_loss:0.0319 train_time:1823708ms step_avg:538.13ms step:3400/6250 train_loss:2.2555 aux_loss:0.0318 train_time:1824241ms step_avg:538.12ms step:3401/6250 train_loss:2.2084 aux_loss:0.0321 train_time:1824772ms step_avg:538.12ms step:3402/6250 train_loss:2.2790 aux_loss:0.0319 train_time:1825307ms step_avg:538.12ms step:3403/6250 train_loss:2.2645 aux_loss:0.0314 train_time:1825838ms step_avg:538.12ms step:3404/6250 train_loss:2.1581 aux_loss:0.0322 train_time:1826370ms step_avg:538.12ms step:3405/6250 train_loss:2.2087 aux_loss:0.0326 train_time:1826901ms step_avg:538.12ms step:3406/6250 train_loss:2.2901 aux_loss:0.0322 train_time:1827433ms step_avg:538.11ms step:3407/6250 train_loss:2.2744 aux_loss:0.0317 train_time:1827964ms step_avg:538.11ms step:3408/6250 train_loss:2.1406 aux_loss:0.0318 train_time:1828498ms step_avg:538.11ms step:3409/6250 train_loss:2.1554 aux_loss:0.0321 train_time:1829028ms step_avg:538.11ms step:3410/6250 train_loss:2.2143 aux_loss:0.0319 train_time:1829561ms step_avg:538.11ms step:3411/6250 train_loss:2.2381 aux_loss:0.0320 train_time:1830096ms step_avg:538.11ms step:3412/6250 train_loss:2.2714 aux_loss:0.0323 train_time:1830626ms step_avg:538.10ms step:3413/6250 train_loss:2.3794 aux_loss:0.0323 train_time:1831155ms step_avg:538.10ms step:3414/6250 train_loss:2.2719 aux_loss:0.0321 train_time:1831688ms step_avg:538.10ms step:3415/6250 train_loss:2.2012 aux_loss:0.0320 train_time:1832219ms step_avg:538.10ms step:3416/6250 train_loss:2.1878 aux_loss:0.0319 train_time:1832750ms step_avg:538.09ms step:3417/6250 train_loss:2.2914 aux_loss:0.0320 train_time:1833290ms step_avg:538.10ms step:3418/6250 train_loss:2.2490 aux_loss:0.0318 train_time:1833859ms step_avg:538.10ms step:3419/6250 train_loss:2.2415 aux_loss:0.0319 train_time:1834428ms step_avg:538.11ms step:3420/6250 train_loss:2.2735 aux_loss:0.0324 train_time:1834960ms step_avg:538.11ms step:3421/6250 train_loss:2.3045 aux_loss:0.0318 train_time:1835501ms step_avg:538.11ms step:3422/6250 train_loss:2.2475 aux_loss:0.0318 train_time:1836072ms step_avg:538.12ms step:3423/6250 train_loss:2.2515 aux_loss:0.0321 train_time:1836633ms step_avg:538.13ms step:3424/6250 train_loss:2.2798 aux_loss:0.0325 train_time:1837169ms step_avg:538.13ms step:3425/6250 train_loss:2.2628 aux_loss:0.0316 train_time:1837702ms step_avg:538.13ms step:3426/6250 train_loss:2.3614 aux_loss:0.0318 train_time:1838235ms step_avg:538.12ms step:3427/6250 train_loss:2.1949 aux_loss:0.0318 train_time:1838767ms step_avg:538.12ms step:3428/6250 train_loss:2.2445 aux_loss:0.0320 train_time:1839297ms step_avg:538.12ms step:3429/6250 train_loss:2.2133 aux_loss:0.0320 train_time:1839829ms step_avg:538.12ms step:3430/6250 train_loss:2.2887 aux_loss:0.0317 train_time:1840360ms step_avg:538.12ms step:3431/6250 train_loss:2.2883 aux_loss:0.0318 train_time:1840892ms step_avg:538.12ms step:3432/6250 train_loss:2.2887 aux_loss:0.0325 train_time:1841424ms step_avg:538.11ms step:3433/6250 train_loss:2.3180 aux_loss:0.0325 train_time:1841955ms step_avg:538.11ms step:3434/6250 train_loss:2.2687 aux_loss:0.0320 train_time:1842488ms step_avg:538.11ms step:3435/6250 train_loss:2.2760 aux_loss:0.0321 train_time:1843018ms step_avg:538.11ms step:3436/6250 train_loss:2.2493 aux_loss:0.0321 train_time:1843552ms step_avg:538.11ms step:3437/6250 train_loss:2.2375 aux_loss:0.0317 train_time:1844082ms step_avg:538.10ms step:3438/6250 train_loss:2.1914 aux_loss:0.0319 train_time:1844614ms step_avg:538.10ms step:3439/6250 train_loss:2.1808 aux_loss:0.0320 train_time:1845144ms step_avg:538.10ms step:3440/6250 train_loss:2.3241 aux_loss:0.0322 train_time:1845674ms step_avg:538.10ms step:3441/6250 train_loss:2.2597 aux_loss:0.0319 train_time:1846206ms step_avg:538.10ms step:3442/6250 train_loss:2.2554 aux_loss:0.0316 train_time:1846738ms step_avg:538.09ms step:3443/6250 train_loss:2.2401 aux_loss:0.0321 train_time:1847269ms step_avg:538.09ms step:3444/6250 train_loss:2.3169 aux_loss:0.0320 train_time:1847799ms step_avg:538.09ms step:3445/6250 train_loss:2.1296 aux_loss:0.0317 train_time:1848330ms step_avg:538.09ms step:3446/6250 train_loss:2.2911 aux_loss:0.0317 train_time:1848861ms step_avg:538.09ms step:3447/6250 train_loss:2.2673 aux_loss:0.0320 train_time:1849394ms step_avg:538.08ms step:3448/6250 train_loss:2.1093 aux_loss:0.0318 train_time:1849924ms step_avg:538.08ms step:3449/6250 train_loss:2.1936 aux_loss:0.0323 train_time:1850456ms step_avg:538.08ms step:3450/6250 train_loss:2.2440 aux_loss:0.0317 train_time:1850989ms step_avg:538.08ms step:3451/6250 train_loss:2.1292 aux_loss:0.0318 train_time:1851520ms step_avg:538.08ms step:3452/6250 train_loss:2.2112 aux_loss:0.0321 train_time:1852055ms step_avg:538.08ms step:3453/6250 train_loss:2.2153 aux_loss:0.0317 train_time:1852586ms step_avg:538.07ms step:3454/6250 train_loss:2.2309 aux_loss:0.0318 train_time:1853116ms step_avg:538.07ms step:3455/6250 train_loss:2.3664 aux_loss:0.0319 train_time:1853668ms step_avg:538.07ms step:3456/6250 train_loss:2.2776 aux_loss:0.0317 train_time:1854227ms step_avg:538.08ms step:3457/6250 train_loss:2.2989 aux_loss:0.0323 train_time:1854760ms step_avg:538.08ms step:3458/6250 train_loss:2.3650 aux_loss:0.0324 train_time:1855290ms step_avg:538.08ms step:3459/6250 train_loss:2.2348 aux_loss:0.0316 train_time:1855821ms step_avg:538.08ms step:3460/6250 train_loss:2.2915 aux_loss:0.0315 train_time:1856352ms step_avg:538.07ms step:3461/6250 train_loss:2.1838 aux_loss:0.0317 train_time:1856882ms step_avg:538.07ms step:3462/6250 train_loss:2.1711 aux_loss:0.0319 train_time:1857414ms step_avg:538.07ms step:3463/6250 train_loss:2.3011 aux_loss:0.0317 train_time:1857945ms step_avg:538.07ms step:3464/6250 train_loss:2.2304 aux_loss:0.0317 train_time:1858476ms step_avg:538.06ms step:3465/6250 train_loss:2.2497 aux_loss:0.0316 train_time:1859007ms step_avg:538.06ms step:3466/6250 train_loss:2.2228 aux_loss:0.0318 train_time:1859539ms step_avg:538.06ms step:3467/6250 train_loss:2.2358 aux_loss:0.0317 train_time:1860069ms step_avg:538.06ms step:3468/6250 train_loss:2.2587 aux_loss:0.0319 train_time:1860602ms step_avg:538.06ms step:3469/6250 train_loss:2.3186 aux_loss:0.0319 train_time:1861140ms step_avg:538.06ms step:3470/6250 train_loss:2.2027 aux_loss:0.0319 train_time:1861672ms step_avg:538.06ms step:3471/6250 train_loss:2.1621 aux_loss:0.0316 train_time:1862203ms step_avg:538.05ms step:3472/6250 train_loss:2.1714 aux_loss:0.0317 train_time:1862733ms step_avg:538.05ms step:3473/6250 train_loss:2.2301 aux_loss:0.0318 train_time:1863264ms step_avg:538.05ms step:3474/6250 train_loss:2.2514 aux_loss:0.0315 train_time:1863796ms step_avg:538.05ms step:3475/6250 train_loss:2.1490 aux_loss:0.0317 train_time:1864329ms step_avg:538.05ms step:3476/6250 train_loss:2.2368 aux_loss:0.0319 train_time:1864860ms step_avg:538.04ms step:3477/6250 train_loss:2.3517 aux_loss:0.0318 train_time:1865391ms step_avg:538.04ms step:3478/6250 train_loss:2.1888 aux_loss:0.0314 train_time:1865922ms step_avg:538.04ms step:3479/6250 train_loss:2.1314 aux_loss:0.0315 train_time:1866455ms step_avg:538.04ms step:3480/6250 train_loss:2.1512 aux_loss:0.0318 train_time:1866988ms step_avg:538.04ms step:3481/6250 train_loss:2.3312 aux_loss:0.0321 train_time:1867520ms step_avg:538.04ms step:3482/6250 train_loss:2.1850 aux_loss:0.0318 train_time:1868050ms step_avg:538.03ms step:3483/6250 train_loss:2.2825 aux_loss:0.0321 train_time:1868582ms step_avg:538.03ms step:3484/6250 train_loss:2.3749 aux_loss:0.0323 train_time:1869112ms step_avg:538.03ms step:3485/6250 train_loss:2.2854 aux_loss:0.0318 train_time:1869643ms step_avg:538.03ms step:3486/6250 train_loss:2.1873 aux_loss:0.0314 train_time:1870174ms step_avg:538.02ms step:3487/6250 train_loss:2.4455 aux_loss:0.0326 train_time:1870706ms step_avg:538.02ms step:3488/6250 train_loss:2.2481 aux_loss:0.0316 train_time:1871238ms step_avg:538.02ms step:3489/6250 train_loss:2.2521 aux_loss:0.0316 train_time:1871770ms step_avg:538.02ms step:3490/6250 train_loss:2.2242 aux_loss:0.0318 train_time:1872301ms step_avg:538.02ms step:3491/6250 train_loss:2.3456 aux_loss:0.0321 train_time:1872834ms step_avg:538.02ms step:3492/6250 train_loss:2.2602 aux_loss:0.0318 train_time:1873365ms step_avg:538.01ms step:3493/6250 train_loss:2.3268 aux_loss:0.0313 train_time:1873925ms step_avg:538.02ms step:3494/6250 train_loss:2.1682 aux_loss:0.0315 train_time:1874481ms step_avg:538.03ms step:3495/6250 train_loss:2.2346 aux_loss:0.0319 train_time:1875013ms step_avg:538.02ms step:3496/6250 train_loss:2.1852 aux_loss:0.0319 train_time:1875547ms step_avg:538.02ms step:3497/6250 train_loss:2.2749 aux_loss:0.0314 train_time:1876079ms step_avg:538.02ms step:3498/6250 train_loss:2.2828 aux_loss:0.0319 train_time:1876617ms step_avg:538.02ms step:3499/6250 train_loss:2.1804 aux_loss:0.0323 train_time:1877151ms step_avg:538.02ms step:3500/6250 train_loss:2.2245 aux_loss:0.0323 train_time:1877682ms step_avg:538.02ms step:3501/6250 train_loss:2.2190 aux_loss:0.0319 train_time:1878223ms step_avg:538.02ms step:3502/6250 train_loss:2.2234 aux_loss:0.0318 train_time:1878783ms step_avg:538.02ms step:3503/6250 train_loss:2.2221 aux_loss:0.0319 train_time:1879319ms step_avg:538.02ms step:3504/6250 train_loss:2.2561 aux_loss:0.0322 train_time:1879851ms step_avg:538.02ms step:3505/6250 train_loss:2.2695 aux_loss:0.0318 train_time:1880394ms step_avg:538.02ms step:3506/6250 train_loss:2.3642 aux_loss:0.0322 train_time:1880927ms step_avg:538.02ms step:3507/6250 train_loss:2.2182 aux_loss:0.0320 train_time:1881458ms step_avg:538.02ms step:3508/6250 train_loss:2.2053 aux_loss:0.0321 train_time:1881992ms step_avg:538.02ms step:3509/6250 train_loss:2.2787 aux_loss:0.0320 train_time:1882524ms step_avg:538.02ms step:3510/6250 train_loss:2.2991 aux_loss:0.0321 train_time:1883055ms step_avg:538.02ms step:3511/6250 train_loss:2.2611 aux_loss:0.0317 train_time:1883586ms step_avg:538.01ms step:3512/6250 train_loss:2.2586 aux_loss:0.0315 train_time:1884117ms step_avg:538.01ms step:3513/6250 train_loss:2.3047 aux_loss:0.0317 train_time:1884650ms step_avg:538.01ms step:3514/6250 train_loss:2.2948 aux_loss:0.0318 train_time:1885181ms step_avg:538.01ms step:3515/6250 train_loss:2.2286 aux_loss:0.0321 train_time:1885713ms step_avg:538.01ms step:3516/6250 train_loss:2.1944 aux_loss:0.0316 train_time:1886243ms step_avg:538.00ms step:3517/6250 train_loss:2.2913 aux_loss:0.0315 train_time:1886772ms step_avg:538.00ms step:3518/6250 train_loss:2.2250 aux_loss:0.0314 train_time:1887305ms step_avg:538.00ms step:3519/6250 train_loss:2.1964 aux_loss:0.0316 train_time:1887836ms step_avg:538.00ms step:3520/6250 train_loss:2.1861 aux_loss:0.0309 train_time:1888371ms step_avg:538.00ms step:3521/6250 train_loss:2.2120 aux_loss:0.0312 train_time:1888916ms step_avg:538.00ms step:3522/6250 train_loss:2.3022 aux_loss:0.0315 train_time:1889454ms step_avg:538.00ms step:3523/6250 train_loss:2.2350 aux_loss:0.0315 train_time:1889996ms step_avg:538.00ms step:3524/6250 train_loss:2.2294 aux_loss:0.0316 train_time:1890530ms step_avg:538.00ms step:3525/6250 train_loss:2.2750 aux_loss:0.0317 train_time:1891069ms step_avg:538.00ms step:3526/6250 train_loss:2.2766 aux_loss:0.0312 train_time:1891604ms step_avg:538.00ms step:3527/6250 train_loss:2.2415 aux_loss:0.0318 train_time:1892139ms step_avg:538.00ms step:3528/6250 train_loss:2.2360 aux_loss:0.0318 train_time:1892672ms step_avg:538.00ms step:3529/6250 train_loss:2.3347 aux_loss:0.0314 train_time:1893203ms step_avg:537.99ms step:3530/6250 train_loss:2.1700 aux_loss:0.0309 train_time:1893761ms step_avg:538.00ms step:3531/6250 train_loss:2.2926 aux_loss:0.0313 train_time:1894323ms step_avg:538.01ms step:3532/6250 train_loss:2.2277 aux_loss:0.0314 train_time:1894875ms step_avg:538.01ms step:3533/6250 train_loss:2.1761 aux_loss:0.0311 train_time:1895410ms step_avg:538.01ms step:3534/6250 train_loss:2.2960 aux_loss:0.0315 train_time:1895945ms step_avg:538.01ms step:3535/6250 train_loss:2.2083 aux_loss:0.0313 train_time:1896477ms step_avg:538.01ms step:3536/6250 train_loss:2.2772 aux_loss:0.0318 train_time:1897008ms step_avg:538.01ms step:3537/6250 train_loss:2.1235 aux_loss:0.0316 train_time:1897544ms step_avg:538.01ms step:3538/6250 train_loss:2.2457 aux_loss:0.0317 train_time:1898075ms step_avg:538.00ms step:3539/6250 train_loss:2.2452 aux_loss:0.0316 train_time:1898608ms step_avg:538.00ms step:3540/6250 train_loss:2.2389 aux_loss:0.0318 train_time:1899139ms step_avg:538.00ms step:3541/6250 train_loss:2.3158 aux_loss:0.0322 train_time:1899670ms step_avg:538.00ms step:3542/6250 train_loss:2.2311 aux_loss:0.0318 train_time:1900201ms step_avg:538.00ms step:3543/6250 train_loss:2.3018 aux_loss:0.0315 train_time:1900733ms step_avg:537.99ms step:3544/6250 train_loss:2.1782 aux_loss:0.0318 train_time:1901264ms step_avg:537.99ms step:3545/6250 train_loss:2.2311 aux_loss:0.0316 train_time:1901798ms step_avg:537.99ms step:3546/6250 train_loss:2.2449 aux_loss:0.0318 train_time:1902334ms step_avg:537.99ms step:3547/6250 train_loss:2.2507 aux_loss:0.0317 train_time:1902865ms step_avg:537.99ms step:3548/6250 train_loss:2.2720 aux_loss:0.0317 train_time:1903397ms step_avg:537.99ms step:3549/6250 train_loss:2.3454 aux_loss:0.0319 train_time:1903926ms step_avg:537.98ms step:3550/6250 train_loss:2.2595 aux_loss:0.0314 train_time:1904456ms step_avg:537.98ms step:3551/6250 train_loss:2.2460 aux_loss:0.0317 train_time:1904988ms step_avg:537.98ms step:3552/6250 train_loss:2.2647 aux_loss:0.0316 train_time:1905518ms step_avg:537.98ms step:3553/6250 train_loss:2.2746 aux_loss:0.0312 train_time:1906053ms step_avg:537.98ms step:3554/6250 train_loss:2.1820 aux_loss:0.0315 train_time:1906586ms step_avg:537.98ms step:3555/6250 train_loss:2.1977 aux_loss:0.0315 train_time:1907117ms step_avg:537.97ms step:3556/6250 train_loss:2.2535 aux_loss:0.0313 train_time:1907648ms step_avg:537.97ms step:3557/6250 train_loss:2.1743 aux_loss:0.0313 train_time:1908179ms step_avg:537.97ms step:3558/6250 train_loss:2.2807 aux_loss:0.0316 train_time:1908714ms step_avg:537.97ms step:3559/6250 train_loss:2.1472 aux_loss:0.0311 train_time:1909248ms step_avg:537.97ms step:3560/6250 train_loss:2.2000 aux_loss:0.0316 train_time:1909780ms step_avg:537.97ms step:3561/6250 train_loss:2.2644 aux_loss:0.0315 train_time:1910320ms step_avg:537.97ms step:3562/6250 train_loss:2.1812 aux_loss:0.0307 train_time:1910860ms step_avg:537.97ms step:3563/6250 train_loss:2.2488 aux_loss:0.0313 train_time:1911397ms step_avg:537.97ms step:3564/6250 train_loss:2.1993 aux_loss:0.0318 train_time:1911928ms step_avg:537.97ms step:3565/6250 train_loss:2.2312 aux_loss:0.0315 train_time:1912459ms step_avg:537.96ms step:3566/6250 train_loss:2.2814 aux_loss:0.0314 train_time:1912990ms step_avg:537.96ms step:3567/6250 train_loss:2.2110 aux_loss:0.0316 train_time:1913522ms step_avg:537.96ms step:3568/6250 train_loss:2.1483 aux_loss:0.0341 train_time:1914079ms step_avg:537.96ms step:3569/6250 train_loss:2.1819 aux_loss:0.0311 train_time:1914633ms step_avg:537.97ms step:3570/6250 train_loss:2.1663 aux_loss:0.0312 train_time:1915165ms step_avg:537.97ms step:3571/6250 train_loss:2.2288 aux_loss:0.0314 train_time:1915697ms step_avg:537.97ms step:3572/6250 train_loss:2.3779 aux_loss:0.0318 train_time:1916228ms step_avg:537.96ms step:3573/6250 train_loss:2.2333 aux_loss:0.0315 train_time:1916759ms step_avg:537.96ms step:3574/6250 train_loss:2.2796 aux_loss:0.0316 train_time:1917291ms step_avg:537.96ms step:3575/6250 train_loss:2.1658 aux_loss:0.0318 train_time:1917821ms step_avg:537.96ms step:3576/6250 train_loss:2.2682 aux_loss:0.0316 train_time:1918352ms step_avg:537.96ms step:3577/6250 train_loss:2.2886 aux_loss:0.0318 train_time:1918883ms step_avg:537.95ms step:3578/6250 train_loss:2.1946 aux_loss:0.0313 train_time:1919414ms step_avg:537.95ms step:3579/6250 train_loss:2.2436 aux_loss:0.0317 train_time:1919944ms step_avg:537.95ms step:3580/6250 train_loss:2.2371 aux_loss:0.0318 train_time:1920475ms step_avg:537.95ms step:3581/6250 train_loss:2.2636 aux_loss:0.0314 train_time:1921007ms step_avg:537.95ms step:3582/6250 train_loss:2.1692 aux_loss:0.0316 train_time:1921539ms step_avg:537.94ms step:3583/6250 train_loss:2.2739 aux_loss:0.0321 train_time:1922071ms step_avg:537.94ms step:3584/6250 train_loss:2.1672 aux_loss:0.0316 train_time:1922601ms step_avg:537.94ms step:3585/6250 train_loss:2.0948 aux_loss:0.0315 train_time:1923132ms step_avg:537.94ms step:3586/6250 train_loss:2.2590 aux_loss:0.0317 train_time:1923661ms step_avg:537.94ms step:3587/6250 train_loss:2.1722 aux_loss:0.0314 train_time:1924193ms step_avg:537.93ms step:3588/6250 train_loss:2.1321 aux_loss:0.0314 train_time:1924724ms step_avg:537.93ms step:3589/6250 train_loss:2.2316 aux_loss:0.0315 train_time:1925255ms step_avg:537.93ms step:3590/6250 train_loss:2.2545 aux_loss:0.0313 train_time:1925786ms step_avg:537.93ms step:3591/6250 train_loss:2.2044 aux_loss:0.0312 train_time:1926318ms step_avg:537.93ms step:3592/6250 train_loss:2.3071 aux_loss:0.0316 train_time:1926847ms step_avg:537.92ms step:3593/6250 train_loss:2.1794 aux_loss:0.0321 train_time:1927387ms step_avg:537.93ms step:3594/6250 train_loss:2.1808 aux_loss:0.0314 train_time:1927917ms step_avg:537.92ms step:3595/6250 train_loss:2.2174 aux_loss:0.0311 train_time:1928448ms step_avg:537.92ms step:3596/6250 train_loss:2.1964 aux_loss:0.0314 train_time:1928979ms step_avg:537.92ms step:3597/6250 train_loss:2.1732 aux_loss:0.0315 train_time:1929509ms step_avg:537.92ms step:3598/6250 train_loss:2.1840 aux_loss:0.0311 train_time:1930042ms step_avg:537.92ms step:3599/6250 train_loss:2.1287 aux_loss:0.0312 train_time:1930574ms step_avg:537.91ms step:3600/6250 train_loss:2.1349 aux_loss:0.0317 train_time:1931107ms step_avg:537.91ms step:3601/6250 train_loss:2.2573 aux_loss:0.0319 train_time:1931638ms step_avg:537.91ms step:3602/6250 train_loss:2.2023 aux_loss:0.0314 train_time:1932170ms step_avg:537.91ms step:3603/6250 train_loss:2.1642 aux_loss:0.0313 train_time:1932700ms step_avg:537.91ms step:3604/6250 train_loss:2.1609 aux_loss:0.0314 train_time:1933231ms step_avg:537.91ms step:3605/6250 train_loss:2.2613 aux_loss:0.0315 train_time:1933769ms step_avg:537.91ms step:3606/6250 train_loss:2.1846 aux_loss:0.0312 train_time:1934329ms step_avg:537.91ms step:3607/6250 train_loss:2.1768 aux_loss:0.0310 train_time:1934879ms step_avg:537.91ms step:3608/6250 train_loss:2.3445 aux_loss:0.0313 train_time:1935414ms step_avg:537.91ms step:3609/6250 train_loss:2.2132 aux_loss:0.0312 train_time:1935945ms step_avg:537.91ms step:3610/6250 train_loss:2.2057 aux_loss:0.0313 train_time:1936479ms step_avg:537.91ms step:3611/6250 train_loss:2.2856 aux_loss:0.0309 train_time:1937010ms step_avg:537.91ms step:3612/6250 train_loss:2.3148 aux_loss:0.0312 train_time:1937541ms step_avg:537.91ms step:3613/6250 train_loss:2.1736 aux_loss:0.0311 train_time:1938072ms step_avg:537.90ms step:3614/6250 train_loss:2.1322 aux_loss:0.0312 train_time:1938602ms step_avg:537.90ms step:3615/6250 train_loss:2.1689 aux_loss:0.0315 train_time:1939133ms step_avg:537.90ms step:3616/6250 train_loss:2.2526 aux_loss:0.0311 train_time:1939664ms step_avg:537.90ms step:3617/6250 train_loss:2.2184 aux_loss:0.0308 train_time:1940197ms step_avg:537.90ms step:3618/6250 train_loss:2.2345 aux_loss:0.0312 train_time:1940727ms step_avg:537.90ms step:3619/6250 train_loss:2.2936 aux_loss:0.0315 train_time:1941259ms step_avg:537.89ms step:3620/6250 train_loss:2.2220 aux_loss:0.0313 train_time:1941789ms step_avg:537.89ms step:3621/6250 train_loss:2.2895 aux_loss:0.0314 train_time:1942321ms step_avg:537.89ms step:3622/6250 train_loss:2.2241 aux_loss:0.0314 train_time:1942851ms step_avg:537.89ms step:3623/6250 train_loss:2.2694 aux_loss:0.0315 train_time:1943384ms step_avg:537.89ms step:3624/6250 train_loss:2.2110 aux_loss:0.0312 train_time:1943916ms step_avg:537.88ms step:3625/6250 train_loss:2.2916 aux_loss:0.0314 train_time:1944447ms step_avg:537.88ms step:3626/6250 train_loss:2.3720 aux_loss:0.0315 train_time:1944977ms step_avg:537.88ms step:3627/6250 train_loss:2.2491 aux_loss:0.0311 train_time:1945509ms step_avg:537.88ms step:3628/6250 train_loss:2.1922 aux_loss:0.0309 train_time:1946039ms step_avg:537.88ms step:3629/6250 train_loss:2.1358 aux_loss:0.0311 train_time:1946571ms step_avg:537.88ms step:3630/6250 train_loss:2.2472 aux_loss:0.0314 train_time:1947104ms step_avg:537.87ms step:3631/6250 train_loss:2.2601 aux_loss:0.0313 train_time:1947642ms step_avg:537.87ms step:3632/6250 train_loss:2.1806 aux_loss:0.0317 train_time:1948173ms step_avg:537.87ms step:3633/6250 train_loss:2.1537 aux_loss:0.0315 train_time:1948704ms step_avg:537.87ms step:3634/6250 train_loss:2.2388 aux_loss:0.0316 train_time:1949241ms step_avg:537.87ms step:3635/6250 train_loss:2.2114 aux_loss:0.0316 train_time:1949773ms step_avg:537.87ms step:3636/6250 train_loss:2.2607 aux_loss:0.0322 train_time:1950310ms step_avg:537.87ms step:3637/6250 train_loss:2.2781 aux_loss:0.0318 train_time:1950847ms step_avg:537.87ms step:3638/6250 train_loss:2.2464 aux_loss:0.0314 train_time:1951379ms step_avg:537.87ms step:3639/6250 train_loss:2.2963 aux_loss:0.0311 train_time:1951910ms step_avg:537.86ms step:3640/6250 train_loss:2.2017 aux_loss:0.0310 train_time:1952440ms step_avg:537.86ms step:3641/6250 train_loss:2.2768 aux_loss:0.0313 train_time:1952970ms step_avg:537.86ms step:3642/6250 train_loss:2.2087 aux_loss:0.0310 train_time:1953503ms step_avg:537.86ms step:3643/6250 train_loss:2.3655 aux_loss:0.0311 train_time:1954054ms step_avg:537.86ms step:3644/6250 train_loss:2.2193 aux_loss:0.0314 train_time:1954617ms step_avg:537.87ms step:3645/6250 train_loss:2.2094 aux_loss:0.0314 train_time:1955159ms step_avg:537.87ms step:3646/6250 train_loss:2.2346 aux_loss:0.0309 train_time:1955708ms step_avg:537.87ms step:3647/6250 train_loss:2.2189 aux_loss:0.0308 train_time:1956241ms step_avg:537.87ms step:3648/6250 train_loss:2.2297 aux_loss:0.0310 train_time:1956773ms step_avg:537.87ms step:3649/6250 train_loss:2.2343 aux_loss:0.0310 train_time:1957305ms step_avg:537.87ms step:3650/6250 train_loss:2.1507 aux_loss:0.0311 train_time:1957839ms step_avg:537.87ms step:3651/6250 train_loss:2.2669 aux_loss:0.0317 train_time:1958375ms step_avg:537.87ms step:3652/6250 train_loss:2.1544 aux_loss:0.0308 train_time:1958912ms step_avg:537.87ms step:3653/6250 train_loss:2.1524 aux_loss:0.0314 train_time:1959448ms step_avg:537.87ms step:3654/6250 train_loss:2.1946 aux_loss:0.0316 train_time:1959992ms step_avg:537.87ms step:3655/6250 train_loss:2.2517 aux_loss:0.0312 train_time:1960532ms step_avg:537.87ms step:3656/6250 train_loss:2.2760 aux_loss:0.0311 train_time:1961069ms step_avg:537.87ms step:3657/6250 train_loss:2.2422 aux_loss:0.0310 train_time:1961603ms step_avg:537.87ms step:3658/6250 train_loss:2.2506 aux_loss:0.0315 train_time:1962135ms step_avg:537.87ms step:3659/6250 train_loss:2.3364 aux_loss:0.0314 train_time:1962666ms step_avg:537.86ms step:3660/6250 train_loss:2.1215 aux_loss:0.0310 train_time:1963198ms step_avg:537.86ms step:3661/6250 train_loss:2.2241 aux_loss:0.0313 train_time:1963729ms step_avg:537.86ms step:3662/6250 train_loss:2.1913 aux_loss:0.0308 train_time:1964263ms step_avg:537.86ms step:3663/6250 train_loss:2.1634 aux_loss:0.0311 train_time:1964793ms step_avg:537.86ms step:3664/6250 train_loss:2.2080 aux_loss:0.0315 train_time:1965329ms step_avg:537.86ms step:3665/6250 train_loss:2.2139 aux_loss:0.0306 train_time:1965860ms step_avg:537.85ms step:3666/6250 train_loss:2.2299 aux_loss:0.0306 train_time:1966390ms step_avg:537.85ms step:3667/6250 train_loss:2.2344 aux_loss:0.0311 train_time:1966920ms step_avg:537.85ms step:3668/6250 train_loss:2.2491 aux_loss:0.0312 train_time:1967452ms step_avg:537.85ms step:3669/6250 train_loss:2.1972 aux_loss:0.0309 train_time:1967982ms step_avg:537.85ms step:3670/6250 train_loss:2.2752 aux_loss:0.0310 train_time:1968513ms step_avg:537.85ms step:3671/6250 train_loss:2.2105 aux_loss:0.0312 train_time:1969044ms step_avg:537.84ms step:3672/6250 train_loss:2.1941 aux_loss:0.0312 train_time:1969577ms step_avg:537.84ms step:3673/6250 train_loss:2.1310 aux_loss:0.0308 train_time:1970137ms step_avg:537.85ms step:3674/6250 train_loss:2.2661 aux_loss:0.0311 train_time:1970703ms step_avg:537.86ms step:3675/6250 train_loss:2.2749 aux_loss:0.0315 train_time:1971241ms step_avg:537.86ms step:3676/6250 train_loss:2.2555 aux_loss:0.0311 train_time:1971775ms step_avg:537.85ms step:3677/6250 train_loss:2.2569 aux_loss:0.0310 train_time:1972308ms step_avg:537.85ms step:3678/6250 train_loss:2.2214 aux_loss:0.0315 train_time:1972841ms step_avg:537.85ms step:3679/6250 train_loss:2.1983 aux_loss:0.0312 train_time:1973375ms step_avg:537.85ms step:3680/6250 train_loss:2.2538 aux_loss:0.0310 train_time:1973935ms step_avg:537.86ms step:3681/6250 train_loss:2.2805 aux_loss:0.0313 train_time:1974504ms step_avg:537.87ms step:3682/6250 train_loss:2.2221 aux_loss:0.0313 train_time:1975070ms step_avg:537.87ms step:3683/6250 train_loss:2.2514 aux_loss:0.0312 train_time:1975602ms step_avg:537.87ms step:3684/6250 train_loss:2.2021 aux_loss:0.0308 train_time:1976133ms step_avg:537.87ms step:3685/6250 train_loss:2.2617 aux_loss:0.0312 train_time:1976666ms step_avg:537.87ms step:3686/6250 train_loss:2.1773 aux_loss:0.0313 train_time:1977198ms step_avg:537.87ms step:3687/6250 train_loss:2.2300 aux_loss:0.0307 train_time:1977729ms step_avg:537.86ms step:3688/6250 train_loss:2.2649 aux_loss:0.0307 train_time:1978262ms step_avg:537.86ms step:3689/6250 train_loss:2.1965 aux_loss:0.0310 train_time:1978801ms step_avg:537.86ms step:3690/6250 train_loss:2.1104 aux_loss:0.0309 train_time:1979335ms step_avg:537.86ms step:3691/6250 train_loss:2.1404 aux_loss:0.0311 train_time:1979865ms step_avg:537.86ms step:3692/6250 train_loss:2.2540 aux_loss:0.0311 train_time:1980397ms step_avg:537.86ms step:3693/6250 train_loss:2.2377 aux_loss:0.0312 train_time:1980928ms step_avg:537.86ms step:3694/6250 train_loss:2.2046 aux_loss:0.0306 train_time:1981460ms step_avg:537.86ms step:3695/6250 train_loss:2.1504 aux_loss:0.0310 train_time:1981991ms step_avg:537.85ms step:3696/6250 train_loss:2.2069 aux_loss:0.0309 train_time:1982523ms step_avg:537.85ms step:3697/6250 train_loss:2.3282 aux_loss:0.0309 train_time:1983055ms step_avg:537.85ms step:3698/6250 train_loss:2.2239 aux_loss:0.0308 train_time:1983587ms step_avg:537.85ms step:3699/6250 train_loss:2.2951 aux_loss:0.0306 train_time:1984116ms step_avg:537.85ms step:3700/6250 train_loss:2.2044 aux_loss:0.0309 train_time:1984649ms step_avg:537.85ms step:3701/6250 train_loss:2.1710 aux_loss:0.0307 train_time:1985179ms step_avg:537.84ms step:3702/6250 train_loss:2.2111 aux_loss:0.0308 train_time:1985709ms step_avg:537.84ms step:3703/6250 train_loss:2.1395 aux_loss:0.0307 train_time:1986242ms step_avg:537.84ms step:3704/6250 train_loss:2.1591 aux_loss:0.0310 train_time:1986775ms step_avg:537.84ms step:3705/6250 train_loss:2.2832 aux_loss:0.0308 train_time:1987307ms step_avg:537.84ms step:3706/6250 train_loss:2.2030 aux_loss:0.0311 train_time:1987839ms step_avg:537.84ms step:3707/6250 train_loss:2.1591 aux_loss:0.0308 train_time:1988371ms step_avg:537.83ms step:3708/6250 train_loss:2.2128 aux_loss:0.0305 train_time:1988901ms step_avg:537.83ms step:3709/6250 train_loss:2.2119 aux_loss:0.0309 train_time:1989434ms step_avg:537.83ms step:3710/6250 train_loss:2.2576 aux_loss:0.0315 train_time:1989966ms step_avg:537.83ms step:3711/6250 train_loss:2.2377 aux_loss:0.0309 train_time:1990499ms step_avg:537.83ms step:3712/6250 train_loss:2.2087 aux_loss:0.0304 train_time:1991030ms step_avg:537.83ms step:3713/6250 train_loss:2.2402 aux_loss:0.0306 train_time:1991571ms step_avg:537.83ms step:3714/6250 train_loss:2.2428 aux_loss:0.0310 train_time:1992104ms step_avg:537.82ms step:3715/6250 train_loss:2.2265 aux_loss:0.0308 train_time:1992635ms step_avg:537.82ms step:3716/6250 train_loss:2.3034 aux_loss:0.0306 train_time:1993166ms step_avg:537.82ms step:3717/6250 train_loss:2.2505 aux_loss:0.0309 train_time:1993698ms step_avg:537.82ms step:3718/6250 train_loss:2.2176 aux_loss:0.0309 train_time:1994245ms step_avg:537.82ms step:3719/6250 train_loss:2.2030 aux_loss:0.0308 train_time:1994807ms step_avg:537.83ms step:3720/6250 train_loss:2.2861 aux_loss:0.0307 train_time:1995349ms step_avg:537.83ms step:3721/6250 train_loss:2.1946 aux_loss:0.0305 train_time:1995880ms step_avg:537.83ms step:3722/6250 train_loss:2.2580 aux_loss:0.0312 train_time:1996413ms step_avg:537.83ms step:3723/6250 train_loss:2.1990 aux_loss:0.0308 train_time:1996944ms step_avg:537.82ms step:3724/6250 train_loss:2.2042 aux_loss:0.0307 train_time:1997477ms step_avg:537.82ms step:3725/6250 train_loss:2.2936 aux_loss:0.0309 train_time:1998023ms step_avg:537.83ms step:3726/6250 train_loss:2.1772 aux_loss:0.0308 train_time:1998588ms step_avg:537.83ms step:3727/6250 train_loss:2.2863 aux_loss:0.0307 train_time:1999133ms step_avg:537.83ms step:3728/6250 train_loss:2.1776 aux_loss:0.0310 train_time:1999664ms step_avg:537.83ms step:3729/6250 train_loss:2.1832 aux_loss:0.0309 train_time:2000196ms step_avg:537.83ms step:3730/6250 train_loss:2.2275 aux_loss:0.0306 train_time:2000737ms step_avg:537.83ms step:3731/6250 train_loss:2.2009 aux_loss:0.0310 train_time:2001295ms step_avg:537.84ms step:3732/6250 train_loss:2.3382 aux_loss:0.0312 train_time:2001851ms step_avg:537.84ms step:3733/6250 train_loss:2.1803 aux_loss:0.0309 train_time:2002383ms step_avg:537.84ms step:3734/6250 train_loss:2.2812 aux_loss:0.0306 train_time:2002913ms step_avg:537.84ms step:3735/6250 train_loss:2.2833 aux_loss:0.0313 train_time:2003447ms step_avg:537.84ms step:3736/6250 train_loss:2.2204 aux_loss:0.0313 train_time:2003985ms step_avg:537.84ms step:3737/6250 train_loss:2.2324 aux_loss:0.0310 train_time:2004518ms step_avg:537.84ms step:3738/6250 train_loss:2.2816 aux_loss:0.0311 train_time:2005059ms step_avg:537.84ms step:3739/6250 train_loss:2.1486 aux_loss:0.0309 train_time:2005597ms step_avg:537.84ms step:3740/6250 train_loss:2.2252 aux_loss:0.0311 train_time:2006128ms step_avg:537.84ms step:3741/6250 train_loss:2.2407 aux_loss:0.0309 train_time:2006659ms step_avg:537.83ms step:3742/6250 train_loss:2.2072 aux_loss:0.0307 train_time:2007190ms step_avg:537.83ms step:3743/6250 train_loss:2.1441 aux_loss:0.0306 train_time:2007723ms step_avg:537.83ms step:3744/6250 train_loss:2.2572 aux_loss:0.0309 train_time:2008254ms step_avg:537.83ms step:3745/6250 train_loss:2.2153 aux_loss:0.0308 train_time:2008787ms step_avg:537.83ms step:3746/6250 train_loss:2.2599 aux_loss:0.0310 train_time:2009317ms step_avg:537.83ms step:3747/6250 train_loss:2.2489 aux_loss:0.0309 train_time:2009847ms step_avg:537.82ms step:3748/6250 train_loss:2.2164 aux_loss:0.0308 train_time:2010378ms step_avg:537.82ms step:3749/6250 train_loss:2.1403 aux_loss:0.0309 train_time:2010910ms step_avg:537.82ms step:3750/6250 train_loss:2.2706 aux_loss:0.0305 train_time:2011442ms step_avg:537.82ms step:3751/6250 train_loss:2.2556 aux_loss:0.0309 train_time:2011973ms step_avg:537.82ms step:3752/6250 train_loss:2.1609 aux_loss:0.0308 train_time:2012505ms step_avg:537.82ms step:3753/6250 train_loss:2.2070 aux_loss:0.0307 train_time:2013036ms step_avg:537.81ms step:3754/6250 train_loss:2.1671 aux_loss:0.0307 train_time:2013577ms step_avg:537.81ms step:3755/6250 train_loss:2.2566 aux_loss:0.0307 train_time:2014141ms step_avg:537.82ms step:3756/6250 train_loss:2.3571 aux_loss:0.0309 train_time:2014701ms step_avg:537.83ms step:3757/6250 train_loss:2.2346 aux_loss:0.0303 train_time:2015262ms step_avg:537.83ms step:3758/6250 train_loss:2.2122 aux_loss:0.0308 train_time:2015799ms step_avg:537.83ms step:3759/6250 train_loss:2.1331 aux_loss:0.0308 train_time:2016332ms step_avg:537.83ms step:3760/6250 train_loss:2.2252 aux_loss:0.0306 train_time:2016880ms step_avg:537.83ms step:3761/6250 train_loss:2.1712 aux_loss:0.0305 train_time:2017448ms step_avg:537.84ms step:3762/6250 train_loss:2.2599 aux_loss:0.0309 train_time:2017984ms step_avg:537.84ms step:3763/6250 train_loss:2.3302 aux_loss:0.0309 train_time:2018523ms step_avg:537.84ms step:3764/6250 train_loss:2.2489 aux_loss:0.0311 train_time:2019064ms step_avg:537.84ms step:3765/6250 train_loss:2.1717 aux_loss:0.0311 train_time:2019602ms step_avg:537.84ms step:3766/6250 train_loss:2.1794 aux_loss:0.0309 train_time:2020139ms step_avg:537.84ms step:3767/6250 train_loss:2.1845 aux_loss:0.0307 train_time:2020678ms step_avg:537.84ms step:3768/6250 train_loss:2.2379 aux_loss:0.0308 train_time:2021212ms step_avg:537.84ms step:3769/6250 train_loss:2.2745 aux_loss:0.0311 train_time:2021758ms step_avg:537.84ms step:3770/6250 train_loss:2.3374 aux_loss:0.0311 train_time:2022297ms step_avg:537.85ms step:3771/6250 train_loss:2.1673 aux_loss:0.0309 train_time:2022840ms step_avg:537.85ms step:3772/6250 train_loss:2.1858 aux_loss:0.0309 train_time:2023377ms step_avg:537.85ms step:3773/6250 train_loss:2.2843 aux_loss:0.0308 train_time:2023915ms step_avg:537.85ms step:3774/6250 train_loss:2.2034 aux_loss:0.0311 train_time:2024458ms step_avg:537.85ms step:3775/6250 train_loss:2.1869 aux_loss:0.0305 train_time:2024993ms step_avg:537.85ms step:3776/6250 train_loss:2.1946 aux_loss:0.0303 train_time:2025546ms step_avg:537.85ms step:3777/6250 train_loss:2.1744 aux_loss:0.0313 train_time:2026096ms step_avg:537.85ms step:3778/6250 train_loss:2.2342 aux_loss:0.0314 train_time:2026634ms step_avg:537.85ms step:3779/6250 train_loss:2.2792 aux_loss:0.0307 train_time:2027167ms step_avg:537.85ms step:3780/6250 train_loss:2.1800 aux_loss:0.0303 train_time:2027698ms step_avg:537.85ms step:3781/6250 train_loss:2.1611 aux_loss:0.0307 train_time:2028228ms step_avg:537.85ms step:3782/6250 train_loss:2.3014 aux_loss:0.0305 train_time:2028758ms step_avg:537.85ms step:3783/6250 train_loss:2.1202 aux_loss:0.0303 train_time:2029292ms step_avg:537.85ms step:3784/6250 train_loss:2.1628 aux_loss:0.0305 train_time:2029825ms step_avg:537.84ms step:3785/6250 train_loss:2.2693 aux_loss:0.0307 train_time:2030361ms step_avg:537.84ms step:3786/6250 train_loss:2.2339 aux_loss:0.0308 train_time:2030897ms step_avg:537.84ms step:3787/6250 train_loss:2.1803 aux_loss:0.0309 train_time:2031429ms step_avg:537.84ms step:3788/6250 train_loss:2.2253 aux_loss:0.0302 train_time:2031958ms step_avg:537.84ms step:3789/6250 train_loss:2.2066 aux_loss:0.0299 train_time:2032489ms step_avg:537.84ms step:3790/6250 train_loss:2.1797 aux_loss:0.0305 train_time:2033027ms step_avg:537.84ms step:3791/6250 train_loss:2.2386 aux_loss:0.0311 train_time:2033584ms step_avg:537.84ms step:3792/6250 train_loss:2.2544 aux_loss:0.0308 train_time:2034139ms step_avg:537.85ms step:3793/6250 train_loss:2.2184 aux_loss:0.0303 train_time:2034708ms step_avg:537.86ms step:3794/6250 train_loss:2.1907 aux_loss:0.0303 train_time:2035283ms step_avg:537.87ms step:3795/6250 train_loss:2.2197 aux_loss:0.0310 train_time:2035821ms step_avg:537.87ms step:3796/6250 train_loss:2.2571 aux_loss:0.0313 train_time:2036349ms step_avg:537.86ms step:3797/6250 train_loss:2.2906 aux_loss:0.0305 train_time:2036881ms step_avg:537.86ms step:3798/6250 train_loss:2.2103 aux_loss:0.0307 train_time:2037411ms step_avg:537.86ms step:3799/6250 train_loss:2.2648 aux_loss:0.0314 train_time:2037941ms step_avg:537.86ms step:3800/6250 train_loss:2.1828 aux_loss:0.0313 train_time:2038469ms step_avg:537.85ms step:3801/6250 train_loss:2.2545 aux_loss:0.0306 train_time:2038998ms step_avg:537.85ms step:3802/6250 train_loss:2.2429 aux_loss:0.0303 train_time:2039529ms step_avg:537.85ms step:3803/6250 train_loss:2.1918 aux_loss:0.0307 train_time:2040059ms step_avg:537.85ms step:3804/6250 train_loss:2.1253 aux_loss:0.0306 train_time:2040589ms step_avg:537.85ms step:3805/6250 train_loss:2.1628 aux_loss:0.0305 train_time:2041119ms step_avg:537.84ms step:3806/6250 train_loss:2.2601 aux_loss:0.0303 train_time:2041649ms step_avg:537.84ms step:3807/6250 train_loss:2.2215 aux_loss:0.0304 train_time:2042178ms step_avg:537.84ms step:3808/6250 train_loss:2.2479 aux_loss:0.0303 train_time:2042708ms step_avg:537.84ms step:3809/6250 train_loss:2.2135 aux_loss:0.0305 train_time:2043237ms step_avg:537.84ms step:3810/6250 train_loss:2.2400 aux_loss:0.0306 train_time:2043767ms step_avg:537.83ms step:3811/6250 train_loss:2.2971 aux_loss:0.0302 train_time:2044297ms step_avg:537.83ms step:3812/6250 train_loss:2.2518 aux_loss:0.0306 train_time:2044826ms step_avg:537.83ms step:3813/6250 train_loss:2.2013 aux_loss:0.0308 train_time:2045356ms step_avg:537.83ms step:3814/6250 train_loss:2.1932 aux_loss:0.0302 train_time:2045886ms step_avg:537.82ms step:3815/6250 train_loss:2.1997 aux_loss:0.0304 train_time:2046416ms step_avg:537.82ms step:3816/6250 train_loss:2.1564 aux_loss:0.0306 train_time:2046946ms step_avg:537.82ms step:3817/6250 train_loss:2.2425 aux_loss:0.0308 train_time:2047477ms step_avg:537.82ms step:3818/6250 train_loss:2.1770 aux_loss:0.0307 train_time:2048029ms step_avg:537.82ms step:3819/6250 train_loss:2.2134 aux_loss:0.0306 train_time:2048574ms step_avg:537.82ms step:3820/6250 train_loss:2.1789 aux_loss:0.0307 train_time:2049102ms step_avg:537.82ms step:3821/6250 train_loss:2.2155 aux_loss:0.0304 train_time:2049635ms step_avg:537.82ms step:3822/6250 train_loss:2.2486 aux_loss:0.0304 train_time:2050185ms step_avg:537.82ms step:3823/6250 train_loss:2.1524 aux_loss:0.0308 train_time:2050714ms step_avg:537.82ms step:3824/6250 train_loss:2.2537 aux_loss:0.0311 train_time:2051247ms step_avg:537.82ms step:3825/6250 train_loss:2.2168 aux_loss:0.0307 train_time:2051784ms step_avg:537.82ms step:3826/6250 train_loss:2.2023 aux_loss:0.0308 train_time:2052316ms step_avg:537.82ms step:3827/6250 train_loss:2.2808 aux_loss:0.0310 train_time:2052852ms step_avg:537.82ms step:3828/6250 train_loss:2.1543 aux_loss:0.0303 train_time:2053392ms step_avg:537.82ms step:3829/6250 train_loss:2.1514 aux_loss:0.0299 train_time:2053933ms step_avg:537.82ms step:3830/6250 train_loss:2.2430 aux_loss:0.0307 train_time:2054469ms step_avg:537.82ms step:3831/6250 train_loss:2.2086 aux_loss:0.0307 train_time:2055030ms step_avg:537.83ms step:3832/6250 train_loss:2.2159 aux_loss:0.0305 train_time:2055603ms step_avg:537.83ms step:3833/6250 train_loss:2.0986 aux_loss:0.0307 train_time:2056163ms step_avg:537.84ms step:3834/6250 train_loss:2.2659 aux_loss:0.0304 train_time:2056734ms step_avg:537.85ms step:3835/6250 train_loss:2.2200 aux_loss:0.0303 train_time:2057309ms step_avg:537.86ms step:3836/6250 train_loss:2.1746 aux_loss:0.0306 train_time:2057876ms step_avg:537.87ms step:3837/6250 train_loss:2.2290 aux_loss:0.0306 train_time:2058465ms step_avg:537.88ms step:3838/6250 train_loss:2.2074 aux_loss:0.0307 train_time:2059047ms step_avg:537.89ms step:3839/6250 train_loss:2.2620 aux_loss:0.0302 train_time:2059608ms step_avg:537.90ms step:3840/6250 train_loss:2.1442 aux_loss:0.0306 train_time:2060169ms step_avg:537.90ms step:3841/6250 train_loss:2.1754 aux_loss:0.0305 train_time:2060710ms step_avg:537.90ms step:3842/6250 train_loss:2.3087 aux_loss:0.0302 train_time:2061271ms step_avg:537.91ms step:3843/6250 train_loss:2.2341 aux_loss:0.0306 train_time:2061836ms step_avg:537.92ms step:3844/6250 train_loss:2.2726 aux_loss:0.0306 train_time:2062396ms step_avg:537.92ms step:3845/6250 train_loss:2.2556 aux_loss:0.0303 train_time:2062957ms step_avg:537.93ms step:3846/6250 train_loss:2.2319 aux_loss:0.0303 train_time:2063503ms step_avg:537.93ms step:3847/6250 train_loss:2.2085 aux_loss:0.0303 train_time:2064037ms step_avg:537.93ms step:3848/6250 train_loss:2.1637 aux_loss:0.0305 train_time:2064571ms step_avg:537.93ms step:3849/6250 train_loss:2.2307 aux_loss:0.0303 train_time:2065102ms step_avg:537.93ms step:3850/6250 train_loss:2.3142 aux_loss:0.0307 train_time:2065633ms step_avg:537.93ms step:3851/6250 train_loss:2.2966 aux_loss:0.0306 train_time:2066162ms step_avg:537.92ms step:3852/6250 train_loss:2.2781 aux_loss:0.0304 train_time:2066693ms step_avg:537.92ms step:3853/6250 train_loss:2.3209 aux_loss:0.0304 train_time:2067223ms step_avg:537.92ms step:3854/6250 train_loss:2.2310 aux_loss:0.0309 train_time:2067754ms step_avg:537.92ms step:3855/6250 train_loss:2.2340 aux_loss:0.0309 train_time:2068284ms step_avg:537.92ms step:3856/6250 train_loss:2.1950 aux_loss:0.0305 train_time:2068814ms step_avg:537.91ms step:3857/6250 train_loss:2.1400 aux_loss:0.0303 train_time:2069343ms step_avg:537.91ms step:3858/6250 train_loss:2.2277 aux_loss:0.0306 train_time:2069871ms step_avg:537.91ms step:3859/6250 train_loss:2.2006 aux_loss:0.0304 train_time:2070401ms step_avg:537.91ms step:3860/6250 train_loss:2.2304 aux_loss:0.0306 train_time:2070931ms step_avg:537.90ms step:3861/6250 train_loss:2.2603 aux_loss:0.0310 train_time:2071461ms step_avg:537.90ms step:3862/6250 train_loss:2.1534 aux_loss:0.0306 train_time:2071991ms step_avg:537.90ms step:3863/6250 train_loss:2.2191 aux_loss:0.0314 train_time:2072521ms step_avg:537.90ms step:3864/6250 train_loss:2.1229 aux_loss:0.0307 train_time:2073050ms step_avg:537.90ms step:3865/6250 train_loss:2.1611 aux_loss:0.0306 train_time:2073581ms step_avg:537.89ms step:3866/6250 train_loss:2.2372 aux_loss:0.0309 train_time:2074110ms step_avg:537.89ms step:3867/6250 train_loss:2.1804 aux_loss:0.0305 train_time:2074641ms step_avg:537.89ms step:3868/6250 train_loss:2.1869 aux_loss:0.0304 train_time:2075200ms step_avg:537.90ms step:3869/6250 train_loss:2.2655 aux_loss:0.0303 train_time:2075760ms step_avg:537.90ms step:3870/6250 train_loss:2.1177 aux_loss:0.0305 train_time:2076294ms step_avg:537.90ms step:3871/6250 train_loss:2.2640 aux_loss:0.0310 train_time:2076823ms step_avg:537.90ms step:3872/6250 train_loss:2.1767 aux_loss:0.0309 train_time:2077351ms step_avg:537.90ms step:3873/6250 train_loss:2.2186 aux_loss:0.0305 train_time:2077880ms step_avg:537.89ms step:3874/6250 train_loss:2.1696 aux_loss:0.0301 train_time:2078409ms step_avg:537.89ms step:3875/6250 train_loss:2.1988 aux_loss:0.0310 train_time:2078937ms step_avg:537.89ms step:3876/6250 train_loss:2.1746 aux_loss:0.0309 train_time:2079466ms step_avg:537.89ms step:3877/6250 train_loss:2.2208 aux_loss:0.0302 train_time:2079994ms step_avg:537.88ms step:3878/6250 train_loss:2.2609 aux_loss:0.0306 train_time:2080523ms step_avg:537.88ms step:3879/6250 train_loss:2.2031 aux_loss:0.0309 train_time:2081054ms step_avg:537.88ms step:3880/6250 train_loss:2.1560 aux_loss:0.0305 train_time:2081581ms step_avg:537.88ms step:3881/6250 train_loss:2.2857 aux_loss:0.0302 train_time:2082110ms step_avg:537.87ms step:3882/6250 train_loss:2.3134 aux_loss:0.0307 train_time:2082644ms step_avg:537.87ms step:3883/6250 train_loss:2.1773 aux_loss:0.0306 train_time:2083218ms step_avg:537.88ms step:3884/6250 train_loss:2.1796 aux_loss:0.0301 train_time:2083797ms step_avg:537.89ms step:3885/6250 train_loss:2.1870 aux_loss:0.0300 train_time:2084368ms step_avg:537.90ms step:3886/6250 train_loss:2.2659 aux_loss:0.0304 train_time:2084937ms step_avg:537.91ms step:3887/6250 train_loss:2.2591 aux_loss:0.0304 train_time:2085477ms step_avg:537.91ms step:3888/6250 train_loss:2.2321 aux_loss:0.0304 train_time:2086005ms step_avg:537.91ms step:3889/6250 train_loss:2.2842 aux_loss:0.0305 train_time:2086534ms step_avg:537.91ms step:3890/6250 train_loss:2.2556 aux_loss:0.0309 train_time:2087062ms step_avg:537.90ms step:3891/6250 train_loss:2.2001 aux_loss:0.0305 train_time:2087592ms step_avg:537.90ms step:3892/6250 train_loss:2.1722 aux_loss:0.0305 train_time:2088121ms step_avg:537.90ms step:3893/6250 train_loss:2.1980 aux_loss:0.0304 train_time:2088649ms step_avg:537.90ms step:3894/6250 train_loss:2.1771 aux_loss:0.0302 train_time:2089177ms step_avg:537.89ms step:3895/6250 train_loss:2.1627 aux_loss:0.0305 train_time:2089704ms step_avg:537.89ms step:3896/6250 train_loss:2.1397 aux_loss:0.0304 train_time:2090233ms step_avg:537.89ms step:3897/6250 train_loss:2.1650 aux_loss:0.0306 train_time:2090761ms step_avg:537.89ms step:3898/6250 train_loss:2.1801 aux_loss:0.0305 train_time:2091290ms step_avg:537.88ms step:3899/6250 train_loss:2.3253 aux_loss:0.0307 train_time:2091819ms step_avg:537.88ms step:3900/6250 train_loss:2.2821 aux_loss:0.0304 train_time:2092348ms step_avg:537.88ms step:3901/6250 train_loss:2.2462 aux_loss:0.0302 train_time:2092878ms step_avg:537.88ms step:3902/6250 train_loss:2.2603 aux_loss:0.0304 train_time:2093406ms step_avg:537.87ms step:3903/6250 train_loss:2.1959 aux_loss:0.0304 train_time:2093934ms step_avg:537.87ms step:3904/6250 train_loss:2.2019 aux_loss:0.0304 train_time:2094463ms step_avg:537.87ms step:3905/6250 train_loss:2.2514 aux_loss:0.0308 train_time:2095000ms step_avg:537.87ms step:3906/6250 train_loss:2.2003 aux_loss:0.0308 train_time:2095551ms step_avg:537.87ms step:3907/6250 train_loss:2.1867 aux_loss:0.0307 train_time:2096089ms step_avg:537.87ms step:3908/6250 train_loss:2.2702 aux_loss:0.0306 train_time:2096618ms step_avg:537.87ms step:3909/6250 train_loss:2.1672 aux_loss:0.0305 train_time:2097147ms step_avg:537.87ms step:3910/6250 train_loss:2.2222 aux_loss:0.0304 train_time:2097675ms step_avg:537.87ms step:3911/6250 train_loss:2.1604 aux_loss:0.0301 train_time:2098203ms step_avg:537.86ms step:3912/6250 train_loss:2.1397 aux_loss:0.0311 train_time:2098732ms step_avg:537.86ms step:3913/6250 train_loss:2.1496 aux_loss:0.0311 train_time:2099261ms step_avg:537.86ms step:3914/6250 train_loss:2.1841 aux_loss:0.0305 train_time:2099790ms step_avg:537.86ms step:3915/6250 train_loss:2.2304 aux_loss:0.0301 train_time:2100318ms step_avg:537.85ms step:3916/6250 train_loss:2.2343 aux_loss:0.0299 train_time:2100847ms step_avg:537.85ms step:3917/6250 train_loss:2.2368 aux_loss:0.0303 train_time:2101376ms step_avg:537.85ms step:3918/6250 train_loss:2.2347 aux_loss:0.0307 train_time:2101905ms step_avg:537.85ms step:3919/6250 train_loss:2.1015 aux_loss:0.0303 train_time:2102435ms step_avg:537.84ms step:3920/6250 train_loss:2.2601 aux_loss:0.0307 train_time:2102965ms step_avg:537.84ms step:3921/6250 train_loss:2.1849 aux_loss:0.0311 train_time:2103494ms step_avg:537.84ms step:3922/6250 train_loss:2.2241 aux_loss:0.0306 train_time:2104022ms step_avg:537.84ms step:3923/6250 train_loss:2.2505 aux_loss:0.0299 train_time:2104550ms step_avg:537.84ms step:3924/6250 train_loss:2.2296 aux_loss:0.0305 train_time:2105078ms step_avg:537.83ms step:3925/6250 train_loss:2.1184 aux_loss:0.0310 train_time:2105607ms step_avg:537.83ms step:3926/6250 train_loss:2.2385 aux_loss:0.0307 train_time:2106136ms step_avg:537.83ms step:3927/6250 train_loss:2.1506 aux_loss:0.0305 train_time:2106663ms step_avg:537.83ms step:3928/6250 train_loss:2.2307 aux_loss:0.0310 train_time:2107192ms step_avg:537.82ms step:3929/6250 train_loss:2.3047 aux_loss:0.0309 train_time:2107720ms step_avg:537.82ms step:3930/6250 train_loss:2.2396 aux_loss:0.0302 train_time:2108272ms step_avg:537.82ms step:3931/6250 train_loss:2.2278 aux_loss:0.0305 train_time:2108811ms step_avg:537.82ms step:3932/6250 train_loss:2.2845 aux_loss:0.0308 train_time:2109369ms step_avg:537.83ms step:3933/6250 train_loss:2.2932 aux_loss:0.0304 train_time:2109897ms step_avg:537.83ms step:3934/6250 train_loss:2.1549 aux_loss:0.0301 train_time:2110425ms step_avg:537.82ms step:3935/6250 train_loss:2.1951 aux_loss:0.0302 train_time:2110952ms step_avg:537.82ms step:3936/6250 train_loss:2.2597 aux_loss:0.0305 train_time:2111483ms step_avg:537.82ms step:3937/6250 train_loss:2.1951 aux_loss:0.0308 train_time:2112010ms step_avg:537.82ms step:3938/6250 train_loss:2.2501 aux_loss:0.0306 train_time:2112539ms step_avg:537.82ms step:3939/6250 train_loss:2.1782 aux_loss:0.0300 train_time:2113068ms step_avg:537.81ms step:3940/6250 train_loss:2.2164 aux_loss:0.0305 train_time:2113597ms step_avg:537.81ms step:3941/6250 train_loss:2.3006 aux_loss:0.0304 train_time:2114140ms step_avg:537.81ms step:3942/6250 train_loss:2.2180 aux_loss:0.0304 train_time:2114669ms step_avg:537.81ms step:3943/6250 train_loss:2.1930 aux_loss:0.0301 train_time:2115209ms step_avg:537.81ms step:3944/6250 train_loss:2.2451 aux_loss:0.0305 train_time:2115768ms step_avg:537.82ms step:3945/6250 train_loss:2.2543 aux_loss:0.0306 train_time:2116303ms step_avg:537.82ms step:3946/6250 train_loss:2.2122 aux_loss:0.0299 train_time:2116831ms step_avg:537.81ms step:3947/6250 train_loss:2.2267 aux_loss:0.0298 train_time:2117360ms step_avg:537.81ms step:3948/6250 train_loss:2.2396 aux_loss:0.0302 train_time:2117889ms step_avg:537.81ms step:3949/6250 train_loss:2.1659 aux_loss:0.0304 train_time:2118418ms step_avg:537.81ms step:3950/6250 train_loss:2.2637 aux_loss:0.0302 train_time:2118947ms step_avg:537.80ms step:3951/6250 train_loss:2.1679 aux_loss:0.0301 train_time:2119493ms step_avg:537.81ms step:3952/6250 train_loss:2.1974 aux_loss:0.0302 train_time:2120023ms step_avg:537.80ms step:3953/6250 train_loss:2.2299 aux_loss:0.0300 train_time:2120553ms step_avg:537.80ms step:3954/6250 train_loss:2.2520 aux_loss:0.0302 train_time:2121081ms step_avg:537.80ms step:3955/6250 train_loss:2.2107 aux_loss:0.0303 train_time:2121610ms step_avg:537.80ms step:3956/6250 train_loss:2.2213 aux_loss:0.0299 train_time:2122139ms step_avg:537.80ms step:3957/6250 train_loss:2.2228 aux_loss:0.0302 train_time:2122668ms step_avg:537.79ms step:3958/6250 train_loss:2.1117 aux_loss:0.0301 train_time:2123196ms step_avg:537.79ms step:3959/6250 train_loss:2.1798 aux_loss:0.0303 train_time:2123724ms step_avg:537.79ms step:3960/6250 train_loss:2.2716 aux_loss:0.0294 train_time:2124274ms step_avg:537.79ms step:3961/6250 train_loss:2.1922 aux_loss:0.0299 train_time:2124802ms step_avg:537.79ms step:3962/6250 train_loss:2.1585 aux_loss:0.0303 train_time:2125333ms step_avg:537.79ms step:3963/6250 train_loss:2.2014 aux_loss:0.0303 train_time:2125863ms step_avg:537.78ms step:3964/6250 train_loss:2.2299 aux_loss:0.0301 train_time:2126410ms step_avg:537.79ms step:3965/6250 train_loss:2.1934 aux_loss:0.0304 train_time:2126940ms step_avg:537.79ms step:3966/6250 train_loss:2.1655 aux_loss:0.0306 train_time:2127476ms step_avg:537.78ms step:3967/6250 train_loss:2.1818 aux_loss:0.0303 train_time:2128004ms step_avg:537.78ms step:3968/6250 train_loss:2.2483 aux_loss:0.0300 train_time:2128533ms step_avg:537.78ms step:3969/6250 train_loss:2.1470 aux_loss:0.0297 train_time:2129062ms step_avg:537.78ms step:3970/6250 train_loss:2.2444 aux_loss:0.0302 train_time:2129591ms step_avg:537.78ms step:3971/6250 train_loss:2.2324 aux_loss:0.0302 train_time:2130120ms step_avg:537.77ms step:3972/6250 train_loss:2.2995 aux_loss:0.0297 train_time:2130652ms step_avg:537.77ms step:3973/6250 train_loss:2.2263 aux_loss:0.0302 train_time:2131196ms step_avg:537.77ms step:3974/6250 train_loss:2.2410 aux_loss:0.0306 train_time:2131735ms step_avg:537.77ms step:3975/6250 train_loss:2.2645 aux_loss:0.0305 train_time:2132267ms step_avg:537.77ms step:3976/6250 train_loss:2.2485 aux_loss:0.0304 train_time:2132808ms step_avg:537.77ms step:3977/6250 train_loss:2.2640 aux_loss:0.0299 train_time:2133350ms step_avg:537.77ms step:3978/6250 train_loss:2.2379 aux_loss:0.0300 train_time:2133901ms step_avg:537.78ms step:3979/6250 train_loss:2.1661 aux_loss:0.0301 train_time:2134435ms step_avg:537.78ms step:3980/6250 train_loss:2.2289 aux_loss:0.0304 train_time:2134965ms step_avg:537.77ms step:3981/6250 train_loss:2.2029 aux_loss:0.0298 train_time:2135513ms step_avg:537.78ms step:3982/6250 train_loss:2.1811 aux_loss:0.0299 train_time:2136064ms step_avg:537.78ms step:3983/6250 train_loss:2.1556 aux_loss:0.0302 train_time:2136593ms step_avg:537.78ms step:3984/6250 train_loss:2.2651 aux_loss:0.0307 train_time:2137120ms step_avg:537.78ms step:3985/6250 train_loss:2.1902 aux_loss:0.0303 train_time:2137647ms step_avg:537.77ms step:3986/6250 train_loss:2.2429 aux_loss:0.0306 train_time:2138175ms step_avg:537.77ms step:3987/6250 train_loss:2.2226 aux_loss:0.0301 train_time:2138705ms step_avg:537.77ms step:3988/6250 train_loss:2.2770 aux_loss:0.0301 train_time:2139234ms step_avg:537.77ms step:3989/6250 train_loss:2.2321 aux_loss:0.0304 train_time:2139763ms step_avg:537.76ms step:3990/6250 train_loss:2.2805 aux_loss:0.0305 train_time:2140294ms step_avg:537.76ms step:3991/6250 train_loss:2.3091 aux_loss:0.0301 train_time:2140825ms step_avg:537.76ms step:3992/6250 train_loss:2.2550 aux_loss:0.0308 train_time:2141356ms step_avg:537.76ms step:3993/6250 train_loss:2.2071 aux_loss:0.0307 train_time:2141885ms step_avg:537.76ms step:3994/6250 train_loss:2.2048 aux_loss:0.0301 train_time:2142414ms step_avg:537.75ms step:3995/6250 train_loss:2.1990 aux_loss:0.0303 train_time:2142944ms step_avg:537.75ms step:3996/6250 train_loss:2.1267 aux_loss:0.0303 train_time:2143476ms step_avg:537.75ms step:3997/6250 train_loss:2.1953 aux_loss:0.0303 train_time:2144005ms step_avg:537.75ms step:3998/6250 train_loss:2.1965 aux_loss:0.0305 train_time:2144534ms step_avg:537.75ms step:3999/6250 train_loss:2.2828 aux_loss:0.0301 train_time:2145061ms step_avg:537.74ms step:4000/6250 train_loss:2.2364 aux_loss:0.0303 train_time:2145591ms step_avg:537.74ms step:4000/6250 val_loss:2.2047 val_aux_loss:0.0301 train_time:2145591ms step_avg:537.74ms step:4001/6250 train_loss:2.1824 aux_loss:0.0300 train_time:2146118ms step_avg:537.74ms step:4002/6250 train_loss:2.1731 aux_loss:0.0301 train_time:2146647ms step_avg:537.74ms step:4003/6250 train_loss:2.2107 aux_loss:0.0304 train_time:2147176ms step_avg:537.74ms step:4004/6250 train_loss:2.2081 aux_loss:0.0299 train_time:2147708ms step_avg:537.73ms step:4005/6250 train_loss:2.2091 aux_loss:0.0301 train_time:2148236ms step_avg:537.73ms step:4006/6250 train_loss:2.1120 aux_loss:0.0303 train_time:2148767ms step_avg:537.73ms step:4007/6250 train_loss:2.2950 aux_loss:0.0305 train_time:2149294ms step_avg:537.73ms step:4008/6250 train_loss:2.2117 aux_loss:0.0302 train_time:2149824ms step_avg:537.72ms step:4009/6250 train_loss:2.3141 aux_loss:0.0307 train_time:2150352ms step_avg:537.72ms step:4010/6250 train_loss:2.2317 aux_loss:0.0301 train_time:2150883ms step_avg:537.72ms step:4011/6250 train_loss:2.2381 aux_loss:0.0302 train_time:2151409ms step_avg:537.72ms step:4012/6250 train_loss:2.1891 aux_loss:0.0309 train_time:2151940ms step_avg:537.72ms step:4013/6250 train_loss:2.3271 aux_loss:0.0302 train_time:2152468ms step_avg:537.71ms step:4014/6250 train_loss:2.1601 aux_loss:0.0298 train_time:2152998ms step_avg:537.71ms step:4015/6250 train_loss:2.1849 aux_loss:0.0304 train_time:2153526ms step_avg:537.71ms step:4016/6250 train_loss:2.2252 aux_loss:0.0298 train_time:2154055ms step_avg:537.71ms step:4017/6250 train_loss:2.2293 aux_loss:0.0296 train_time:2154582ms step_avg:537.70ms step:4018/6250 train_loss:2.1948 aux_loss:0.0300 train_time:2155113ms step_avg:537.70ms step:4019/6250 train_loss:2.1419 aux_loss:0.0303 train_time:2155641ms step_avg:537.70ms step:4020/6250 train_loss:2.2454 aux_loss:0.0299 train_time:2156170ms step_avg:537.70ms step:4021/6250 train_loss:2.2917 aux_loss:0.0300 train_time:2156699ms step_avg:537.70ms step:4022/6250 train_loss:2.2100 aux_loss:0.0302 train_time:2157229ms step_avg:537.69ms step:4023/6250 train_loss:2.1500 aux_loss:0.0299 train_time:2157767ms step_avg:537.69ms step:4024/6250 train_loss:2.1928 aux_loss:0.0296 train_time:2158317ms step_avg:537.70ms step:4025/6250 train_loss:2.2214 aux_loss:0.0303 train_time:2158855ms step_avg:537.70ms step:4026/6250 train_loss:2.2670 aux_loss:0.0306 train_time:2159383ms step_avg:537.69ms step:4027/6250 train_loss:2.1480 aux_loss:0.0298 train_time:2159913ms step_avg:537.69ms step:4028/6250 train_loss:2.1167 aux_loss:0.0296 train_time:2160441ms step_avg:537.69ms step:4029/6250 train_loss:2.2070 aux_loss:0.0303 train_time:2160971ms step_avg:537.69ms step:4030/6250 train_loss:2.1189 aux_loss:0.0303 train_time:2161498ms step_avg:537.69ms step:4031/6250 train_loss:2.2806 aux_loss:0.0299 train_time:2162028ms step_avg:537.68ms step:4032/6250 train_loss:2.2787 aux_loss:0.0299 train_time:2162578ms step_avg:537.69ms step:4033/6250 train_loss:2.0939 aux_loss:0.0294 train_time:2163132ms step_avg:537.69ms step:4034/6250 train_loss:2.1441 aux_loss:0.0294 train_time:2163661ms step_avg:537.69ms step:4035/6250 train_loss:2.2071 aux_loss:0.0299 train_time:2164189ms step_avg:537.69ms step:4036/6250 train_loss:2.2266 aux_loss:0.0298 train_time:2164721ms step_avg:537.69ms step:4037/6250 train_loss:2.2566 aux_loss:0.0297 train_time:2165250ms step_avg:537.68ms step:4038/6250 train_loss:2.2428 aux_loss:0.0294 train_time:2165780ms step_avg:537.68ms step:4039/6250 train_loss:2.1441 aux_loss:0.0296 train_time:2166309ms step_avg:537.68ms step:4040/6250 train_loss:2.2124 aux_loss:0.0301 train_time:2166837ms step_avg:537.68ms step:4041/6250 train_loss:2.2262 aux_loss:0.0297 train_time:2167365ms step_avg:537.67ms step:4042/6250 train_loss:2.2108 aux_loss:0.0294 train_time:2167895ms step_avg:537.67ms step:4043/6250 train_loss:2.1808 aux_loss:0.0297 train_time:2168421ms step_avg:537.67ms step:4044/6250 train_loss:2.2611 aux_loss:0.0303 train_time:2168950ms step_avg:537.67ms step:4045/6250 train_loss:2.1525 aux_loss:0.0297 train_time:2169478ms step_avg:537.66ms step:4046/6250 train_loss:2.1871 aux_loss:0.0296 train_time:2170007ms step_avg:537.66ms step:4047/6250 train_loss:2.1527 aux_loss:0.0303 train_time:2170535ms step_avg:537.66ms step:4048/6250 train_loss:2.1992 aux_loss:0.0299 train_time:2171062ms step_avg:537.66ms step:4049/6250 train_loss:2.2189 aux_loss:0.0297 train_time:2171593ms step_avg:537.66ms step:4050/6250 train_loss:2.3567 aux_loss:0.0297 train_time:2172122ms step_avg:537.65ms step:4051/6250 train_loss:2.2541 aux_loss:0.0298 train_time:2172650ms step_avg:537.65ms step:4052/6250 train_loss:2.2602 aux_loss:0.0299 train_time:2173179ms step_avg:537.65ms step:4053/6250 train_loss:2.1403 aux_loss:0.0297 train_time:2173710ms step_avg:537.65ms step:4054/6250 train_loss:2.2396 aux_loss:0.0297 train_time:2174237ms step_avg:537.65ms step:4055/6250 train_loss:2.2046 aux_loss:0.0296 train_time:2174767ms step_avg:537.64ms step:4056/6250 train_loss:2.2278 aux_loss:0.0297 train_time:2175296ms step_avg:537.64ms step:4057/6250 train_loss:2.2476 aux_loss:0.0301 train_time:2175824ms step_avg:537.64ms step:4058/6250 train_loss:2.1502 aux_loss:0.0297 train_time:2176353ms step_avg:537.64ms step:4059/6250 train_loss:2.3232 aux_loss:0.0302 train_time:2176883ms step_avg:537.63ms step:4060/6250 train_loss:2.2269 aux_loss:0.0301 train_time:2177412ms step_avg:537.63ms step:4061/6250 train_loss:2.2315 aux_loss:0.0295 train_time:2177951ms step_avg:537.63ms step:4062/6250 train_loss:2.1785 aux_loss:0.0287 train_time:2178500ms step_avg:537.64ms step:4063/6250 train_loss:2.1845 aux_loss:0.0291 train_time:2179035ms step_avg:537.64ms step:4064/6250 train_loss:2.1397 aux_loss:0.0298 train_time:2179563ms step_avg:537.63ms step:4065/6250 train_loss:2.2197 aux_loss:0.0298 train_time:2180094ms step_avg:537.63ms step:4066/6250 train_loss:2.1390 aux_loss:0.0295 train_time:2180623ms step_avg:537.63ms step:4067/6250 train_loss:2.1107 aux_loss:0.0296 train_time:2181154ms step_avg:537.63ms step:4068/6250 train_loss:2.2555 aux_loss:0.0292 train_time:2181682ms step_avg:537.62ms step:4069/6250 train_loss:2.2454 aux_loss:0.0293 train_time:2182210ms step_avg:537.62ms step:4070/6250 train_loss:2.2185 aux_loss:0.0302 train_time:2182739ms step_avg:537.62ms step:4071/6250 train_loss:2.1586 aux_loss:0.0299 train_time:2183266ms step_avg:537.62ms step:4072/6250 train_loss:2.1303 aux_loss:0.0295 train_time:2183796ms step_avg:537.62ms step:4073/6250 train_loss:2.1546 aux_loss:0.0296 train_time:2184324ms step_avg:537.61ms step:4074/6250 train_loss:2.1598 aux_loss:0.0298 train_time:2184853ms step_avg:537.61ms step:4075/6250 train_loss:2.1904 aux_loss:0.0293 train_time:2185381ms step_avg:537.61ms step:4076/6250 train_loss:2.2029 aux_loss:0.0299 train_time:2185911ms step_avg:537.61ms step:4077/6250 train_loss:2.2789 aux_loss:0.0298 train_time:2186439ms step_avg:537.60ms step:4078/6250 train_loss:2.1570 aux_loss:0.0291 train_time:2186969ms step_avg:537.60ms step:4079/6250 train_loss:2.2646 aux_loss:0.0297 train_time:2187497ms step_avg:537.60ms step:4080/6250 train_loss:2.1459 aux_loss:0.0300 train_time:2188026ms step_avg:537.60ms step:4081/6250 train_loss:2.2133 aux_loss:0.0298 train_time:2188582ms step_avg:537.60ms step:4082/6250 train_loss:2.2000 aux_loss:0.0292 train_time:2189112ms step_avg:537.60ms step:4083/6250 train_loss:2.1481 aux_loss:0.0299 train_time:2189640ms step_avg:537.60ms step:4084/6250 train_loss:2.1856 aux_loss:0.0300 train_time:2190168ms step_avg:537.60ms step:4085/6250 train_loss:2.3364 aux_loss:0.0297 train_time:2190697ms step_avg:537.59ms step:4086/6250 train_loss:2.3247 aux_loss:0.0299 train_time:2191224ms step_avg:537.59ms step:4087/6250 train_loss:2.1920 aux_loss:0.0302 train_time:2191753ms step_avg:537.59ms step:4088/6250 train_loss:2.1852 aux_loss:0.0298 train_time:2192282ms step_avg:537.59ms step:4089/6250 train_loss:2.1670 aux_loss:0.0298 train_time:2192848ms step_avg:537.59ms step:4090/6250 train_loss:2.2421 aux_loss:0.0303 train_time:2193381ms step_avg:537.59ms step:4091/6250 train_loss:2.1891 aux_loss:0.0301 train_time:2193911ms step_avg:537.59ms step:4092/6250 train_loss:2.1418 aux_loss:0.0299 train_time:2194441ms step_avg:537.59ms step:4093/6250 train_loss:2.2460 aux_loss:0.0300 train_time:2194970ms step_avg:537.59ms step:4094/6250 train_loss:2.1286 aux_loss:0.0300 train_time:2195529ms step_avg:537.59ms step:4095/6250 train_loss:2.1748 aux_loss:0.0305 train_time:2196059ms step_avg:537.59ms step:4096/6250 train_loss:2.1586 aux_loss:0.0302 train_time:2196628ms step_avg:537.60ms step:4097/6250 train_loss:2.2455 aux_loss:0.0298 train_time:2197184ms step_avg:537.60ms step:4098/6250 train_loss:2.2156 aux_loss:0.0302 train_time:2197727ms step_avg:537.60ms step:4099/6250 train_loss:2.1992 aux_loss:0.0299 train_time:2198292ms step_avg:537.61ms step:4100/6250 train_loss:2.2544 aux_loss:0.0306 train_time:2198843ms step_avg:537.61ms step:4101/6250 train_loss:2.1509 aux_loss:0.0301 train_time:2199380ms step_avg:537.61ms step:4102/6250 train_loss:2.1642 aux_loss:0.0298 train_time:2199923ms step_avg:537.62ms step:4103/6250 train_loss:2.2032 aux_loss:0.0301 train_time:2200469ms step_avg:537.62ms step:4104/6250 train_loss:2.2153 aux_loss:0.0297 train_time:2201030ms step_avg:537.62ms step:4105/6250 train_loss:2.2158 aux_loss:0.0293 train_time:2201573ms step_avg:537.62ms step:4106/6250 train_loss:2.0994 aux_loss:0.0295 train_time:2202117ms step_avg:537.63ms step:4107/6250 train_loss:2.1318 aux_loss:0.0296 train_time:2202663ms step_avg:537.63ms step:4108/6250 train_loss:2.2838 aux_loss:0.0299 train_time:2203209ms step_avg:537.63ms step:4109/6250 train_loss:2.1181 aux_loss:0.0298 train_time:2203758ms step_avg:537.63ms step:4110/6250 train_loss:2.1235 aux_loss:0.0294 train_time:2204306ms step_avg:537.64ms step:4111/6250 train_loss:2.2199 aux_loss:0.0293 train_time:2204855ms step_avg:537.64ms step:4112/6250 train_loss:2.2049 aux_loss:0.0299 train_time:2205402ms step_avg:537.64ms step:4113/6250 train_loss:2.2017 aux_loss:0.0301 train_time:2205952ms step_avg:537.64ms step:4114/6250 train_loss:2.1773 aux_loss:0.0297 train_time:2206495ms step_avg:537.65ms step:4115/6250 train_loss:2.1437 aux_loss:0.0299 train_time:2207042ms step_avg:537.65ms step:4116/6250 train_loss:2.2054 aux_loss:0.0304 train_time:2207590ms step_avg:537.65ms step:4117/6250 train_loss:2.2017 aux_loss:0.0300 train_time:2208135ms step_avg:537.65ms step:4118/6250 train_loss:2.2276 aux_loss:0.0297 train_time:2208683ms step_avg:537.65ms step:4119/6250 train_loss:2.1849 aux_loss:0.0297 train_time:2209231ms step_avg:537.66ms step:4120/6250 train_loss:2.1944 aux_loss:0.0299 train_time:2209806ms step_avg:537.67ms step:4121/6250 train_loss:2.1186 aux_loss:0.0295 train_time:2210355ms step_avg:537.67ms step:4122/6250 train_loss:2.2933 aux_loss:0.0297 train_time:2210903ms step_avg:537.67ms step:4123/6250 train_loss:2.2359 aux_loss:0.0296 train_time:2211445ms step_avg:537.67ms step:4124/6250 train_loss:2.1574 aux_loss:0.0297 train_time:2211991ms step_avg:537.67ms step:4125/6250 train_loss:2.2203 aux_loss:0.0298 train_time:2212534ms step_avg:537.68ms step:4126/6250 train_loss:2.2099 aux_loss:0.0299 train_time:2213077ms step_avg:537.68ms step:4127/6250 train_loss:2.1943 aux_loss:0.0300 train_time:2213620ms step_avg:537.68ms step:4128/6250 train_loss:2.2029 aux_loss:0.0298 train_time:2214162ms step_avg:537.68ms step:4129/6250 train_loss:2.1885 aux_loss:0.0297 train_time:2214707ms step_avg:537.68ms step:4130/6250 train_loss:2.1828 aux_loss:0.0300 train_time:2215250ms step_avg:537.68ms step:4131/6250 train_loss:2.1881 aux_loss:0.0296 train_time:2215784ms step_avg:537.68ms step:4132/6250 train_loss:2.1664 aux_loss:0.0297 train_time:2216326ms step_avg:537.68ms step:4133/6250 train_loss:2.2011 aux_loss:0.0299 train_time:2216872ms step_avg:537.68ms step:4134/6250 train_loss:2.2607 aux_loss:0.0300 train_time:2217410ms step_avg:537.68ms step:4135/6250 train_loss:2.1575 aux_loss:0.0296 train_time:2217954ms step_avg:537.69ms step:4136/6250 train_loss:2.2369 aux_loss:0.0295 train_time:2218520ms step_avg:537.69ms step:4137/6250 train_loss:2.2701 aux_loss:0.0297 train_time:2219101ms step_avg:537.70ms step:4138/6250 train_loss:2.1928 aux_loss:0.0302 train_time:2219681ms step_avg:537.71ms step:4139/6250 train_loss:2.1914 aux_loss:0.0294 train_time:2220247ms step_avg:537.72ms step:4140/6250 train_loss:2.1954 aux_loss:0.0294 train_time:2220796ms step_avg:537.72ms step:4141/6250 train_loss:2.1424 aux_loss:0.0296 train_time:2221344ms step_avg:537.73ms step:4142/6250 train_loss:2.0486 aux_loss:0.0296 train_time:2221892ms step_avg:537.73ms step:4143/6250 train_loss:2.1687 aux_loss:0.0299 train_time:2222440ms step_avg:537.73ms step:4144/6250 train_loss:2.1918 aux_loss:0.0302 train_time:2222989ms step_avg:537.73ms step:4145/6250 train_loss:2.2066 aux_loss:0.0298 train_time:2223539ms step_avg:537.74ms step:4146/6250 train_loss:2.1856 aux_loss:0.0292 train_time:2224087ms step_avg:537.74ms step:4147/6250 train_loss:2.2548 aux_loss:0.0300 train_time:2224636ms step_avg:537.74ms step:4148/6250 train_loss:2.1946 aux_loss:0.0303 train_time:2225177ms step_avg:537.74ms step:4149/6250 train_loss:2.1675 aux_loss:0.0297 train_time:2225726ms step_avg:537.74ms step:4150/6250 train_loss:2.2219 aux_loss:0.0295 train_time:2226274ms step_avg:537.75ms step:4151/6250 train_loss:2.3371 aux_loss:0.0299 train_time:2226823ms step_avg:537.75ms step:4152/6250 train_loss:2.2039 aux_loss:0.0295 train_time:2227374ms step_avg:537.75ms step:4153/6250 train_loss:2.2426 aux_loss:0.0294 train_time:2227923ms step_avg:537.76ms step:4154/6250 train_loss:2.1243 aux_loss:0.0297 train_time:2228473ms step_avg:537.76ms step:4155/6250 train_loss:2.2476 aux_loss:0.0299 train_time:2229031ms step_avg:537.76ms step:4156/6250 train_loss:2.1232 aux_loss:0.0293 train_time:2229583ms step_avg:537.77ms step:4157/6250 train_loss:2.2059 aux_loss:0.0291 train_time:2230128ms step_avg:537.77ms step:4158/6250 train_loss:2.1990 aux_loss:0.0297 train_time:2230671ms step_avg:537.77ms step:4159/6250 train_loss:2.1080 aux_loss:0.0299 train_time:2231253ms step_avg:537.78ms step:4160/6250 train_loss:2.1583 aux_loss:0.0294 train_time:2231803ms step_avg:537.78ms step:4161/6250 train_loss:2.3263 aux_loss:0.0297 train_time:2232330ms step_avg:537.78ms step:4162/6250 train_loss:2.1715 aux_loss:0.0303 train_time:2232858ms step_avg:537.78ms step:4163/6250 train_loss:2.2038 aux_loss:0.0299 train_time:2233394ms step_avg:537.78ms step:4164/6250 train_loss:2.1980 aux_loss:0.0292 train_time:2233938ms step_avg:537.78ms step:4165/6250 train_loss:2.1414 aux_loss:0.0293 train_time:2234484ms step_avg:537.78ms step:4166/6250 train_loss:2.1891 aux_loss:0.0301 train_time:2235029ms step_avg:537.78ms step:4167/6250 train_loss:2.2456 aux_loss:0.0299 train_time:2235572ms step_avg:537.78ms step:4168/6250 train_loss:2.1424 aux_loss:0.0295 train_time:2236117ms step_avg:537.79ms step:4169/6250 train_loss:2.1761 aux_loss:0.0298 train_time:2236651ms step_avg:537.79ms step:4170/6250 train_loss:2.2445 aux_loss:0.0296 train_time:2237180ms step_avg:537.78ms step:4171/6250 train_loss:2.2298 aux_loss:0.0294 train_time:2237713ms step_avg:537.78ms step:4172/6250 train_loss:2.0691 aux_loss:0.0294 train_time:2238251ms step_avg:537.78ms step:4173/6250 train_loss:2.2329 aux_loss:0.0297 train_time:2238804ms step_avg:537.79ms step:4174/6250 train_loss:2.1733 aux_loss:0.0293 train_time:2239360ms step_avg:537.79ms step:4175/6250 train_loss:2.1528 aux_loss:0.0294 train_time:2239910ms step_avg:537.79ms step:4176/6250 train_loss:2.1787 aux_loss:0.0294 train_time:2240460ms step_avg:537.80ms step:4177/6250 train_loss:2.1748 aux_loss:0.0295 train_time:2241011ms step_avg:537.80ms step:4178/6250 train_loss:2.1766 aux_loss:0.0294 train_time:2241561ms step_avg:537.80ms step:4179/6250 train_loss:2.2204 aux_loss:0.0296 train_time:2242111ms step_avg:537.81ms step:4180/6250 train_loss:2.2327 aux_loss:0.0295 train_time:2242661ms step_avg:537.81ms step:4181/6250 train_loss:2.2354 aux_loss:0.0297 train_time:2243211ms step_avg:537.81ms step:4182/6250 train_loss:2.1966 aux_loss:0.0300 train_time:2243781ms step_avg:537.82ms step:4183/6250 train_loss:2.1599 aux_loss:0.0296 train_time:2244342ms step_avg:537.82ms step:4184/6250 train_loss:2.2131 aux_loss:0.0299 train_time:2244894ms step_avg:537.83ms step:4185/6250 train_loss:2.1035 aux_loss:0.0294 train_time:2245448ms step_avg:537.83ms step:4186/6250 train_loss:2.1492 aux_loss:0.0296 train_time:2246013ms step_avg:537.84ms step:4187/6250 train_loss:2.1785 aux_loss:0.0293 train_time:2246571ms step_avg:537.84ms step:4188/6250 train_loss:2.1162 aux_loss:0.0304 train_time:2247123ms step_avg:537.85ms step:4189/6250 train_loss:2.1558 aux_loss:0.0291 train_time:2247670ms step_avg:537.85ms step:4190/6250 train_loss:2.1520 aux_loss:0.0289 train_time:2248217ms step_avg:537.85ms step:4191/6250 train_loss:2.2417 aux_loss:0.0291 train_time:2248763ms step_avg:537.85ms step:4192/6250 train_loss:2.1741 aux_loss:0.0291 train_time:2249308ms step_avg:537.85ms step:4193/6250 train_loss:2.1462 aux_loss:0.0289 train_time:2249854ms step_avg:537.86ms step:4194/6250 train_loss:2.2130 aux_loss:0.0292 train_time:2250401ms step_avg:537.86ms step:4195/6250 train_loss:2.1724 aux_loss:0.0292 train_time:2250946ms step_avg:537.86ms step:4196/6250 train_loss:2.2004 aux_loss:0.0289 train_time:2251492ms step_avg:537.86ms step:4197/6250 train_loss:2.2271 aux_loss:0.0296 train_time:2252035ms step_avg:537.86ms step:4198/6250 train_loss:2.1641 aux_loss:0.0291 train_time:2252581ms step_avg:537.87ms step:4199/6250 train_loss:2.1674 aux_loss:0.0289 train_time:2253127ms step_avg:537.87ms step:4200/6250 train_loss:2.2163 aux_loss:0.0294 train_time:2253671ms step_avg:537.87ms step:4201/6250 train_loss:2.1124 aux_loss:0.0292 train_time:2254218ms step_avg:537.87ms step:4202/6250 train_loss:2.2557 aux_loss:0.0299 train_time:2254763ms step_avg:537.87ms step:4203/6250 train_loss:2.1529 aux_loss:0.0295 train_time:2255309ms step_avg:537.87ms step:4204/6250 train_loss:2.0961 aux_loss:0.0291 train_time:2255855ms step_avg:537.88ms step:4205/6250 train_loss:2.1765 aux_loss:0.0294 train_time:2256383ms step_avg:537.87ms step:4206/6250 train_loss:2.1885 aux_loss:0.0294 train_time:2256915ms step_avg:537.87ms step:4207/6250 train_loss:2.2415 aux_loss:0.0294 train_time:2257444ms step_avg:537.87ms step:4208/6250 train_loss:2.0743 aux_loss:0.0295 train_time:2257972ms step_avg:537.87ms step:4209/6250 train_loss:2.1391 aux_loss:0.0294 train_time:2258513ms step_avg:537.87ms step:4210/6250 train_loss:2.1969 aux_loss:0.0292 train_time:2259063ms step_avg:537.87ms step:4211/6250 train_loss:2.1884 aux_loss:0.0290 train_time:2259596ms step_avg:537.87ms step:4212/6250 train_loss:2.0886 aux_loss:0.0289 train_time:2260126ms step_avg:537.87ms step:4213/6250 train_loss:2.2762 aux_loss:0.0293 train_time:2260655ms step_avg:537.87ms step:4214/6250 train_loss:2.2070 aux_loss:0.0295 train_time:2261184ms step_avg:537.86ms step:4215/6250 train_loss:2.2062 aux_loss:0.0292 train_time:2261713ms step_avg:537.86ms step:4216/6250 train_loss:2.2151 aux_loss:0.0296 train_time:2262242ms step_avg:537.86ms step:4217/6250 train_loss:2.1821 aux_loss:0.0296 train_time:2262772ms step_avg:537.86ms step:4218/6250 train_loss:2.2457 aux_loss:0.0288 train_time:2263300ms step_avg:537.86ms step:4219/6250 train_loss:2.1204 aux_loss:0.0295 train_time:2263827ms step_avg:537.85ms step:4220/6250 train_loss:2.2123 aux_loss:0.0303 train_time:2264355ms step_avg:537.85ms step:4221/6250 train_loss:2.1652 aux_loss:0.0294 train_time:2264885ms step_avg:537.85ms step:4222/6250 train_loss:2.3254 aux_loss:0.0291 train_time:2265413ms step_avg:537.85ms step:4223/6250 train_loss:2.2154 aux_loss:0.0294 train_time:2265942ms step_avg:537.85ms step:4224/6250 train_loss:2.1850 aux_loss:0.0296 train_time:2266470ms step_avg:537.84ms step:4225/6250 train_loss:2.2596 aux_loss:0.0291 train_time:2267000ms step_avg:537.84ms step:4226/6250 train_loss:2.1446 aux_loss:0.0290 train_time:2267529ms step_avg:537.84ms step:4227/6250 train_loss:2.2222 aux_loss:0.0294 train_time:2268057ms step_avg:537.84ms step:4228/6250 train_loss:2.2557 aux_loss:0.0293 train_time:2268586ms step_avg:537.83ms step:4229/6250 train_loss:2.2393 aux_loss:0.0291 train_time:2269114ms step_avg:537.83ms step:4230/6250 train_loss:2.2801 aux_loss:0.0293 train_time:2269644ms step_avg:537.83ms step:4231/6250 train_loss:2.1735 aux_loss:0.0290 train_time:2270171ms step_avg:537.83ms step:4232/6250 train_loss:2.1703 aux_loss:0.0290 train_time:2270701ms step_avg:537.83ms step:4233/6250 train_loss:2.2234 aux_loss:0.0290 train_time:2271231ms step_avg:537.82ms step:4234/6250 train_loss:2.2073 aux_loss:0.0289 train_time:2271764ms step_avg:537.82ms step:4235/6250 train_loss:2.2085 aux_loss:0.0297 train_time:2272293ms step_avg:537.82ms step:4236/6250 train_loss:2.1536 aux_loss:0.0296 train_time:2272822ms step_avg:537.82ms step:4237/6250 train_loss:2.1441 aux_loss:0.0292 train_time:2273350ms step_avg:537.82ms step:4238/6250 train_loss:2.2118 aux_loss:0.0289 train_time:2273879ms step_avg:537.81ms step:4239/6250 train_loss:2.2300 aux_loss:0.0292 train_time:2274407ms step_avg:537.81ms step:4240/6250 train_loss:2.1740 aux_loss:0.0291 train_time:2274936ms step_avg:537.81ms step:4241/6250 train_loss:2.2750 aux_loss:0.0292 train_time:2275466ms step_avg:537.81ms step:4242/6250 train_loss:2.1647 aux_loss:0.0287 train_time:2275999ms step_avg:537.81ms step:4243/6250 train_loss:2.1991 aux_loss:0.0294 train_time:2276540ms step_avg:537.81ms step:4244/6250 train_loss:2.1570 aux_loss:0.0294 train_time:2277119ms step_avg:537.82ms step:4245/6250 train_loss:2.2518 aux_loss:0.0293 train_time:2277667ms step_avg:537.82ms step:4246/6250 train_loss:2.2031 aux_loss:0.0295 train_time:2278224ms step_avg:537.82ms step:4247/6250 train_loss:2.2916 aux_loss:0.0294 train_time:2278780ms step_avg:537.83ms step:4248/6250 train_loss:2.2645 aux_loss:0.0292 train_time:2279362ms step_avg:537.84ms step:4249/6250 train_loss:2.2174 aux_loss:0.0298 train_time:2279916ms step_avg:537.84ms step:4250/6250 train_loss:2.1400 aux_loss:0.0297 train_time:2280467ms step_avg:537.85ms step:4251/6250 train_loss:2.2360 aux_loss:0.0291 train_time:2281019ms step_avg:537.85ms step:4252/6250 train_loss:2.2290 aux_loss:0.0297 train_time:2281576ms step_avg:537.85ms step:4253/6250 train_loss:2.2222 aux_loss:0.0297 train_time:2282128ms step_avg:537.86ms step:4254/6250 train_loss:2.2175 aux_loss:0.0296 train_time:2282677ms step_avg:537.86ms step:4255/6250 train_loss:2.1937 aux_loss:0.0301 train_time:2283206ms step_avg:537.86ms step:4256/6250 train_loss:2.2037 aux_loss:0.0301 train_time:2283735ms step_avg:537.86ms step:4257/6250 train_loss:2.1480 aux_loss:0.0290 train_time:2284294ms step_avg:537.86ms step:4258/6250 train_loss:2.2054 aux_loss:0.0289 train_time:2284835ms step_avg:537.86ms step:4259/6250 train_loss:2.2343 aux_loss:0.0295 train_time:2285380ms step_avg:537.86ms step:4260/6250 train_loss:2.2646 aux_loss:0.0301 train_time:2285927ms step_avg:537.87ms step:4261/6250 train_loss:2.1145 aux_loss:0.0295 train_time:2286472ms step_avg:537.87ms step:4262/6250 train_loss:2.1609 aux_loss:0.0291 train_time:2287020ms step_avg:537.87ms step:4263/6250 train_loss:2.0773 aux_loss:0.0292 train_time:2287567ms step_avg:537.87ms step:4264/6250 train_loss:2.2514 aux_loss:0.0296 train_time:2288116ms step_avg:537.87ms step:4265/6250 train_loss:2.1931 aux_loss:0.0292 train_time:2288657ms step_avg:537.87ms step:4266/6250 train_loss:2.1861 aux_loss:0.0290 train_time:2289186ms step_avg:537.87ms step:4267/6250 train_loss:2.2737 aux_loss:0.0291 train_time:2289715ms step_avg:537.87ms step:4268/6250 train_loss:2.2468 aux_loss:0.0294 train_time:2290243ms step_avg:537.87ms step:4269/6250 train_loss:2.2122 aux_loss:0.0294 train_time:2290773ms step_avg:537.87ms step:4270/6250 train_loss:2.2803 aux_loss:0.0291 train_time:2291301ms step_avg:537.86ms step:4271/6250 train_loss:2.1822 aux_loss:0.0289 train_time:2291831ms step_avg:537.86ms step:4272/6250 train_loss:2.2677 aux_loss:0.0292 train_time:2292360ms step_avg:537.86ms step:4273/6250 train_loss:2.1173 aux_loss:0.0290 train_time:2292888ms step_avg:537.86ms step:4274/6250 train_loss:2.2435 aux_loss:0.0289 train_time:2293417ms step_avg:537.86ms step:4275/6250 train_loss:2.1079 aux_loss:0.0288 train_time:2293948ms step_avg:537.85ms step:4276/6250 train_loss:2.2138 aux_loss:0.0292 train_time:2294476ms step_avg:537.85ms step:4277/6250 train_loss:2.1248 aux_loss:0.0296 train_time:2295007ms step_avg:537.85ms step:4278/6250 train_loss:2.2669 aux_loss:0.0296 train_time:2295537ms step_avg:537.85ms step:4279/6250 train_loss:2.2083 aux_loss:0.0285 train_time:2296067ms step_avg:537.85ms step:4280/6250 train_loss:2.1646 aux_loss:0.0283 train_time:2296596ms step_avg:537.84ms step:4281/6250 train_loss:2.1640 aux_loss:0.0290 train_time:2297142ms step_avg:537.85ms step:4282/6250 train_loss:2.1875 aux_loss:0.0291 train_time:2297670ms step_avg:537.84ms step:4283/6250 train_loss:2.2390 aux_loss:0.0291 train_time:2298198ms step_avg:537.84ms step:4284/6250 train_loss:2.1985 aux_loss:0.0288 train_time:2298740ms step_avg:537.84ms step:4285/6250 train_loss:2.1254 aux_loss:0.0289 train_time:2299291ms step_avg:537.85ms step:4286/6250 train_loss:2.1843 aux_loss:0.0294 train_time:2299824ms step_avg:537.84ms step:4287/6250 train_loss:2.1903 aux_loss:0.0288 train_time:2300352ms step_avg:537.84ms step:4288/6250 train_loss:2.2395 aux_loss:0.0291 train_time:2300879ms step_avg:537.84ms step:4289/6250 train_loss:2.1119 aux_loss:0.0290 train_time:2301410ms step_avg:537.84ms step:4290/6250 train_loss:2.1825 aux_loss:0.0288 train_time:2301939ms step_avg:537.84ms step:4291/6250 train_loss:2.2131 aux_loss:0.0289 train_time:2302486ms step_avg:537.84ms step:4292/6250 train_loss:2.1734 aux_loss:0.0287 train_time:2303015ms step_avg:537.84ms step:4293/6250 train_loss:2.1593 aux_loss:0.0287 train_time:2303541ms step_avg:537.83ms step:4294/6250 train_loss:2.1566 aux_loss:0.0292 train_time:2304071ms step_avg:537.83ms step:4295/6250 train_loss:2.1791 aux_loss:0.0289 train_time:2304601ms step_avg:537.83ms step:4296/6250 train_loss:2.2402 aux_loss:0.0285 train_time:2305129ms step_avg:537.83ms step:4297/6250 train_loss:2.1512 aux_loss:0.0290 train_time:2305659ms step_avg:537.83ms step:4298/6250 train_loss:2.2169 aux_loss:0.0294 train_time:2306187ms step_avg:537.82ms step:4299/6250 train_loss:2.1308 aux_loss:0.0290 train_time:2306716ms step_avg:537.82ms step:4300/6250 train_loss:2.2848 aux_loss:0.0289 train_time:2307244ms step_avg:537.82ms step:4301/6250 train_loss:2.1625 aux_loss:0.0295 train_time:2307774ms step_avg:537.82ms step:4302/6250 train_loss:2.1925 aux_loss:0.0296 train_time:2308303ms step_avg:537.82ms step:4303/6250 train_loss:2.2210 aux_loss:0.0294 train_time:2308832ms step_avg:537.81ms step:4304/6250 train_loss:2.1558 aux_loss:0.0287 train_time:2309361ms step_avg:537.81ms step:4305/6250 train_loss:2.1958 aux_loss:0.0289 train_time:2309909ms step_avg:537.81ms step:4306/6250 train_loss:2.1281 aux_loss:0.0290 train_time:2310448ms step_avg:537.81ms step:4307/6250 train_loss:2.2130 aux_loss:0.0290 train_time:2310978ms step_avg:537.81ms step:4308/6250 train_loss:2.2245 aux_loss:0.0294 train_time:2311514ms step_avg:537.81ms step:4309/6250 train_loss:2.2081 aux_loss:0.0296 train_time:2312057ms step_avg:537.81ms step:4310/6250 train_loss:2.2218 aux_loss:0.0291 train_time:2312590ms step_avg:537.81ms step:4311/6250 train_loss:2.1107 aux_loss:0.0289 train_time:2313120ms step_avg:537.81ms step:4312/6250 train_loss:2.1906 aux_loss:0.0293 train_time:2313648ms step_avg:537.81ms step:4313/6250 train_loss:2.2461 aux_loss:0.0291 train_time:2314178ms step_avg:537.81ms step:4314/6250 train_loss:2.2142 aux_loss:0.0292 train_time:2314708ms step_avg:537.80ms step:4315/6250 train_loss:2.3200 aux_loss:0.0295 train_time:2315235ms step_avg:537.80ms step:4316/6250 train_loss:2.1742 aux_loss:0.0294 train_time:2315765ms step_avg:537.80ms step:4317/6250 train_loss:2.2542 aux_loss:0.0291 train_time:2316294ms step_avg:537.80ms step:4318/6250 train_loss:2.0855 aux_loss:0.0287 train_time:2316823ms step_avg:537.80ms step:4319/6250 train_loss:2.2131 aux_loss:0.0288 train_time:2317351ms step_avg:537.79ms step:4320/6250 train_loss:2.2112 aux_loss:0.0290 train_time:2317881ms step_avg:537.79ms step:4321/6250 train_loss:2.1177 aux_loss:0.0293 train_time:2318417ms step_avg:537.79ms step:4322/6250 train_loss:2.2075 aux_loss:0.0296 train_time:2318969ms step_avg:537.79ms step:4323/6250 train_loss:2.1237 aux_loss:0.0288 train_time:2319521ms step_avg:537.80ms step:4324/6250 train_loss:2.2721 aux_loss:0.0289 train_time:2320062ms step_avg:537.80ms step:4325/6250 train_loss:2.2570 aux_loss:0.0294 train_time:2320597ms step_avg:537.80ms step:4326/6250 train_loss:2.2146 aux_loss:0.0295 train_time:2321140ms step_avg:537.80ms step:4327/6250 train_loss:2.2835 aux_loss:0.0292 train_time:2321712ms step_avg:537.81ms step:4328/6250 train_loss:2.2165 aux_loss:0.0289 train_time:2322249ms step_avg:537.81ms step:4329/6250 train_loss:2.1202 aux_loss:0.0290 train_time:2322803ms step_avg:537.81ms step:4330/6250 train_loss:2.1971 aux_loss:0.0294 train_time:2323371ms step_avg:537.82ms step:4331/6250 train_loss:2.2223 aux_loss:0.0291 train_time:2323916ms step_avg:537.82ms step:4332/6250 train_loss:2.1841 aux_loss:0.0290 train_time:2324445ms step_avg:537.82ms step:4333/6250 train_loss:2.1751 aux_loss:0.0289 train_time:2325000ms step_avg:537.82ms step:4334/6250 train_loss:2.0928 aux_loss:0.0288 train_time:2325550ms step_avg:537.82ms step:4335/6250 train_loss:2.2896 aux_loss:0.0295 train_time:2326080ms step_avg:537.82ms step:4336/6250 train_loss:2.1553 aux_loss:0.0290 train_time:2326616ms step_avg:537.82ms step:4337/6250 train_loss:2.1542 aux_loss:0.0287 train_time:2327152ms step_avg:537.82ms step:4338/6250 train_loss:2.2685 aux_loss:0.0299 train_time:2327681ms step_avg:537.82ms step:4339/6250 train_loss:2.2755 aux_loss:0.0290 train_time:2328209ms step_avg:537.82ms step:4340/6250 train_loss:2.1946 aux_loss:0.0282 train_time:2328740ms step_avg:537.82ms step:4341/6250 train_loss:2.1621 aux_loss:0.0290 train_time:2329269ms step_avg:537.81ms step:4342/6250 train_loss:2.1965 aux_loss:0.0291 train_time:2329797ms step_avg:537.81ms step:4343/6250 train_loss:2.2355 aux_loss:0.0288 train_time:2330326ms step_avg:537.81ms step:4344/6250 train_loss:2.1994 aux_loss:0.0286 train_time:2330855ms step_avg:537.81ms step:4345/6250 train_loss:2.1959 aux_loss:0.0287 train_time:2331385ms step_avg:537.81ms step:4346/6250 train_loss:2.1895 aux_loss:0.0288 train_time:2331920ms step_avg:537.80ms step:4347/6250 train_loss:2.2424 aux_loss:0.0291 train_time:2332453ms step_avg:537.80ms step:4348/6250 train_loss:2.1405 aux_loss:0.0287 train_time:2332984ms step_avg:537.80ms step:4349/6250 train_loss:2.1993 aux_loss:0.0288 train_time:2333512ms step_avg:537.80ms step:4350/6250 train_loss:2.2956 aux_loss:0.0292 train_time:2334041ms step_avg:537.80ms step:4351/6250 train_loss:2.1862 aux_loss:0.0290 train_time:2334573ms step_avg:537.80ms step:4352/6250 train_loss:2.1744 aux_loss:0.0287 train_time:2335103ms step_avg:537.79ms step:4353/6250 train_loss:2.1512 aux_loss:0.0290 train_time:2335631ms step_avg:537.79ms step:4354/6250 train_loss:2.1005 aux_loss:0.0288 train_time:2336158ms step_avg:537.79ms step:4355/6250 train_loss:2.2098 aux_loss:0.0291 train_time:2336689ms step_avg:537.79ms step:4356/6250 train_loss:2.0632 aux_loss:0.0292 train_time:2337220ms step_avg:537.79ms step:4357/6250 train_loss:2.1535 aux_loss:0.0288 train_time:2337751ms step_avg:537.78ms step:4358/6250 train_loss:2.2742 aux_loss:0.0291 train_time:2338280ms step_avg:537.78ms step:4359/6250 train_loss:2.2544 aux_loss:0.0288 train_time:2338815ms step_avg:537.78ms step:4360/6250 train_loss:2.1741 aux_loss:0.0289 train_time:2339368ms step_avg:537.79ms step:4361/6250 train_loss:2.2399 aux_loss:0.0291 train_time:2339908ms step_avg:537.79ms step:4362/6250 train_loss:2.1531 aux_loss:0.0292 train_time:2340438ms step_avg:537.78ms step:4363/6250 train_loss:2.2061 aux_loss:0.0289 train_time:2340975ms step_avg:537.78ms step:4364/6250 train_loss:2.2486 aux_loss:0.0288 train_time:2341525ms step_avg:537.79ms step:4365/6250 train_loss:2.2036 aux_loss:0.0287 train_time:2342080ms step_avg:537.79ms step:4366/6250 train_loss:2.2412 aux_loss:0.0287 train_time:2342622ms step_avg:537.79ms step:4367/6250 train_loss:2.0956 aux_loss:0.0289 train_time:2343162ms step_avg:537.79ms step:4368/6250 train_loss:2.2121 aux_loss:0.0289 train_time:2343715ms step_avg:537.80ms step:4369/6250 train_loss:2.0777 aux_loss:0.0284 train_time:2344250ms step_avg:537.80ms step:4370/6250 train_loss:2.1626 aux_loss:0.0284 train_time:2344783ms step_avg:537.79ms step:4371/6250 train_loss:2.2568 aux_loss:0.0288 train_time:2345320ms step_avg:537.79ms step:4372/6250 train_loss:2.1960 aux_loss:0.0283 train_time:2345865ms step_avg:537.80ms step:4373/6250 train_loss:2.2390 aux_loss:0.0290 train_time:2346394ms step_avg:537.79ms step:4374/6250 train_loss:2.1577 aux_loss:0.0291 train_time:2346925ms step_avg:537.79ms step:4375/6250 train_loss:2.1321 aux_loss:0.0286 train_time:2347462ms step_avg:537.79ms step:4376/6250 train_loss:2.1261 aux_loss:0.0292 train_time:2347992ms step_avg:537.79ms step:4377/6250 train_loss:2.1935 aux_loss:0.0289 train_time:2348521ms step_avg:537.79ms step:4378/6250 train_loss:2.1681 aux_loss:0.0284 train_time:2349051ms step_avg:537.79ms step:4379/6250 train_loss:2.1832 aux_loss:0.0288 train_time:2349580ms step_avg:537.78ms step:4380/6250 train_loss:2.1487 aux_loss:0.0284 train_time:2350109ms step_avg:537.78ms step:4381/6250 train_loss:2.1734 aux_loss:0.0285 train_time:2350640ms step_avg:537.78ms step:4382/6250 train_loss:2.1482 aux_loss:0.0289 train_time:2351168ms step_avg:537.78ms step:4383/6250 train_loss:2.0609 aux_loss:0.0288 train_time:2351697ms step_avg:537.78ms step:4384/6250 train_loss:2.1733 aux_loss:0.0286 train_time:2352225ms step_avg:537.77ms step:4385/6250 train_loss:2.1795 aux_loss:0.0291 train_time:2352756ms step_avg:537.77ms step:4386/6250 train_loss:2.1592 aux_loss:0.0286 train_time:2353285ms step_avg:537.77ms step:4387/6250 train_loss:2.0659 aux_loss:0.0283 train_time:2353817ms step_avg:537.77ms step:4388/6250 train_loss:2.2446 aux_loss:0.0293 train_time:2354346ms step_avg:537.77ms step:4389/6250 train_loss:2.1790 aux_loss:0.0288 train_time:2354877ms step_avg:537.77ms step:4390/6250 train_loss:2.1821 aux_loss:0.0287 train_time:2355404ms step_avg:537.76ms step:4391/6250 train_loss:2.2246 aux_loss:0.0292 train_time:2355934ms step_avg:537.76ms step:4392/6250 train_loss:2.2154 aux_loss:0.0290 train_time:2356461ms step_avg:537.76ms step:4393/6250 train_loss:2.1860 aux_loss:0.0287 train_time:2356990ms step_avg:537.76ms step:4394/6250 train_loss:2.2853 aux_loss:0.0291 train_time:2357520ms step_avg:537.76ms step:4395/6250 train_loss:2.2068 aux_loss:0.0292 train_time:2358050ms step_avg:537.75ms step:4396/6250 train_loss:2.2076 aux_loss:0.0289 train_time:2358580ms step_avg:537.75ms step:4397/6250 train_loss:2.1701 aux_loss:0.0284 train_time:2359123ms step_avg:537.75ms step:4398/6250 train_loss:2.2247 aux_loss:0.0291 train_time:2359674ms step_avg:537.76ms step:4399/6250 train_loss:2.1605 aux_loss:0.0288 train_time:2360209ms step_avg:537.76ms step:4400/6250 train_loss:2.2241 aux_loss:0.0285 train_time:2360738ms step_avg:537.75ms step:4401/6250 train_loss:2.2777 aux_loss:0.0291 train_time:2361266ms step_avg:537.75ms step:4402/6250 train_loss:2.2185 aux_loss:0.0291 train_time:2361796ms step_avg:537.75ms step:4403/6250 train_loss:2.1656 aux_loss:0.0288 train_time:2362326ms step_avg:537.75ms step:4404/6250 train_loss:2.1542 aux_loss:0.0284 train_time:2362856ms step_avg:537.75ms step:4405/6250 train_loss:2.2562 aux_loss:0.0288 train_time:2363384ms step_avg:537.74ms step:4406/6250 train_loss:2.3154 aux_loss:0.0290 train_time:2363914ms step_avg:537.74ms step:4407/6250 train_loss:2.2146 aux_loss:0.0286 train_time:2364441ms step_avg:537.74ms step:4408/6250 train_loss:2.1820 aux_loss:0.0289 train_time:2364970ms step_avg:537.74ms step:4409/6250 train_loss:2.1714 aux_loss:0.0289 train_time:2365500ms step_avg:537.74ms step:4410/6250 train_loss:2.2374 aux_loss:0.0287 train_time:2366032ms step_avg:537.73ms step:4411/6250 train_loss:2.2231 aux_loss:0.0286 train_time:2366562ms step_avg:537.73ms step:4412/6250 train_loss:2.2550 aux_loss:0.0288 train_time:2367090ms step_avg:537.73ms step:4413/6250 train_loss:2.0967 aux_loss:0.0288 train_time:2367619ms step_avg:537.73ms step:4414/6250 train_loss:2.1695 aux_loss:0.0282 train_time:2368148ms step_avg:537.73ms step:4415/6250 train_loss:2.2428 aux_loss:0.0288 train_time:2368678ms step_avg:537.72ms step:4416/6250 train_loss:2.1614 aux_loss:0.0290 train_time:2369209ms step_avg:537.72ms step:4417/6250 train_loss:2.2201 aux_loss:0.0288 train_time:2369738ms step_avg:537.72ms step:4418/6250 train_loss:2.1232 aux_loss:0.0285 train_time:2370267ms step_avg:537.72ms step:4419/6250 train_loss:2.1942 aux_loss:0.0290 train_time:2370799ms step_avg:537.72ms step:4420/6250 train_loss:2.3010 aux_loss:0.0288 train_time:2371327ms step_avg:537.72ms step:4421/6250 train_loss:2.3161 aux_loss:0.0290 train_time:2371856ms step_avg:537.71ms step:4422/6250 train_loss:2.2713 aux_loss:0.0290 train_time:2372384ms step_avg:537.71ms step:4423/6250 train_loss:2.2820 aux_loss:0.0290 train_time:2372914ms step_avg:537.71ms step:4424/6250 train_loss:2.1661 aux_loss:0.0284 train_time:2373443ms step_avg:537.71ms step:4425/6250 train_loss:2.1658 aux_loss:0.0288 train_time:2373973ms step_avg:537.71ms step:4426/6250 train_loss:2.0596 aux_loss:0.0282 train_time:2374502ms step_avg:537.70ms step:4427/6250 train_loss:2.2110 aux_loss:0.0287 train_time:2375033ms step_avg:537.70ms step:4428/6250 train_loss:2.1509 aux_loss:0.0286 train_time:2375562ms step_avg:537.70ms step:4429/6250 train_loss:2.2988 aux_loss:0.0284 train_time:2376092ms step_avg:537.70ms step:4430/6250 train_loss:2.1673 aux_loss:0.0286 train_time:2376621ms step_avg:537.70ms step:4431/6250 train_loss:2.1482 aux_loss:0.0292 train_time:2377150ms step_avg:537.70ms step:4432/6250 train_loss:2.3209 aux_loss:0.0287 train_time:2377679ms step_avg:537.69ms step:4433/6250 train_loss:2.1450 aux_loss:0.0278 train_time:2378208ms step_avg:537.69ms step:4434/6250 train_loss:2.2399 aux_loss:0.0289 train_time:2378737ms step_avg:537.69ms step:4435/6250 train_loss:2.2270 aux_loss:0.0291 train_time:2379282ms step_avg:537.69ms step:4436/6250 train_loss:2.1882 aux_loss:0.0289 train_time:2379832ms step_avg:537.69ms step:4437/6250 train_loss:2.1230 aux_loss:0.0288 train_time:2380362ms step_avg:537.69ms step:4438/6250 train_loss:2.1550 aux_loss:0.0289 train_time:2380892ms step_avg:537.69ms step:4439/6250 train_loss:2.1040 aux_loss:0.0286 train_time:2381420ms step_avg:537.69ms step:4440/6250 train_loss:2.2850 aux_loss:0.0286 train_time:2381949ms step_avg:537.69ms step:4441/6250 train_loss:2.1354 aux_loss:0.0288 train_time:2382477ms step_avg:537.68ms step:4442/6250 train_loss:2.1122 aux_loss:0.0285 train_time:2383006ms step_avg:537.68ms step:4443/6250 train_loss:2.2428 aux_loss:0.0284 train_time:2383535ms step_avg:537.68ms step:4444/6250 train_loss:2.2555 aux_loss:0.0284 train_time:2384063ms step_avg:537.68ms step:4445/6250 train_loss:2.1697 aux_loss:0.0291 train_time:2384593ms step_avg:537.68ms step:4446/6250 train_loss:2.1441 aux_loss:0.0286 train_time:2385122ms step_avg:537.67ms step:4447/6250 train_loss:2.2871 aux_loss:0.0286 train_time:2385652ms step_avg:537.67ms step:4448/6250 train_loss:2.1090 aux_loss:0.0285 train_time:2386181ms step_avg:537.67ms step:4449/6250 train_loss:2.0859 aux_loss:0.0286 train_time:2386710ms step_avg:537.67ms step:4450/6250 train_loss:2.1059 aux_loss:0.0286 train_time:2387238ms step_avg:537.67ms step:4451/6250 train_loss:2.1308 aux_loss:0.0281 train_time:2387767ms step_avg:537.66ms step:4452/6250 train_loss:2.3962 aux_loss:0.0285 train_time:2388295ms step_avg:537.66ms step:4453/6250 train_loss:2.3285 aux_loss:0.0286 train_time:2388825ms step_avg:537.66ms step:4454/6250 train_loss:2.2367 aux_loss:0.0289 train_time:2389355ms step_avg:537.66ms step:4455/6250 train_loss:2.1964 aux_loss:0.0285 train_time:2389885ms step_avg:537.66ms step:4456/6250 train_loss:2.2435 aux_loss:0.0283 train_time:2390414ms step_avg:537.65ms step:4457/6250 train_loss:2.1790 aux_loss:0.0285 train_time:2390942ms step_avg:537.65ms step:4458/6250 train_loss:2.2782 aux_loss:0.0294 train_time:2391471ms step_avg:537.65ms step:4459/6250 train_loss:2.1914 aux_loss:0.0286 train_time:2391999ms step_avg:537.65ms step:4460/6250 train_loss:2.1823 aux_loss:0.0282 train_time:2392529ms step_avg:537.65ms step:4461/6250 train_loss:2.0998 aux_loss:0.0286 train_time:2393058ms step_avg:537.65ms step:4462/6250 train_loss:2.2571 aux_loss:0.0288 train_time:2393587ms step_avg:537.64ms step:4463/6250 train_loss:2.2033 aux_loss:0.0286 train_time:2394116ms step_avg:537.64ms step:4464/6250 train_loss:2.1497 aux_loss:0.0286 train_time:2394645ms step_avg:537.64ms step:4465/6250 train_loss:2.3226 aux_loss:0.0286 train_time:2395176ms step_avg:537.64ms step:4466/6250 train_loss:2.2578 aux_loss:0.0286 train_time:2395707ms step_avg:537.64ms step:4467/6250 train_loss:2.1511 aux_loss:0.0284 train_time:2396236ms step_avg:537.63ms step:4468/6250 train_loss:2.1801 aux_loss:0.0289 train_time:2396766ms step_avg:537.63ms step:4469/6250 train_loss:2.1910 aux_loss:0.0290 train_time:2397298ms step_avg:537.63ms step:4470/6250 train_loss:2.1830 aux_loss:0.0285 train_time:2397829ms step_avg:537.63ms step:4471/6250 train_loss:2.1471 aux_loss:0.0286 train_time:2398357ms step_avg:537.63ms step:4472/6250 train_loss:2.2205 aux_loss:0.0288 train_time:2398887ms step_avg:537.63ms step:4473/6250 train_loss:2.2246 aux_loss:0.0287 train_time:2399433ms step_avg:537.63ms step:4474/6250 train_loss:2.2324 aux_loss:0.0287 train_time:2399982ms step_avg:537.63ms step:4475/6250 train_loss:2.1546 aux_loss:0.0288 train_time:2400512ms step_avg:537.63ms step:4476/6250 train_loss:2.2330 aux_loss:0.0293 train_time:2401041ms step_avg:537.63ms step:4477/6250 train_loss:2.1953 aux_loss:0.0286 train_time:2401567ms step_avg:537.62ms step:4478/6250 train_loss:2.2909 aux_loss:0.0285 train_time:2402096ms step_avg:537.62ms step:4479/6250 train_loss:2.1587 aux_loss:0.0284 train_time:2402626ms step_avg:537.62ms step:4480/6250 train_loss:2.1735 aux_loss:0.0287 train_time:2403155ms step_avg:537.62ms step:4481/6250 train_loss:2.1880 aux_loss:0.0285 train_time:2403684ms step_avg:537.62ms step:4482/6250 train_loss:2.1959 aux_loss:0.0287 train_time:2404213ms step_avg:537.61ms step:4483/6250 train_loss:2.2153 aux_loss:0.0287 train_time:2404743ms step_avg:537.61ms step:4484/6250 train_loss:2.2286 aux_loss:0.0282 train_time:2405272ms step_avg:537.61ms step:4485/6250 train_loss:2.1666 aux_loss:0.0285 train_time:2405802ms step_avg:537.61ms step:4486/6250 train_loss:2.1132 aux_loss:0.0288 train_time:2406332ms step_avg:537.61ms step:4487/6250 train_loss:2.3026 aux_loss:0.0289 train_time:2406862ms step_avg:537.61ms step:4488/6250 train_loss:2.2623 aux_loss:0.0285 train_time:2407390ms step_avg:537.60ms step:4489/6250 train_loss:2.2364 aux_loss:0.0288 train_time:2407919ms step_avg:537.60ms step:4490/6250 train_loss:2.1707 aux_loss:0.0289 train_time:2408448ms step_avg:537.60ms step:4491/6250 train_loss:2.1298 aux_loss:0.0285 train_time:2408979ms step_avg:537.60ms step:4492/6250 train_loss:2.2321 aux_loss:0.0291 train_time:2409507ms step_avg:537.60ms step:4493/6250 train_loss:2.2204 aux_loss:0.0289 train_time:2410037ms step_avg:537.59ms step:4494/6250 train_loss:2.2465 aux_loss:0.0284 train_time:2410565ms step_avg:537.59ms step:4495/6250 train_loss:2.2216 aux_loss:0.0283 train_time:2411095ms step_avg:537.59ms step:4496/6250 train_loss:2.2485 aux_loss:0.0289 train_time:2411624ms step_avg:537.59ms step:4497/6250 train_loss:2.2465 aux_loss:0.0294 train_time:2412153ms step_avg:537.59ms step:4498/6250 train_loss:2.2121 aux_loss:0.0288 train_time:2412682ms step_avg:537.59ms step:4499/6250 train_loss:2.2351 aux_loss:0.0285 train_time:2413211ms step_avg:537.58ms step:4500/6250 train_loss:2.1812 aux_loss:0.0287 train_time:2413741ms step_avg:537.58ms step:4501/6250 train_loss:2.1810 aux_loss:0.0289 train_time:2414271ms step_avg:537.58ms step:4502/6250 train_loss:2.1051 aux_loss:0.0285 train_time:2414801ms step_avg:537.58ms step:4503/6250 train_loss:2.0912 aux_loss:0.0286 train_time:2415331ms step_avg:537.58ms step:4504/6250 train_loss:2.1886 aux_loss:0.0285 train_time:2415860ms step_avg:537.57ms step:4505/6250 train_loss:2.2151 aux_loss:0.0286 train_time:2416389ms step_avg:537.57ms step:4506/6250 train_loss:2.1226 aux_loss:0.0284 train_time:2416920ms step_avg:537.57ms step:4507/6250 train_loss:2.1899 aux_loss:0.0288 train_time:2417447ms step_avg:537.57ms step:4508/6250 train_loss:2.1459 aux_loss:0.0281 train_time:2417976ms step_avg:537.57ms step:4509/6250 train_loss:2.1481 aux_loss:0.0286 train_time:2418504ms step_avg:537.56ms step:4510/6250 train_loss:2.1569 aux_loss:0.0285 train_time:2419033ms step_avg:537.56ms step:4511/6250 train_loss:2.2133 aux_loss:0.0284 train_time:2419584ms step_avg:537.57ms step:4512/6250 train_loss:2.1533 aux_loss:0.0292 train_time:2420132ms step_avg:537.57ms step:4513/6250 train_loss:2.2775 aux_loss:0.0288 train_time:2420662ms step_avg:537.57ms step:4514/6250 train_loss:2.1128 aux_loss:0.0283 train_time:2421192ms step_avg:537.56ms step:4515/6250 train_loss:2.2546 aux_loss:0.0290 train_time:2421722ms step_avg:537.56ms step:4516/6250 train_loss:2.2172 aux_loss:0.0292 train_time:2422248ms step_avg:537.56ms step:4517/6250 train_loss:2.2436 aux_loss:0.0286 train_time:2422780ms step_avg:537.56ms step:4518/6250 train_loss:2.1374 aux_loss:0.0282 train_time:2423308ms step_avg:537.56ms step:4519/6250 train_loss:2.1140 aux_loss:0.0287 train_time:2423839ms step_avg:537.56ms step:4520/6250 train_loss:2.2054 aux_loss:0.0285 train_time:2424368ms step_avg:537.55ms step:4521/6250 train_loss:2.2492 aux_loss:0.0281 train_time:2424898ms step_avg:537.55ms step:4522/6250 train_loss:2.2466 aux_loss:0.0288 train_time:2425428ms step_avg:537.55ms step:4523/6250 train_loss:2.2988 aux_loss:0.0285 train_time:2425957ms step_avg:537.55ms step:4524/6250 train_loss:2.3146 aux_loss:0.0285 train_time:2426486ms step_avg:537.55ms step:4525/6250 train_loss:2.1986 aux_loss:0.0286 train_time:2427016ms step_avg:537.54ms step:4526/6250 train_loss:2.2025 aux_loss:0.0288 train_time:2427544ms step_avg:537.54ms step:4527/6250 train_loss:2.2585 aux_loss:0.0291 train_time:2428074ms step_avg:537.54ms step:4528/6250 train_loss:2.1857 aux_loss:0.0285 train_time:2428602ms step_avg:537.54ms step:4529/6250 train_loss:2.1952 aux_loss:0.0286 train_time:2429132ms step_avg:537.54ms step:4530/6250 train_loss:2.1490 aux_loss:0.0287 train_time:2429662ms step_avg:537.54ms step:4531/6250 train_loss:2.1783 aux_loss:0.0287 train_time:2430191ms step_avg:537.53ms step:4532/6250 train_loss:2.1939 aux_loss:0.0286 train_time:2430722ms step_avg:537.53ms step:4533/6250 train_loss:2.1574 aux_loss:0.0286 train_time:2431251ms step_avg:537.53ms step:4534/6250 train_loss:2.1646 aux_loss:0.0290 train_time:2431781ms step_avg:537.53ms step:4535/6250 train_loss:2.1588 aux_loss:0.0287 train_time:2432310ms step_avg:537.53ms step:4536/6250 train_loss:2.1833 aux_loss:0.0286 train_time:2432839ms step_avg:537.53ms step:4537/6250 train_loss:2.1815 aux_loss:0.0285 train_time:2433368ms step_avg:537.52ms step:4538/6250 train_loss:2.2385 aux_loss:0.0286 train_time:2433897ms step_avg:537.52ms step:4539/6250 train_loss:2.2423 aux_loss:0.0282 train_time:2434426ms step_avg:537.52ms step:4540/6250 train_loss:2.2233 aux_loss:0.0287 train_time:2434955ms step_avg:537.52ms step:4541/6250 train_loss:2.1548 aux_loss:0.0285 train_time:2435482ms step_avg:537.52ms step:4542/6250 train_loss:2.2699 aux_loss:0.0280 train_time:2436014ms step_avg:537.51ms step:4543/6250 train_loss:2.2469 aux_loss:0.0283 train_time:2436544ms step_avg:537.51ms step:4544/6250 train_loss:2.2304 aux_loss:0.0286 train_time:2437073ms step_avg:537.51ms step:4545/6250 train_loss:2.2568 aux_loss:0.0286 train_time:2437604ms step_avg:537.51ms step:4546/6250 train_loss:2.2045 aux_loss:0.0283 train_time:2438135ms step_avg:537.51ms step:4547/6250 train_loss:2.2400 aux_loss:0.0284 train_time:2438671ms step_avg:537.51ms step:4548/6250 train_loss:2.1686 aux_loss:0.0284 train_time:2439223ms step_avg:537.51ms step:4549/6250 train_loss:2.2495 aux_loss:0.0284 train_time:2439828ms step_avg:537.53ms step:4550/6250 train_loss:2.2004 aux_loss:0.0284 train_time:2440392ms step_avg:537.53ms step:4551/6250 train_loss:2.2244 aux_loss:0.0289 train_time:2440946ms step_avg:537.53ms step:4552/6250 train_loss:2.2137 aux_loss:0.0282 train_time:2441489ms step_avg:537.54ms step:4553/6250 train_loss:2.1068 aux_loss:0.0282 train_time:2442042ms step_avg:537.54ms step:4554/6250 train_loss:2.1349 aux_loss:0.0282 train_time:2442579ms step_avg:537.54ms step:4555/6250 train_loss:2.2354 aux_loss:0.0284 train_time:2443112ms step_avg:537.54ms step:4556/6250 train_loss:2.2533 aux_loss:0.0286 train_time:2443664ms step_avg:537.54ms step:4557/6250 train_loss:2.1001 aux_loss:0.0287 train_time:2444223ms step_avg:537.55ms step:4558/6250 train_loss:2.0905 aux_loss:0.0282 train_time:2444759ms step_avg:537.55ms step:4559/6250 train_loss:2.2549 aux_loss:0.0283 train_time:2445298ms step_avg:537.55ms step:4560/6250 train_loss:2.2352 aux_loss:0.0283 train_time:2445838ms step_avg:537.55ms step:4561/6250 train_loss:2.1761 aux_loss:0.0282 train_time:2446378ms step_avg:537.55ms step:4562/6250 train_loss:2.1806 aux_loss:0.0281 train_time:2446908ms step_avg:537.55ms step:4563/6250 train_loss:2.1537 aux_loss:0.0281 train_time:2447437ms step_avg:537.54ms step:4564/6250 train_loss:2.2415 aux_loss:0.0289 train_time:2447990ms step_avg:537.55ms step:4565/6250 train_loss:2.2123 aux_loss:0.0284 train_time:2448547ms step_avg:537.55ms step:4566/6250 train_loss:2.2527 aux_loss:0.0282 train_time:2449080ms step_avg:537.55ms step:4567/6250 train_loss:2.2722 aux_loss:0.0279 train_time:2449611ms step_avg:537.55ms step:4568/6250 train_loss:2.1358 aux_loss:0.0277 train_time:2450161ms step_avg:537.55ms step:4569/6250 train_loss:2.2382 aux_loss:0.0284 train_time:2450728ms step_avg:537.56ms step:4570/6250 train_loss:2.2001 aux_loss:0.0285 train_time:2451278ms step_avg:537.56ms step:4571/6250 train_loss:2.1913 aux_loss:0.0282 train_time:2451815ms step_avg:537.56ms step:4572/6250 train_loss:2.1249 aux_loss:0.0278 train_time:2452346ms step_avg:537.56ms step:4573/6250 train_loss:2.2152 aux_loss:0.0285 train_time:2452881ms step_avg:537.56ms step:4574/6250 train_loss:2.2058 aux_loss:0.0282 train_time:2453412ms step_avg:537.56ms step:4575/6250 train_loss:2.1138 aux_loss:0.0281 train_time:2453954ms step_avg:537.56ms step:4576/6250 train_loss:2.2566 aux_loss:0.0283 train_time:2454511ms step_avg:537.56ms step:4577/6250 train_loss:2.1643 aux_loss:0.0284 train_time:2455054ms step_avg:537.56ms step:4578/6250 train_loss:2.1320 aux_loss:0.0283 train_time:2455728ms step_avg:537.59ms step:4579/6250 train_loss:2.1260 aux_loss:0.0282 train_time:2456287ms step_avg:537.60ms step:4580/6250 train_loss:2.1987 aux_loss:0.0281 train_time:2456858ms step_avg:537.61ms step:4581/6250 train_loss:2.0744 aux_loss:0.0283 train_time:2457394ms step_avg:537.61ms step:4582/6250 train_loss:2.2307 aux_loss:0.0286 train_time:2457921ms step_avg:537.60ms step:4583/6250 train_loss:2.2967 aux_loss:0.0282 train_time:2458450ms step_avg:537.60ms step:4584/6250 train_loss:2.1254 aux_loss:0.0284 train_time:2458979ms step_avg:537.60ms step:4585/6250 train_loss:2.1779 aux_loss:0.0283 train_time:2459519ms step_avg:537.60ms step:4586/6250 train_loss:2.1924 aux_loss:0.0281 train_time:2460069ms step_avg:537.60ms step:4587/6250 train_loss:2.1777 aux_loss:0.0282 train_time:2460632ms step_avg:537.61ms step:4588/6250 train_loss:2.2275 aux_loss:0.0287 train_time:2461182ms step_avg:537.61ms step:4589/6250 train_loss:2.2220 aux_loss:0.0284 train_time:2461723ms step_avg:537.61ms step:4590/6250 train_loss:2.2180 aux_loss:0.0282 train_time:2462254ms step_avg:537.61ms step:4591/6250 train_loss:2.1465 aux_loss:0.0282 train_time:2462791ms step_avg:537.61ms step:4592/6250 train_loss:2.1774 aux_loss:0.0284 train_time:2463319ms step_avg:537.61ms step:4593/6250 train_loss:2.1855 aux_loss:0.0277 train_time:2463856ms step_avg:537.61ms step:4594/6250 train_loss:2.1244 aux_loss:0.0282 train_time:2464398ms step_avg:537.61ms step:4595/6250 train_loss:2.1990 aux_loss:0.0287 train_time:2464930ms step_avg:537.61ms step:4596/6250 train_loss:2.1012 aux_loss:0.0283 train_time:2465474ms step_avg:537.61ms step:4597/6250 train_loss:2.1439 aux_loss:0.0282 train_time:2466022ms step_avg:537.61ms step:4598/6250 train_loss:2.2531 aux_loss:0.0286 train_time:2466571ms step_avg:537.61ms step:4599/6250 train_loss:2.1587 aux_loss:0.0285 train_time:2467117ms step_avg:537.62ms step:4600/6250 train_loss:2.2495 aux_loss:0.0282 train_time:2467654ms step_avg:537.62ms step:4601/6250 train_loss:2.2355 aux_loss:0.0284 train_time:2468184ms step_avg:537.61ms step:4602/6250 train_loss:2.2295 aux_loss:0.0278 train_time:2468714ms step_avg:537.61ms step:4603/6250 train_loss:2.1950 aux_loss:0.0279 train_time:2469241ms step_avg:537.61ms step:4604/6250 train_loss:2.1708 aux_loss:0.0283 train_time:2469771ms step_avg:537.61ms step:4605/6250 train_loss:2.2648 aux_loss:0.0282 train_time:2470319ms step_avg:537.61ms step:4606/6250 train_loss:2.2663 aux_loss:0.0278 train_time:2470857ms step_avg:537.61ms step:4607/6250 train_loss:2.1451 aux_loss:0.0285 train_time:2471386ms step_avg:537.61ms step:4608/6250 train_loss:2.1535 aux_loss:0.0281 train_time:2471915ms step_avg:537.61ms step:4609/6250 train_loss:2.1684 aux_loss:0.0281 train_time:2472445ms step_avg:537.60ms step:4610/6250 train_loss:2.2097 aux_loss:0.0286 train_time:2472988ms step_avg:537.61ms step:4611/6250 train_loss:2.1764 aux_loss:0.0283 train_time:2473526ms step_avg:537.61ms step:4612/6250 train_loss:2.2205 aux_loss:0.0278 train_time:2474075ms step_avg:537.61ms step:4613/6250 train_loss:2.2052 aux_loss:0.0285 train_time:2474621ms step_avg:537.61ms step:4614/6250 train_loss:2.1358 aux_loss:0.0283 train_time:2475157ms step_avg:537.61ms step:4615/6250 train_loss:2.3379 aux_loss:0.0280 train_time:2475690ms step_avg:537.61ms step:4616/6250 train_loss:2.2161 aux_loss:0.0281 train_time:2476230ms step_avg:537.61ms step:4617/6250 train_loss:2.1498 aux_loss:0.0286 train_time:2476782ms step_avg:537.61ms step:4618/6250 train_loss:2.1907 aux_loss:0.0283 train_time:2477312ms step_avg:537.61ms step:4619/6250 train_loss:2.1709 aux_loss:0.0281 train_time:2477842ms step_avg:537.61ms step:4620/6250 train_loss:2.2640 aux_loss:0.0284 train_time:2478383ms step_avg:537.61ms step:4621/6250 train_loss:2.1947 aux_loss:0.0281 train_time:2478922ms step_avg:537.61ms step:4622/6250 train_loss:2.2021 aux_loss:0.0280 train_time:2479459ms step_avg:537.61ms step:4623/6250 train_loss:2.1724 aux_loss:0.0286 train_time:2480008ms step_avg:537.61ms step:4624/6250 train_loss:2.1797 aux_loss:0.0285 train_time:2480550ms step_avg:537.61ms step:4625/6250 train_loss:2.1161 aux_loss:0.0278 train_time:2481079ms step_avg:537.61ms step:4626/6250 train_loss:2.1608 aux_loss:0.0282 train_time:2481608ms step_avg:537.61ms step:4627/6250 train_loss:2.1708 aux_loss:0.0281 train_time:2482138ms step_avg:537.61ms step:4628/6250 train_loss:2.2085 aux_loss:0.0277 train_time:2482666ms step_avg:537.61ms step:4629/6250 train_loss:2.2114 aux_loss:0.0282 train_time:2483199ms step_avg:537.61ms step:4630/6250 train_loss:2.2928 aux_loss:0.0283 train_time:2483742ms step_avg:537.61ms step:4631/6250 train_loss:2.1308 aux_loss:0.0280 train_time:2484284ms step_avg:537.61ms step:4632/6250 train_loss:2.1889 aux_loss:0.0279 train_time:2484828ms step_avg:537.61ms step:4633/6250 train_loss:2.0941 aux_loss:0.0280 train_time:2485364ms step_avg:537.61ms step:4634/6250 train_loss:2.1355 aux_loss:0.0284 train_time:2485896ms step_avg:537.61ms step:4635/6250 train_loss:2.2504 aux_loss:0.0285 train_time:2486429ms step_avg:537.61ms step:4636/6250 train_loss:2.1541 aux_loss:0.0283 train_time:2486972ms step_avg:537.61ms step:4637/6250 train_loss:2.1356 aux_loss:0.0282 train_time:2487511ms step_avg:537.61ms step:4638/6250 train_loss:2.1359 aux_loss:0.0283 train_time:2488053ms step_avg:537.61ms step:4639/6250 train_loss:2.1848 aux_loss:0.0278 train_time:2488589ms step_avg:537.61ms step:4640/6250 train_loss:2.1774 aux_loss:0.0276 train_time:2489126ms step_avg:537.61ms step:4641/6250 train_loss:2.0122 aux_loss:0.0277 train_time:2489663ms step_avg:537.61ms step:4642/6250 train_loss:2.3000 aux_loss:0.0280 train_time:2490201ms step_avg:537.61ms step:4643/6250 train_loss:2.1129 aux_loss:0.0278 train_time:2490742ms step_avg:537.61ms step:4644/6250 train_loss:2.1172 aux_loss:0.0280 train_time:2491302ms step_avg:537.61ms step:4645/6250 train_loss:2.2434 aux_loss:0.0282 train_time:2491840ms step_avg:537.61ms step:4646/6250 train_loss:2.1252 aux_loss:0.0281 train_time:2492382ms step_avg:537.61ms step:4647/6250 train_loss:2.0988 aux_loss:0.0279 train_time:2492933ms step_avg:537.62ms step:4648/6250 train_loss:2.2081 aux_loss:0.0282 train_time:2493490ms step_avg:537.62ms step:4649/6250 train_loss:2.1815 aux_loss:0.0279 train_time:2494040ms step_avg:537.62ms step:4650/6250 train_loss:2.1242 aux_loss:0.0280 train_time:2494598ms step_avg:537.63ms step:4651/6250 train_loss:2.1194 aux_loss:0.0278 train_time:2495154ms step_avg:537.63ms step:4652/6250 train_loss:2.2588 aux_loss:0.0280 train_time:2495715ms step_avg:537.64ms step:4653/6250 train_loss:2.2212 aux_loss:0.0282 train_time:2496247ms step_avg:537.64ms step:4654/6250 train_loss:2.2167 aux_loss:0.0283 train_time:2496799ms step_avg:537.64ms step:4655/6250 train_loss:2.2045 aux_loss:0.0278 train_time:2497353ms step_avg:537.64ms step:4656/6250 train_loss:2.2603 aux_loss:0.0282 train_time:2497906ms step_avg:537.65ms step:4657/6250 train_loss:2.2014 aux_loss:0.0277 train_time:2498458ms step_avg:537.65ms step:4658/6250 train_loss:2.1958 aux_loss:0.0278 train_time:2499027ms step_avg:537.66ms step:4659/6250 train_loss:2.1811 aux_loss:0.0284 train_time:2499617ms step_avg:537.67ms step:4660/6250 train_loss:2.1574 aux_loss:0.0282 train_time:2500213ms step_avg:537.68ms step:4661/6250 train_loss:2.1600 aux_loss:0.0283 train_time:2500775ms step_avg:537.69ms step:4662/6250 train_loss:2.1570 aux_loss:0.0283 train_time:2501344ms step_avg:537.69ms step:4663/6250 train_loss:2.2845 aux_loss:0.0280 train_time:2501905ms step_avg:537.70ms step:4664/6250 train_loss:2.1907 aux_loss:0.0283 train_time:2502469ms step_avg:537.70ms step:4665/6250 train_loss:2.1191 aux_loss:0.0284 train_time:2503032ms step_avg:537.71ms step:4666/6250 train_loss:2.1840 aux_loss:0.0280 train_time:2503594ms step_avg:537.71ms step:4667/6250 train_loss:2.1332 aux_loss:0.0275 train_time:2504153ms step_avg:537.72ms step:4668/6250 train_loss:2.1897 aux_loss:0.0284 train_time:2504709ms step_avg:537.72ms step:4669/6250 train_loss:2.1752 aux_loss:0.0283 train_time:2505271ms step_avg:537.73ms step:4670/6250 train_loss:2.1187 aux_loss:0.0275 train_time:2505832ms step_avg:537.73ms step:4671/6250 train_loss:2.1839 aux_loss:0.0284 train_time:2506391ms step_avg:537.74ms step:4672/6250 train_loss:2.1110 aux_loss:0.0285 train_time:2506943ms step_avg:537.74ms step:4673/6250 train_loss:2.2317 aux_loss:0.0282 train_time:2507498ms step_avg:537.74ms step:4674/6250 train_loss:2.1897 aux_loss:0.0281 train_time:2508058ms step_avg:537.75ms step:4675/6250 train_loss:2.2454 aux_loss:0.0285 train_time:2508616ms step_avg:537.75ms step:4676/6250 train_loss:2.2342 aux_loss:0.0284 train_time:2509173ms step_avg:537.76ms step:4677/6250 train_loss:2.2899 aux_loss:0.0283 train_time:2509733ms step_avg:537.76ms step:4678/6250 train_loss:2.1235 aux_loss:0.0280 train_time:2510284ms step_avg:537.76ms step:4679/6250 train_loss:2.1230 aux_loss:0.0279 train_time:2510844ms step_avg:537.77ms step:4680/6250 train_loss:2.2214 aux_loss:0.0282 train_time:2511395ms step_avg:537.77ms step:4681/6250 train_loss:2.1800 aux_loss:0.0278 train_time:2511946ms step_avg:537.77ms step:4682/6250 train_loss:2.1536 aux_loss:0.0276 train_time:2512500ms step_avg:537.78ms step:4683/6250 train_loss:2.1591 aux_loss:0.0280 train_time:2513051ms step_avg:537.78ms step:4684/6250 train_loss:2.1000 aux_loss:0.0278 train_time:2513602ms step_avg:537.78ms step:4685/6250 train_loss:2.2240 aux_loss:0.0279 train_time:2514153ms step_avg:537.79ms step:4686/6250 train_loss:2.2099 aux_loss:0.0274 train_time:2514704ms step_avg:537.79ms step:4687/6250 train_loss:2.1886 aux_loss:0.0282 train_time:2515255ms step_avg:537.79ms step:4688/6250 train_loss:2.2185 aux_loss:0.0281 train_time:2515804ms step_avg:537.79ms step:4689/6250 train_loss:2.2327 aux_loss:0.0277 train_time:2516355ms step_avg:537.80ms step:4690/6250 train_loss:2.1610 aux_loss:0.0281 train_time:2516907ms step_avg:537.80ms step:4691/6250 train_loss:2.1503 aux_loss:0.0282 train_time:2517458ms step_avg:537.80ms step:4692/6250 train_loss:2.1927 aux_loss:0.0280 train_time:2518009ms step_avg:537.81ms step:4693/6250 train_loss:2.1838 aux_loss:0.0278 train_time:2518561ms step_avg:537.81ms step:4694/6250 train_loss:2.1436 aux_loss:0.0283 train_time:2519111ms step_avg:537.81ms step:4695/6250 train_loss:2.2327 aux_loss:0.0284 train_time:2519663ms step_avg:537.81ms step:4696/6250 train_loss:2.2099 aux_loss:0.0280 train_time:2520213ms step_avg:537.82ms step:4697/6250 train_loss:2.1951 aux_loss:0.0279 train_time:2520766ms step_avg:537.82ms step:4698/6250 train_loss:2.2776 aux_loss:0.0280 train_time:2521320ms step_avg:537.82ms step:4699/6250 train_loss:2.2262 aux_loss:0.0279 train_time:2521874ms step_avg:537.83ms step:4700/6250 train_loss:2.2348 aux_loss:0.0283 train_time:2522423ms step_avg:537.83ms step:4701/6250 train_loss:2.2113 aux_loss:0.0285 train_time:2522978ms step_avg:537.83ms step:4702/6250 train_loss:2.0942 aux_loss:0.0283 train_time:2523535ms step_avg:537.84ms step:4703/6250 train_loss:2.2031 aux_loss:0.0282 train_time:2524085ms step_avg:537.84ms step:4704/6250 train_loss:2.2381 aux_loss:0.0278 train_time:2524638ms step_avg:537.84ms step:4705/6250 train_loss:2.2118 aux_loss:0.0280 train_time:2525187ms step_avg:537.85ms step:4706/6250 train_loss:2.3632 aux_loss:0.0287 train_time:2525738ms step_avg:537.85ms step:4707/6250 train_loss:2.2325 aux_loss:0.0280 train_time:2526287ms step_avg:537.85ms step:4708/6250 train_loss:2.2714 aux_loss:0.0283 train_time:2526838ms step_avg:537.85ms step:4709/6250 train_loss:2.1414 aux_loss:0.0277 train_time:2527388ms step_avg:537.86ms step:4710/6250 train_loss:2.0647 aux_loss:0.0272 train_time:2527942ms step_avg:537.86ms step:4711/6250 train_loss:2.1212 aux_loss:0.0278 train_time:2528502ms step_avg:537.86ms step:4712/6250 train_loss:2.1509 aux_loss:0.0281 train_time:2529084ms step_avg:537.87ms step:4713/6250 train_loss:2.2464 aux_loss:0.0275 train_time:2529645ms step_avg:537.88ms step:4714/6250 train_loss:2.1782 aux_loss:0.0280 train_time:2530194ms step_avg:537.88ms step:4715/6250 train_loss:2.2697 aux_loss:0.0281 train_time:2530757ms step_avg:537.89ms step:4716/6250 train_loss:2.1717 aux_loss:0.0275 train_time:2531313ms step_avg:537.89ms step:4717/6250 train_loss:2.2003 aux_loss:0.0280 train_time:2531869ms step_avg:537.89ms step:4718/6250 train_loss:2.0930 aux_loss:0.0276 train_time:2532447ms step_avg:537.90ms step:4719/6250 train_loss:2.2123 aux_loss:0.0273 train_time:2533013ms step_avg:537.91ms step:4720/6250 train_loss:2.3464 aux_loss:0.0280 train_time:2533580ms step_avg:537.91ms step:4721/6250 train_loss:2.2048 aux_loss:0.0281 train_time:2534142ms step_avg:537.92ms step:4722/6250 train_loss:2.2827 aux_loss:0.0277 train_time:2534704ms step_avg:537.93ms step:4723/6250 train_loss:2.2429 aux_loss:0.0280 train_time:2535259ms step_avg:537.93ms step:4724/6250 train_loss:2.1443 aux_loss:0.0279 train_time:2535811ms step_avg:537.93ms step:4725/6250 train_loss:2.1359 aux_loss:0.0279 train_time:2536364ms step_avg:537.94ms step:4726/6250 train_loss:2.1080 aux_loss:0.0283 train_time:2536924ms step_avg:537.94ms step:4727/6250 train_loss:2.1523 aux_loss:0.0282 train_time:2537491ms step_avg:537.95ms step:4728/6250 train_loss:2.1849 aux_loss:0.0276 train_time:2538067ms step_avg:537.95ms step:4729/6250 train_loss:2.1085 aux_loss:0.0277 train_time:2538641ms step_avg:537.96ms step:4730/6250 train_loss:2.1911 aux_loss:0.0282 train_time:2539217ms step_avg:537.97ms step:4731/6250 train_loss:2.1524 aux_loss:0.0280 train_time:2539796ms step_avg:537.98ms step:4732/6250 train_loss:2.0652 aux_loss:0.0277 train_time:2540382ms step_avg:537.99ms step:4733/6250 train_loss:2.2614 aux_loss:0.0284 train_time:2540972ms step_avg:538.00ms step:4734/6250 train_loss:2.0546 aux_loss:0.0281 train_time:2541541ms step_avg:538.01ms step:4735/6250 train_loss:2.2101 aux_loss:0.0273 train_time:2542120ms step_avg:538.01ms step:4736/6250 train_loss:2.0873 aux_loss:0.0273 train_time:2542694ms step_avg:538.02ms step:4737/6250 train_loss:2.2239 aux_loss:0.0283 train_time:2543264ms step_avg:538.03ms step:4738/6250 train_loss:2.1586 aux_loss:0.0276 train_time:2543829ms step_avg:538.03ms step:4739/6250 train_loss:2.1500 aux_loss:0.0278 train_time:2544385ms step_avg:538.04ms step:4740/6250 train_loss:2.1019 aux_loss:0.0281 train_time:2544937ms step_avg:538.04ms step:4741/6250 train_loss:2.1700 aux_loss:0.0277 train_time:2545497ms step_avg:538.05ms step:4742/6250 train_loss:2.2458 aux_loss:0.0278 train_time:2546059ms step_avg:538.05ms step:4743/6250 train_loss:2.1647 aux_loss:0.0282 train_time:2546624ms step_avg:538.06ms step:4744/6250 train_loss:2.3239 aux_loss:0.0282 train_time:2547193ms step_avg:538.06ms step:4745/6250 train_loss:2.3401 aux_loss:0.0282 train_time:2547767ms step_avg:538.07ms step:4746/6250 train_loss:2.1322 aux_loss:0.0277 train_time:2548330ms step_avg:538.08ms step:4747/6250 train_loss:2.1968 aux_loss:0.0276 train_time:2548896ms step_avg:538.08ms step:4748/6250 train_loss:2.1736 aux_loss:0.0276 train_time:2549463ms step_avg:538.09ms step:4749/6250 train_loss:2.2657 aux_loss:0.0275 train_time:2550030ms step_avg:538.09ms step:4750/6250 train_loss:2.1505 aux_loss:0.0279 train_time:2550598ms step_avg:538.10ms step:4751/6250 train_loss:2.1690 aux_loss:0.0278 train_time:2551164ms step_avg:538.11ms step:4752/6250 train_loss:2.1435 aux_loss:0.0281 train_time:2551730ms step_avg:538.11ms step:4753/6250 train_loss:2.1817 aux_loss:0.0280 train_time:2552302ms step_avg:538.12ms step:4754/6250 train_loss:2.1678 aux_loss:0.0276 train_time:2552871ms step_avg:538.13ms step:4755/6250 train_loss:2.1120 aux_loss:0.0276 train_time:2553432ms step_avg:538.13ms step:4756/6250 train_loss:2.1573 aux_loss:0.0281 train_time:2553991ms step_avg:538.14ms step:4757/6250 train_loss:2.2049 aux_loss:0.0279 train_time:2554561ms step_avg:538.14ms step:4758/6250 train_loss:2.1641 aux_loss:0.0280 train_time:2555115ms step_avg:538.15ms step:4759/6250 train_loss:2.1323 aux_loss:0.0280 train_time:2555694ms step_avg:538.15ms step:4760/6250 train_loss:2.1842 aux_loss:0.0278 train_time:2556266ms step_avg:538.16ms step:4761/6250 train_loss:2.0874 aux_loss:0.0280 train_time:2556839ms step_avg:538.17ms step:4762/6250 train_loss:2.1438 aux_loss:0.0276 train_time:2557411ms step_avg:538.18ms step:4763/6250 train_loss:2.1433 aux_loss:0.0278 train_time:2557980ms step_avg:538.18ms step:4764/6250 train_loss:2.2226 aux_loss:0.0281 train_time:2558558ms step_avg:538.19ms step:4765/6250 train_loss:2.1463 aux_loss:0.0280 train_time:2559121ms step_avg:538.20ms step:4766/6250 train_loss:2.0598 aux_loss:0.0275 train_time:2559675ms step_avg:538.20ms step:4767/6250 train_loss:2.1860 aux_loss:0.0280 train_time:2560228ms step_avg:538.20ms step:4768/6250 train_loss:2.2311 aux_loss:0.0271 train_time:2560792ms step_avg:538.21ms step:4769/6250 train_loss:2.1906 aux_loss:0.0276 train_time:2561375ms step_avg:538.22ms step:4770/6250 train_loss:2.0999 aux_loss:0.0278 train_time:2561947ms step_avg:538.22ms step:4771/6250 train_loss:2.2019 aux_loss:0.0276 train_time:2562499ms step_avg:538.23ms step:4772/6250 train_loss:2.1204 aux_loss:0.0271 train_time:2563052ms step_avg:538.23ms step:4773/6250 train_loss:2.2616 aux_loss:0.0283 train_time:2563607ms step_avg:538.23ms step:4774/6250 train_loss:2.1997 aux_loss:0.0280 train_time:2564164ms step_avg:538.24ms step:4775/6250 train_loss:2.1103 aux_loss:0.0274 train_time:2564721ms step_avg:538.24ms step:4776/6250 train_loss:2.0995 aux_loss:0.0276 train_time:2565296ms step_avg:538.25ms step:4777/6250 train_loss:2.1300 aux_loss:0.0278 train_time:2565861ms step_avg:538.25ms step:4778/6250 train_loss:2.2106 aux_loss:0.0277 train_time:2566411ms step_avg:538.26ms step:4779/6250 train_loss:2.1497 aux_loss:0.0277 train_time:2567008ms step_avg:538.27ms step:4780/6250 train_loss:2.1321 aux_loss:0.0282 train_time:2567603ms step_avg:538.28ms step:4781/6250 train_loss:2.1592 aux_loss:0.0283 train_time:2568165ms step_avg:538.29ms step:4782/6250 train_loss:2.1575 aux_loss:0.0278 train_time:2568747ms step_avg:538.30ms step:4783/6250 train_loss:2.1347 aux_loss:0.0276 train_time:2569306ms step_avg:538.30ms step:4784/6250 train_loss:2.1298 aux_loss:0.0276 train_time:2569859ms step_avg:538.30ms step:4785/6250 train_loss:2.1368 aux_loss:0.0277 train_time:2570410ms step_avg:538.31ms step:4786/6250 train_loss:2.1156 aux_loss:0.0275 train_time:2570962ms step_avg:538.31ms step:4787/6250 train_loss:2.1146 aux_loss:0.0276 train_time:2571514ms step_avg:538.31ms step:4788/6250 train_loss:2.1665 aux_loss:0.0278 train_time:2572068ms step_avg:538.31ms step:4789/6250 train_loss:2.1309 aux_loss:0.0275 train_time:2572619ms step_avg:538.32ms step:4790/6250 train_loss:2.1580 aux_loss:0.0276 train_time:2573171ms step_avg:538.32ms step:4791/6250 train_loss:2.2204 aux_loss:0.0279 train_time:2573745ms step_avg:538.33ms step:4792/6250 train_loss:2.2439 aux_loss:0.0278 train_time:2574350ms step_avg:538.34ms step:4793/6250 train_loss:2.2183 aux_loss:0.0277 train_time:2574958ms step_avg:538.36ms step:4794/6250 train_loss:2.1982 aux_loss:0.0279 train_time:2575518ms step_avg:538.36ms step:4795/6250 train_loss:2.1634 aux_loss:0.0278 train_time:2576092ms step_avg:538.37ms step:4796/6250 train_loss:2.0801 aux_loss:0.0272 train_time:2576650ms step_avg:538.37ms step:4797/6250 train_loss:2.2483 aux_loss:0.0276 train_time:2577213ms step_avg:538.38ms step:4798/6250 train_loss:2.2294 aux_loss:0.0280 train_time:2577767ms step_avg:538.38ms step:4799/6250 train_loss:2.1967 aux_loss:0.0280 train_time:2578330ms step_avg:538.39ms step:4800/6250 train_loss:2.1217 aux_loss:0.0278 train_time:2578925ms step_avg:538.40ms step:4801/6250 train_loss:2.1466 aux_loss:0.0276 train_time:2579515ms step_avg:538.41ms step:4802/6250 train_loss:2.1396 aux_loss:0.0276 train_time:2580069ms step_avg:538.41ms step:4803/6250 train_loss:2.1678 aux_loss:0.0274 train_time:2580624ms step_avg:538.42ms step:4804/6250 train_loss:2.0860 aux_loss:0.0277 train_time:2581184ms step_avg:538.42ms step:4805/6250 train_loss:2.2295 aux_loss:0.0277 train_time:2581741ms step_avg:538.42ms step:4806/6250 train_loss:2.2009 aux_loss:0.0271 train_time:2582292ms step_avg:538.43ms step:4807/6250 train_loss:2.1888 aux_loss:0.0278 train_time:2582858ms step_avg:538.43ms step:4808/6250 train_loss:2.1617 aux_loss:0.0277 train_time:2583420ms step_avg:538.44ms step:4809/6250 train_loss:2.1907 aux_loss:0.0274 train_time:2583979ms step_avg:538.44ms step:4810/6250 train_loss:2.2234 aux_loss:0.0279 train_time:2584531ms step_avg:538.44ms step:4811/6250 train_loss:2.1832 aux_loss:0.0280 train_time:2585080ms step_avg:538.45ms step:4812/6250 train_loss:2.0977 aux_loss:0.0275 train_time:2585630ms step_avg:538.45ms step:4813/6250 train_loss:2.1683 aux_loss:0.0274 train_time:2586180ms step_avg:538.45ms step:4814/6250 train_loss:2.1691 aux_loss:0.0282 train_time:2586731ms step_avg:538.45ms step:4815/6250 train_loss:2.2058 aux_loss:0.0282 train_time:2587283ms step_avg:538.46ms step:4816/6250 train_loss:2.1996 aux_loss:0.0273 train_time:2587839ms step_avg:538.46ms step:4817/6250 train_loss:2.2054 aux_loss:0.0274 train_time:2588390ms step_avg:538.46ms step:4818/6250 train_loss:2.1680 aux_loss:0.0278 train_time:2588941ms step_avg:538.47ms step:4819/6250 train_loss:2.2305 aux_loss:0.0282 train_time:2589491ms step_avg:538.47ms step:4820/6250 train_loss:2.2515 aux_loss:0.0274 train_time:2590041ms step_avg:538.47ms step:4821/6250 train_loss:2.2130 aux_loss:0.0272 train_time:2590591ms step_avg:538.47ms step:4822/6250 train_loss:2.1689 aux_loss:0.0276 train_time:2591146ms step_avg:538.48ms step:4823/6250 train_loss:2.1714 aux_loss:0.0276 train_time:2591699ms step_avg:538.48ms step:4824/6250 train_loss:2.1654 aux_loss:0.0274 train_time:2592250ms step_avg:538.48ms step:4825/6250 train_loss:2.1997 aux_loss:0.0273 train_time:2592804ms step_avg:538.48ms step:4826/6250 train_loss:2.1321 aux_loss:0.0277 train_time:2593358ms step_avg:538.49ms step:4827/6250 train_loss:2.1240 aux_loss:0.0277 train_time:2593917ms step_avg:538.49ms step:4828/6250 train_loss:2.1543 aux_loss:0.0274 train_time:2594471ms step_avg:538.50ms step:4829/6250 train_loss:2.1741 aux_loss:0.0271 train_time:2595027ms step_avg:538.50ms step:4830/6250 train_loss:2.2311 aux_loss:0.0281 train_time:2595592ms step_avg:538.50ms step:4831/6250 train_loss:2.1950 aux_loss:0.0281 train_time:2596165ms step_avg:538.51ms step:4832/6250 train_loss:2.2624 aux_loss:0.0283 train_time:2596732ms step_avg:538.52ms step:4833/6250 train_loss:2.1943 aux_loss:0.0275 train_time:2597295ms step_avg:538.52ms step:4834/6250 train_loss:2.1506 aux_loss:0.0278 train_time:2597851ms step_avg:538.53ms step:4835/6250 train_loss:2.2459 aux_loss:0.0281 train_time:2598406ms step_avg:538.53ms step:4836/6250 train_loss:2.1529 aux_loss:0.0276 train_time:2598958ms step_avg:538.53ms step:4837/6250 train_loss:2.2001 aux_loss:0.0273 train_time:2599511ms step_avg:538.54ms step:4838/6250 train_loss:2.1632 aux_loss:0.0277 train_time:2600067ms step_avg:538.54ms step:4839/6250 train_loss:2.1369 aux_loss:0.0282 train_time:2600624ms step_avg:538.54ms step:4840/6250 train_loss:2.1573 aux_loss:0.0278 train_time:2601175ms step_avg:538.55ms step:4841/6250 train_loss:2.1588 aux_loss:0.0274 train_time:2601731ms step_avg:538.55ms step:4842/6250 train_loss:2.2334 aux_loss:0.0277 train_time:2602282ms step_avg:538.55ms step:4843/6250 train_loss:2.1650 aux_loss:0.0279 train_time:2602836ms step_avg:538.55ms step:4844/6250 train_loss:2.2540 aux_loss:0.0279 train_time:2603393ms step_avg:538.56ms step:4845/6250 train_loss:2.1501 aux_loss:0.0275 train_time:2603952ms step_avg:538.56ms step:4846/6250 train_loss:2.0596 aux_loss:0.0277 train_time:2604513ms step_avg:538.57ms step:4847/6250 train_loss:2.1479 aux_loss:0.0278 train_time:2605080ms step_avg:538.57ms step:4848/6250 train_loss:2.1035 aux_loss:0.0276 train_time:2605647ms step_avg:538.58ms step:4849/6250 train_loss:2.2055 aux_loss:0.0273 train_time:2606207ms step_avg:538.58ms step:4850/6250 train_loss:2.1026 aux_loss:0.0271 train_time:2606761ms step_avg:538.59ms step:4851/6250 train_loss:2.2605 aux_loss:0.0278 train_time:2607314ms step_avg:538.59ms step:4852/6250 train_loss:2.1847 aux_loss:0.0277 train_time:2607869ms step_avg:538.59ms step:4853/6250 train_loss:2.1008 aux_loss:0.0274 train_time:2608425ms step_avg:538.60ms step:4854/6250 train_loss:2.1228 aux_loss:0.0275 train_time:2608982ms step_avg:538.60ms step:4855/6250 train_loss:2.1848 aux_loss:0.0274 train_time:2609557ms step_avg:538.61ms step:4856/6250 train_loss:2.1898 aux_loss:0.0273 train_time:2610109ms step_avg:538.61ms step:4857/6250 train_loss:2.2094 aux_loss:0.0275 train_time:2610662ms step_avg:538.61ms step:4858/6250 train_loss:2.0706 aux_loss:0.0276 train_time:2611201ms step_avg:538.61ms step:4859/6250 train_loss:2.2618 aux_loss:0.0277 train_time:2611748ms step_avg:538.62ms step:4860/6250 train_loss:2.2170 aux_loss:0.0276 train_time:2612297ms step_avg:538.62ms step:4861/6250 train_loss:2.0723 aux_loss:0.0275 train_time:2612858ms step_avg:538.62ms step:4862/6250 train_loss:2.1069 aux_loss:0.0276 train_time:2613401ms step_avg:538.62ms step:4863/6250 train_loss:2.1970 aux_loss:0.0276 train_time:2613944ms step_avg:538.62ms step:4864/6250 train_loss:2.1266 aux_loss:0.0281 train_time:2614483ms step_avg:538.62ms step:4865/6250 train_loss:2.2119 aux_loss:0.0277 train_time:2615031ms step_avg:538.63ms step:4866/6250 train_loss:2.1693 aux_loss:0.0273 train_time:2615585ms step_avg:538.63ms step:4867/6250 train_loss:2.2009 aux_loss:0.0277 train_time:2616125ms step_avg:538.63ms step:4868/6250 train_loss:2.2103 aux_loss:0.0279 train_time:2616672ms step_avg:538.63ms step:4869/6250 train_loss:2.2872 aux_loss:0.0277 train_time:2617231ms step_avg:538.64ms step:4870/6250 train_loss:2.1519 aux_loss:0.0277 train_time:2617784ms step_avg:538.64ms step:4871/6250 train_loss:2.1655 aux_loss:0.0278 train_time:2618323ms step_avg:538.64ms step:4872/6250 train_loss:2.1151 aux_loss:0.0275 train_time:2618867ms step_avg:538.64ms step:4873/6250 train_loss:2.2096 aux_loss:0.0275 train_time:2619402ms step_avg:538.64ms step:4874/6250 train_loss:2.2245 aux_loss:0.0280 train_time:2619949ms step_avg:538.64ms step:4875/6250 train_loss:2.1335 aux_loss:0.0275 train_time:2620510ms step_avg:538.65ms step:4876/6250 train_loss:2.1906 aux_loss:0.0277 train_time:2621055ms step_avg:538.65ms step:4877/6250 train_loss:2.1336 aux_loss:0.0281 train_time:2621590ms step_avg:538.65ms step:4878/6250 train_loss:2.1537 aux_loss:0.0281 train_time:2622127ms step_avg:538.65ms step:4879/6250 train_loss:2.2374 aux_loss:0.0276 train_time:2622666ms step_avg:538.65ms step:4880/6250 train_loss:2.1277 aux_loss:0.0277 train_time:2623206ms step_avg:538.65ms step:4881/6250 train_loss:2.2436 aux_loss:0.0281 train_time:2623767ms step_avg:538.65ms step:4882/6250 train_loss:2.1607 aux_loss:0.0273 train_time:2624305ms step_avg:538.65ms step:4883/6250 train_loss:2.2117 aux_loss:0.0273 train_time:2624853ms step_avg:538.65ms step:4884/6250 train_loss:2.2578 aux_loss:0.0273 train_time:2625402ms step_avg:538.65ms step:4885/6250 train_loss:2.0969 aux_loss:0.0274 train_time:2625944ms step_avg:538.66ms step:4886/6250 train_loss:2.0830 aux_loss:0.0274 train_time:2626504ms step_avg:538.66ms step:4887/6250 train_loss:2.1717 aux_loss:0.0275 train_time:2627081ms step_avg:538.67ms step:4888/6250 train_loss:2.1299 aux_loss:0.0275 train_time:2627629ms step_avg:538.67ms step:4889/6250 train_loss:2.1366 aux_loss:0.0269 train_time:2628168ms step_avg:538.67ms step:4890/6250 train_loss:2.0980 aux_loss:0.0279 train_time:2628706ms step_avg:538.67ms step:4891/6250 train_loss:2.0782 aux_loss:0.0280 train_time:2629254ms step_avg:538.67ms step:4892/6250 train_loss:2.2451 aux_loss:0.0274 train_time:2629789ms step_avg:538.67ms step:4893/6250 train_loss:2.1583 aux_loss:0.0278 train_time:2630337ms step_avg:538.67ms step:4894/6250 train_loss:2.1654 aux_loss:0.0281 train_time:2630891ms step_avg:538.68ms step:4895/6250 train_loss:2.1319 aux_loss:0.0275 train_time:2631432ms step_avg:538.68ms step:4896/6250 train_loss:2.0938 aux_loss:0.0273 train_time:2631985ms step_avg:538.68ms step:4897/6250 train_loss:2.2264 aux_loss:0.0276 train_time:2632549ms step_avg:538.68ms step:4898/6250 train_loss:2.2056 aux_loss:0.0275 train_time:2633088ms step_avg:538.68ms step:4899/6250 train_loss:2.2287 aux_loss:0.0278 train_time:2633625ms step_avg:538.68ms step:4900/6250 train_loss:2.1329 aux_loss:0.0279 train_time:2634162ms step_avg:538.68ms step:4901/6250 train_loss:2.2562 aux_loss:0.0282 train_time:2634701ms step_avg:538.68ms step:4902/6250 train_loss:2.1705 aux_loss:0.0278 train_time:2635237ms step_avg:538.68ms step:4903/6250 train_loss:2.1378 aux_loss:0.0274 train_time:2635786ms step_avg:538.69ms step:4904/6250 train_loss:2.2465 aux_loss:0.0279 train_time:2636336ms step_avg:538.69ms step:4905/6250 train_loss:2.1151 aux_loss:0.0277 train_time:2636884ms step_avg:538.69ms step:4906/6250 train_loss:2.1618 aux_loss:0.0282 train_time:2637436ms step_avg:538.69ms step:4907/6250 train_loss:2.1672 aux_loss:0.0280 train_time:2637974ms step_avg:538.69ms step:4908/6250 train_loss:2.1578 aux_loss:0.0277 train_time:2638539ms step_avg:538.70ms step:4909/6250 train_loss:2.1641 aux_loss:0.0274 train_time:2639099ms step_avg:538.70ms step:4910/6250 train_loss:2.1655 aux_loss:0.0275 train_time:2639642ms step_avg:538.70ms step:4911/6250 train_loss:2.1781 aux_loss:0.0276 train_time:2640200ms step_avg:538.71ms step:4912/6250 train_loss:2.0912 aux_loss:0.0273 train_time:2640754ms step_avg:538.71ms step:4913/6250 train_loss:2.1865 aux_loss:0.0272 train_time:2641297ms step_avg:538.71ms step:4914/6250 train_loss:2.1080 aux_loss:0.0273 train_time:2641834ms step_avg:538.71ms step:4915/6250 train_loss:2.2005 aux_loss:0.0273 train_time:2642370ms step_avg:538.71ms step:4916/6250 train_loss:2.0842 aux_loss:0.0271 train_time:2642915ms step_avg:538.71ms step:4917/6250 train_loss:2.1349 aux_loss:0.0273 train_time:2643461ms step_avg:538.71ms step:4918/6250 train_loss:2.1753 aux_loss:0.0271 train_time:2644023ms step_avg:538.72ms step:4919/6250 train_loss:2.2564 aux_loss:0.0272 train_time:2644561ms step_avg:538.72ms step:4920/6250 train_loss:2.1036 aux_loss:0.0276 train_time:2645115ms step_avg:538.72ms step:4921/6250 train_loss:2.1529 aux_loss:0.0276 train_time:2645654ms step_avg:538.72ms step:4922/6250 train_loss:2.2020 aux_loss:0.0273 train_time:2646191ms step_avg:538.72ms step:4923/6250 train_loss:2.1466 aux_loss:0.0277 train_time:2646729ms step_avg:538.72ms step:4924/6250 train_loss:2.1565 aux_loss:0.0275 train_time:2647273ms step_avg:538.72ms step:4925/6250 train_loss:2.1793 aux_loss:0.0271 train_time:2647812ms step_avg:538.72ms step:4926/6250 train_loss:2.1404 aux_loss:0.0279 train_time:2648353ms step_avg:538.72ms step:4927/6250 train_loss:2.1519 aux_loss:0.0281 train_time:2648900ms step_avg:538.72ms step:4928/6250 train_loss:2.1524 aux_loss:0.0277 train_time:2649441ms step_avg:538.72ms step:4929/6250 train_loss:2.1780 aux_loss:0.0273 train_time:2649980ms step_avg:538.72ms step:4930/6250 train_loss:2.2081 aux_loss:0.0275 train_time:2650518ms step_avg:538.72ms step:4931/6250 train_loss:2.2514 aux_loss:0.0274 train_time:2651054ms step_avg:538.72ms step:4932/6250 train_loss:2.2914 aux_loss:0.0273 train_time:2651591ms step_avg:538.72ms step:4933/6250 train_loss:2.1610 aux_loss:0.0274 train_time:2652128ms step_avg:538.72ms step:4934/6250 train_loss:2.1043 aux_loss:0.0279 train_time:2652664ms step_avg:538.72ms step:4935/6250 train_loss:2.1152 aux_loss:0.0272 train_time:2653202ms step_avg:538.72ms step:4936/6250 train_loss:2.1380 aux_loss:0.0270 train_time:2653737ms step_avg:538.72ms step:4937/6250 train_loss:2.1581 aux_loss:0.0273 train_time:2654274ms step_avg:538.72ms step:4938/6250 train_loss:2.2061 aux_loss:0.0274 train_time:2654832ms step_avg:538.72ms step:4939/6250 train_loss:2.1608 aux_loss:0.0276 train_time:2655367ms step_avg:538.72ms step:4940/6250 train_loss:2.0807 aux_loss:0.0275 train_time:2655921ms step_avg:538.73ms step:4941/6250 train_loss:2.1151 aux_loss:0.0275 train_time:2656466ms step_avg:538.73ms step:4942/6250 train_loss:2.1213 aux_loss:0.0276 train_time:2657003ms step_avg:538.73ms step:4943/6250 train_loss:2.1641 aux_loss:0.0276 train_time:2657538ms step_avg:538.73ms step:4944/6250 train_loss:2.0741 aux_loss:0.0275 train_time:2658076ms step_avg:538.73ms step:4945/6250 train_loss:2.3179 aux_loss:0.0276 train_time:2658611ms step_avg:538.73ms step:4946/6250 train_loss:2.2047 aux_loss:0.0275 train_time:2659147ms step_avg:538.73ms step:4947/6250 train_loss:2.1992 aux_loss:0.0277 train_time:2659681ms step_avg:538.72ms step:4948/6250 train_loss:2.1482 aux_loss:0.0274 train_time:2660217ms step_avg:538.72ms step:4949/6250 train_loss:2.0887 aux_loss:0.0271 train_time:2660753ms step_avg:538.72ms step:4950/6250 train_loss:2.0817 aux_loss:0.0275 train_time:2661290ms step_avg:538.72ms step:4951/6250 train_loss:2.2132 aux_loss:0.0272 train_time:2661827ms step_avg:538.72ms step:4952/6250 train_loss:2.2629 aux_loss:0.0276 train_time:2662363ms step_avg:538.72ms step:4953/6250 train_loss:2.1759 aux_loss:0.0275 train_time:2662899ms step_avg:538.72ms step:4954/6250 train_loss:2.2441 aux_loss:0.0272 train_time:2663443ms step_avg:538.72ms step:4955/6250 train_loss:2.1625 aux_loss:0.0278 train_time:2663978ms step_avg:538.72ms step:4956/6250 train_loss:2.2390 aux_loss:0.0273 train_time:2664515ms step_avg:538.72ms step:4957/6250 train_loss:2.1821 aux_loss:0.0278 train_time:2665053ms step_avg:538.72ms step:4958/6250 train_loss:2.1169 aux_loss:0.0280 train_time:2665588ms step_avg:538.72ms step:4959/6250 train_loss:2.1612 aux_loss:0.0277 train_time:2666140ms step_avg:538.72ms step:4960/6250 train_loss:2.1419 aux_loss:0.0272 train_time:2666694ms step_avg:538.73ms step:4961/6250 train_loss:2.2140 aux_loss:0.0279 train_time:2667232ms step_avg:538.73ms step:4962/6250 train_loss:2.1802 aux_loss:0.0274 train_time:2667768ms step_avg:538.73ms step:4963/6250 train_loss:2.1701 aux_loss:0.0274 train_time:2668302ms step_avg:538.72ms step:4964/6250 train_loss:2.1535 aux_loss:0.0271 train_time:2668840ms step_avg:538.72ms step:4965/6250 train_loss:2.1473 aux_loss:0.0271 train_time:2669375ms step_avg:538.72ms step:4966/6250 train_loss:2.1547 aux_loss:0.0277 train_time:2669913ms step_avg:538.72ms step:4967/6250 train_loss:2.1183 aux_loss:0.0276 train_time:2670471ms step_avg:538.73ms step:4968/6250 train_loss:2.2124 aux_loss:0.0274 train_time:2671032ms step_avg:538.73ms step:4969/6250 train_loss:2.2625 aux_loss:0.0277 train_time:2671570ms step_avg:538.73ms step:4970/6250 train_loss:2.2127 aux_loss:0.0276 train_time:2672111ms step_avg:538.73ms step:4971/6250 train_loss:2.1002 aux_loss:0.0274 train_time:2672650ms step_avg:538.73ms step:4972/6250 train_loss:2.1769 aux_loss:0.0276 train_time:2673193ms step_avg:538.73ms step:4973/6250 train_loss:2.1236 aux_loss:0.0270 train_time:2673730ms step_avg:538.73ms step:4974/6250 train_loss:2.1232 aux_loss:0.0268 train_time:2674267ms step_avg:538.73ms step:4975/6250 train_loss:2.1917 aux_loss:0.0279 train_time:2674803ms step_avg:538.73ms step:4976/6250 train_loss:2.2921 aux_loss:0.0277 train_time:2675339ms step_avg:538.73ms step:4977/6250 train_loss:2.1476 aux_loss:0.0273 train_time:2675875ms step_avg:538.73ms step:4978/6250 train_loss:2.2047 aux_loss:0.0282 train_time:2676418ms step_avg:538.73ms step:4979/6250 train_loss:2.1577 aux_loss:0.0278 train_time:2676953ms step_avg:538.73ms step:4980/6250 train_loss:2.1575 aux_loss:0.0275 train_time:2677490ms step_avg:538.73ms step:4981/6250 train_loss:2.0684 aux_loss:0.0274 train_time:2678026ms step_avg:538.73ms step:4982/6250 train_loss:2.1092 aux_loss:0.0276 train_time:2678592ms step_avg:538.74ms step:4983/6250 train_loss:2.1794 aux_loss:0.0276 train_time:2679137ms step_avg:538.74ms step:4984/6250 train_loss:2.1270 aux_loss:0.0273 train_time:2679673ms step_avg:538.74ms step:4985/6250 train_loss:2.1084 aux_loss:0.0277 train_time:2680223ms step_avg:538.74ms step:4986/6250 train_loss:2.2222 aux_loss:0.0277 train_time:2680763ms step_avg:538.74ms step:4987/6250 train_loss:2.1257 aux_loss:0.0273 train_time:2681324ms step_avg:538.74ms step:4988/6250 train_loss:2.1310 aux_loss:0.0275 train_time:2681869ms step_avg:538.74ms step:4989/6250 train_loss:2.1306 aux_loss:0.0272 train_time:2682429ms step_avg:538.75ms step:4990/6250 train_loss:2.1759 aux_loss:0.0274 train_time:2682966ms step_avg:538.75ms step:4991/6250 train_loss:2.1726 aux_loss:0.0271 train_time:2683503ms step_avg:538.75ms step:4992/6250 train_loss:2.1294 aux_loss:0.0271 train_time:2684039ms step_avg:538.75ms step:4993/6250 train_loss:2.1407 aux_loss:0.0270 train_time:2684575ms step_avg:538.75ms step:4994/6250 train_loss:2.0929 aux_loss:0.0268 train_time:2685113ms step_avg:538.75ms step:4995/6250 train_loss:2.1979 aux_loss:0.0272 train_time:2685652ms step_avg:538.75ms step:4996/6250 train_loss:2.2761 aux_loss:0.0274 train_time:2686190ms step_avg:538.75ms step:4997/6250 train_loss:2.2398 aux_loss:0.0274 train_time:2686726ms step_avg:538.75ms step:4998/6250 train_loss:2.1934 aux_loss:0.0276 train_time:2687262ms step_avg:538.75ms step:4999/6250 train_loss:2.1714 aux_loss:0.0271 train_time:2687799ms step_avg:538.74ms step:5000/6250 train_loss:2.1559 aux_loss:0.0269 train_time:2688335ms step_avg:538.74ms step:5001/6250 train_loss:2.1054 aux_loss:0.0276 train_time:2688871ms step_avg:538.74ms step:5002/6250 train_loss:2.2633 aux_loss:0.0276 train_time:2689411ms step_avg:538.74ms step:5003/6250 train_loss:2.1922 aux_loss:0.0271 train_time:2689948ms step_avg:538.74ms step:5004/6250 train_loss:2.1521 aux_loss:0.0270 train_time:2690483ms step_avg:538.74ms step:5005/6250 train_loss:2.1906 aux_loss:0.0269 train_time:2691020ms step_avg:538.74ms step:5006/6250 train_loss:2.2372 aux_loss:0.0276 train_time:2691564ms step_avg:538.74ms step:5007/6250 train_loss:2.1498 aux_loss:0.0277 train_time:2692099ms step_avg:538.74ms step:5008/6250 train_loss:2.1444 aux_loss:0.0270 train_time:2692635ms step_avg:538.74ms step:5009/6250 train_loss:2.1592 aux_loss:0.0273 train_time:2693173ms step_avg:538.74ms step:5010/6250 train_loss:2.2164 aux_loss:0.0272 train_time:2693716ms step_avg:538.74ms step:5011/6250 train_loss:2.0985 aux_loss:0.0276 train_time:2694255ms step_avg:538.74ms step:5012/6250 train_loss:2.1518 aux_loss:0.0276 train_time:2694797ms step_avg:538.74ms step:5013/6250 train_loss:2.1703 aux_loss:0.0274 train_time:2695334ms step_avg:538.74ms step:5014/6250 train_loss:2.1630 aux_loss:0.0273 train_time:2695881ms step_avg:538.75ms step:5015/6250 train_loss:2.2357 aux_loss:0.0271 train_time:2696434ms step_avg:538.75ms step:5016/6250 train_loss:2.1566 aux_loss:0.0272 train_time:2696979ms step_avg:538.75ms step:5017/6250 train_loss:2.1843 aux_loss:0.0281 train_time:2697523ms step_avg:538.75ms step:5018/6250 train_loss:2.2147 aux_loss:0.0267 train_time:2698059ms step_avg:538.75ms step:5019/6250 train_loss:2.1887 aux_loss:0.0277 train_time:2698597ms step_avg:538.75ms step:5020/6250 train_loss:2.1179 aux_loss:0.0274 train_time:2699135ms step_avg:538.75ms step:5021/6250 train_loss:2.1945 aux_loss:0.0267 train_time:2699672ms step_avg:538.75ms step:5022/6250 train_loss:2.1531 aux_loss:0.0272 train_time:2700209ms step_avg:538.75ms step:5023/6250 train_loss:2.0668 aux_loss:0.0275 train_time:2700747ms step_avg:538.75ms step:5024/6250 train_loss:2.2016 aux_loss:0.0273 train_time:2701286ms step_avg:538.75ms step:5025/6250 train_loss:2.1345 aux_loss:0.0270 train_time:2701828ms step_avg:538.75ms step:5026/6250 train_loss:2.1034 aux_loss:0.0271 train_time:2702366ms step_avg:538.75ms step:5027/6250 train_loss:2.1829 aux_loss:0.0269 train_time:2702902ms step_avg:538.75ms step:5028/6250 train_loss:2.1812 aux_loss:0.0265 train_time:2703445ms step_avg:538.75ms step:5029/6250 train_loss:2.1114 aux_loss:0.0271 train_time:2703984ms step_avg:538.75ms step:5030/6250 train_loss:2.1384 aux_loss:0.0273 train_time:2704522ms step_avg:538.75ms step:5031/6250 train_loss:2.1021 aux_loss:0.0269 train_time:2705066ms step_avg:538.75ms step:5032/6250 train_loss:2.2025 aux_loss:0.0271 train_time:2705603ms step_avg:538.75ms step:5033/6250 train_loss:2.0885 aux_loss:0.0274 train_time:2706141ms step_avg:538.75ms step:5034/6250 train_loss:2.2003 aux_loss:0.0277 train_time:2706679ms step_avg:538.75ms step:5035/6250 train_loss:2.1909 aux_loss:0.0275 train_time:2707216ms step_avg:538.75ms step:5036/6250 train_loss:2.1625 aux_loss:0.0273 train_time:2707755ms step_avg:538.75ms step:5037/6250 train_loss:2.1803 aux_loss:0.0273 train_time:2708291ms step_avg:538.75ms step:5038/6250 train_loss:2.1703 aux_loss:0.0273 train_time:2708828ms step_avg:538.75ms step:5039/6250 train_loss:2.1364 aux_loss:0.0268 train_time:2709366ms step_avg:538.75ms step:5040/6250 train_loss:2.1111 aux_loss:0.0269 train_time:2709904ms step_avg:538.75ms step:5041/6250 train_loss:2.0861 aux_loss:0.0271 train_time:2710441ms step_avg:538.75ms step:5042/6250 train_loss:2.1501 aux_loss:0.0270 train_time:2710981ms step_avg:538.75ms step:5043/6250 train_loss:2.1350 aux_loss:0.0269 train_time:2711518ms step_avg:538.75ms step:5044/6250 train_loss:2.1012 aux_loss:0.0270 train_time:2712055ms step_avg:538.75ms step:5045/6250 train_loss:2.1554 aux_loss:0.0269 train_time:2712595ms step_avg:538.75ms step:5046/6250 train_loss:2.0979 aux_loss:0.0267 train_time:2713133ms step_avg:538.75ms step:5047/6250 train_loss:2.1712 aux_loss:0.0269 train_time:2713680ms step_avg:538.75ms step:5048/6250 train_loss:2.1014 aux_loss:0.0274 train_time:2714217ms step_avg:538.75ms step:5049/6250 train_loss:2.1208 aux_loss:0.0271 train_time:2714756ms step_avg:538.75ms step:5050/6250 train_loss:2.0960 aux_loss:0.0269 train_time:2715293ms step_avg:538.75ms step:5051/6250 train_loss:2.1106 aux_loss:0.0271 train_time:2715844ms step_avg:538.75ms step:5052/6250 train_loss:2.1266 aux_loss:0.0274 train_time:2716388ms step_avg:538.75ms step:5053/6250 train_loss:2.2097 aux_loss:0.0267 train_time:2716927ms step_avg:538.75ms step:5054/6250 train_loss:2.1998 aux_loss:0.0271 train_time:2717471ms step_avg:538.75ms step:5055/6250 train_loss:2.1788 aux_loss:0.0271 train_time:2718008ms step_avg:538.75ms step:5056/6250 train_loss:2.1965 aux_loss:0.0269 train_time:2718547ms step_avg:538.75ms step:5057/6250 train_loss:2.1102 aux_loss:0.0266 train_time:2719113ms step_avg:538.76ms step:5058/6250 train_loss:2.0629 aux_loss:0.0273 train_time:2719708ms step_avg:538.77ms step:5059/6250 train_loss:2.1845 aux_loss:0.0271 train_time:2720247ms step_avg:538.77ms step:5060/6250 train_loss:2.1504 aux_loss:0.0268 train_time:2720781ms step_avg:538.77ms step:5061/6250 train_loss:2.1511 aux_loss:0.0275 train_time:2721314ms step_avg:538.77ms step:5062/6250 train_loss:2.1223 aux_loss:0.0275 train_time:2721850ms step_avg:538.77ms step:5063/6250 train_loss:2.1856 aux_loss:0.0269 train_time:2722383ms step_avg:538.77ms step:5064/6250 train_loss:2.0865 aux_loss:0.0269 train_time:2722923ms step_avg:538.77ms step:5065/6250 train_loss:2.1729 aux_loss:0.0273 train_time:2723452ms step_avg:538.76ms step:5066/6250 train_loss:2.1881 aux_loss:0.0269 train_time:2723984ms step_avg:538.76ms step:5067/6250 train_loss:2.2142 aux_loss:0.0268 train_time:2724513ms step_avg:538.76ms step:5068/6250 train_loss:2.1312 aux_loss:0.0274 train_time:2725044ms step_avg:538.76ms step:5069/6250 train_loss:2.1982 aux_loss:0.0274 train_time:2725575ms step_avg:538.76ms step:5070/6250 train_loss:2.1552 aux_loss:0.0267 train_time:2726144ms step_avg:538.76ms step:5071/6250 train_loss:2.0613 aux_loss:0.0266 train_time:2726730ms step_avg:538.77ms step:5072/6250 train_loss:2.1290 aux_loss:0.0272 train_time:2727315ms step_avg:538.78ms step:5073/6250 train_loss:2.1756 aux_loss:0.0271 train_time:2727903ms step_avg:538.79ms step:5074/6250 train_loss:2.1288 aux_loss:0.0270 train_time:2728488ms step_avg:538.80ms step:5075/6250 train_loss:2.1507 aux_loss:0.0269 train_time:2729052ms step_avg:538.81ms step:5076/6250 train_loss:2.1522 aux_loss:0.0273 train_time:2729590ms step_avg:538.81ms step:5077/6250 train_loss:2.0963 aux_loss:0.0269 train_time:2730125ms step_avg:538.81ms step:5078/6250 train_loss:2.2560 aux_loss:0.0271 train_time:2730659ms step_avg:538.80ms step:5079/6250 train_loss:2.1477 aux_loss:0.0269 train_time:2731204ms step_avg:538.81ms step:5080/6250 train_loss:2.2160 aux_loss:0.0270 train_time:2731773ms step_avg:538.81ms step:5081/6250 train_loss:2.0950 aux_loss:0.0269 train_time:2732328ms step_avg:538.81ms step:5082/6250 train_loss:2.0649 aux_loss:0.0270 train_time:2732877ms step_avg:538.82ms step:5083/6250 train_loss:2.1954 aux_loss:0.0272 train_time:2733422ms step_avg:538.82ms step:5084/6250 train_loss:2.1256 aux_loss:0.0273 train_time:2733954ms step_avg:538.82ms step:5085/6250 train_loss:2.1252 aux_loss:0.0274 train_time:2734516ms step_avg:538.82ms step:5086/6250 train_loss:2.2097 aux_loss:0.0269 train_time:2735113ms step_avg:538.83ms step:5087/6250 train_loss:2.2024 aux_loss:0.0272 train_time:2735698ms step_avg:538.84ms step:5088/6250 train_loss:2.2686 aux_loss:0.0274 train_time:2736282ms step_avg:538.85ms step:5089/6250 train_loss:2.0925 aux_loss:0.0273 train_time:2736867ms step_avg:538.86ms step:5090/6250 train_loss:2.2356 aux_loss:0.0272 train_time:2737454ms step_avg:538.87ms step:5091/6250 train_loss:2.1091 aux_loss:0.0274 train_time:2738037ms step_avg:538.88ms step:5092/6250 train_loss:2.1156 aux_loss:0.0278 train_time:2738632ms step_avg:538.89ms step:5093/6250 train_loss:2.1535 aux_loss:0.0271 train_time:2739173ms step_avg:538.89ms step:5094/6250 train_loss:2.1558 aux_loss:0.0268 train_time:2739708ms step_avg:538.89ms step:5095/6250 train_loss:2.1333 aux_loss:0.0273 train_time:2740249ms step_avg:538.89ms step:5096/6250 train_loss:2.2022 aux_loss:0.0271 train_time:2740789ms step_avg:538.89ms step:5097/6250 train_loss:2.1259 aux_loss:0.0267 train_time:2741321ms step_avg:538.89ms step:5098/6250 train_loss:2.1683 aux_loss:0.0269 train_time:2741863ms step_avg:538.89ms step:5099/6250 train_loss:2.1478 aux_loss:0.0268 train_time:2742415ms step_avg:538.89ms step:5100/6250 train_loss:2.2165 aux_loss:0.0264 train_time:2742960ms step_avg:538.89ms step:5101/6250 train_loss:2.1355 aux_loss:0.0273 train_time:2743497ms step_avg:538.89ms step:5102/6250 train_loss:2.1099 aux_loss:0.0269 train_time:2744053ms step_avg:538.89ms step:5103/6250 train_loss:2.1273 aux_loss:0.0271 train_time:2744590ms step_avg:538.89ms step:5104/6250 train_loss:2.0987 aux_loss:0.0269 train_time:2745135ms step_avg:538.90ms step:5105/6250 train_loss:2.1168 aux_loss:0.0263 train_time:2745669ms step_avg:538.89ms step:5106/6250 train_loss:2.1432 aux_loss:0.0266 train_time:2746221ms step_avg:538.90ms step:5107/6250 train_loss:2.2690 aux_loss:0.0273 train_time:2746760ms step_avg:538.90ms step:5108/6250 train_loss:2.1342 aux_loss:0.0267 train_time:2747330ms step_avg:538.90ms step:5109/6250 train_loss:2.1738 aux_loss:0.0269 train_time:2747869ms step_avg:538.90ms step:5110/6250 train_loss:2.1520 aux_loss:0.0266 train_time:2748419ms step_avg:538.91ms step:5111/6250 train_loss:2.1449 aux_loss:0.0269 train_time:2748978ms step_avg:538.91ms step:5112/6250 train_loss:2.1641 aux_loss:0.0270 train_time:2749512ms step_avg:538.91ms step:5113/6250 train_loss:2.1409 aux_loss:0.0268 train_time:2750063ms step_avg:538.91ms step:5114/6250 train_loss:2.2439 aux_loss:0.0267 train_time:2750623ms step_avg:538.92ms step:5115/6250 train_loss:2.2076 aux_loss:0.0266 train_time:2751184ms step_avg:538.92ms step:5116/6250 train_loss:2.2094 aux_loss:0.0272 train_time:2751733ms step_avg:538.92ms step:5117/6250 train_loss:2.2312 aux_loss:0.0271 train_time:2752269ms step_avg:538.92ms step:5118/6250 train_loss:2.1259 aux_loss:0.0267 train_time:2752813ms step_avg:538.92ms step:5119/6250 train_loss:2.1807 aux_loss:0.0268 train_time:2753343ms step_avg:538.92ms step:5120/6250 train_loss:2.1325 aux_loss:0.0272 train_time:2753881ms step_avg:538.92ms step:5121/6250 train_loss:2.1225 aux_loss:0.0270 train_time:2754416ms step_avg:538.92ms step:5122/6250 train_loss:2.2555 aux_loss:0.0270 train_time:2754954ms step_avg:538.92ms step:5123/6250 train_loss:2.1726 aux_loss:0.0271 train_time:2755486ms step_avg:538.92ms step:5124/6250 train_loss:2.0696 aux_loss:0.0270 train_time:2756017ms step_avg:538.92ms step:5125/6250 train_loss:2.1176 aux_loss:0.0266 train_time:2756548ms step_avg:538.91ms step:5126/6250 train_loss:2.1556 aux_loss:0.0274 train_time:2757080ms step_avg:538.91ms step:5127/6250 train_loss:2.0954 aux_loss:0.0271 train_time:2757613ms step_avg:538.91ms step:5128/6250 train_loss:2.1243 aux_loss:0.0270 train_time:2758149ms step_avg:538.91ms step:5129/6250 train_loss:2.1417 aux_loss:0.0275 train_time:2758691ms step_avg:538.91ms step:5130/6250 train_loss:2.1305 aux_loss:0.0265 train_time:2759224ms step_avg:538.91ms step:5131/6250 train_loss:2.1247 aux_loss:0.0269 train_time:2759760ms step_avg:538.91ms step:5132/6250 train_loss:2.1097 aux_loss:0.0268 train_time:2760296ms step_avg:538.91ms step:5133/6250 train_loss:2.1661 aux_loss:0.0268 train_time:2760832ms step_avg:538.91ms step:5134/6250 train_loss:2.1121 aux_loss:0.0268 train_time:2761368ms step_avg:538.91ms step:5135/6250 train_loss:2.1801 aux_loss:0.0268 train_time:2761900ms step_avg:538.91ms step:5136/6250 train_loss:2.0942 aux_loss:0.0266 train_time:2762431ms step_avg:538.91ms step:5137/6250 train_loss:2.1237 aux_loss:0.0272 train_time:2762963ms step_avg:538.90ms step:5138/6250 train_loss:2.1526 aux_loss:0.0274 train_time:2763497ms step_avg:538.90ms step:5139/6250 train_loss:2.1574 aux_loss:0.0266 train_time:2764030ms step_avg:538.90ms step:5140/6250 train_loss:2.0721 aux_loss:0.0263 train_time:2764568ms step_avg:538.90ms step:5141/6250 train_loss:2.1984 aux_loss:0.0272 train_time:2765099ms step_avg:538.90ms step:5142/6250 train_loss:2.1520 aux_loss:0.0268 train_time:2765630ms step_avg:538.90ms step:5143/6250 train_loss:2.1112 aux_loss:0.0267 train_time:2766160ms step_avg:538.90ms step:5144/6250 train_loss:2.1821 aux_loss:0.0264 train_time:2766694ms step_avg:538.90ms step:5145/6250 train_loss:2.1239 aux_loss:0.0264 train_time:2767237ms step_avg:538.90ms step:5146/6250 train_loss:2.1452 aux_loss:0.0270 train_time:2767773ms step_avg:538.90ms step:5147/6250 train_loss:2.1663 aux_loss:0.0263 train_time:2768304ms step_avg:538.90ms step:5148/6250 train_loss:2.2174 aux_loss:0.0269 train_time:2768836ms step_avg:538.89ms step:5149/6250 train_loss:2.0949 aux_loss:0.0267 train_time:2769368ms step_avg:538.89ms step:5150/6250 train_loss:2.0597 aux_loss:0.0267 train_time:2769907ms step_avg:538.89ms step:5151/6250 train_loss:2.0794 aux_loss:0.0269 train_time:2770443ms step_avg:538.89ms step:5152/6250 train_loss:2.1544 aux_loss:0.0271 train_time:2770980ms step_avg:538.89ms step:5153/6250 train_loss:2.1926 aux_loss:0.0280 train_time:2771519ms step_avg:538.89ms step:5154/6250 train_loss:2.2171 aux_loss:0.0268 train_time:2772053ms step_avg:538.89ms step:5155/6250 train_loss:2.1557 aux_loss:0.0272 train_time:2772585ms step_avg:538.89ms step:5156/6250 train_loss:2.1785 aux_loss:0.0271 train_time:2773118ms step_avg:538.89ms step:5157/6250 train_loss:2.1819 aux_loss:0.0261 train_time:2773649ms step_avg:538.89ms step:5158/6250 train_loss:2.1615 aux_loss:0.0264 train_time:2774179ms step_avg:538.88ms step:5159/6250 train_loss:2.2605 aux_loss:0.0269 train_time:2774712ms step_avg:538.88ms step:5160/6250 train_loss:2.1595 aux_loss:0.0267 train_time:2775241ms step_avg:538.88ms step:5161/6250 train_loss:2.1392 aux_loss:0.0266 train_time:2775772ms step_avg:538.88ms step:5162/6250 train_loss:2.2829 aux_loss:0.0269 train_time:2776303ms step_avg:538.88ms step:5163/6250 train_loss:2.2230 aux_loss:0.0266 train_time:2776833ms step_avg:538.88ms step:5164/6250 train_loss:2.2539 aux_loss:0.0268 train_time:2777364ms step_avg:538.88ms step:5165/6250 train_loss:2.1692 aux_loss:0.0271 train_time:2777895ms step_avg:538.87ms step:5166/6250 train_loss:2.1880 aux_loss:0.0269 train_time:2778427ms step_avg:538.87ms step:5167/6250 train_loss:2.1917 aux_loss:0.0268 train_time:2778962ms step_avg:538.87ms step:5168/6250 train_loss:2.0737 aux_loss:0.0271 train_time:2779502ms step_avg:538.87ms step:5169/6250 train_loss:2.1185 aux_loss:0.0269 train_time:2780038ms step_avg:538.87ms step:5170/6250 train_loss:2.1819 aux_loss:0.0268 train_time:2780569ms step_avg:538.87ms step:5171/6250 train_loss:2.0904 aux_loss:0.0269 train_time:2781102ms step_avg:538.87ms step:5172/6250 train_loss:2.1800 aux_loss:0.0271 train_time:2781635ms step_avg:538.87ms step:5173/6250 train_loss:2.2269 aux_loss:0.0268 train_time:2782176ms step_avg:538.87ms step:5174/6250 train_loss:2.1219 aux_loss:0.0272 train_time:2782713ms step_avg:538.87ms step:5175/6250 train_loss:2.1714 aux_loss:0.0267 train_time:2783249ms step_avg:538.87ms step:5176/6250 train_loss:2.0828 aux_loss:0.0261 train_time:2783781ms step_avg:538.87ms step:5177/6250 train_loss:2.1143 aux_loss:0.0269 train_time:2784309ms step_avg:538.86ms step:5178/6250 train_loss:2.0907 aux_loss:0.0269 train_time:2784841ms step_avg:538.86ms step:5179/6250 train_loss:2.1377 aux_loss:0.0267 train_time:2785371ms step_avg:538.86ms step:5180/6250 train_loss:2.1489 aux_loss:0.0269 train_time:2785903ms step_avg:538.86ms step:5181/6250 train_loss:2.1790 aux_loss:0.0269 train_time:2786432ms step_avg:538.86ms step:5182/6250 train_loss:2.0570 aux_loss:0.0266 train_time:2786963ms step_avg:538.86ms step:5183/6250 train_loss:2.2308 aux_loss:0.0267 train_time:2787495ms step_avg:538.85ms step:5184/6250 train_loss:2.0434 aux_loss:0.0268 train_time:2788029ms step_avg:538.85ms step:5185/6250 train_loss:2.1040 aux_loss:0.0270 train_time:2788561ms step_avg:538.85ms step:5186/6250 train_loss:2.0963 aux_loss:0.0264 train_time:2789093ms step_avg:538.85ms step:5187/6250 train_loss:2.2089 aux_loss:0.0271 train_time:2789622ms step_avg:538.85ms step:5188/6250 train_loss:2.1933 aux_loss:0.0269 train_time:2790153ms step_avg:538.85ms step:5189/6250 train_loss:2.1378 aux_loss:0.0268 train_time:2790683ms step_avg:538.85ms step:5190/6250 train_loss:2.1869 aux_loss:0.0270 train_time:2791214ms step_avg:538.84ms step:5191/6250 train_loss:2.0816 aux_loss:0.0270 train_time:2791741ms step_avg:538.84ms step:5192/6250 train_loss:2.1736 aux_loss:0.0264 train_time:2792271ms step_avg:538.84ms step:5193/6250 train_loss:2.2260 aux_loss:0.0274 train_time:2792802ms step_avg:538.84ms step:5194/6250 train_loss:2.0766 aux_loss:0.0270 train_time:2793333ms step_avg:538.84ms step:5195/6250 train_loss:2.0111 aux_loss:0.0266 train_time:2793866ms step_avg:538.84ms step:5196/6250 train_loss:2.1506 aux_loss:0.0267 train_time:2794401ms step_avg:538.84ms step:5197/6250 train_loss:2.2175 aux_loss:0.0270 train_time:2794933ms step_avg:538.83ms step:5198/6250 train_loss:2.0980 aux_loss:0.0266 train_time:2795471ms step_avg:538.83ms step:5199/6250 train_loss:2.1585 aux_loss:0.0271 train_time:2796007ms step_avg:538.83ms step:5200/6250 train_loss:2.1194 aux_loss:0.0269 train_time:2796539ms step_avg:538.83ms step:5201/6250 train_loss:2.1555 aux_loss:0.0266 train_time:2797073ms step_avg:538.83ms step:5202/6250 train_loss:2.2177 aux_loss:0.0268 train_time:2797608ms step_avg:538.83ms step:5203/6250 train_loss:2.1867 aux_loss:0.0272 train_time:2798141ms step_avg:538.83ms step:5204/6250 train_loss:2.1104 aux_loss:0.0266 train_time:2798674ms step_avg:538.83ms step:5205/6250 train_loss:2.1399 aux_loss:0.0265 train_time:2799207ms step_avg:538.83ms step:5206/6250 train_loss:2.0823 aux_loss:0.0269 train_time:2799739ms step_avg:538.83ms step:5207/6250 train_loss:2.3158 aux_loss:0.0273 train_time:2800272ms step_avg:538.82ms step:5208/6250 train_loss:2.2077 aux_loss:0.0265 train_time:2800812ms step_avg:538.82ms step:5209/6250 train_loss:2.2516 aux_loss:0.0265 train_time:2801347ms step_avg:538.82ms step:5210/6250 train_loss:2.1866 aux_loss:0.0267 train_time:2801881ms step_avg:538.82ms step:5211/6250 train_loss:2.1336 aux_loss:0.0267 train_time:2802416ms step_avg:538.82ms step:5212/6250 train_loss:2.1572 aux_loss:0.0264 train_time:2802952ms step_avg:538.82ms step:5213/6250 train_loss:2.0507 aux_loss:0.0270 train_time:2803486ms step_avg:538.82ms step:5214/6250 train_loss:2.1153 aux_loss:0.0269 train_time:2804020ms step_avg:538.82ms step:5215/6250 train_loss:2.0418 aux_loss:0.0262 train_time:2804554ms step_avg:538.82ms step:5216/6250 train_loss:2.1491 aux_loss:0.0260 train_time:2805090ms step_avg:538.82ms step:5217/6250 train_loss:2.1156 aux_loss:0.0268 train_time:2805625ms step_avg:538.82ms step:5218/6250 train_loss:2.1057 aux_loss:0.0272 train_time:2806160ms step_avg:538.82ms step:5219/6250 train_loss:2.1083 aux_loss:0.0267 train_time:2806693ms step_avg:538.82ms step:5220/6250 train_loss:2.1219 aux_loss:0.0268 train_time:2807224ms step_avg:538.81ms step:5221/6250 train_loss:2.1740 aux_loss:0.0272 train_time:2807761ms step_avg:538.81ms step:5222/6250 train_loss:2.1819 aux_loss:0.0269 train_time:2808300ms step_avg:538.81ms step:5223/6250 train_loss:2.2090 aux_loss:0.0264 train_time:2808838ms step_avg:538.81ms step:5224/6250 train_loss:2.1304 aux_loss:0.0267 train_time:2809374ms step_avg:538.81ms step:5225/6250 train_loss:2.1001 aux_loss:0.0265 train_time:2809909ms step_avg:538.81ms step:5226/6250 train_loss:2.1619 aux_loss:0.0263 train_time:2810441ms step_avg:538.81ms step:5227/6250 train_loss:2.2749 aux_loss:0.0269 train_time:2810973ms step_avg:538.81ms step:5228/6250 train_loss:2.0984 aux_loss:0.0267 train_time:2811507ms step_avg:538.81ms step:5229/6250 train_loss:2.2108 aux_loss:0.0264 train_time:2812042ms step_avg:538.81ms step:5230/6250 train_loss:2.1097 aux_loss:0.0267 train_time:2812578ms step_avg:538.81ms step:5231/6250 train_loss:2.1542 aux_loss:0.0269 train_time:2813109ms step_avg:538.81ms step:5232/6250 train_loss:2.1218 aux_loss:0.0269 train_time:2813642ms step_avg:538.81ms step:5233/6250 train_loss:2.1228 aux_loss:0.0265 train_time:2814172ms step_avg:538.80ms step:5234/6250 train_loss:2.1836 aux_loss:0.0269 train_time:2814705ms step_avg:538.80ms step:5235/6250 train_loss:2.1921 aux_loss:0.0267 train_time:2815234ms step_avg:538.80ms step:5236/6250 train_loss:2.2108 aux_loss:0.0265 train_time:2815766ms step_avg:538.80ms step:5237/6250 train_loss:2.0938 aux_loss:0.0269 train_time:2816296ms step_avg:538.80ms step:5238/6250 train_loss:2.1820 aux_loss:0.0265 train_time:2816829ms step_avg:538.80ms step:5239/6250 train_loss:2.1346 aux_loss:0.0258 train_time:2817365ms step_avg:538.80ms step:5240/6250 train_loss:2.0886 aux_loss:0.0266 train_time:2817899ms step_avg:538.80ms step:5241/6250 train_loss:2.0529 aux_loss:0.0264 train_time:2818432ms step_avg:538.79ms step:5242/6250 train_loss:2.1482 aux_loss:0.0261 train_time:2818967ms step_avg:538.79ms step:5243/6250 train_loss:2.1569 aux_loss:0.0263 train_time:2819501ms step_avg:538.79ms step:5244/6250 train_loss:2.1523 aux_loss:0.0268 train_time:2820035ms step_avg:538.79ms step:5245/6250 train_loss:2.1466 aux_loss:0.0265 train_time:2820569ms step_avg:538.79ms step:5246/6250 train_loss:2.1870 aux_loss:0.0262 train_time:2821099ms step_avg:538.79ms step:5247/6250 train_loss:2.1451 aux_loss:0.0267 train_time:2821632ms step_avg:538.79ms step:5248/6250 train_loss:2.1153 aux_loss:0.0265 train_time:2822163ms step_avg:538.79ms step:5249/6250 train_loss:2.1151 aux_loss:0.0263 train_time:2822702ms step_avg:538.79ms step:5250/6250 train_loss:2.1459 aux_loss:0.0265 train_time:2823233ms step_avg:538.78ms step:5251/6250 train_loss:2.2369 aux_loss:0.0267 train_time:2823765ms step_avg:538.78ms step:5252/6250 train_loss:2.1563 aux_loss:0.0261 train_time:2824295ms step_avg:538.78ms step:5253/6250 train_loss:2.1134 aux_loss:0.0263 train_time:2824826ms step_avg:538.78ms step:5254/6250 train_loss:2.0829 aux_loss:0.0264 train_time:2825356ms step_avg:538.78ms step:5255/6250 train_loss:2.1767 aux_loss:0.0263 train_time:2825886ms step_avg:538.78ms step:5256/6250 train_loss:2.1377 aux_loss:0.0266 train_time:2826420ms step_avg:538.78ms step:5257/6250 train_loss:2.0788 aux_loss:0.0265 train_time:2826957ms step_avg:538.78ms step:5258/6250 train_loss:2.2238 aux_loss:0.0262 train_time:2827488ms step_avg:538.77ms step:5259/6250 train_loss:2.2149 aux_loss:0.0267 train_time:2828022ms step_avg:538.77ms step:5260/6250 train_loss:2.1370 aux_loss:0.0267 train_time:2828557ms step_avg:538.77ms step:5261/6250 train_loss:2.1096 aux_loss:0.0259 train_time:2829091ms step_avg:538.77ms step:5262/6250 train_loss:2.2064 aux_loss:0.0269 train_time:2829642ms step_avg:538.77ms step:5263/6250 train_loss:2.1674 aux_loss:0.0265 train_time:2830181ms step_avg:538.77ms step:5264/6250 train_loss:2.1462 aux_loss:0.0260 train_time:2830738ms step_avg:538.78ms step:5265/6250 train_loss:2.1970 aux_loss:0.0265 train_time:2831304ms step_avg:538.78ms step:5266/6250 train_loss:2.1866 aux_loss:0.0271 train_time:2831845ms step_avg:538.78ms step:5267/6250 train_loss:2.1830 aux_loss:0.0265 train_time:2832380ms step_avg:538.78ms step:5268/6250 train_loss:2.2316 aux_loss:0.0266 train_time:2832922ms step_avg:538.78ms step:5269/6250 train_loss:2.1738 aux_loss:0.0269 train_time:2833458ms step_avg:538.78ms step:5270/6250 train_loss:2.1254 aux_loss:0.0266 train_time:2833998ms step_avg:538.78ms step:5271/6250 train_loss:2.1591 aux_loss:0.0261 train_time:2834546ms step_avg:538.78ms step:5272/6250 train_loss:2.1336 aux_loss:0.0268 train_time:2835083ms step_avg:538.78ms step:5273/6250 train_loss:2.1904 aux_loss:0.0272 train_time:2835619ms step_avg:538.78ms step:5274/6250 train_loss:2.1737 aux_loss:0.0268 train_time:2836157ms step_avg:538.78ms step:5275/6250 train_loss:2.1953 aux_loss:0.0267 train_time:2836693ms step_avg:538.78ms step:5276/6250 train_loss:2.1210 aux_loss:0.0265 train_time:2837227ms step_avg:538.78ms step:5277/6250 train_loss:2.0928 aux_loss:0.0266 train_time:2837768ms step_avg:538.78ms step:5278/6250 train_loss:2.2272 aux_loss:0.0268 train_time:2838322ms step_avg:538.79ms step:5279/6250 train_loss:2.1746 aux_loss:0.0266 train_time:2838866ms step_avg:538.79ms step:5280/6250 train_loss:2.0855 aux_loss:0.0266 train_time:2839414ms step_avg:538.79ms step:5281/6250 train_loss:2.1572 aux_loss:0.0268 train_time:2839961ms step_avg:538.79ms step:5282/6250 train_loss:2.1353 aux_loss:0.0265 train_time:2840497ms step_avg:538.79ms step:5283/6250 train_loss:2.1490 aux_loss:0.0264 train_time:2841037ms step_avg:538.79ms step:5284/6250 train_loss:2.1277 aux_loss:0.0270 train_time:2841575ms step_avg:538.79ms step:5285/6250 train_loss:2.1081 aux_loss:0.0268 train_time:2842110ms step_avg:538.79ms step:5286/6250 train_loss:2.2146 aux_loss:0.0263 train_time:2842655ms step_avg:538.79ms step:5287/6250 train_loss:2.1346 aux_loss:0.0267 train_time:2843203ms step_avg:538.79ms step:5288/6250 train_loss:2.2042 aux_loss:0.0266 train_time:2843740ms step_avg:538.79ms step:5289/6250 train_loss:2.3395 aux_loss:0.0265 train_time:2844279ms step_avg:538.79ms step:5290/6250 train_loss:2.2238 aux_loss:0.0269 train_time:2844820ms step_avg:538.79ms step:5291/6250 train_loss:2.1497 aux_loss:0.0268 train_time:2845357ms step_avg:538.79ms step:5292/6250 train_loss:2.0670 aux_loss:0.0264 train_time:2845894ms step_avg:538.79ms step:5293/6250 train_loss:2.1625 aux_loss:0.0263 train_time:2846429ms step_avg:538.79ms step:5294/6250 train_loss:2.2377 aux_loss:0.0266 train_time:2846965ms step_avg:538.79ms step:5295/6250 train_loss:2.0746 aux_loss:0.0263 train_time:2847502ms step_avg:538.79ms step:5296/6250 train_loss:2.1814 aux_loss:0.0267 train_time:2848039ms step_avg:538.79ms step:5297/6250 train_loss:2.1815 aux_loss:0.0269 train_time:2848578ms step_avg:538.79ms step:5298/6250 train_loss:2.2255 aux_loss:0.0268 train_time:2849117ms step_avg:538.79ms step:5299/6250 train_loss:2.2071 aux_loss:0.0263 train_time:2849663ms step_avg:538.79ms step:5300/6250 train_loss:2.1709 aux_loss:0.0261 train_time:2850203ms step_avg:538.79ms step:5301/6250 train_loss:2.1068 aux_loss:0.0268 train_time:2850740ms step_avg:538.79ms step:5302/6250 train_loss:2.1181 aux_loss:0.0267 train_time:2851279ms step_avg:538.79ms step:5303/6250 train_loss:2.2394 aux_loss:0.0266 train_time:2851822ms step_avg:538.79ms step:5304/6250 train_loss:2.1408 aux_loss:0.0263 train_time:2852357ms step_avg:538.79ms step:5305/6250 train_loss:2.2208 aux_loss:0.0264 train_time:2852895ms step_avg:538.79ms step:5306/6250 train_loss:2.1937 aux_loss:0.0273 train_time:2853432ms step_avg:538.79ms step:5307/6250 train_loss:2.0849 aux_loss:0.0263 train_time:2853968ms step_avg:538.79ms step:5308/6250 train_loss:2.0816 aux_loss:0.0263 train_time:2854502ms step_avg:538.79ms step:5309/6250 train_loss:2.1146 aux_loss:0.0265 train_time:2855038ms step_avg:538.79ms step:5310/6250 train_loss:2.2029 aux_loss:0.0266 train_time:2855572ms step_avg:538.79ms step:5311/6250 train_loss:2.1306 aux_loss:0.0263 train_time:2856110ms step_avg:538.79ms step:5312/6250 train_loss:2.1599 aux_loss:0.0267 train_time:2856648ms step_avg:538.79ms step:5313/6250 train_loss:2.1627 aux_loss:0.0263 train_time:2857184ms step_avg:538.79ms step:5314/6250 train_loss:2.1609 aux_loss:0.0260 train_time:2857720ms step_avg:538.79ms step:5315/6250 train_loss:2.1154 aux_loss:0.0267 train_time:2858258ms step_avg:538.79ms step:5316/6250 train_loss:2.1316 aux_loss:0.0267 train_time:2858797ms step_avg:538.79ms step:5317/6250 train_loss:2.1153 aux_loss:0.0264 train_time:2859334ms step_avg:538.79ms step:5318/6250 train_loss:2.2268 aux_loss:0.0264 train_time:2859870ms step_avg:538.78ms step:5319/6250 train_loss:2.0776 aux_loss:0.0268 train_time:2860407ms step_avg:538.78ms step:5320/6250 train_loss:2.1133 aux_loss:0.0262 train_time:2860942ms step_avg:538.78ms step:5321/6250 train_loss:2.1969 aux_loss:0.0268 train_time:2861482ms step_avg:538.78ms step:5322/6250 train_loss:2.1321 aux_loss:0.0265 train_time:2862016ms step_avg:538.78ms step:5323/6250 train_loss:2.1717 aux_loss:0.0266 train_time:2862552ms step_avg:538.78ms step:5324/6250 train_loss:2.2024 aux_loss:0.0261 train_time:2863100ms step_avg:538.78ms step:5325/6250 train_loss:2.1099 aux_loss:0.0266 train_time:2863635ms step_avg:538.78ms step:5326/6250 train_loss:2.2180 aux_loss:0.0268 train_time:2864172ms step_avg:538.78ms step:5327/6250 train_loss:2.1010 aux_loss:0.0260 train_time:2864710ms step_avg:538.78ms step:5328/6250 train_loss:2.1551 aux_loss:0.0267 train_time:2865250ms step_avg:538.78ms step:5329/6250 train_loss:2.1521 aux_loss:0.0262 train_time:2865788ms step_avg:538.78ms step:5330/6250 train_loss:2.1426 aux_loss:0.0261 train_time:2866324ms step_avg:538.78ms step:5331/6250 train_loss:2.1354 aux_loss:0.0266 train_time:2866861ms step_avg:538.78ms step:5332/6250 train_loss:2.1355 aux_loss:0.0266 train_time:2867399ms step_avg:538.78ms step:5333/6250 train_loss:2.1655 aux_loss:0.0264 train_time:2867939ms step_avg:538.78ms step:5334/6250 train_loss:2.1354 aux_loss:0.0266 train_time:2868474ms step_avg:538.78ms step:5335/6250 train_loss:2.1526 aux_loss:0.0265 train_time:2869017ms step_avg:538.78ms step:5336/6250 train_loss:2.1496 aux_loss:0.0261 train_time:2869562ms step_avg:538.78ms step:5337/6250 train_loss:2.1610 aux_loss:0.0261 train_time:2870099ms step_avg:538.78ms step:5338/6250 train_loss:2.1938 aux_loss:0.0269 train_time:2870636ms step_avg:538.78ms step:5339/6250 train_loss:2.1468 aux_loss:0.0266 train_time:2871177ms step_avg:538.78ms step:5340/6250 train_loss:2.2180 aux_loss:0.0261 train_time:2871713ms step_avg:538.78ms step:5341/6250 train_loss:2.1923 aux_loss:0.0267 train_time:2872259ms step_avg:538.78ms step:5342/6250 train_loss:2.1614 aux_loss:0.0268 train_time:2872799ms step_avg:538.78ms step:5343/6250 train_loss:2.1582 aux_loss:0.0264 train_time:2873334ms step_avg:538.78ms step:5344/6250 train_loss:2.1630 aux_loss:0.0268 train_time:2873870ms step_avg:538.78ms step:5345/6250 train_loss:2.1570 aux_loss:0.0269 train_time:2874409ms step_avg:538.78ms step:5346/6250 train_loss:2.1679 aux_loss:0.0268 train_time:2874989ms step_avg:538.79ms step:5347/6250 train_loss:2.0912 aux_loss:0.0263 train_time:2875528ms step_avg:538.79ms step:5348/6250 train_loss:2.1666 aux_loss:0.0266 train_time:2876071ms step_avg:538.79ms step:5349/6250 train_loss:2.0986 aux_loss:0.0265 train_time:2876612ms step_avg:538.79ms step:5350/6250 train_loss:2.1861 aux_loss:0.0266 train_time:2877159ms step_avg:538.79ms step:5351/6250 train_loss:2.1087 aux_loss:0.0268 train_time:2877704ms step_avg:538.79ms step:5352/6250 train_loss:2.1768 aux_loss:0.0268 train_time:2878244ms step_avg:538.80ms step:5353/6250 train_loss:2.0933 aux_loss:0.0263 train_time:2878782ms step_avg:538.79ms step:5354/6250 train_loss:2.1622 aux_loss:0.0269 train_time:2879318ms step_avg:538.79ms step:5355/6250 train_loss:2.0472 aux_loss:0.0263 train_time:2879854ms step_avg:538.79ms step:5356/6250 train_loss:2.1701 aux_loss:0.0262 train_time:2880391ms step_avg:538.79ms step:5357/6250 train_loss:2.1591 aux_loss:0.0270 train_time:2880930ms step_avg:538.79ms step:5358/6250 train_loss:2.0981 aux_loss:0.0265 train_time:2881466ms step_avg:538.79ms step:5359/6250 train_loss:2.1175 aux_loss:0.0257 train_time:2882002ms step_avg:538.79ms step:5360/6250 train_loss:2.1660 aux_loss:0.0263 train_time:2882538ms step_avg:538.79ms step:5361/6250 train_loss:2.0991 aux_loss:0.0262 train_time:2883074ms step_avg:538.79ms step:5362/6250 train_loss:2.2037 aux_loss:0.0265 train_time:2883609ms step_avg:538.79ms step:5363/6250 train_loss:2.1428 aux_loss:0.0266 train_time:2884146ms step_avg:538.79ms step:5364/6250 train_loss:2.0710 aux_loss:0.0262 train_time:2884680ms step_avg:538.79ms step:5365/6250 train_loss:2.2239 aux_loss:0.0261 train_time:2885228ms step_avg:538.79ms step:5366/6250 train_loss:2.1334 aux_loss:0.0255 train_time:2885765ms step_avg:538.79ms step:5367/6250 train_loss:2.2334 aux_loss:0.0262 train_time:2886303ms step_avg:538.79ms step:5368/6250 train_loss:2.1160 aux_loss:0.0260 train_time:2886842ms step_avg:538.79ms step:5369/6250 train_loss:2.1490 aux_loss:0.0255 train_time:2887384ms step_avg:538.79ms step:5370/6250 train_loss:2.1434 aux_loss:0.0263 train_time:2887921ms step_avg:538.79ms step:5371/6250 train_loss:2.0985 aux_loss:0.0265 train_time:2888462ms step_avg:538.79ms step:5372/6250 train_loss:2.1298 aux_loss:0.0260 train_time:2888999ms step_avg:538.79ms step:5373/6250 train_loss:2.0573 aux_loss:0.0259 train_time:2889558ms step_avg:538.79ms step:5374/6250 train_loss:2.1525 aux_loss:0.0263 train_time:2890096ms step_avg:538.79ms step:5375/6250 train_loss:2.1525 aux_loss:0.0263 train_time:2890632ms step_avg:538.79ms step:5376/6250 train_loss:2.1626 aux_loss:0.0257 train_time:2891169ms step_avg:538.79ms step:5377/6250 train_loss:2.1578 aux_loss:0.0259 train_time:2891708ms step_avg:538.79ms step:5378/6250 train_loss:2.2106 aux_loss:0.0265 train_time:2892258ms step_avg:538.80ms step:5379/6250 train_loss:2.1291 aux_loss:0.0265 train_time:2892797ms step_avg:538.80ms step:5380/6250 train_loss:2.1090 aux_loss:0.0262 train_time:2893335ms step_avg:538.80ms step:5381/6250 train_loss:2.0631 aux_loss:0.0258 train_time:2893871ms step_avg:538.80ms step:5382/6250 train_loss:2.1139 aux_loss:0.0261 train_time:2894407ms step_avg:538.80ms step:5383/6250 train_loss:2.2273 aux_loss:0.0260 train_time:2894945ms step_avg:538.79ms step:5384/6250 train_loss:2.1282 aux_loss:0.0257 train_time:2895479ms step_avg:538.79ms step:5385/6250 train_loss:2.1672 aux_loss:0.0262 train_time:2896014ms step_avg:538.79ms step:5386/6250 train_loss:2.0531 aux_loss:0.0259 train_time:2896551ms step_avg:538.79ms step:5387/6250 train_loss:2.1508 aux_loss:0.0258 train_time:2897086ms step_avg:538.79ms step:5388/6250 train_loss:2.2163 aux_loss:0.0267 train_time:2897632ms step_avg:538.79ms step:5389/6250 train_loss:2.2135 aux_loss:0.0266 train_time:2898196ms step_avg:538.80ms step:5390/6250 train_loss:2.0957 aux_loss:0.0255 train_time:2898736ms step_avg:538.80ms step:5391/6250 train_loss:2.1526 aux_loss:0.0259 train_time:2899271ms step_avg:538.80ms step:5392/6250 train_loss:2.1275 aux_loss:0.0268 train_time:2899807ms step_avg:538.80ms step:5393/6250 train_loss:2.1405 aux_loss:0.0261 train_time:2900339ms step_avg:538.80ms step:5394/6250 train_loss:2.0786 aux_loss:0.0262 train_time:2900874ms step_avg:538.80ms step:5395/6250 train_loss:2.2730 aux_loss:0.0267 train_time:2901410ms step_avg:538.79ms step:5396/6250 train_loss:2.1377 aux_loss:0.0262 train_time:2901946ms step_avg:538.79ms step:5397/6250 train_loss:2.1383 aux_loss:0.0261 train_time:2902481ms step_avg:538.79ms step:5398/6250 train_loss:2.1700 aux_loss:0.0263 train_time:2903016ms step_avg:538.79ms step:5399/6250 train_loss:2.1121 aux_loss:0.0258 train_time:2903551ms step_avg:538.79ms step:5400/6250 train_loss:2.2070 aux_loss:0.0264 train_time:2904088ms step_avg:538.79ms step:5401/6250 train_loss:2.1272 aux_loss:0.0266 train_time:2904629ms step_avg:538.79ms step:5402/6250 train_loss:2.1027 aux_loss:0.0258 train_time:2905165ms step_avg:538.79ms step:5403/6250 train_loss:2.1338 aux_loss:0.0260 train_time:2905700ms step_avg:538.79ms step:5404/6250 train_loss:2.2838 aux_loss:0.0264 train_time:2906235ms step_avg:538.79ms step:5405/6250 train_loss:2.0995 aux_loss:0.0260 train_time:2906771ms step_avg:538.79ms step:5406/6250 train_loss:2.1668 aux_loss:0.0268 train_time:2907307ms step_avg:538.79ms step:5407/6250 train_loss:2.0487 aux_loss:0.0264 train_time:2907866ms step_avg:538.79ms step:5408/6250 train_loss:2.1625 aux_loss:0.0260 train_time:2908422ms step_avg:538.80ms step:5409/6250 train_loss:2.2186 aux_loss:0.0269 train_time:2908966ms step_avg:538.80ms step:5410/6250 train_loss:2.0744 aux_loss:0.0266 train_time:2909502ms step_avg:538.80ms step:5411/6250 train_loss:2.1391 aux_loss:0.0264 train_time:2910037ms step_avg:538.80ms step:5412/6250 train_loss:2.1681 aux_loss:0.0268 train_time:2910572ms step_avg:538.80ms step:5413/6250 train_loss:2.0280 aux_loss:0.0267 train_time:2911107ms step_avg:538.79ms step:5414/6250 train_loss:2.2316 aux_loss:0.0269 train_time:2911642ms step_avg:538.79ms step:5415/6250 train_loss:2.1247 aux_loss:0.0267 train_time:2912178ms step_avg:538.79ms step:5416/6250 train_loss:2.0236 aux_loss:0.0267 train_time:2912713ms step_avg:538.79ms step:5417/6250 train_loss:2.2143 aux_loss:0.0266 train_time:2913248ms step_avg:538.79ms step:5418/6250 train_loss:2.0872 aux_loss:0.0265 train_time:2913785ms step_avg:538.79ms step:5419/6250 train_loss:2.1669 aux_loss:0.0265 train_time:2914324ms step_avg:538.79ms step:5420/6250 train_loss:2.1264 aux_loss:0.0262 train_time:2914861ms step_avg:538.79ms step:5421/6250 train_loss:2.1669 aux_loss:0.0262 train_time:2915396ms step_avg:538.79ms step:5422/6250 train_loss:2.0972 aux_loss:0.0267 train_time:2915933ms step_avg:538.79ms step:5423/6250 train_loss:2.2168 aux_loss:0.0270 train_time:2916469ms step_avg:538.79ms step:5424/6250 train_loss:2.1320 aux_loss:0.0267 train_time:2917005ms step_avg:538.79ms step:5425/6250 train_loss:2.1882 aux_loss:0.0271 train_time:2917541ms step_avg:538.79ms step:5426/6250 train_loss:2.2047 aux_loss:0.0262 train_time:2918078ms step_avg:538.79ms step:5427/6250 train_loss:2.1168 aux_loss:0.0267 train_time:2918614ms step_avg:538.79ms step:5428/6250 train_loss:2.1657 aux_loss:0.0263 train_time:2919149ms step_avg:538.79ms step:5429/6250 train_loss:2.1923 aux_loss:0.0261 train_time:2919686ms step_avg:538.79ms step:5430/6250 train_loss:2.1402 aux_loss:0.0265 train_time:2920225ms step_avg:538.79ms step:5431/6250 train_loss:2.0301 aux_loss:0.0263 train_time:2920763ms step_avg:538.79ms step:5432/6250 train_loss:2.1202 aux_loss:0.0263 train_time:2921303ms step_avg:538.79ms step:5433/6250 train_loss:2.0909 aux_loss:0.0260 train_time:2921839ms step_avg:538.79ms step:5434/6250 train_loss:2.2468 aux_loss:0.0268 train_time:2922374ms step_avg:538.79ms step:5435/6250 train_loss:2.1734 aux_loss:0.0266 train_time:2922909ms step_avg:538.79ms step:5436/6250 train_loss:2.1405 aux_loss:0.0263 train_time:2923445ms step_avg:538.78ms step:5437/6250 train_loss:2.1744 aux_loss:0.0263 train_time:2923980ms step_avg:538.78ms step:5438/6250 train_loss:2.1344 aux_loss:0.0261 train_time:2924521ms step_avg:538.78ms step:5439/6250 train_loss:2.0861 aux_loss:0.0258 train_time:2925056ms step_avg:538.78ms step:5440/6250 train_loss:2.1287 aux_loss:0.0266 train_time:2925591ms step_avg:538.78ms step:5441/6250 train_loss:2.1857 aux_loss:0.0266 train_time:2926126ms step_avg:538.78ms step:5442/6250 train_loss:2.1187 aux_loss:0.0264 train_time:2926660ms step_avg:538.78ms step:5443/6250 train_loss:2.0703 aux_loss:0.0262 train_time:2927194ms step_avg:538.78ms step:5444/6250 train_loss:2.1285 aux_loss:0.0262 train_time:2927728ms step_avg:538.78ms step:5445/6250 train_loss:2.1347 aux_loss:0.0264 train_time:2928262ms step_avg:538.78ms step:5446/6250 train_loss:2.1295 aux_loss:0.0261 train_time:2928797ms step_avg:538.78ms step:5447/6250 train_loss:2.1558 aux_loss:0.0266 train_time:2929331ms step_avg:538.78ms step:5448/6250 train_loss:2.1495 aux_loss:0.0262 train_time:2929867ms step_avg:538.78ms step:5449/6250 train_loss:2.1847 aux_loss:0.0261 train_time:2930400ms step_avg:538.78ms step:5450/6250 train_loss:2.2133 aux_loss:0.0266 train_time:2930935ms step_avg:538.77ms step:5451/6250 train_loss:2.1909 aux_loss:0.0263 train_time:2931471ms step_avg:538.77ms step:5452/6250 train_loss:2.1416 aux_loss:0.0262 train_time:2932007ms step_avg:538.77ms step:5453/6250 train_loss:2.1522 aux_loss:0.0264 train_time:2932543ms step_avg:538.77ms step:5454/6250 train_loss:2.1079 aux_loss:0.0261 train_time:2933078ms step_avg:538.77ms step:5455/6250 train_loss:2.1023 aux_loss:0.0267 train_time:2933614ms step_avg:538.77ms step:5456/6250 train_loss:2.0955 aux_loss:0.0264 train_time:2934150ms step_avg:538.77ms step:5457/6250 train_loss:2.1900 aux_loss:0.0265 train_time:2934684ms step_avg:538.77ms step:5458/6250 train_loss:2.0960 aux_loss:0.0259 train_time:2935218ms step_avg:538.77ms step:5459/6250 train_loss:2.1109 aux_loss:0.0262 train_time:2935753ms step_avg:538.77ms step:5460/6250 train_loss:2.1499 aux_loss:0.0263 train_time:2936287ms step_avg:538.77ms step:5461/6250 train_loss:2.1324 aux_loss:0.0270 train_time:2936825ms step_avg:538.77ms step:5462/6250 train_loss:2.1226 aux_loss:0.0261 train_time:2937361ms step_avg:538.77ms step:5463/6250 train_loss:2.1465 aux_loss:0.0262 train_time:2937896ms step_avg:538.77ms step:5464/6250 train_loss:2.1115 aux_loss:0.0262 train_time:2938432ms step_avg:538.77ms step:5465/6250 train_loss:2.1454 aux_loss:0.0264 train_time:2938967ms step_avg:538.77ms step:5466/6250 train_loss:2.0997 aux_loss:0.0260 train_time:2939503ms step_avg:538.77ms step:5467/6250 train_loss:2.2178 aux_loss:0.0268 train_time:2940038ms step_avg:538.76ms step:5468/6250 train_loss:2.0540 aux_loss:0.0262 train_time:2940574ms step_avg:538.76ms step:5469/6250 train_loss:2.1174 aux_loss:0.0259 train_time:2941109ms step_avg:538.76ms step:5470/6250 train_loss:2.1368 aux_loss:0.0261 train_time:2941643ms step_avg:538.76ms step:5471/6250 train_loss:2.1415 aux_loss:0.0260 train_time:2942179ms step_avg:538.76ms step:5472/6250 train_loss:2.1684 aux_loss:0.0266 train_time:2942719ms step_avg:538.76ms step:5473/6250 train_loss:2.2051 aux_loss:0.0264 train_time:2943254ms step_avg:538.76ms step:5474/6250 train_loss:2.1785 aux_loss:0.0264 train_time:2943790ms step_avg:538.76ms step:5475/6250 train_loss:2.1863 aux_loss:0.0264 train_time:2944325ms step_avg:538.76ms step:5476/6250 train_loss:2.1830 aux_loss:0.0259 train_time:2944862ms step_avg:538.76ms step:5477/6250 train_loss:2.1394 aux_loss:0.0262 train_time:2945398ms step_avg:538.76ms step:5478/6250 train_loss:2.2492 aux_loss:0.0260 train_time:2945933ms step_avg:538.76ms step:5479/6250 train_loss:2.1748 aux_loss:0.0260 train_time:2946480ms step_avg:538.76ms step:5480/6250 train_loss:2.1183 aux_loss:0.0268 train_time:2947015ms step_avg:538.76ms step:5481/6250 train_loss:2.1335 aux_loss:0.0261 train_time:2947550ms step_avg:538.76ms step:5482/6250 train_loss:2.1520 aux_loss:0.0263 train_time:2948086ms step_avg:538.76ms step:5483/6250 train_loss:2.1691 aux_loss:0.0256 train_time:2948621ms step_avg:538.76ms step:5484/6250 train_loss:2.0098 aux_loss:0.0255 train_time:2949156ms step_avg:538.76ms step:5485/6250 train_loss:2.0805 aux_loss:0.0260 train_time:2949693ms step_avg:538.76ms step:5486/6250 train_loss:2.0552 aux_loss:0.0259 train_time:2950228ms step_avg:538.76ms step:5487/6250 train_loss:2.2042 aux_loss:0.0260 train_time:2950763ms step_avg:538.76ms step:5488/6250 train_loss:2.0841 aux_loss:0.0254 train_time:2951299ms step_avg:538.75ms step:5489/6250 train_loss:2.0974 aux_loss:0.0259 train_time:2951836ms step_avg:538.75ms step:5490/6250 train_loss:2.1014 aux_loss:0.0257 train_time:2952372ms step_avg:538.75ms step:5491/6250 train_loss:2.1912 aux_loss:0.0257 train_time:2952914ms step_avg:538.75ms step:5492/6250 train_loss:2.1132 aux_loss:0.0267 train_time:2953449ms step_avg:538.75ms step:5493/6250 train_loss:2.1682 aux_loss:0.0264 train_time:2953985ms step_avg:538.75ms step:5494/6250 train_loss:2.1837 aux_loss:0.0263 train_time:2954521ms step_avg:538.75ms step:5495/6250 train_loss:2.0924 aux_loss:0.0265 train_time:2955057ms step_avg:538.75ms step:5496/6250 train_loss:2.1065 aux_loss:0.0258 train_time:2955594ms step_avg:538.75ms step:5497/6250 train_loss:2.1316 aux_loss:0.0257 train_time:2956128ms step_avg:538.75ms step:5498/6250 train_loss:2.1097 aux_loss:0.0262 train_time:2956670ms step_avg:538.75ms step:5499/6250 train_loss:2.0536 aux_loss:0.0265 train_time:2957205ms step_avg:538.75ms step:5500/6250 train_loss:2.2202 aux_loss:0.0264 train_time:2957743ms step_avg:538.75ms step:5501/6250 train_loss:2.2006 aux_loss:0.0264 train_time:2958277ms step_avg:538.75ms step:5502/6250 train_loss:2.1237 aux_loss:0.0261 train_time:2958818ms step_avg:538.75ms step:5503/6250 train_loss:2.2373 aux_loss:0.0260 train_time:2959369ms step_avg:538.75ms step:5504/6250 train_loss:2.1347 aux_loss:0.0259 train_time:2959918ms step_avg:538.75ms step:5505/6250 train_loss:2.0960 aux_loss:0.0256 train_time:2960459ms step_avg:538.76ms step:5506/6250 train_loss:2.1874 aux_loss:0.0259 train_time:2961022ms step_avg:538.76ms step:5507/6250 train_loss:2.1180 aux_loss:0.0263 train_time:2961562ms step_avg:538.76ms step:5508/6250 train_loss:2.1707 aux_loss:0.0259 train_time:2962106ms step_avg:538.76ms step:5509/6250 train_loss:2.2068 aux_loss:0.0259 train_time:2962656ms step_avg:538.76ms step:5510/6250 train_loss:2.0436 aux_loss:0.0263 train_time:2963193ms step_avg:538.76ms step:5511/6250 train_loss:2.2402 aux_loss:0.0258 train_time:2963731ms step_avg:538.76ms step:5512/6250 train_loss:2.1305 aux_loss:0.0263 train_time:2964302ms step_avg:538.77ms step:5513/6250 train_loss:2.0200 aux_loss:0.0257 train_time:2964856ms step_avg:538.77ms step:5514/6250 train_loss:2.2296 aux_loss:0.0262 train_time:2965410ms step_avg:538.77ms step:5515/6250 train_loss:2.1219 aux_loss:0.0265 train_time:2965957ms step_avg:538.78ms step:5516/6250 train_loss:2.1093 aux_loss:0.0259 train_time:2966516ms step_avg:538.78ms step:5517/6250 train_loss:2.2095 aux_loss:0.0259 train_time:2967057ms step_avg:538.78ms step:5518/6250 train_loss:2.0920 aux_loss:0.0264 train_time:2967590ms step_avg:538.78ms step:5519/6250 train_loss:2.2693 aux_loss:0.0263 train_time:2968122ms step_avg:538.78ms step:5520/6250 train_loss:2.1370 aux_loss:0.0258 train_time:2968652ms step_avg:538.78ms step:5521/6250 train_loss:2.1214 aux_loss:0.0266 train_time:2969182ms step_avg:538.77ms step:5522/6250 train_loss:2.1502 aux_loss:0.0263 train_time:2969714ms step_avg:538.77ms step:5523/6250 train_loss:2.1015 aux_loss:0.0255 train_time:2970244ms step_avg:538.77ms step:5524/6250 train_loss:2.1675 aux_loss:0.0261 train_time:2970776ms step_avg:538.77ms step:5525/6250 train_loss:2.1444 aux_loss:0.0267 train_time:2971305ms step_avg:538.77ms step:5526/6250 train_loss:2.1166 aux_loss:0.0259 train_time:2971838ms step_avg:538.77ms step:5527/6250 train_loss:2.1711 aux_loss:0.0257 train_time:2972368ms step_avg:538.77ms step:5528/6250 train_loss:2.2070 aux_loss:0.0260 train_time:2972937ms step_avg:538.77ms step:5529/6250 train_loss:2.0696 aux_loss:0.0261 train_time:2973490ms step_avg:538.77ms step:5530/6250 train_loss:2.0769 aux_loss:0.0258 train_time:2974022ms step_avg:538.77ms step:5531/6250 train_loss:2.0805 aux_loss:0.0259 train_time:2974552ms step_avg:538.77ms step:5532/6250 train_loss:2.1508 aux_loss:0.0260 train_time:2975085ms step_avg:538.77ms step:5533/6250 train_loss:2.1113 aux_loss:0.0263 train_time:2975615ms step_avg:538.77ms step:5534/6250 train_loss:2.1406 aux_loss:0.0257 train_time:2976144ms step_avg:538.77ms step:5535/6250 train_loss:2.1049 aux_loss:0.0255 train_time:2976676ms step_avg:538.76ms step:5536/6250 train_loss:2.1497 aux_loss:0.0263 train_time:2977213ms step_avg:538.76ms step:5537/6250 train_loss:2.1678 aux_loss:0.0261 train_time:2977745ms step_avg:538.76ms step:5538/6250 train_loss:2.1623 aux_loss:0.0260 train_time:2978275ms step_avg:538.76ms step:5539/6250 train_loss:2.1572 aux_loss:0.0260 train_time:2978812ms step_avg:538.76ms step:5540/6250 train_loss:2.2254 aux_loss:0.0263 train_time:2979357ms step_avg:538.76ms step:5541/6250 train_loss:2.1201 aux_loss:0.0262 train_time:2979921ms step_avg:538.77ms step:5542/6250 train_loss:2.1548 aux_loss:0.0260 train_time:2980477ms step_avg:538.77ms step:5543/6250 train_loss:2.1435 aux_loss:0.0257 train_time:2981009ms step_avg:538.77ms step:5544/6250 train_loss:2.1480 aux_loss:0.0263 train_time:2981541ms step_avg:538.77ms step:5545/6250 train_loss:2.2148 aux_loss:0.0266 train_time:2982081ms step_avg:538.77ms step:5546/6250 train_loss:2.1489 aux_loss:0.0261 train_time:2982612ms step_avg:538.77ms step:5547/6250 train_loss:2.1750 aux_loss:0.0267 train_time:2983142ms step_avg:538.77ms step:5548/6250 train_loss:2.1572 aux_loss:0.0263 train_time:2983674ms step_avg:538.76ms step:5549/6250 train_loss:2.1792 aux_loss:0.0255 train_time:2984224ms step_avg:538.77ms step:5550/6250 train_loss:2.1262 aux_loss:0.0263 train_time:2984791ms step_avg:538.77ms step:5551/6250 train_loss:2.2728 aux_loss:0.0267 train_time:2985345ms step_avg:538.77ms step:5552/6250 train_loss:2.1374 aux_loss:0.0257 train_time:2985902ms step_avg:538.78ms step:5553/6250 train_loss:2.1257 aux_loss:0.0261 train_time:2986453ms step_avg:538.78ms step:5554/6250 train_loss:2.1621 aux_loss:0.0262 train_time:2987007ms step_avg:538.78ms step:5555/6250 train_loss:2.0966 aux_loss:0.0262 train_time:2987560ms step_avg:538.78ms step:5556/6250 train_loss:2.1245 aux_loss:0.0260 train_time:2988115ms step_avg:538.79ms step:5557/6250 train_loss:2.2984 aux_loss:0.0265 train_time:2988648ms step_avg:538.79ms step:5558/6250 train_loss:2.1658 aux_loss:0.0270 train_time:2989183ms step_avg:538.79ms step:5559/6250 train_loss:2.1142 aux_loss:0.0264 train_time:2989762ms step_avg:538.79ms step:5560/6250 train_loss:2.2046 aux_loss:0.0261 train_time:2990303ms step_avg:538.79ms step:5561/6250 train_loss:2.1435 aux_loss:0.0260 train_time:2990833ms step_avg:538.79ms step:5562/6250 train_loss:2.1437 aux_loss:0.0262 train_time:2991364ms step_avg:538.79ms step:5563/6250 train_loss:2.2465 aux_loss:0.0264 train_time:2991896ms step_avg:538.79ms step:5564/6250 train_loss:2.1833 aux_loss:0.0264 train_time:2992443ms step_avg:538.79ms step:5565/6250 train_loss:2.1469 aux_loss:0.0269 train_time:2992996ms step_avg:538.79ms step:5566/6250 train_loss:2.1668 aux_loss:0.0261 train_time:2993542ms step_avg:538.79ms step:5567/6250 train_loss:2.1969 aux_loss:0.0262 train_time:2994073ms step_avg:538.79ms step:5568/6250 train_loss:2.1277 aux_loss:0.0259 train_time:2994602ms step_avg:538.79ms step:5569/6250 train_loss:2.1277 aux_loss:0.0262 train_time:2995151ms step_avg:538.79ms step:5570/6250 train_loss:2.1175 aux_loss:0.0258 train_time:2995708ms step_avg:538.80ms step:5571/6250 train_loss:2.1154 aux_loss:0.0260 train_time:2996258ms step_avg:538.80ms step:5572/6250 train_loss:2.1568 aux_loss:0.0261 train_time:2996824ms step_avg:538.80ms step:5573/6250 train_loss:2.1750 aux_loss:0.0261 train_time:2997385ms step_avg:538.81ms step:5574/6250 train_loss:2.2121 aux_loss:0.0269 train_time:2997915ms step_avg:538.81ms step:5575/6250 train_loss:2.1090 aux_loss:0.0263 train_time:2998445ms step_avg:538.80ms step:5576/6250 train_loss:2.1027 aux_loss:0.0263 train_time:2998976ms step_avg:538.80ms step:5577/6250 train_loss:2.1542 aux_loss:0.0264 train_time:2999508ms step_avg:538.80ms step:5578/6250 train_loss:2.1720 aux_loss:0.0263 train_time:3000058ms step_avg:538.80ms step:5579/6250 train_loss:2.0967 aux_loss:0.0257 train_time:3000591ms step_avg:538.80ms step:5580/6250 train_loss:2.2194 aux_loss:0.0263 train_time:3001121ms step_avg:538.80ms step:5581/6250 train_loss:2.1782 aux_loss:0.0268 train_time:3001664ms step_avg:538.80ms step:5582/6250 train_loss:2.1994 aux_loss:0.0266 train_time:3002213ms step_avg:538.80ms step:5583/6250 train_loss:2.0473 aux_loss:0.0263 train_time:3002783ms step_avg:538.81ms step:5584/6250 train_loss:2.2357 aux_loss:0.0260 train_time:3003341ms step_avg:538.81ms step:5585/6250 train_loss:2.1150 aux_loss:0.0259 train_time:3003888ms step_avg:538.81ms step:5586/6250 train_loss:2.1130 aux_loss:0.0264 train_time:3004455ms step_avg:538.82ms step:5587/6250 train_loss:2.1300 aux_loss:0.0256 train_time:3005023ms step_avg:538.82ms step:5588/6250 train_loss:2.1731 aux_loss:0.0256 train_time:3005593ms step_avg:538.83ms step:5589/6250 train_loss:2.0795 aux_loss:0.0261 train_time:3006144ms step_avg:538.83ms step:5590/6250 train_loss:2.1910 aux_loss:0.0259 train_time:3006687ms step_avg:538.83ms step:5591/6250 train_loss:2.0967 aux_loss:0.0259 train_time:3007230ms step_avg:538.83ms step:5592/6250 train_loss:2.1780 aux_loss:0.0260 train_time:3007774ms step_avg:538.83ms step:5593/6250 train_loss:2.2017 aux_loss:0.0259 train_time:3008320ms step_avg:538.84ms step:5594/6250 train_loss:2.1309 aux_loss:0.0262 train_time:3008866ms step_avg:538.84ms step:5595/6250 train_loss:2.1420 aux_loss:0.0258 train_time:3009411ms step_avg:538.84ms step:5596/6250 train_loss:2.1382 aux_loss:0.0252 train_time:3009977ms step_avg:538.84ms step:5597/6250 train_loss:2.1471 aux_loss:0.0261 train_time:3010529ms step_avg:538.85ms step:5598/6250 train_loss:2.1924 aux_loss:0.0265 train_time:3011092ms step_avg:538.85ms step:5599/6250 train_loss:2.1336 aux_loss:0.0257 train_time:3011625ms step_avg:538.85ms step:5600/6250 train_loss:2.0973 aux_loss:0.0259 train_time:3012176ms step_avg:538.85ms step:5601/6250 train_loss:2.0634 aux_loss:0.0261 train_time:3012706ms step_avg:538.85ms step:5602/6250 train_loss:2.0857 aux_loss:0.0260 train_time:3013236ms step_avg:538.85ms step:5603/6250 train_loss:2.0980 aux_loss:0.0257 train_time:3013791ms step_avg:538.85ms step:5604/6250 train_loss:2.2198 aux_loss:0.0268 train_time:3014339ms step_avg:538.85ms step:5605/6250 train_loss:2.0820 aux_loss:0.0264 train_time:3014898ms step_avg:538.86ms step:5606/6250 train_loss:2.0982 aux_loss:0.0259 train_time:3015441ms step_avg:538.86ms step:5607/6250 train_loss:2.1580 aux_loss:0.0262 train_time:3015989ms step_avg:538.86ms step:5608/6250 train_loss:2.1852 aux_loss:0.0267 train_time:3016547ms step_avg:538.86ms step:5609/6250 train_loss:2.1570 aux_loss:0.0262 train_time:3017085ms step_avg:538.86ms step:5610/6250 train_loss:2.1308 aux_loss:0.0253 train_time:3017621ms step_avg:538.86ms step:5611/6250 train_loss:2.1443 aux_loss:0.0259 train_time:3018183ms step_avg:538.86ms step:5612/6250 train_loss:2.1314 aux_loss:0.0263 train_time:3018741ms step_avg:538.87ms step:5613/6250 train_loss:2.1356 aux_loss:0.0262 train_time:3019279ms step_avg:538.87ms step:5614/6250 train_loss:2.0492 aux_loss:0.0254 train_time:3019822ms step_avg:538.87ms step:5615/6250 train_loss:2.1013 aux_loss:0.0259 train_time:3020371ms step_avg:538.87ms step:5616/6250 train_loss:2.0795 aux_loss:0.0258 train_time:3020903ms step_avg:538.87ms step:5617/6250 train_loss:2.1999 aux_loss:0.0257 train_time:3021469ms step_avg:538.87ms step:5618/6250 train_loss:2.1155 aux_loss:0.0257 train_time:3022027ms step_avg:538.88ms step:5619/6250 train_loss:2.1970 aux_loss:0.0256 train_time:3022586ms step_avg:538.88ms step:5620/6250 train_loss:2.1620 aux_loss:0.0256 train_time:3023142ms step_avg:538.88ms step:5621/6250 train_loss:2.1524 aux_loss:0.0260 train_time:3023673ms step_avg:538.88ms step:5622/6250 train_loss:2.1825 aux_loss:0.0266 train_time:3024221ms step_avg:538.88ms step:5623/6250 train_loss:2.1392 aux_loss:0.0259 train_time:3024752ms step_avg:538.88ms step:5624/6250 train_loss:2.0791 aux_loss:0.0256 train_time:3025283ms step_avg:538.88ms step:5625/6250 train_loss:2.1043 aux_loss:0.0261 train_time:3025827ms step_avg:538.88ms step:5626/6250 train_loss:2.1654 aux_loss:0.0261 train_time:3026377ms step_avg:538.88ms step:5627/6250 train_loss:2.1682 aux_loss:0.0254 train_time:3026923ms step_avg:538.89ms step:5628/6250 train_loss:2.1198 aux_loss:0.0259 train_time:3027472ms step_avg:538.89ms step:5629/6250 train_loss:2.1877 aux_loss:0.0263 train_time:3028006ms step_avg:538.89ms step:5630/6250 train_loss:2.1879 aux_loss:0.0263 train_time:3028536ms step_avg:538.89ms step:5631/6250 train_loss:2.1757 aux_loss:0.0261 train_time:3029080ms step_avg:538.89ms step:5632/6250 train_loss:2.0124 aux_loss:0.0264 train_time:3029631ms step_avg:538.89ms step:5633/6250 train_loss:2.0642 aux_loss:0.0262 train_time:3030180ms step_avg:538.89ms step:5634/6250 train_loss:2.1089 aux_loss:0.0256 train_time:3030717ms step_avg:538.89ms step:5635/6250 train_loss:2.1075 aux_loss:0.0258 train_time:3031275ms step_avg:538.89ms step:5636/6250 train_loss:2.1120 aux_loss:0.0259 train_time:3031825ms step_avg:538.90ms step:5637/6250 train_loss:2.0804 aux_loss:0.0255 train_time:3032366ms step_avg:538.90ms step:5638/6250 train_loss:2.1499 aux_loss:0.0257 train_time:3032896ms step_avg:538.89ms step:5639/6250 train_loss:2.0687 aux_loss:0.0264 train_time:3033426ms step_avg:538.89ms step:5640/6250 train_loss:2.0982 aux_loss:0.0258 train_time:3033957ms step_avg:538.89ms step:5641/6250 train_loss:2.2942 aux_loss:0.0258 train_time:3034489ms step_avg:538.89ms step:5642/6250 train_loss:2.1631 aux_loss:0.0260 train_time:3035018ms step_avg:538.89ms step:5643/6250 train_loss:2.1279 aux_loss:0.0260 train_time:3035562ms step_avg:538.89ms step:5644/6250 train_loss:2.1093 aux_loss:0.0257 train_time:3036090ms step_avg:538.89ms step:5645/6250 train_loss:2.1889 aux_loss:0.0259 train_time:3036630ms step_avg:538.89ms step:5646/6250 train_loss:2.1474 aux_loss:0.0261 train_time:3037168ms step_avg:538.89ms step:5647/6250 train_loss:2.0993 aux_loss:0.0257 train_time:3037720ms step_avg:538.89ms step:5648/6250 train_loss:2.1824 aux_loss:0.0256 train_time:3038287ms step_avg:538.89ms step:5649/6250 train_loss:2.1151 aux_loss:0.0261 train_time:3038818ms step_avg:538.89ms step:5650/6250 train_loss:2.1662 aux_loss:0.0260 train_time:3039348ms step_avg:538.89ms step:5651/6250 train_loss:2.1541 aux_loss:0.0256 train_time:3039878ms step_avg:538.89ms step:5652/6250 train_loss:2.0843 aux_loss:0.0259 train_time:3040409ms step_avg:538.89ms step:5653/6250 train_loss:2.2032 aux_loss:0.0262 train_time:3040940ms step_avg:538.89ms step:5654/6250 train_loss:2.1482 aux_loss:0.0263 train_time:3041470ms step_avg:538.89ms step:5655/6250 train_loss:2.1601 aux_loss:0.0260 train_time:3042001ms step_avg:538.88ms step:5656/6250 train_loss:2.1997 aux_loss:0.0258 train_time:3042531ms step_avg:538.88ms step:5657/6250 train_loss:2.1412 aux_loss:0.0258 train_time:3043062ms step_avg:538.88ms step:5658/6250 train_loss:2.1026 aux_loss:0.0257 train_time:3043591ms step_avg:538.88ms step:5659/6250 train_loss:2.1154 aux_loss:0.0257 train_time:3044121ms step_avg:538.88ms step:5660/6250 train_loss:2.0975 aux_loss:0.0255 train_time:3044662ms step_avg:538.88ms step:5661/6250 train_loss:2.1308 aux_loss:0.0257 train_time:3045193ms step_avg:538.88ms step:5662/6250 train_loss:2.2791 aux_loss:0.0256 train_time:3045734ms step_avg:538.88ms step:5663/6250 train_loss:2.0823 aux_loss:0.0258 train_time:3046270ms step_avg:538.88ms step:5664/6250 train_loss:2.1483 aux_loss:0.0259 train_time:3046802ms step_avg:538.88ms step:5665/6250 train_loss:2.1679 aux_loss:0.0253 train_time:3047336ms step_avg:538.87ms step:5666/6250 train_loss:2.0963 aux_loss:0.0254 train_time:3047868ms step_avg:538.87ms step:5667/6250 train_loss:2.1006 aux_loss:0.0255 train_time:3048399ms step_avg:538.87ms step:5668/6250 train_loss:2.1825 aux_loss:0.0254 train_time:3048929ms step_avg:538.87ms step:5669/6250 train_loss:2.1640 aux_loss:0.0256 train_time:3049460ms step_avg:538.87ms step:5670/6250 train_loss:2.0376 aux_loss:0.0258 train_time:3049990ms step_avg:538.87ms step:5671/6250 train_loss:2.1065 aux_loss:0.0253 train_time:3050520ms step_avg:538.87ms step:5672/6250 train_loss:2.1328 aux_loss:0.0252 train_time:3051051ms step_avg:538.86ms step:5673/6250 train_loss:2.2216 aux_loss:0.0262 train_time:3051582ms step_avg:538.86ms step:5674/6250 train_loss:2.1495 aux_loss:0.0251 train_time:3052112ms step_avg:538.86ms step:5675/6250 train_loss:2.1869 aux_loss:0.0254 train_time:3052646ms step_avg:538.86ms step:5676/6250 train_loss:2.1958 aux_loss:0.0260 train_time:3053175ms step_avg:538.86ms step:5677/6250 train_loss:2.1415 aux_loss:0.0256 train_time:3053706ms step_avg:538.86ms step:5678/6250 train_loss:2.2126 aux_loss:0.0254 train_time:3054235ms step_avg:538.86ms step:5679/6250 train_loss:2.2072 aux_loss:0.0262 train_time:3054768ms step_avg:538.85ms step:5680/6250 train_loss:2.1618 aux_loss:0.0261 train_time:3055297ms step_avg:538.85ms step:5681/6250 train_loss:2.1898 aux_loss:0.0252 train_time:3055827ms step_avg:538.85ms step:5682/6250 train_loss:2.0803 aux_loss:0.0253 train_time:3056357ms step_avg:538.85ms step:5683/6250 train_loss:2.1550 aux_loss:0.0265 train_time:3056890ms step_avg:538.85ms step:5684/6250 train_loss:2.1670 aux_loss:0.0263 train_time:3057420ms step_avg:538.85ms step:5685/6250 train_loss:2.2157 aux_loss:0.0256 train_time:3057950ms step_avg:538.85ms step:5686/6250 train_loss:2.1559 aux_loss:0.0257 train_time:3058492ms step_avg:538.85ms step:5687/6250 train_loss:2.2350 aux_loss:0.0261 train_time:3059042ms step_avg:538.85ms step:5688/6250 train_loss:2.1637 aux_loss:0.0257 train_time:3059589ms step_avg:538.85ms step:5689/6250 train_loss:2.0428 aux_loss:0.0254 train_time:3060119ms step_avg:538.85ms step:5690/6250 train_loss:2.2029 aux_loss:0.0258 train_time:3060649ms step_avg:538.85ms step:5691/6250 train_loss:2.1983 aux_loss:0.0261 train_time:3061180ms step_avg:538.85ms step:5692/6250 train_loss:2.1683 aux_loss:0.0257 train_time:3061712ms step_avg:538.84ms step:5693/6250 train_loss:2.1472 aux_loss:0.0258 train_time:3062253ms step_avg:538.84ms step:5694/6250 train_loss:2.1024 aux_loss:0.0261 train_time:3062800ms step_avg:538.85ms step:5695/6250 train_loss:2.1527 aux_loss:0.0258 train_time:3063353ms step_avg:538.85ms step:5696/6250 train_loss:2.1531 aux_loss:0.0256 train_time:3063907ms step_avg:538.85ms step:5697/6250 train_loss:2.1609 aux_loss:0.0252 train_time:3064446ms step_avg:538.85ms step:5698/6250 train_loss:2.1183 aux_loss:0.0257 train_time:3064976ms step_avg:538.85ms step:5699/6250 train_loss:2.1584 aux_loss:0.0257 train_time:3065507ms step_avg:538.85ms step:5700/6250 train_loss:2.1320 aux_loss:0.0254 train_time:3066036ms step_avg:538.85ms step:5701/6250 train_loss:2.1653 aux_loss:0.0259 train_time:3066567ms step_avg:538.85ms step:5702/6250 train_loss:2.1902 aux_loss:0.0259 train_time:3067097ms step_avg:538.84ms step:5703/6250 train_loss:2.1256 aux_loss:0.0256 train_time:3067627ms step_avg:538.84ms step:5704/6250 train_loss:2.2026 aux_loss:0.0255 train_time:3068158ms step_avg:538.84ms step:5705/6250 train_loss:2.1016 aux_loss:0.0261 train_time:3068689ms step_avg:538.84ms step:5706/6250 train_loss:2.0041 aux_loss:0.0258 train_time:3069219ms step_avg:538.84ms step:5707/6250 train_loss:2.1005 aux_loss:0.0256 train_time:3069751ms step_avg:538.84ms step:5708/6250 train_loss:2.1961 aux_loss:0.0262 train_time:3070281ms step_avg:538.83ms step:5709/6250 train_loss:2.1416 aux_loss:0.0259 train_time:3070812ms step_avg:538.83ms step:5710/6250 train_loss:2.1173 aux_loss:0.0253 train_time:3071343ms step_avg:538.83ms step:5711/6250 train_loss:2.1347 aux_loss:0.0258 train_time:3071876ms step_avg:538.83ms step:5712/6250 train_loss:2.0822 aux_loss:0.0258 train_time:3072406ms step_avg:538.83ms step:5713/6250 train_loss:2.2656 aux_loss:0.0259 train_time:3072935ms step_avg:538.83ms step:5714/6250 train_loss:2.0857 aux_loss:0.0258 train_time:3073465ms step_avg:538.83ms step:5715/6250 train_loss:2.1511 aux_loss:0.0259 train_time:3073993ms step_avg:538.82ms step:5716/6250 train_loss:2.1920 aux_loss:0.0255 train_time:3074548ms step_avg:538.83ms step:5717/6250 train_loss:2.1210 aux_loss:0.0253 train_time:3075098ms step_avg:538.83ms step:5718/6250 train_loss:2.0707 aux_loss:0.0260 train_time:3075635ms step_avg:538.83ms step:5719/6250 train_loss:2.0813 aux_loss:0.0254 train_time:3076171ms step_avg:538.83ms step:5720/6250 train_loss:2.1996 aux_loss:0.0255 train_time:3076720ms step_avg:538.83ms step:5721/6250 train_loss:2.1167 aux_loss:0.0255 train_time:3077267ms step_avg:538.83ms step:5722/6250 train_loss:2.0589 aux_loss:0.0253 train_time:3077823ms step_avg:538.83ms step:5723/6250 train_loss:2.2024 aux_loss:0.0258 train_time:3078363ms step_avg:538.83ms step:5724/6250 train_loss:2.1651 aux_loss:0.0264 train_time:3078898ms step_avg:538.83ms step:5725/6250 train_loss:2.0661 aux_loss:0.0257 train_time:3079460ms step_avg:538.84ms step:5726/6250 train_loss:2.0339 aux_loss:0.0251 train_time:3079996ms step_avg:538.84ms step:5727/6250 train_loss:2.1370 aux_loss:0.0262 train_time:3080532ms step_avg:538.84ms step:5728/6250 train_loss:2.0568 aux_loss:0.0255 train_time:3081085ms step_avg:538.84ms step:5729/6250 train_loss:2.0570 aux_loss:0.0255 train_time:3081626ms step_avg:538.84ms step:5730/6250 train_loss:2.0176 aux_loss:0.0258 train_time:3082169ms step_avg:538.84ms step:5731/6250 train_loss:2.1000 aux_loss:0.0257 train_time:3082722ms step_avg:538.84ms step:5732/6250 train_loss:2.0770 aux_loss:0.0251 train_time:3083291ms step_avg:538.85ms step:5733/6250 train_loss:2.1178 aux_loss:0.0260 train_time:3083856ms step_avg:538.85ms step:5734/6250 train_loss:2.1096 aux_loss:0.0259 train_time:3084400ms step_avg:538.85ms step:5735/6250 train_loss:2.0604 aux_loss:0.0252 train_time:3084937ms step_avg:538.85ms step:5736/6250 train_loss:2.1680 aux_loss:0.0256 train_time:3085473ms step_avg:538.85ms step:5737/6250 train_loss:2.1920 aux_loss:0.0261 train_time:3086009ms step_avg:538.85ms step:5738/6250 train_loss:2.0881 aux_loss:0.0256 train_time:3086543ms step_avg:538.85ms step:5739/6250 train_loss:2.2441 aux_loss:0.0259 train_time:3087078ms step_avg:538.85ms step:5740/6250 train_loss:2.1755 aux_loss:0.0262 train_time:3087610ms step_avg:538.85ms step:5741/6250 train_loss:2.1065 aux_loss:0.0258 train_time:3088152ms step_avg:538.85ms step:5742/6250 train_loss:2.0876 aux_loss:0.0255 train_time:3088689ms step_avg:538.85ms step:5743/6250 train_loss:2.1232 aux_loss:0.0263 train_time:3089223ms step_avg:538.85ms step:5744/6250 train_loss:2.1149 aux_loss:0.0262 train_time:3089759ms step_avg:538.85ms step:5745/6250 train_loss:2.1018 aux_loss:0.0255 train_time:3090316ms step_avg:538.85ms step:5746/6250 train_loss:2.1101 aux_loss:0.0255 train_time:3090867ms step_avg:538.85ms step:5747/6250 train_loss:2.1875 aux_loss:0.0259 train_time:3091398ms step_avg:538.85ms step:5748/6250 train_loss:2.1612 aux_loss:0.0260 train_time:3091949ms step_avg:538.85ms step:5749/6250 train_loss:2.1821 aux_loss:0.0254 train_time:3092486ms step_avg:538.85ms step:5750/6250 train_loss:2.1218 aux_loss:0.0255 train_time:3093016ms step_avg:538.85ms step:5751/6250 train_loss:2.1478 aux_loss:0.0258 train_time:3093547ms step_avg:538.85ms step:5752/6250 train_loss:2.1429 aux_loss:0.0257 train_time:3094107ms step_avg:538.86ms step:5753/6250 train_loss:2.0655 aux_loss:0.0255 train_time:3094644ms step_avg:538.85ms step:5754/6250 train_loss:2.0446 aux_loss:0.0256 train_time:3095175ms step_avg:538.85ms step:5755/6250 train_loss:2.1147 aux_loss:0.0257 train_time:3095708ms step_avg:538.85ms step:5756/6250 train_loss:2.1491 aux_loss:0.0253 train_time:3096241ms step_avg:538.85ms step:5757/6250 train_loss:2.1129 aux_loss:0.0253 train_time:3096773ms step_avg:538.85ms step:5758/6250 train_loss:2.2269 aux_loss:0.0258 train_time:3097306ms step_avg:538.85ms step:5759/6250 train_loss:2.1564 aux_loss:0.0257 train_time:3097838ms step_avg:538.85ms step:5760/6250 train_loss:2.1410 aux_loss:0.0260 train_time:3098370ms step_avg:538.85ms step:5761/6250 train_loss:2.2121 aux_loss:0.0261 train_time:3098900ms step_avg:538.85ms step:5762/6250 train_loss:2.1730 aux_loss:0.0253 train_time:3099430ms step_avg:538.84ms step:5763/6250 train_loss:2.1559 aux_loss:0.0254 train_time:3099960ms step_avg:538.84ms step:5764/6250 train_loss:2.0701 aux_loss:0.0255 train_time:3100488ms step_avg:538.84ms step:5765/6250 train_loss:2.0603 aux_loss:0.0257 train_time:3101018ms step_avg:538.84ms step:5766/6250 train_loss:2.1514 aux_loss:0.0257 train_time:3101555ms step_avg:538.84ms step:5767/6250 train_loss:2.0752 aux_loss:0.0259 train_time:3102094ms step_avg:538.84ms step:5768/6250 train_loss:2.2002 aux_loss:0.0257 train_time:3102623ms step_avg:538.84ms step:5769/6250 train_loss:2.0935 aux_loss:0.0256 train_time:3103153ms step_avg:538.84ms step:5770/6250 train_loss:2.1493 aux_loss:0.0259 train_time:3103683ms step_avg:538.83ms step:5771/6250 train_loss:2.1216 aux_loss:0.0256 train_time:3104223ms step_avg:538.83ms step:5772/6250 train_loss:2.1069 aux_loss:0.0256 train_time:3104754ms step_avg:538.83ms step:5773/6250 train_loss:2.3003 aux_loss:0.0262 train_time:3105286ms step_avg:538.83ms step:5774/6250 train_loss:2.1890 aux_loss:0.0262 train_time:3105818ms step_avg:538.83ms step:5775/6250 train_loss:2.0686 aux_loss:0.0252 train_time:3106346ms step_avg:538.83ms step:5776/6250 train_loss:2.0576 aux_loss:0.0253 train_time:3106889ms step_avg:538.83ms step:5777/6250 train_loss:2.1984 aux_loss:0.0253 train_time:3107419ms step_avg:538.83ms step:5778/6250 train_loss:2.1449 aux_loss:0.0254 train_time:3107950ms step_avg:538.83ms step:5779/6250 train_loss:2.1785 aux_loss:0.0263 train_time:3108502ms step_avg:538.83ms step:5780/6250 train_loss:2.0944 aux_loss:0.0260 train_time:3109035ms step_avg:538.83ms step:5781/6250 train_loss:2.1426 aux_loss:0.0253 train_time:3109576ms step_avg:538.83ms step:5782/6250 train_loss:2.1256 aux_loss:0.0258 train_time:3110127ms step_avg:538.83ms step:5783/6250 train_loss:2.1119 aux_loss:0.0254 train_time:3110678ms step_avg:538.83ms step:5784/6250 train_loss:2.0492 aux_loss:0.0244 train_time:3111227ms step_avg:538.83ms step:5785/6250 train_loss:2.1353 aux_loss:0.0258 train_time:3111761ms step_avg:538.83ms step:5786/6250 train_loss:2.1188 aux_loss:0.0261 train_time:3112291ms step_avg:538.83ms step:5787/6250 train_loss:2.1120 aux_loss:0.0255 train_time:3112821ms step_avg:538.83ms step:5788/6250 train_loss:2.0733 aux_loss:0.0253 train_time:3113351ms step_avg:538.83ms step:5789/6250 train_loss:2.1606 aux_loss:0.0259 train_time:3113882ms step_avg:538.83ms step:5790/6250 train_loss:2.0932 aux_loss:0.0255 train_time:3114412ms step_avg:538.83ms step:5791/6250 train_loss:2.1102 aux_loss:0.0253 train_time:3114940ms step_avg:538.82ms step:5792/6250 train_loss:2.2746 aux_loss:0.0253 train_time:3115471ms step_avg:538.82ms step:5793/6250 train_loss:2.1605 aux_loss:0.0262 train_time:3116001ms step_avg:538.82ms step:5794/6250 train_loss:2.2416 aux_loss:0.0260 train_time:3116537ms step_avg:538.82ms step:5795/6250 train_loss:2.1840 aux_loss:0.0260 train_time:3117071ms step_avg:538.82ms step:5796/6250 train_loss:2.1033 aux_loss:0.0256 train_time:3117601ms step_avg:538.82ms step:5797/6250 train_loss:2.1613 aux_loss:0.0258 train_time:3118132ms step_avg:538.82ms step:5798/6250 train_loss:2.1596 aux_loss:0.0257 train_time:3118661ms step_avg:538.82ms step:5799/6250 train_loss:2.1172 aux_loss:0.0254 train_time:3119190ms step_avg:538.81ms step:5800/6250 train_loss:2.1857 aux_loss:0.0258 train_time:3119722ms step_avg:538.81ms step:5801/6250 train_loss:2.1223 aux_loss:0.0258 train_time:3120251ms step_avg:538.81ms step:5802/6250 train_loss:2.1312 aux_loss:0.0251 train_time:3120797ms step_avg:538.81ms step:5803/6250 train_loss:2.0295 aux_loss:0.0254 train_time:3121331ms step_avg:538.81ms step:5804/6250 train_loss:2.1660 aux_loss:0.0258 train_time:3121862ms step_avg:538.81ms step:5805/6250 train_loss:2.1038 aux_loss:0.0256 train_time:3122391ms step_avg:538.81ms step:5806/6250 train_loss:2.2221 aux_loss:0.0257 train_time:3122922ms step_avg:538.81ms step:5807/6250 train_loss:2.1382 aux_loss:0.0255 train_time:3123453ms step_avg:538.81ms step:5808/6250 train_loss:2.2019 aux_loss:0.0255 train_time:3123985ms step_avg:538.80ms step:5809/6250 train_loss:2.1237 aux_loss:0.0265 train_time:3124514ms step_avg:538.80ms step:5810/6250 train_loss:2.0903 aux_loss:0.0257 train_time:3125044ms step_avg:538.80ms step:5811/6250 train_loss:2.1228 aux_loss:0.0258 train_time:3125576ms step_avg:538.80ms step:5812/6250 train_loss:2.1015 aux_loss:0.0259 train_time:3126108ms step_avg:538.80ms step:5813/6250 train_loss:2.1556 aux_loss:0.0260 train_time:3126640ms step_avg:538.80ms step:5814/6250 train_loss:2.1336 aux_loss:0.0258 train_time:3127201ms step_avg:538.80ms step:5815/6250 train_loss:2.1799 aux_loss:0.0254 train_time:3127740ms step_avg:538.80ms step:5816/6250 train_loss:2.2812 aux_loss:0.0258 train_time:3128271ms step_avg:538.80ms step:5817/6250 train_loss:2.1010 aux_loss:0.0257 train_time:3128801ms step_avg:538.80ms step:5818/6250 train_loss:2.1291 aux_loss:0.0256 train_time:3129332ms step_avg:538.80ms step:5819/6250 train_loss:2.1659 aux_loss:0.0257 train_time:3129867ms step_avg:538.80ms step:5820/6250 train_loss:2.1795 aux_loss:0.0260 train_time:3130432ms step_avg:538.80ms step:5821/6250 train_loss:2.1014 aux_loss:0.0252 train_time:3130974ms step_avg:538.80ms step:5822/6250 train_loss:2.1399 aux_loss:0.0251 train_time:3131504ms step_avg:538.80ms step:5823/6250 train_loss:2.1709 aux_loss:0.0258 train_time:3132033ms step_avg:538.80ms step:5824/6250 train_loss:2.0952 aux_loss:0.0253 train_time:3132562ms step_avg:538.80ms step:5825/6250 train_loss:2.1534 aux_loss:0.0253 train_time:3133093ms step_avg:538.79ms step:5826/6250 train_loss:2.1466 aux_loss:0.0259 train_time:3133623ms step_avg:538.79ms step:5827/6250 train_loss:2.0647 aux_loss:0.0259 train_time:3134154ms step_avg:538.79ms step:5828/6250 train_loss:2.1858 aux_loss:0.0251 train_time:3134684ms step_avg:538.79ms step:5829/6250 train_loss:2.0588 aux_loss:0.0254 train_time:3135214ms step_avg:538.79ms step:5830/6250 train_loss:2.1947 aux_loss:0.0256 train_time:3135746ms step_avg:538.79ms step:5831/6250 train_loss:2.1296 aux_loss:0.0253 train_time:3136276ms step_avg:538.79ms step:5832/6250 train_loss:2.1992 aux_loss:0.0255 train_time:3136807ms step_avg:538.79ms step:5833/6250 train_loss:2.1830 aux_loss:0.0256 train_time:3137338ms step_avg:538.78ms step:5834/6250 train_loss:2.0388 aux_loss:0.0249 train_time:3137869ms step_avg:538.78ms step:5835/6250 train_loss:2.0622 aux_loss:0.0257 train_time:3138400ms step_avg:538.78ms step:5836/6250 train_loss:2.1036 aux_loss:0.0259 train_time:3138930ms step_avg:538.78ms step:5837/6250 train_loss:2.0713 aux_loss:0.0252 train_time:3139461ms step_avg:538.78ms step:5838/6250 train_loss:2.1910 aux_loss:0.0252 train_time:3139991ms step_avg:538.78ms step:5839/6250 train_loss:2.1607 aux_loss:0.0260 train_time:3140522ms step_avg:538.78ms step:5840/6250 train_loss:2.2072 aux_loss:0.0257 train_time:3141053ms step_avg:538.77ms step:5841/6250 train_loss:2.0760 aux_loss:0.0252 train_time:3141584ms step_avg:538.77ms step:5842/6250 train_loss:2.0593 aux_loss:0.0253 train_time:3142113ms step_avg:538.77ms step:5843/6250 train_loss:2.1247 aux_loss:0.0258 train_time:3142644ms step_avg:538.77ms step:5844/6250 train_loss:2.1107 aux_loss:0.0249 train_time:3143174ms step_avg:538.77ms step:5845/6250 train_loss:2.1539 aux_loss:0.0255 train_time:3143703ms step_avg:538.77ms step:5846/6250 train_loss:2.1633 aux_loss:0.0254 train_time:3144232ms step_avg:538.76ms step:5847/6250 train_loss:2.1610 aux_loss:0.0255 train_time:3144762ms step_avg:538.76ms step:5848/6250 train_loss:2.0652 aux_loss:0.0255 train_time:3145292ms step_avg:538.76ms step:5849/6250 train_loss:2.1042 aux_loss:0.0252 train_time:3145823ms step_avg:538.76ms step:5850/6250 train_loss:2.0974 aux_loss:0.0252 train_time:3146353ms step_avg:538.76ms step:5851/6250 train_loss:2.1098 aux_loss:0.0250 train_time:3146886ms step_avg:538.76ms step:5852/6250 train_loss:2.1485 aux_loss:0.0258 train_time:3147417ms step_avg:538.76ms step:5853/6250 train_loss:2.2143 aux_loss:0.0261 train_time:3147958ms step_avg:538.76ms step:5854/6250 train_loss:2.1178 aux_loss:0.0254 train_time:3148499ms step_avg:538.76ms step:5855/6250 train_loss:2.1656 aux_loss:0.0255 train_time:3149037ms step_avg:538.76ms step:5856/6250 train_loss:2.1308 aux_loss:0.0251 train_time:3149578ms step_avg:538.76ms step:5857/6250 train_loss:2.1581 aux_loss:0.0247 train_time:3150122ms step_avg:538.76ms step:5858/6250 train_loss:2.2005 aux_loss:0.0251 train_time:3150652ms step_avg:538.76ms step:5859/6250 train_loss:2.1469 aux_loss:0.0252 train_time:3151183ms step_avg:538.76ms step:5860/6250 train_loss:2.1817 aux_loss:0.0257 train_time:3151717ms step_avg:538.75ms step:5861/6250 train_loss:2.0606 aux_loss:0.0252 train_time:3152245ms step_avg:538.75ms step:5862/6250 train_loss:2.0933 aux_loss:0.0251 train_time:3152778ms step_avg:538.75ms step:5863/6250 train_loss:2.2541 aux_loss:0.0253 train_time:3153308ms step_avg:538.75ms step:5864/6250 train_loss:2.1535 aux_loss:0.0255 train_time:3153842ms step_avg:538.75ms step:5865/6250 train_loss:2.1278 aux_loss:0.0254 train_time:3154371ms step_avg:538.75ms step:5866/6250 train_loss:2.2049 aux_loss:0.0253 train_time:3154902ms step_avg:538.75ms step:5867/6250 train_loss:2.0667 aux_loss:0.0255 train_time:3155454ms step_avg:538.75ms step:5868/6250 train_loss:2.1056 aux_loss:0.0250 train_time:3155987ms step_avg:538.75ms step:5869/6250 train_loss:2.1470 aux_loss:0.0257 train_time:3156536ms step_avg:538.75ms step:5870/6250 train_loss:2.1442 aux_loss:0.0256 train_time:3157080ms step_avg:538.75ms step:5871/6250 train_loss:2.1346 aux_loss:0.0255 train_time:3157624ms step_avg:538.75ms step:5872/6250 train_loss:2.1363 aux_loss:0.0256 train_time:3158180ms step_avg:538.75ms step:5873/6250 train_loss:2.0115 aux_loss:0.0258 train_time:3158748ms step_avg:538.76ms step:5874/6250 train_loss:2.1558 aux_loss:0.0256 train_time:3159290ms step_avg:538.76ms step:5875/6250 train_loss:2.1103 aux_loss:0.0258 train_time:3159821ms step_avg:538.76ms step:5876/6250 train_loss:2.1930 aux_loss:0.0252 train_time:3160351ms step_avg:538.76ms step:5877/6250 train_loss:2.1839 aux_loss:0.0260 train_time:3160882ms step_avg:538.76ms step:5878/6250 train_loss:2.0739 aux_loss:0.0255 train_time:3161413ms step_avg:538.75ms step:5879/6250 train_loss:2.0799 aux_loss:0.0252 train_time:3161944ms step_avg:538.75ms step:5880/6250 train_loss:2.1238 aux_loss:0.0264 train_time:3162474ms step_avg:538.75ms step:5881/6250 train_loss:2.0911 aux_loss:0.0255 train_time:3163004ms step_avg:538.75ms step:5882/6250 train_loss:2.1935 aux_loss:0.0252 train_time:3163543ms step_avg:538.75ms step:5883/6250 train_loss:2.1265 aux_loss:0.0261 train_time:3164073ms step_avg:538.75ms step:5884/6250 train_loss:2.0657 aux_loss:0.0262 train_time:3164605ms step_avg:538.75ms step:5885/6250 train_loss:2.1716 aux_loss:0.0255 train_time:3165137ms step_avg:538.75ms step:5886/6250 train_loss:2.1552 aux_loss:0.0260 train_time:3165670ms step_avg:538.75ms step:5887/6250 train_loss:2.0917 aux_loss:0.0261 train_time:3166202ms step_avg:538.74ms step:5888/6250 train_loss:2.1619 aux_loss:0.0262 train_time:3166732ms step_avg:538.74ms step:5889/6250 train_loss:2.1164 aux_loss:0.0254 train_time:3167263ms step_avg:538.74ms step:5890/6250 train_loss:2.0632 aux_loss:0.0250 train_time:3167794ms step_avg:538.74ms step:5891/6250 train_loss:2.1068 aux_loss:0.0255 train_time:3168323ms step_avg:538.74ms step:5892/6250 train_loss:2.1428 aux_loss:0.0259 train_time:3168854ms step_avg:538.74ms step:5893/6250 train_loss:2.0828 aux_loss:0.0257 train_time:3169384ms step_avg:538.74ms step:5894/6250 train_loss:2.2196 aux_loss:0.0258 train_time:3169915ms step_avg:538.73ms step:5895/6250 train_loss:2.1546 aux_loss:0.0259 train_time:3170447ms step_avg:538.73ms step:5896/6250 train_loss:2.2132 aux_loss:0.0258 train_time:3170976ms step_avg:538.73ms step:5897/6250 train_loss:2.1273 aux_loss:0.0251 train_time:3171528ms step_avg:538.73ms step:5898/6250 train_loss:2.0995 aux_loss:0.0253 train_time:3172080ms step_avg:538.74ms step:5899/6250 train_loss:2.1690 aux_loss:0.0257 train_time:3172639ms step_avg:538.74ms step:5900/6250 train_loss:2.0832 aux_loss:0.0252 train_time:3173199ms step_avg:538.74ms step:5901/6250 train_loss:2.0729 aux_loss:0.0249 train_time:3173764ms step_avg:538.75ms step:5902/6250 train_loss:2.1188 aux_loss:0.0254 train_time:3174298ms step_avg:538.75ms step:5903/6250 train_loss:2.1530 aux_loss:0.0257 train_time:3174829ms step_avg:538.75ms step:5904/6250 train_loss:2.1266 aux_loss:0.0255 train_time:3175358ms step_avg:538.74ms step:5905/6250 train_loss:2.0561 aux_loss:0.0254 train_time:3175890ms step_avg:538.74ms step:5906/6250 train_loss:2.0789 aux_loss:0.0256 train_time:3176419ms step_avg:538.74ms step:5907/6250 train_loss:2.2958 aux_loss:0.0260 train_time:3176951ms step_avg:538.74ms step:5908/6250 train_loss:2.0956 aux_loss:0.0254 train_time:3177483ms step_avg:538.74ms step:5909/6250 train_loss:2.1668 aux_loss:0.0255 train_time:3178014ms step_avg:538.74ms step:5910/6250 train_loss:2.0778 aux_loss:0.0257 train_time:3178544ms step_avg:538.74ms step:5911/6250 train_loss:2.0711 aux_loss:0.0255 train_time:3179071ms step_avg:538.73ms step:5912/6250 train_loss:2.1612 aux_loss:0.0254 train_time:3179601ms step_avg:538.73ms step:5913/6250 train_loss:2.0776 aux_loss:0.0257 train_time:3180131ms step_avg:538.73ms step:5914/6250 train_loss:2.1206 aux_loss:0.0257 train_time:3180669ms step_avg:538.73ms step:5915/6250 train_loss:2.1825 aux_loss:0.0253 train_time:3181199ms step_avg:538.73ms step:5916/6250 train_loss:2.1043 aux_loss:0.0253 train_time:3181731ms step_avg:538.73ms step:5917/6250 train_loss:2.1907 aux_loss:0.0253 train_time:3182262ms step_avg:538.73ms step:5918/6250 train_loss:2.1783 aux_loss:0.0257 train_time:3182794ms step_avg:538.73ms step:5919/6250 train_loss:2.2443 aux_loss:0.0258 train_time:3183324ms step_avg:538.72ms step:5920/6250 train_loss:2.0934 aux_loss:0.0259 train_time:3183855ms step_avg:538.72ms step:5921/6250 train_loss:2.0891 aux_loss:0.0255 train_time:3184384ms step_avg:538.72ms step:5922/6250 train_loss:2.0719 aux_loss:0.0250 train_time:3184915ms step_avg:538.72ms step:5923/6250 train_loss:2.1594 aux_loss:0.0254 train_time:3185447ms step_avg:538.72ms step:5924/6250 train_loss:2.0941 aux_loss:0.0255 train_time:3185978ms step_avg:538.72ms step:5925/6250 train_loss:2.1185 aux_loss:0.0255 train_time:3186508ms step_avg:538.72ms step:5926/6250 train_loss:2.0301 aux_loss:0.0253 train_time:3187039ms step_avg:538.72ms step:5927/6250 train_loss:2.1204 aux_loss:0.0255 train_time:3187579ms step_avg:538.72ms step:5928/6250 train_loss:2.1926 aux_loss:0.0256 train_time:3188120ms step_avg:538.72ms step:5929/6250 train_loss:2.1699 aux_loss:0.0254 train_time:3188665ms step_avg:538.72ms step:5930/6250 train_loss:2.1638 aux_loss:0.0258 train_time:3189197ms step_avg:538.72ms step:5931/6250 train_loss:2.1486 aux_loss:0.0256 train_time:3189728ms step_avg:538.71ms step:5932/6250 train_loss:2.1710 aux_loss:0.0256 train_time:3190268ms step_avg:538.71ms step:5933/6250 train_loss:2.0846 aux_loss:0.0253 train_time:3190820ms step_avg:538.72ms step:5934/6250 train_loss:2.1839 aux_loss:0.0257 train_time:3191369ms step_avg:538.72ms step:5935/6250 train_loss:2.0915 aux_loss:0.0258 train_time:3191919ms step_avg:538.72ms step:5936/6250 train_loss:2.0395 aux_loss:0.0248 train_time:3192461ms step_avg:538.72ms step:5937/6250 train_loss:2.1164 aux_loss:0.0256 train_time:3192993ms step_avg:538.72ms step:5938/6250 train_loss:2.1155 aux_loss:0.0251 train_time:3193522ms step_avg:538.72ms step:5939/6250 train_loss:2.1293 aux_loss:0.0253 train_time:3194053ms step_avg:538.72ms step:5940/6250 train_loss:2.2147 aux_loss:0.0252 train_time:3194584ms step_avg:538.72ms step:5941/6250 train_loss:2.1566 aux_loss:0.0254 train_time:3195114ms step_avg:538.71ms step:5942/6250 train_loss:2.2117 aux_loss:0.0257 train_time:3195645ms step_avg:538.71ms step:5943/6250 train_loss:2.1136 aux_loss:0.0254 train_time:3196174ms step_avg:538.71ms step:5944/6250 train_loss:2.2385 aux_loss:0.0253 train_time:3196705ms step_avg:538.71ms step:5945/6250 train_loss:2.0887 aux_loss:0.0257 train_time:3197238ms step_avg:538.71ms step:5946/6250 train_loss:2.2205 aux_loss:0.0251 train_time:3197791ms step_avg:538.71ms step:5947/6250 train_loss:2.1184 aux_loss:0.0253 train_time:3198334ms step_avg:538.71ms step:5948/6250 train_loss:2.1326 aux_loss:0.0250 train_time:3198867ms step_avg:538.71ms step:5949/6250 train_loss:2.1176 aux_loss:0.0255 train_time:3199424ms step_avg:538.71ms step:5950/6250 train_loss:2.0699 aux_loss:0.0260 train_time:3199985ms step_avg:538.72ms step:5951/6250 train_loss:2.1891 aux_loss:0.0256 train_time:3200530ms step_avg:538.72ms step:5952/6250 train_loss:2.1885 aux_loss:0.0259 train_time:3201082ms step_avg:538.72ms step:5953/6250 train_loss:2.1255 aux_loss:0.0251 train_time:3201614ms step_avg:538.72ms step:5954/6250 train_loss:2.1271 aux_loss:0.0253 train_time:3202163ms step_avg:538.72ms step:5955/6250 train_loss:2.0590 aux_loss:0.0248 train_time:3202700ms step_avg:538.72ms step:5956/6250 train_loss:2.1422 aux_loss:0.0255 train_time:3203246ms step_avg:538.72ms step:5957/6250 train_loss:2.2134 aux_loss:0.0258 train_time:3203793ms step_avg:538.72ms step:5958/6250 train_loss:2.1566 aux_loss:0.0251 train_time:3204333ms step_avg:538.72ms step:5959/6250 train_loss:2.0910 aux_loss:0.0249 train_time:3204882ms step_avg:538.73ms step:5960/6250 train_loss:2.0898 aux_loss:0.0256 train_time:3205427ms step_avg:538.73ms step:5961/6250 train_loss:2.1675 aux_loss:0.0251 train_time:3205963ms step_avg:538.73ms step:5962/6250 train_loss:2.2129 aux_loss:0.0253 train_time:3206507ms step_avg:538.73ms step:5963/6250 train_loss:2.1465 aux_loss:0.0251 train_time:3207063ms step_avg:538.73ms step:5964/6250 train_loss:2.1948 aux_loss:0.0255 train_time:3207624ms step_avg:538.73ms step:5965/6250 train_loss:2.1343 aux_loss:0.0250 train_time:3208184ms step_avg:538.74ms step:5966/6250 train_loss:2.1308 aux_loss:0.0249 train_time:3208724ms step_avg:538.74ms step:5967/6250 train_loss:2.1114 aux_loss:0.0255 train_time:3209253ms step_avg:538.74ms step:5968/6250 train_loss:2.1332 aux_loss:0.0249 train_time:3209782ms step_avg:538.73ms step:5969/6250 train_loss:2.1351 aux_loss:0.0251 train_time:3210313ms step_avg:538.73ms step:5970/6250 train_loss:2.2370 aux_loss:0.0254 train_time:3210844ms step_avg:538.73ms step:5971/6250 train_loss:2.1120 aux_loss:0.0249 train_time:3211374ms step_avg:538.73ms step:5972/6250 train_loss:2.1830 aux_loss:0.0250 train_time:3211904ms step_avg:538.73ms step:5973/6250 train_loss:2.1299 aux_loss:0.0252 train_time:3212432ms step_avg:538.73ms step:5974/6250 train_loss:2.1198 aux_loss:0.0248 train_time:3212964ms step_avg:538.73ms step:5975/6250 train_loss:2.2301 aux_loss:0.0258 train_time:3213494ms step_avg:538.72ms step:5976/6250 train_loss:2.1754 aux_loss:0.0254 train_time:3214025ms step_avg:538.72ms step:5977/6250 train_loss:2.2090 aux_loss:0.0251 train_time:3214590ms step_avg:538.73ms step:5978/6250 train_loss:2.1237 aux_loss:0.0258 train_time:3215129ms step_avg:538.73ms step:5979/6250 train_loss:2.1752 aux_loss:0.0258 train_time:3215689ms step_avg:538.73ms step:5980/6250 train_loss:2.1505 aux_loss:0.0252 train_time:3216224ms step_avg:538.73ms step:5981/6250 train_loss:2.1997 aux_loss:0.0254 train_time:3216756ms step_avg:538.73ms step:5982/6250 train_loss:2.1214 aux_loss:0.0256 train_time:3217287ms step_avg:538.73ms step:5983/6250 train_loss:2.0486 aux_loss:0.0252 train_time:3217817ms step_avg:538.73ms step:5984/6250 train_loss:2.1361 aux_loss:0.0251 train_time:3218351ms step_avg:538.73ms step:5985/6250 train_loss:2.0993 aux_loss:0.0258 train_time:3218899ms step_avg:538.73ms step:5986/6250 train_loss:2.0761 aux_loss:0.0250 train_time:3219430ms step_avg:538.73ms step:5987/6250 train_loss:2.1025 aux_loss:0.0252 train_time:3219962ms step_avg:538.73ms step:5988/6250 train_loss:2.1029 aux_loss:0.0256 train_time:3220494ms step_avg:538.72ms step:5989/6250 train_loss:2.1294 aux_loss:0.0253 train_time:3221032ms step_avg:538.72ms step:5990/6250 train_loss:2.0914 aux_loss:0.0255 train_time:3221567ms step_avg:538.72ms step:5991/6250 train_loss:2.2569 aux_loss:0.0256 train_time:3222115ms step_avg:538.73ms step:5992/6250 train_loss:2.0617 aux_loss:0.0252 train_time:3222648ms step_avg:538.72ms step:5993/6250 train_loss:2.1816 aux_loss:0.0257 train_time:3223179ms step_avg:538.72ms step:5994/6250 train_loss:2.1575 aux_loss:0.0250 train_time:3223710ms step_avg:538.72ms step:5995/6250 train_loss:2.1371 aux_loss:0.0247 train_time:3224241ms step_avg:538.72ms step:5996/6250 train_loss:2.1815 aux_loss:0.0250 train_time:3224772ms step_avg:538.72ms step:5997/6250 train_loss:2.1728 aux_loss:0.0253 train_time:3225303ms step_avg:538.72ms step:5998/6250 train_loss:2.1681 aux_loss:0.0249 train_time:3225834ms step_avg:538.72ms step:5999/6250 train_loss:2.0067 aux_loss:0.0252 train_time:3226382ms step_avg:538.72ms step:6000/6250 train_loss:2.1751 aux_loss:0.0252 train_time:3226934ms step_avg:538.72ms step:6000/6250 val_loss:2.1203 val_aux_loss:0.0248 train_time:3226934ms step_avg:538.72ms step:6001/6250 train_loss:2.1225 aux_loss:0.0249 train_time:3227464ms step_avg:538.72ms step:6002/6250 train_loss:2.1130 aux_loss:0.0253 train_time:3228023ms step_avg:538.72ms step:6003/6250 train_loss:2.1300 aux_loss:0.0251 train_time:3228569ms step_avg:538.72ms step:6004/6250 train_loss:2.0882 aux_loss:0.0247 train_time:3229113ms step_avg:538.72ms step:6005/6250 train_loss:2.1942 aux_loss:0.0259 train_time:3229644ms step_avg:538.72ms step:6006/6250 train_loss:2.0835 aux_loss:0.0251 train_time:3230175ms step_avg:538.72ms step:6007/6250 train_loss:2.1024 aux_loss:0.0245 train_time:3230706ms step_avg:538.72ms step:6008/6250 train_loss:2.1499 aux_loss:0.0253 train_time:3231237ms step_avg:538.72ms step:6009/6250 train_loss:2.1328 aux_loss:0.0253 train_time:3231798ms step_avg:538.72ms step:6010/6250 train_loss:2.1628 aux_loss:0.0246 train_time:3232347ms step_avg:538.72ms step:6011/6250 train_loss:2.0368 aux_loss:0.0247 train_time:3232887ms step_avg:538.72ms step:6012/6250 train_loss:2.1222 aux_loss:0.0257 train_time:3233427ms step_avg:538.72ms step:6013/6250 train_loss:2.1192 aux_loss:0.0254 train_time:3233976ms step_avg:538.73ms step:6014/6250 train_loss:2.1790 aux_loss:0.0247 train_time:3234510ms step_avg:538.73ms step:6015/6250 train_loss:2.1061 aux_loss:0.0251 train_time:3235060ms step_avg:538.73ms step:6016/6250 train_loss:2.1608 aux_loss:0.0259 train_time:3235599ms step_avg:538.73ms step:6017/6250 train_loss:2.1323 aux_loss:0.0260 train_time:3236139ms step_avg:538.73ms step:6018/6250 train_loss:2.1032 aux_loss:0.0250 train_time:3236691ms step_avg:538.73ms step:6019/6250 train_loss:2.1568 aux_loss:0.0252 train_time:3237241ms step_avg:538.73ms step:6020/6250 train_loss:2.1656 aux_loss:0.0256 train_time:3237792ms step_avg:538.73ms step:6021/6250 train_loss:2.1010 aux_loss:0.0258 train_time:3238354ms step_avg:538.74ms step:6022/6250 train_loss:2.1859 aux_loss:0.0254 train_time:3238887ms step_avg:538.74ms step:6023/6250 train_loss:2.1048 aux_loss:0.0250 train_time:3239420ms step_avg:538.74ms step:6024/6250 train_loss:2.1084 aux_loss:0.0253 train_time:3239967ms step_avg:538.74ms step:6025/6250 train_loss:2.1108 aux_loss:0.0251 train_time:3240517ms step_avg:538.74ms step:6026/6250 train_loss:2.1857 aux_loss:0.0249 train_time:3241057ms step_avg:538.74ms step:6027/6250 train_loss:2.0689 aux_loss:0.0248 train_time:3241589ms step_avg:538.74ms step:6028/6250 train_loss:2.1133 aux_loss:0.0252 train_time:3242119ms step_avg:538.74ms step:6029/6250 train_loss:2.1169 aux_loss:0.0256 train_time:3242650ms step_avg:538.74ms step:6030/6250 train_loss:2.0647 aux_loss:0.0253 train_time:3243188ms step_avg:538.74ms step:6031/6250 train_loss:2.0159 aux_loss:0.0251 train_time:3243738ms step_avg:538.74ms step:6032/6250 train_loss:2.1919 aux_loss:0.0249 train_time:3244285ms step_avg:538.74ms step:6033/6250 train_loss:2.1611 aux_loss:0.0251 train_time:3244835ms step_avg:538.74ms step:6034/6250 train_loss:2.1363 aux_loss:0.0249 train_time:3245367ms step_avg:538.74ms step:6035/6250 train_loss:2.1376 aux_loss:0.0254 train_time:3245898ms step_avg:538.74ms step:6036/6250 train_loss:2.1768 aux_loss:0.0250 train_time:3246429ms step_avg:538.74ms step:6037/6250 train_loss:2.1033 aux_loss:0.0250 train_time:3246960ms step_avg:538.74ms step:6038/6250 train_loss:2.1702 aux_loss:0.0252 train_time:3247490ms step_avg:538.73ms step:6039/6250 train_loss:2.1559 aux_loss:0.0253 train_time:3248021ms step_avg:538.73ms step:6040/6250 train_loss:2.1482 aux_loss:0.0252 train_time:3248552ms step_avg:538.73ms step:6041/6250 train_loss:2.2044 aux_loss:0.0258 train_time:3249081ms step_avg:538.73ms step:6042/6250 train_loss:2.1944 aux_loss:0.0255 train_time:3249612ms step_avg:538.73ms step:6043/6250 train_loss:2.1601 aux_loss:0.0255 train_time:3250141ms step_avg:538.73ms step:6044/6250 train_loss:2.0830 aux_loss:0.0253 train_time:3250671ms step_avg:538.73ms step:6045/6250 train_loss:1.9974 aux_loss:0.0253 train_time:3251201ms step_avg:538.72ms step:6046/6250 train_loss:2.0830 aux_loss:0.0255 train_time:3251730ms step_avg:538.72ms step:6047/6250 train_loss:2.1153 aux_loss:0.0250 train_time:3252260ms step_avg:538.72ms step:6048/6250 train_loss:2.0680 aux_loss:0.0255 train_time:3252803ms step_avg:538.72ms step:6049/6250 train_loss:2.0838 aux_loss:0.0254 train_time:3253354ms step_avg:538.72ms step:6050/6250 train_loss:2.1508 aux_loss:0.0250 train_time:3253905ms step_avg:538.73ms step:6051/6250 train_loss:2.0667 aux_loss:0.0254 train_time:3254439ms step_avg:538.73ms step:6052/6250 train_loss:2.0108 aux_loss:0.0252 train_time:3254974ms step_avg:538.72ms step:6053/6250 train_loss:2.1057 aux_loss:0.0248 train_time:3255508ms step_avg:538.72ms step:6054/6250 train_loss:2.0636 aux_loss:0.0252 train_time:3256040ms step_avg:538.72ms step:6055/6250 train_loss:2.1022 aux_loss:0.0255 train_time:3256578ms step_avg:538.72ms step:6056/6250 train_loss:2.0878 aux_loss:0.0248 train_time:3257118ms step_avg:538.72ms step:6057/6250 train_loss:2.2195 aux_loss:0.0250 train_time:3257654ms step_avg:538.72ms step:6058/6250 train_loss:2.0129 aux_loss:0.0252 train_time:3258187ms step_avg:538.72ms step:6059/6250 train_loss:2.1340 aux_loss:0.0253 train_time:3258718ms step_avg:538.72ms step:6060/6250 train_loss:2.0755 aux_loss:0.0250 train_time:3259249ms step_avg:538.72ms step:6061/6250 train_loss:2.1574 aux_loss:0.0251 train_time:3259773ms step_avg:538.72ms step:6062/6250 train_loss:2.1507 aux_loss:0.0254 train_time:3260302ms step_avg:538.71ms step:6063/6250 train_loss:2.1330 aux_loss:0.0252 train_time:3260830ms step_avg:538.71ms step:6064/6250 train_loss:2.0441 aux_loss:0.0251 train_time:3261360ms step_avg:538.71ms step:6065/6250 train_loss:2.0713 aux_loss:0.0253 train_time:3261891ms step_avg:538.71ms step:6066/6250 train_loss:2.0648 aux_loss:0.0251 train_time:3262436ms step_avg:538.71ms step:6067/6250 train_loss:2.1506 aux_loss:0.0257 train_time:3262979ms step_avg:538.71ms step:6068/6250 train_loss:2.0206 aux_loss:0.0250 train_time:3263519ms step_avg:538.71ms step:6069/6250 train_loss:2.1795 aux_loss:0.0256 train_time:3264079ms step_avg:538.72ms step:6070/6250 train_loss:2.0534 aux_loss:0.0252 train_time:3264616ms step_avg:538.72ms step:6071/6250 train_loss:2.1329 aux_loss:0.0250 train_time:3265146ms step_avg:538.71ms step:6072/6250 train_loss:2.0909 aux_loss:0.0252 train_time:3265678ms step_avg:538.71ms step:6073/6250 train_loss:2.0803 aux_loss:0.0255 train_time:3266209ms step_avg:538.71ms step:6074/6250 train_loss:2.1494 aux_loss:0.0253 train_time:3266742ms step_avg:538.71ms step:6075/6250 train_loss:2.1464 aux_loss:0.0251 train_time:3267275ms step_avg:538.71ms step:6076/6250 train_loss:2.0631 aux_loss:0.0252 train_time:3267807ms step_avg:538.71ms step:6077/6250 train_loss:2.1265 aux_loss:0.0254 train_time:3268337ms step_avg:538.71ms step:6078/6250 train_loss:2.0638 aux_loss:0.0250 train_time:3268867ms step_avg:538.71ms step:6079/6250 train_loss:2.1592 aux_loss:0.0249 train_time:3269401ms step_avg:538.71ms step:6080/6250 train_loss:2.1014 aux_loss:0.0252 train_time:3269946ms step_avg:538.71ms step:6081/6250 train_loss:2.1000 aux_loss:0.0252 train_time:3270486ms step_avg:538.71ms step:6082/6250 train_loss:2.0576 aux_loss:0.0249 train_time:3271022ms step_avg:538.71ms step:6083/6250 train_loss:2.0940 aux_loss:0.0253 train_time:3271554ms step_avg:538.70ms step:6084/6250 train_loss:2.0591 aux_loss:0.0253 train_time:3272087ms step_avg:538.70ms step:6085/6250 train_loss:2.0437 aux_loss:0.0249 train_time:3272619ms step_avg:538.70ms step:6086/6250 train_loss:2.2168 aux_loss:0.0250 train_time:3273152ms step_avg:538.70ms step:6087/6250 train_loss:2.1775 aux_loss:0.0250 train_time:3273703ms step_avg:538.70ms step:6088/6250 train_loss:2.1098 aux_loss:0.0249 train_time:3274240ms step_avg:538.70ms step:6089/6250 train_loss:2.1271 aux_loss:0.0249 train_time:3274773ms step_avg:538.70ms step:6090/6250 train_loss:2.0248 aux_loss:0.0251 train_time:3275312ms step_avg:538.70ms step:6091/6250 train_loss:2.0697 aux_loss:0.0255 train_time:3275854ms step_avg:538.70ms step:6092/6250 train_loss:2.0804 aux_loss:0.0254 train_time:3276386ms step_avg:538.70ms step:6093/6250 train_loss:2.0410 aux_loss:0.0248 train_time:3276919ms step_avg:538.70ms step:6094/6250 train_loss:2.1304 aux_loss:0.0251 train_time:3277449ms step_avg:538.70ms step:6095/6250 train_loss:2.0773 aux_loss:0.0251 train_time:3277982ms step_avg:538.70ms step:6096/6250 train_loss:2.1678 aux_loss:0.0251 train_time:3278513ms step_avg:538.70ms step:6097/6250 train_loss:2.0444 aux_loss:0.0253 train_time:3279047ms step_avg:538.70ms step:6098/6250 train_loss:2.0270 aux_loss:0.0254 train_time:3279579ms step_avg:538.70ms step:6099/6250 train_loss:2.0920 aux_loss:0.0252 train_time:3280110ms step_avg:538.69ms step:6100/6250 train_loss:2.1062 aux_loss:0.0250 train_time:3280660ms step_avg:538.70ms step:6101/6250 train_loss:2.0702 aux_loss:0.0250 train_time:3281207ms step_avg:538.70ms step:6102/6250 train_loss:2.0973 aux_loss:0.0253 train_time:3281750ms step_avg:538.70ms step:6103/6250 train_loss:1.9961 aux_loss:0.0250 train_time:3282461ms step_avg:538.73ms step:6104/6250 train_loss:2.1088 aux_loss:0.0249 train_time:3282994ms step_avg:538.73ms step:6105/6250 train_loss:2.1139 aux_loss:0.0251 train_time:3283542ms step_avg:538.73ms step:6106/6250 train_loss:2.1929 aux_loss:0.0256 train_time:3284088ms step_avg:538.73ms step:6107/6250 train_loss:2.1939 aux_loss:0.0254 train_time:3284657ms step_avg:538.73ms step:6108/6250 train_loss:2.1543 aux_loss:0.0250 train_time:3285194ms step_avg:538.73ms step:6109/6250 train_loss:2.1412 aux_loss:0.0253 train_time:3285745ms step_avg:538.73ms step:6110/6250 train_loss:2.1205 aux_loss:0.0251 train_time:3286298ms step_avg:538.74ms step:6111/6250 train_loss:2.0768 aux_loss:0.0246 train_time:3286836ms step_avg:538.74ms step:6112/6250 train_loss:2.1192 aux_loss:0.0250 train_time:3287367ms step_avg:538.74ms step:6113/6250 train_loss:2.1341 aux_loss:0.0252 train_time:3287920ms step_avg:538.74ms step:6114/6250 train_loss:2.0720 aux_loss:0.0250 train_time:3288452ms step_avg:538.74ms step:6115/6250 train_loss:2.1759 aux_loss:0.0249 train_time:3288990ms step_avg:538.74ms step:6116/6250 train_loss:2.0346 aux_loss:0.0250 train_time:3289535ms step_avg:538.74ms step:6117/6250 train_loss:2.1795 aux_loss:0.0254 train_time:3290067ms step_avg:538.74ms step:6118/6250 train_loss:2.0730 aux_loss:0.0249 train_time:3290602ms step_avg:538.74ms step:6119/6250 train_loss:2.0696 aux_loss:0.0250 train_time:3291155ms step_avg:538.74ms step:6120/6250 train_loss:2.1190 aux_loss:0.0253 train_time:3291703ms step_avg:538.74ms step:6121/6250 train_loss:2.0758 aux_loss:0.0250 train_time:3292234ms step_avg:538.74ms step:6122/6250 train_loss:2.1316 aux_loss:0.0249 train_time:3292766ms step_avg:538.74ms step:6123/6250 train_loss:2.0641 aux_loss:0.0250 train_time:3293295ms step_avg:538.74ms step:6124/6250 train_loss:2.0808 aux_loss:0.0252 train_time:3293828ms step_avg:538.74ms step:6125/6250 train_loss:1.9724 aux_loss:0.0248 train_time:3294358ms step_avg:538.73ms step:6126/6250 train_loss:2.0872 aux_loss:0.0249 train_time:3294890ms step_avg:538.73ms step:6127/6250 train_loss:2.0168 aux_loss:0.0248 train_time:3295421ms step_avg:538.73ms step:6128/6250 train_loss:2.0965 aux_loss:0.0250 train_time:3295954ms step_avg:538.73ms step:6129/6250 train_loss:2.0615 aux_loss:0.0249 train_time:3296485ms step_avg:538.73ms step:6130/6250 train_loss:2.1770 aux_loss:0.0247 train_time:3297018ms step_avg:538.73ms step:6131/6250 train_loss:2.0821 aux_loss:0.0250 train_time:3297549ms step_avg:538.73ms step:6132/6250 train_loss:2.0789 aux_loss:0.0251 train_time:3298082ms step_avg:538.73ms step:6133/6250 train_loss:2.0989 aux_loss:0.0253 train_time:3298613ms step_avg:538.72ms step:6134/6250 train_loss:1.9893 aux_loss:0.0247 train_time:3299143ms step_avg:538.72ms step:6135/6250 train_loss:1.9932 aux_loss:0.0247 train_time:3299673ms step_avg:538.72ms step:6136/6250 train_loss:2.0808 aux_loss:0.0250 train_time:3300205ms step_avg:538.72ms step:6137/6250 train_loss:2.1070 aux_loss:0.0249 train_time:3300739ms step_avg:538.72ms step:6138/6250 train_loss:2.0516 aux_loss:0.0253 train_time:3301269ms step_avg:538.72ms step:6139/6250 train_loss:2.1639 aux_loss:0.0250 train_time:3301799ms step_avg:538.72ms step:6140/6250 train_loss:2.0716 aux_loss:0.0248 train_time:3302329ms step_avg:538.72ms step:6141/6250 train_loss:2.0304 aux_loss:0.0251 train_time:3302859ms step_avg:538.71ms step:6142/6250 train_loss:2.1262 aux_loss:0.0252 train_time:3303390ms step_avg:538.71ms step:6143/6250 train_loss:2.1114 aux_loss:0.0249 train_time:3303928ms step_avg:538.71ms step:6144/6250 train_loss:2.0777 aux_loss:0.0251 train_time:3304462ms step_avg:538.71ms step:6145/6250 train_loss:2.0796 aux_loss:0.0249 train_time:3304994ms step_avg:538.71ms step:6146/6250 train_loss:2.0286 aux_loss:0.0248 train_time:3305525ms step_avg:538.71ms step:6147/6250 train_loss:2.0089 aux_loss:0.0246 train_time:3306056ms step_avg:538.71ms step:6148/6250 train_loss:2.0698 aux_loss:0.0256 train_time:3306587ms step_avg:538.71ms step:6149/6250 train_loss:1.9444 aux_loss:0.0249 train_time:3307118ms step_avg:538.71ms step:6150/6250 train_loss:2.1052 aux_loss:0.0252 train_time:3307647ms step_avg:538.70ms step:6151/6250 train_loss:2.0673 aux_loss:0.0249 train_time:3308181ms step_avg:538.70ms step:6152/6250 train_loss:1.9500 aux_loss:0.0248 train_time:3308725ms step_avg:538.70ms step:6153/6250 train_loss:2.0145 aux_loss:0.0244 train_time:3309279ms step_avg:538.71ms step:6154/6250 train_loss:2.0527 aux_loss:0.0246 train_time:3309822ms step_avg:538.71ms step:6155/6250 train_loss:2.1500 aux_loss:0.0249 train_time:3310362ms step_avg:538.71ms step:6156/6250 train_loss:2.0633 aux_loss:0.0247 train_time:3310901ms step_avg:538.71ms step:6157/6250 train_loss:2.0694 aux_loss:0.0250 train_time:3311439ms step_avg:538.71ms step:6158/6250 train_loss:2.1187 aux_loss:0.0250 train_time:3311975ms step_avg:538.71ms step:6159/6250 train_loss:2.1352 aux_loss:0.0245 train_time:3312512ms step_avg:538.71ms step:6160/6250 train_loss:2.0676 aux_loss:0.0247 train_time:3313051ms step_avg:538.71ms step:6161/6250 train_loss:2.0440 aux_loss:0.0248 train_time:3313587ms step_avg:538.71ms step:6162/6250 train_loss:2.0473 aux_loss:0.0252 train_time:3314124ms step_avg:538.71ms step:6163/6250 train_loss:2.1082 aux_loss:0.0248 train_time:3314660ms step_avg:538.71ms step:6164/6250 train_loss:2.0921 aux_loss:0.0248 train_time:3315193ms step_avg:538.71ms step:6165/6250 train_loss:2.0114 aux_loss:0.0246 train_time:3315734ms step_avg:538.71ms step:6166/6250 train_loss:2.0209 aux_loss:0.0254 train_time:3316279ms step_avg:538.71ms step:6167/6250 train_loss:2.0111 aux_loss:0.0250 train_time:3316830ms step_avg:538.71ms step:6168/6250 train_loss:2.1418 aux_loss:0.0250 train_time:3317388ms step_avg:538.71ms step:6169/6250 train_loss:2.0081 aux_loss:0.0248 train_time:3317976ms step_avg:538.72ms step:6170/6250 train_loss:1.9762 aux_loss:0.0247 train_time:3318583ms step_avg:538.73ms step:6171/6250 train_loss:2.1873 aux_loss:0.0253 train_time:3319158ms step_avg:538.74ms step:6172/6250 train_loss:2.0466 aux_loss:0.0249 train_time:3319701ms step_avg:538.74ms step:6173/6250 train_loss:2.0650 aux_loss:0.0249 train_time:3320255ms step_avg:538.74ms step:6174/6250 train_loss:2.0322 aux_loss:0.0246 train_time:3320789ms step_avg:538.74ms step:6175/6250 train_loss:2.0515 aux_loss:0.0249 train_time:3321334ms step_avg:538.74ms step:6176/6250 train_loss:2.0288 aux_loss:0.0251 train_time:3321872ms step_avg:538.74ms step:6177/6250 train_loss:2.0547 aux_loss:0.0249 train_time:3322458ms step_avg:538.75ms step:6178/6250 train_loss:2.0387 aux_loss:0.0251 train_time:3323057ms step_avg:538.76ms step:6179/6250 train_loss:2.0206 aux_loss:0.0249 train_time:3323659ms step_avg:538.77ms step:6180/6250 train_loss:2.1461 aux_loss:0.0251 train_time:3324257ms step_avg:538.78ms step:6181/6250 train_loss:2.0962 aux_loss:0.0251 train_time:3324856ms step_avg:538.79ms step:6182/6250 train_loss:2.0822 aux_loss:0.0252 train_time:3325464ms step_avg:538.80ms step:6183/6250 train_loss:2.0765 aux_loss:0.0253 train_time:3326023ms step_avg:538.80ms step:6184/6250 train_loss:2.0318 aux_loss:0.0248 train_time:3326565ms step_avg:538.80ms step:6185/6250 train_loss:2.0034 aux_loss:0.0245 train_time:3327102ms step_avg:538.80ms step:6186/6250 train_loss:2.0026 aux_loss:0.0246 train_time:3327657ms step_avg:538.80ms step:6187/6250 train_loss:2.0136 aux_loss:0.0246 train_time:3328192ms step_avg:538.80ms step:6188/6250 train_loss:2.0107 aux_loss:0.0245 train_time:3328729ms step_avg:538.80ms step:6189/6250 train_loss:2.0002 aux_loss:0.0246 train_time:3329261ms step_avg:538.80ms step:6190/6250 train_loss:2.1626 aux_loss:0.0248 train_time:3329795ms step_avg:538.80ms step:6191/6250 train_loss:2.1013 aux_loss:0.0250 train_time:3330326ms step_avg:538.80ms step:6192/6250 train_loss:2.0240 aux_loss:0.0249 train_time:3330862ms step_avg:538.80ms step:6193/6250 train_loss:2.0751 aux_loss:0.0251 train_time:3331397ms step_avg:538.80ms step:6194/6250 train_loss:2.0849 aux_loss:0.0249 train_time:3331932ms step_avg:538.80ms step:6195/6250 train_loss:2.0250 aux_loss:0.0249 train_time:3332466ms step_avg:538.80ms step:6196/6250 train_loss:2.0847 aux_loss:0.0250 train_time:3332999ms step_avg:538.80ms step:6197/6250 train_loss:2.0599 aux_loss:0.0249 train_time:3333534ms step_avg:538.80ms step:6198/6250 train_loss:2.0439 aux_loss:0.0247 train_time:3334066ms step_avg:538.80ms step:6199/6250 train_loss:2.1254 aux_loss:0.0248 train_time:3334604ms step_avg:538.80ms step:6200/6250 train_loss:2.0366 aux_loss:0.0248 train_time:3335148ms step_avg:538.80ms step:6201/6250 train_loss:2.0765 aux_loss:0.0247 train_time:3335692ms step_avg:538.80ms step:6202/6250 train_loss:2.1313 aux_loss:0.0247 train_time:3336230ms step_avg:538.80ms step:6203/6250 train_loss:2.0961 aux_loss:0.0251 train_time:3336769ms step_avg:538.80ms step:6204/6250 train_loss:2.0570 aux_loss:0.0251 train_time:3337302ms step_avg:538.80ms step:6205/6250 train_loss:1.9559 aux_loss:0.0250 train_time:3337837ms step_avg:538.80ms step:6206/6250 train_loss:2.1131 aux_loss:0.0250 train_time:3338372ms step_avg:538.79ms step:6207/6250 train_loss:2.0829 aux_loss:0.0248 train_time:3338916ms step_avg:538.80ms step:6208/6250 train_loss:2.0729 aux_loss:0.0248 train_time:3339463ms step_avg:538.80ms step:6209/6250 train_loss:2.0104 aux_loss:0.0247 train_time:3340001ms step_avg:538.80ms step:6210/6250 train_loss:2.0764 aux_loss:0.0251 train_time:3340542ms step_avg:538.80ms step:6211/6250 train_loss:2.0986 aux_loss:0.0250 train_time:3341079ms step_avg:538.80ms step:6212/6250 train_loss:2.0254 aux_loss:0.0248 train_time:3341618ms step_avg:538.80ms step:6213/6250 train_loss:2.0884 aux_loss:0.0247 train_time:3342157ms step_avg:538.80ms step:6214/6250 train_loss:2.0279 aux_loss:0.0249 train_time:3342698ms step_avg:538.80ms step:6215/6250 train_loss:2.0337 aux_loss:0.0249 train_time:3343231ms step_avg:538.80ms step:6216/6250 train_loss:2.0522 aux_loss:0.0249 train_time:3343767ms step_avg:538.80ms step:6217/6250 train_loss:2.1180 aux_loss:0.0250 train_time:3344303ms step_avg:538.80ms step:6218/6250 train_loss:2.0335 aux_loss:0.0247 train_time:3344842ms step_avg:538.80ms step:6219/6250 train_loss:2.0397 aux_loss:0.0244 train_time:3345384ms step_avg:538.80ms step:6220/6250 train_loss:2.0496 aux_loss:0.0246 train_time:3345926ms step_avg:538.80ms step:6221/6250 train_loss:2.0779 aux_loss:0.0249 train_time:3346465ms step_avg:538.80ms step:6222/6250 train_loss:2.0985 aux_loss:0.0248 train_time:3347003ms step_avg:538.80ms step:6223/6250 train_loss:1.9944 aux_loss:0.0250 train_time:3347544ms step_avg:538.80ms step:6224/6250 train_loss:2.0766 aux_loss:0.0251 train_time:3348081ms step_avg:538.80ms step:6225/6250 train_loss:2.0370 aux_loss:0.0248 train_time:3348618ms step_avg:538.80ms step:6226/6250 train_loss:2.0788 aux_loss:0.0249 train_time:3349155ms step_avg:538.80ms step:6227/6250 train_loss:2.0718 aux_loss:0.0249 train_time:3349695ms step_avg:538.80ms step:6228/6250 train_loss:1.9620 aux_loss:0.0246 train_time:3350232ms step_avg:538.80ms step:6229/6250 train_loss:2.0220 aux_loss:0.0247 train_time:3350773ms step_avg:538.80ms step:6230/6250 train_loss:1.9434 aux_loss:0.0248 train_time:3351310ms step_avg:538.80ms step:6231/6250 train_loss:1.9869 aux_loss:0.0248 train_time:3351851ms step_avg:538.80ms step:6232/6250 train_loss:2.1094 aux_loss:0.0250 train_time:3352388ms step_avg:538.80ms step:6233/6250 train_loss:2.0874 aux_loss:0.0248 train_time:3352931ms step_avg:538.80ms step:6234/6250 train_loss:2.1100 aux_loss:0.0247 train_time:3353468ms step_avg:538.80ms step:6235/6250 train_loss:1.9802 aux_loss:0.0249 train_time:3354007ms step_avg:538.80ms step:6236/6250 train_loss:2.1439 aux_loss:0.0249 train_time:3354544ms step_avg:538.80ms step:6237/6250 train_loss:2.0842 aux_loss:0.0248 train_time:3355078ms step_avg:538.80ms step:6238/6250 train_loss:2.0981 aux_loss:0.0251 train_time:3355615ms step_avg:538.79ms step:6239/6250 train_loss:2.0682 aux_loss:0.0251 train_time:3356149ms step_avg:538.79ms step:6240/6250 train_loss:1.9720 aux_loss:0.0252 train_time:3356684ms step_avg:538.79ms step:6241/6250 train_loss:2.0621 aux_loss:0.0248 train_time:3357223ms step_avg:538.79ms step:6242/6250 train_loss:2.0224 aux_loss:0.0247 train_time:3357763ms step_avg:538.79ms step:6243/6250 train_loss:2.0811 aux_loss:0.0247 train_time:3358299ms step_avg:538.79ms step:6244/6250 train_loss:2.0352 aux_loss:0.0246 train_time:3358837ms step_avg:538.79ms step:6245/6250 train_loss:1.9908 aux_loss:0.0245 train_time:3359378ms step_avg:538.79ms step:6246/6250 train_loss:2.0258 aux_loss:0.0246 train_time:3359916ms step_avg:538.79ms step:6247/6250 train_loss:1.9803 aux_loss:0.0247 train_time:3360456ms step_avg:538.79ms step:6248/6250 train_loss:2.0193 aux_loss:0.0246 train_time:3360993ms step_avg:538.79ms step:6249/6250 train_loss:2.1492 aux_loss:0.0248 train_time:3361530ms step_avg:538.79ms step:6250/6250 train_loss:2.0508 aux_loss:0.0248 train_time:3362065ms step_avg:538.79ms step:6250/6250 val_loss:2.0421 val_aux_loss:0.0248 train_time:3362065ms step_avg:538.79ms