Training in progress, step 8000, checkpoint

checkpoint-8000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:bd920ed307800f9676e02b0806cff62fa720bee74840a6b0c339883669378db5
+oid sha256:54857ea4b2765e1fac348ffc35f0ee072db20cf16331ad2d3ee78298488ae5b9
 size 932401285

checkpoint-8000/pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:76a22100d07bea2c4f116417824e2691230d656c0197a2c6949aad03ae04169a
+oid sha256:db88959cf4153c6f9beee27af384c388c511e76998298abf16f8bf5a59595df3
 size 466205145

checkpoint-8000/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:211f987caa5ad563a0b3ef4910b525ce01acc34ffe45c63b8119c5c594446aca
+oid sha256:a1d423a747d5b5bf7a577919b35661a07b4fe7a0608f44b2fc46cfc79c2e2582
 size 14575

checkpoint-8000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f6bb0ff1a7db8ba294d63d8562a057da1c3b31916a36feee45749afe605cb4d2
+oid sha256:0c17c3a677e0eebd8947dbeeef1a40427619e475f0866c6ff1eface6be3236c8
 size 627

checkpoint-8000/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 1.1167163848876953,
+  "best_metric": 1.1031057834625244,
   "best_model_checkpoint": "/kaggle/output/checkpoint-8000",
-  "epoch": 128.0,
+  "epoch": 0.3259452411994785,
   "eval_steps": 1000,
   "global_step": 8000,
   "is_hyper_param_search": false,
@@ -9,135 +9,135 @@
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.02,
+      "epoch": 0.0,
       "learning_rate": 2.7777777777777777e-11,
-      "loss": 1.2395,
+      "loss": 1.029,
       "step": 1
     },
     {
-      "epoch": 16.0,
+      "epoch": 0.04,
       "learning_rate": 2.7638888888888893e-08,
-      "loss": 1.1422,
+      "loss": 1.181,
       "step": 1000
     },
     {
-      "epoch": 16.0,
-      "eval_accuracy": 0.33073852295409184,
-      "eval_loss": 1.1557353734970093,
-      "eval_runtime": 54.1805,
-      "eval_samples_per_second": 92.469,
-      "eval_steps_per_second": 11.572,
+      "epoch": 0.04,
+      "eval_accuracy": 0.3273453093812375,
+      "eval_loss": 1.1529844999313354,
+      "eval_runtime": 54.2837,
+      "eval_samples_per_second": 92.293,
+      "eval_steps_per_second": 11.55,
       "step": 1000
     },
     {
-      "epoch": 32.0,
+      "epoch": 0.08,
       "learning_rate": 5.541666666666667e-08,
-      "loss": 1.1401,
+      "loss": 1.1527,
       "step": 2000
     },
     {
-      "epoch": 32.0,
-      "eval_accuracy": 0.3311377245508982,
-      "eval_loss": 1.1507551670074463,
-      "eval_runtime": 53.98,
-      "eval_samples_per_second": 92.812,
-      "eval_steps_per_second": 11.615,
+      "epoch": 0.08,
+      "eval_accuracy": 0.33013972055888224,
+      "eval_loss": 1.1351025104522705,
+      "eval_runtime": 54.0518,
+      "eval_samples_per_second": 92.689,
+      "eval_steps_per_second": 11.6,
       "step": 2000
     },
     {
-      "epoch": 48.0,
+      "epoch": 0.12,
       "learning_rate": 8.316666666666666e-08,
-      "loss": 1.1287,
+      "loss": 1.142,
       "step": 3000
     },
     {
-      "epoch": 48.0,
-      "eval_accuracy": 0.3321357285429142,
-      "eval_loss": 1.1441665887832642,
-      "eval_runtime": 53.9873,
-      "eval_samples_per_second": 92.8,
-      "eval_steps_per_second": 11.614,
+      "epoch": 0.12,
+      "eval_accuracy": 0.3317365269461078,
+      "eval_loss": 1.127414345741272,
+      "eval_runtime": 54.0871,
+      "eval_samples_per_second": 92.628,
+      "eval_steps_per_second": 11.592,
       "step": 3000
     },
     {
-      "epoch": 64.0,
-      "learning_rate": 1.1094444444444445e-07,
-      "loss": 1.117,
+      "epoch": 0.16,
+      "learning_rate": 1.1091666666666668e-07,
+      "loss": 1.1371,
       "step": 4000
     },
     {
-      "epoch": 64.0,
-      "eval_accuracy": 0.3287425149700599,
-      "eval_loss": 1.134506344795227,
-      "eval_runtime": 54.0067,
-      "eval_samples_per_second": 92.766,
-      "eval_steps_per_second": 11.61,
+      "epoch": 0.16,
+      "eval_accuracy": 0.331936127744511,
+      "eval_loss": 1.1194497346878052,
+      "eval_runtime": 54.3907,
+      "eval_samples_per_second": 92.111,
+      "eval_steps_per_second": 11.528,
       "step": 4000
     },
     {
-      "epoch": 80.0,
-      "learning_rate": 1.3872222222222222e-07,
-      "loss": 1.1111,
+      "epoch": 0.2,
+      "learning_rate": 1.3869444444444447e-07,
+      "loss": 1.1246,
       "step": 5000
     },
     {
-      "epoch": 80.0,
-      "eval_accuracy": 0.32954091816367265,
-      "eval_loss": 1.1283142566680908,
-      "eval_runtime": 54.0921,
-      "eval_samples_per_second": 92.62,
-      "eval_steps_per_second": 11.591,
+      "epoch": 0.2,
+      "eval_accuracy": 0.3345309381237525,
+      "eval_loss": 1.1131775379180908,
+      "eval_runtime": 54.3492,
+      "eval_samples_per_second": 92.182,
+      "eval_steps_per_second": 11.537,
       "step": 5000
     },
     {
-      "epoch": 96.0,
-      "learning_rate": 1.6650000000000002e-07,
-      "loss": 1.1053,
+      "epoch": 0.24,
+      "learning_rate": 1.664722222222222e-07,
+      "loss": 1.1201,
       "step": 6000
     },
     {
-      "epoch": 96.0,
-      "eval_accuracy": 0.32894211576846305,
-      "eval_loss": 1.1218452453613281,
-      "eval_runtime": 54.0607,
-      "eval_samples_per_second": 92.674,
-      "eval_steps_per_second": 11.598,
+      "epoch": 0.24,
+      "eval_accuracy": 0.34151696606786425,
+      "eval_loss": 1.1093512773513794,
+      "eval_runtime": 54.3536,
+      "eval_samples_per_second": 92.174,
+      "eval_steps_per_second": 11.536,
       "step": 6000
     },
     {
-      "epoch": 112.0,
-      "learning_rate": 1.9422222222222223e-07,
-      "loss": 1.0952,
+      "epoch": 0.29,
+      "learning_rate": 1.9425e-07,
+      "loss": 1.115,
       "step": 7000
     },
     {
-      "epoch": 112.0,
-      "eval_accuracy": 0.33073852295409184,
-      "eval_loss": 1.118302583694458,
-      "eval_runtime": 53.9252,
-      "eval_samples_per_second": 92.906,
-      "eval_steps_per_second": 11.627,
+      "epoch": 0.29,
+      "eval_accuracy": 0.34510978043912177,
+      "eval_loss": 1.1066845655441284,
+      "eval_runtime": 53.9977,
+      "eval_samples_per_second": 92.782,
+      "eval_steps_per_second": 11.612,
       "step": 7000
     },
     {
-      "epoch": 128.0,
-      "learning_rate": 2.2197222222222225e-07,
-      "loss": 1.0909,
+      "epoch": 0.33,
+      "learning_rate": 2.2202777777777778e-07,
+      "loss": 1.1125,
       "step": 8000
     },
     {
-      "epoch": 128.0,
-      "eval_accuracy": 0.3281437125748503,
-      "eval_loss": 1.1167163848876953,
-      "eval_runtime": 54.095,
-      "eval_samples_per_second": 92.615,
-      "eval_steps_per_second": 11.591,
+      "epoch": 0.33,
+      "eval_accuracy": 0.3449101796407186,
+      "eval_loss": 1.1031057834625244,
+      "eval_runtime": 54.2562,
+      "eval_samples_per_second": 92.34,
+      "eval_steps_per_second": 11.556,
       "step": 8000
     }
   ],
   "logging_steps": 1000,
   "max_steps": 10000000,
-  "num_train_epochs": 161291,
+  "num_train_epochs": 408,
   "save_steps": 1000,
   "total_flos": 1.6722841042944e+16,
   "trial_name": null,

checkpoint-8000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:cb2cd1b7e648a386ccb79e47a451842e11001981c4363e993b678e0818db5cab
+oid sha256:5520446b18da09f4a43173b75839457981a344c0e709fd4bbcb249e3f60c859e
 size 4091