Training in progress, step 36, checkpoint

last-checkpoint/adapter_config.json

@@ -10,23 +10,23 @@
   "layers_pattern": null,
   "layers_to_transform": null,
   "loftq_config": {},
-  "lora_alpha": 16,
-  "lora_dropout": 0.1,
+  "lora_alpha": 32,
+  "lora_dropout": 0.05,
   "megatron_config": null,
   "megatron_core": "megatron.core",
   "modules_to_save": null,
   "peft_type": "LORA",
-  "r": 8,
+  "r": 16,
   "rank_pattern": {},
   "revision": null,
   "target_modules": [
-    "o_proj",
+    "v_proj",
+    "down_proj",
     "up_proj",
     "k_proj",
-    "q_proj",
-    "down_proj",
+    "o_proj",
     "gate_proj",
-    "v_proj"
+    "q_proj"
   ],
   "task_type": "CAUSAL_LM",
   "use_dora": false,

last-checkpoint/adapter_model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2c646f392aaf5ffe251d79f16f9c5aefbee754426cb4708fe2d5d34551bcc0e6
-size 83945296
+oid sha256:3b6453bdb0395fa8c4d11b15fd7f743cf902cda76f9c56fd48148e48a6c60e32
+size 167832240

last-checkpoint/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:842d68b94184abc7a320914c5753b95302a3a4c76261a2400b65001a749e8d5f
-size 168149074
+oid sha256:2874ab5eb82d434bfc0b3aa8ae48bb0ba8ef59b0416fb0a82bc3ca044821102e
+size 85723284

last-checkpoint/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c03e971f2de1fda980f54f3289028b715b1671cbcd9e283c00c849bd70f82934
+oid sha256:a59f4bef5f627d7afe8f2103773befa851585a000baf07cc5f352d4521942c00
 size 14244

last-checkpoint/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4521b8db9cc205e54aa606d85e707c024abd2d8ad4a20bec4b2cff365dc59cdf
+oid sha256:0aa0f75a1f3e346be25756b578158b09a68943f0b9f1cfe29f97939687f864ef
 size 1064

last-checkpoint/trainer_state.json

@@ -1,266 +1,351 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 0.010778184953653805,
-  "eval_steps": 3,
-  "global_step": 25,
+  "epoch": 0.031041172666522957,
+  "eval_steps": 4,
+  "global_step": 36,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0004311273981461522,
-      "grad_norm": 14.988656044006348,
+      "epoch": 0.0008622547962923044,
+      "grad_norm": 18.349796295166016,
       "learning_rate": 2e-05,
-      "loss": 11.9265,
+      "loss": 12.2382,
       "step": 1
     },
     {
-      "epoch": 0.0004311273981461522,
-      "eval_loss": 12.78434944152832,
-      "eval_runtime": 58.2929,
-      "eval_samples_per_second": 8.389,
-      "eval_steps_per_second": 8.389,
+      "epoch": 0.0008622547962923044,
+      "eval_loss": 12.783363342285156,
+      "eval_runtime": 42.0314,
+      "eval_samples_per_second": 11.634,
+      "eval_steps_per_second": 5.829,
       "step": 1
     },
     {
-      "epoch": 0.0008622547962923044,
-      "grad_norm": 15.124715805053711,
+      "epoch": 0.0017245095925846087,
+      "grad_norm": 18.404314041137695,
       "learning_rate": 4e-05,
-      "loss": 12.5452,
+      "loss": 12.818,
       "step": 2
     },
     {
-      "epoch": 0.0012933821944384565,
-      "grad_norm": 12.851057052612305,
+      "epoch": 0.002586764388876913,
+      "grad_norm": 18.81452178955078,
       "learning_rate": 6e-05,
-      "loss": 13.0835,
+      "loss": 12.2314,
       "step": 3
     },
     {
-      "epoch": 0.0012933821944384565,
-      "eval_loss": 12.626995086669922,
-      "eval_runtime": 58.3375,
-      "eval_samples_per_second": 8.382,
-      "eval_steps_per_second": 8.382,
-      "step": 3
+      "epoch": 0.0034490191851692174,
+      "grad_norm": 18.826942443847656,
+      "learning_rate": 8e-05,
+      "loss": 12.0138,
+      "step": 4
     },
     {
-      "epoch": 0.0017245095925846087,
-      "grad_norm": 15.81566047668457,
-      "learning_rate": 8e-05,
-      "loss": 12.3524,
+      "epoch": 0.0034490191851692174,
+      "eval_loss": 10.487162590026855,
+      "eval_runtime": 41.3199,
+      "eval_samples_per_second": 11.834,
+      "eval_steps_per_second": 5.929,
       "step": 4
     },
     {
-      "epoch": 0.002155636990730761,
-      "grad_norm": 18.372833251953125,
+      "epoch": 0.004311273981461522,
+      "grad_norm": 23.916542053222656,
       "learning_rate": 0.0001,
-      "loss": 11.0613,
+      "loss": 10.8756,
       "step": 5
     },
     {
-      "epoch": 0.002586764388876913,
-      "grad_norm": 21.02024269104004,
+      "epoch": 0.005173528777753826,
+      "grad_norm": 25.37953758239746,
       "learning_rate": 0.00012,
-      "loss": 11.0406,
+      "loss": 7.9933,
       "step": 6
     },
     {
-      "epoch": 0.002586764388876913,
-      "eval_loss": 8.352932929992676,
-      "eval_runtime": 58.4155,
-      "eval_samples_per_second": 8.371,
-      "eval_steps_per_second": 8.371,
-      "step": 6
-    },
-    {
-      "epoch": 0.003017891787023065,
-      "grad_norm": 20.353683471679688,
+      "epoch": 0.00603578357404613,
+      "grad_norm": 36.603450775146484,
       "learning_rate": 0.00014,
-      "loss": 9.0451,
+      "loss": 6.8124,
       "step": 7
     },
     {
-      "epoch": 0.0034490191851692174,
-      "grad_norm": 35.790313720703125,
+      "epoch": 0.006898038370338435,
+      "grad_norm": 39.93467712402344,
       "learning_rate": 0.00016,
-      "loss": 7.4442,
+      "loss": 4.8259,
       "step": 8
     },
     {
-      "epoch": 0.0038801465833153696,
-      "grad_norm": 30.771913528442383,
-      "learning_rate": 0.00018,
-      "loss": 6.2842,
-      "step": 9
+      "epoch": 0.006898038370338435,
+      "eval_loss": 1.5156302452087402,
+      "eval_runtime": 41.4263,
+      "eval_samples_per_second": 11.804,
+      "eval_steps_per_second": 5.914,
+      "step": 8
     },
     {
-      "epoch": 0.0038801465833153696,
-      "eval_loss": 4.413941860198975,
-      "eval_runtime": 58.3924,
-      "eval_samples_per_second": 8.374,
-      "eval_steps_per_second": 8.374,
+      "epoch": 0.007760293166630739,
+      "grad_norm": 24.483657836914062,
+      "learning_rate": 0.00018,
+      "loss": 1.6681,
       "step": 9
     },
     {
-      "epoch": 0.004311273981461522,
-      "grad_norm": 31.092329025268555,
+      "epoch": 0.008622547962923045,
+      "grad_norm": 20.87933921813965,
       "learning_rate": 0.0002,
-      "loss": 4.2657,
+      "loss": 1.6076,
       "step": 10
     },
     {
-      "epoch": 0.004742401379607674,
-      "grad_norm": 24.426212310791016,
-      "learning_rate": 0.00019781476007338058,
-      "loss": 1.2179,
+      "epoch": 0.009484802759215347,
+      "grad_norm": 28.462326049804688,
+      "learning_rate": 0.0001996917333733128,
+      "loss": 0.9284,
       "step": 11
     },
     {
-      "epoch": 0.005173528777753826,
-      "grad_norm": 26.97491455078125,
-      "learning_rate": 0.0001913545457642601,
-      "loss": 1.8928,
+      "epoch": 0.010347057555507652,
+      "grad_norm": 7.382307529449463,
+      "learning_rate": 0.00019876883405951377,
+      "loss": 0.5719,
       "step": 12
     },
     {
-      "epoch": 0.005173528777753826,
-      "eval_loss": 1.3277825117111206,
-      "eval_runtime": 58.4038,
-      "eval_samples_per_second": 8.373,
-      "eval_steps_per_second": 8.373,
+      "epoch": 0.010347057555507652,
+      "eval_loss": 0.8780114650726318,
+      "eval_runtime": 41.4971,
+      "eval_samples_per_second": 11.784,
+      "eval_steps_per_second": 5.904,
       "step": 12
     },
     {
-      "epoch": 0.005604656175899978,
-      "grad_norm": 24.090688705444336,
-      "learning_rate": 0.00018090169943749476,
-      "loss": 2.1082,
+      "epoch": 0.011209312351799956,
+      "grad_norm": 23.78351402282715,
+      "learning_rate": 0.00019723699203976766,
+      "loss": 1.0036,
       "step": 13
     },
     {
-      "epoch": 0.00603578357404613,
-      "grad_norm": 54.64119338989258,
-      "learning_rate": 0.00016691306063588583,
-      "loss": 1.25,
+      "epoch": 0.01207156714809226,
+      "grad_norm": 6.90433406829834,
+      "learning_rate": 0.00019510565162951537,
+      "loss": 0.6748,
       "step": 14
     },
     {
-      "epoch": 0.006466910972192283,
-      "grad_norm": 26.83595848083496,
-      "learning_rate": 0.00015000000000000001,
-      "loss": 1.1507,
+      "epoch": 0.012933821944384565,
+      "grad_norm": 12.185566902160645,
+      "learning_rate": 0.0001923879532511287,
+      "loss": 0.6046,
       "step": 15
     },
     {
-      "epoch": 0.006466910972192283,
-      "eval_loss": 0.8006857633590698,
-      "eval_runtime": 58.359,
-      "eval_samples_per_second": 8.379,
-      "eval_steps_per_second": 8.379,
-      "step": 15
+      "epoch": 0.01379607674067687,
+      "grad_norm": 19.96064567565918,
+      "learning_rate": 0.0001891006524188368,
+      "loss": 0.9423,
+      "step": 16
     },
     {
-      "epoch": 0.006898038370338435,
-      "grad_norm": 18.02296257019043,
-      "learning_rate": 0.00013090169943749476,
-      "loss": 0.9997,
+      "epoch": 0.01379607674067687,
+      "eval_loss": 0.5733368992805481,
+      "eval_runtime": 41.5222,
+      "eval_samples_per_second": 11.777,
+      "eval_steps_per_second": 5.9,
       "step": 16
     },
     {
-      "epoch": 0.007329165768484587,
-      "grad_norm": 21.693586349487305,
-      "learning_rate": 0.00011045284632676536,
-      "loss": 0.8036,
+      "epoch": 0.014658331536969174,
+      "grad_norm": 13.482269287109375,
+      "learning_rate": 0.00018526401643540922,
+      "loss": 0.4834,
       "step": 17
     },
     {
-      "epoch": 0.007760293166630739,
-      "grad_norm": 24.909374237060547,
-      "learning_rate": 8.954715367323468e-05,
-      "loss": 0.8566,
-      "step": 18
-    },
-    {
-      "epoch": 0.007760293166630739,
-      "eval_loss": 0.6211725473403931,
-      "eval_runtime": 58.3093,
-      "eval_samples_per_second": 8.386,
-      "eval_steps_per_second": 8.386,
+      "epoch": 0.015520586333261479,
+      "grad_norm": 22.92808723449707,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.9821,
       "step": 18
     },
     {
-      "epoch": 0.008191420564776892,
-      "grad_norm": 16.71841049194336,
-      "learning_rate": 6.909830056250527e-05,
-      "loss": 0.3249,
+      "epoch": 0.016382841129553783,
+      "grad_norm": 19.03671646118164,
+      "learning_rate": 0.0001760405965600031,
+      "loss": 0.8987,
       "step": 19
     },
     {
-      "epoch": 0.008622547962923045,
-      "grad_norm": 11.01509952545166,
-      "learning_rate": 5.000000000000002e-05,
-      "loss": 0.3442,
+      "epoch": 0.01724509592584609,
+      "grad_norm": 7.670022487640381,
+      "learning_rate": 0.00017071067811865476,
+      "loss": 0.4543,
       "step": 20
     },
     {
-      "epoch": 0.009053675361069196,
-      "grad_norm": 10.443573951721191,
-      "learning_rate": 3.308693936411421e-05,
-      "loss": 0.4075,
-      "step": 21
+      "epoch": 0.01724509592584609,
+      "eval_loss": 0.8335480093955994,
+      "eval_runtime": 41.5035,
+      "eval_samples_per_second": 11.782,
+      "eval_steps_per_second": 5.903,
+      "step": 20
     },
     {
-      "epoch": 0.009053675361069196,
-      "eval_loss": 0.8346744775772095,
-      "eval_runtime": 58.344,
-      "eval_samples_per_second": 8.381,
-      "eval_steps_per_second": 8.381,
+      "epoch": 0.018107350722138392,
+      "grad_norm": 16.6032657623291,
+      "learning_rate": 0.00016494480483301836,
+      "loss": 0.8973,
       "step": 21
     },
     {
-      "epoch": 0.009484802759215347,
-      "grad_norm": 19.735132217407227,
-      "learning_rate": 1.9098300562505266e-05,
-      "loss": 1.2175,
+      "epoch": 0.018969605518430695,
+      "grad_norm": 7.4585185050964355,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 0.4621,
       "step": 22
     },
     {
-      "epoch": 0.0099159301573615,
-      "grad_norm": 15.83383846282959,
-      "learning_rate": 8.645454235739903e-06,
-      "loss": 0.9296,
+      "epoch": 0.019831860314723,
+      "grad_norm": 6.364231586456299,
+      "learning_rate": 0.0001522498564715949,
+      "loss": 0.6574,
       "step": 23
     },
     {
-      "epoch": 0.010347057555507652,
-      "grad_norm": 8.399575233459473,
-      "learning_rate": 2.1852399266194314e-06,
-      "loss": 0.8743,
+      "epoch": 0.020694115111015304,
+      "grad_norm": 20.26422882080078,
+      "learning_rate": 0.00014539904997395468,
+      "loss": 1.3692,
       "step": 24
     },
     {
-      "epoch": 0.010347057555507652,
-      "eval_loss": 0.8398098349571228,
-      "eval_runtime": 58.4574,
-      "eval_samples_per_second": 8.365,
-      "eval_steps_per_second": 8.365,
+      "epoch": 0.020694115111015304,
+      "eval_loss": 1.0001345872879028,
+      "eval_runtime": 41.524,
+      "eval_samples_per_second": 11.776,
+      "eval_steps_per_second": 5.9,
       "step": 24
     },
     {
-      "epoch": 0.010778184953653805,
-      "grad_norm": 19.13532066345215,
-      "learning_rate": 0.0,
-      "loss": 0.8357,
+      "epoch": 0.02155636990730761,
+      "grad_norm": 7.076972961425781,
+      "learning_rate": 0.000138268343236509,
+      "loss": 0.3212,
       "step": 25
+    },
+    {
+      "epoch": 0.022418624703599913,
+      "grad_norm": 6.530824661254883,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.6942,
+      "step": 26
+    },
+    {
+      "epoch": 0.02328087949989222,
+      "grad_norm": 7.959903717041016,
+      "learning_rate": 0.00012334453638559057,
+      "loss": 0.4833,
+      "step": 27
+    },
+    {
+      "epoch": 0.02414313429618452,
+      "grad_norm": 10.797367095947266,
+      "learning_rate": 0.0001156434465040231,
+      "loss": 0.9116,
+      "step": 28
+    },
+    {
+      "epoch": 0.02414313429618452,
+      "eval_loss": 0.6466512680053711,
+      "eval_runtime": 41.5138,
+      "eval_samples_per_second": 11.779,
+      "eval_steps_per_second": 5.902,
+      "step": 28
+    },
+    {
+      "epoch": 0.025005389092476828,
+      "grad_norm": 7.114670276641846,
+      "learning_rate": 0.0001078459095727845,
+      "loss": 0.5407,
+      "step": 29
+    },
+    {
+      "epoch": 0.02586764388876913,
+      "grad_norm": 3.2049789428710938,
+      "learning_rate": 0.0001,
+      "loss": 0.4781,
+      "step": 30
+    },
+    {
+      "epoch": 0.026729898685061437,
+      "grad_norm": 9.187115669250488,
+      "learning_rate": 9.215409042721552e-05,
+      "loss": 0.591,
+      "step": 31
+    },
+    {
+      "epoch": 0.02759215348135374,
+      "grad_norm": 5.867708206176758,
+      "learning_rate": 8.435655349597689e-05,
+      "loss": 0.7579,
+      "step": 32
+    },
+    {
+      "epoch": 0.02759215348135374,
+      "eval_loss": 0.8243234753608704,
+      "eval_runtime": 41.5067,
+      "eval_samples_per_second": 11.781,
+      "eval_steps_per_second": 5.903,
+      "step": 32
+    },
+    {
+      "epoch": 0.028454408277646046,
+      "grad_norm": 10.666952133178711,
+      "learning_rate": 7.66554636144095e-05,
+      "loss": 0.9866,
+      "step": 33
+    },
+    {
+      "epoch": 0.029316663073938348,
+      "grad_norm": 4.174076080322266,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 0.6182,
+      "step": 34
+    },
+    {
+      "epoch": 0.030178917870230654,
+      "grad_norm": 3.677931070327759,
+      "learning_rate": 6.173165676349103e-05,
+      "loss": 0.5182,
+      "step": 35
+    },
+    {
+      "epoch": 0.031041172666522957,
+      "grad_norm": 3.110746145248413,
+      "learning_rate": 5.4600950026045326e-05,
+      "loss": 0.4857,
+      "step": 36
+    },
+    {
+      "epoch": 0.031041172666522957,
+      "eval_loss": 0.6007174849510193,
+      "eval_runtime": 41.5227,
+      "eval_samples_per_second": 11.777,
+      "eval_steps_per_second": 5.9,
+      "step": 36
     }
   ],
   "logging_steps": 1,
-  "max_steps": 25,
+  "max_steps": 50,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 1,
-  "save_steps": 3,
+  "save_steps": 4,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -268,13 +353,13 @@
         "should_evaluate": false,
         "should_log": false,
         "should_save": true,
-        "should_training_stop": true
+        "should_training_stop": false
       },
       "attributes": {}
     }
   },
-  "total_flos": 9247821240729600.0,
-  "train_batch_size": 1,
+  "total_flos": 1.3353971104088064e+16,
+  "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null
 }

last-checkpoint/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c1c25b60f410e4ec79e489229656c7b2a433ea4b96521a9df498c8e899096118
+oid sha256:2068f42f79a870a63a6397e16dda8eebb4ed67ad19583a418d7733a958a03a3d
 size 6776