Training in progress, step 8000, checkpoint

checkpoint-8000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f60884bbf39e24ccbc6d9c6042040a45f79c2e0332696986c2d079150509eb4e
+oid sha256:89086835c0c2d119eef6282f2d5d6448a112611a9fe3fca8da1303e0823095a1
 size 932401285

checkpoint-8000/pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d2c8d1471aa541fbbf6abef8b6cc4c08ab52e4140fbb69bf9b440cfee51ec8b7
+oid sha256:358dead2624a149db3e0c5797d0cdf316b7c21f794b878d61af5f12fe6ea5d9c
 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/trainer_state.json

@@ -1,7 +1,7 @@
 {
-  "best_metric": 1.1056212186813354,
-  "best_model_checkpoint": "/kaggle/output/checkpoint-6000",
-  "epoch": 128.0,
+  "best_metric": 1.094658613204956,
+  "best_model_checkpoint": "/kaggle/output/checkpoint-8000",
+  "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": 1.5477,
+      "loss": 1.1383,
       "step": 1
     },
     {
-      "epoch": 16.0,
+      "epoch": 0.04,
       "learning_rate": 2.7750000000000004e-08,
-      "loss": 1.2028,
+      "loss": 1.1424,
       "step": 1000
     },
     {
-      "epoch": 16.0,
-      "eval_accuracy": 0.3387225548902196,
-      "eval_loss": 1.1078879833221436,
-      "eval_runtime": 54.1057,
-      "eval_samples_per_second": 92.596,
-      "eval_steps_per_second": 11.588,
+      "epoch": 0.04,
+      "eval_accuracy": 0.32375249500998005,
+      "eval_loss": 1.1077626943588257,
+      "eval_runtime": 54.8633,
+      "eval_samples_per_second": 91.318,
+      "eval_steps_per_second": 11.428,
       "step": 1000
     },
     {
-      "epoch": 32.0,
+      "epoch": 0.08,
       "learning_rate": 5.5527777777777784e-08,
-      "loss": 1.1152,
+      "loss": 1.1244,
       "step": 2000
     },
     {
-      "epoch": 32.0,
-      "eval_accuracy": 0.3349301397205589,
-      "eval_loss": 1.1080944538116455,
-      "eval_runtime": 54.2029,
-      "eval_samples_per_second": 92.43,
-      "eval_steps_per_second": 11.568,
+      "epoch": 0.08,
+      "eval_accuracy": 0.33652694610778444,
+      "eval_loss": 1.1080161333084106,
+      "eval_runtime": 54.7384,
+      "eval_samples_per_second": 91.526,
+      "eval_steps_per_second": 11.454,
       "step": 2000
     },
     {
-      "epoch": 48.0,
+      "epoch": 0.12,
       "learning_rate": 8.327777777777778e-08,
-      "loss": 1.1036,
+      "loss": 1.1228,
       "step": 3000
     },
     {
-      "epoch": 48.0,
-      "eval_accuracy": 0.3377245508982036,
-      "eval_loss": 1.1081210374832153,
-      "eval_runtime": 54.0467,
-      "eval_samples_per_second": 92.698,
-      "eval_steps_per_second": 11.601,
+      "epoch": 0.12,
+      "eval_accuracy": 0.34331337325349304,
+      "eval_loss": 1.1084064245224,
+      "eval_runtime": 54.7948,
+      "eval_samples_per_second": 91.432,
+      "eval_steps_per_second": 11.443,
       "step": 3000
     },
     {
-      "epoch": 64.0,
+      "epoch": 0.16,
       "learning_rate": 1.1105555555555557e-07,
-      "loss": 1.0895,
+      "loss": 1.1216,
       "step": 4000
     },
     {
-      "epoch": 64.0,
-      "eval_accuracy": 0.33912175648702597,
-      "eval_loss": 1.1072543859481812,
-      "eval_runtime": 54.1318,
-      "eval_samples_per_second": 92.552,
-      "eval_steps_per_second": 11.583,
+      "epoch": 0.16,
+      "eval_accuracy": 0.3385229540918164,
+      "eval_loss": 1.1014840602874756,
+      "eval_runtime": 54.8508,
+      "eval_samples_per_second": 91.339,
+      "eval_steps_per_second": 11.431,
       "step": 4000
     },
     {
-      "epoch": 80.0,
+      "epoch": 0.2,
       "learning_rate": 1.3880555555555558e-07,
-      "loss": 1.0748,
+      "loss": 1.1181,
       "step": 5000
     },
     {
-      "epoch": 80.0,
-      "eval_accuracy": 0.3435129740518962,
-      "eval_loss": 1.1071943044662476,
-      "eval_runtime": 54.4121,
-      "eval_samples_per_second": 92.075,
-      "eval_steps_per_second": 11.523,
+      "epoch": 0.2,
+      "eval_accuracy": 0.33073852295409184,
+      "eval_loss": 1.1008135080337524,
+      "eval_runtime": 54.8304,
+      "eval_samples_per_second": 91.373,
+      "eval_steps_per_second": 11.435,
       "step": 5000
     },
     {
-      "epoch": 96.0,
+      "epoch": 0.24,
       "learning_rate": 1.6658333333333335e-07,
-      "loss": 1.055,
+      "loss": 1.1132,
       "step": 6000
     },
     {
-      "epoch": 96.0,
-      "eval_accuracy": 0.3459081836327345,
-      "eval_loss": 1.1056212186813354,
-      "eval_runtime": 54.092,
-      "eval_samples_per_second": 92.62,
-      "eval_steps_per_second": 11.591,
+      "epoch": 0.24,
+      "eval_accuracy": 0.3520958083832335,
+      "eval_loss": 1.0993762016296387,
+      "eval_runtime": 54.8804,
+      "eval_samples_per_second": 91.289,
+      "eval_steps_per_second": 11.425,
       "step": 6000
     },
     {
-      "epoch": 112.0,
+      "epoch": 0.29,
       "learning_rate": 1.9433333333333334e-07,
-      "loss": 1.0352,
+      "loss": 1.1113,
       "step": 7000
     },
     {
-      "epoch": 112.0,
-      "eval_accuracy": 0.3469061876247505,
-      "eval_loss": 1.1121071577072144,
-      "eval_runtime": 53.9318,
-      "eval_samples_per_second": 92.895,
-      "eval_steps_per_second": 11.626,
+      "epoch": 0.29,
+      "eval_accuracy": 0.3530938123752495,
+      "eval_loss": 1.0965770483016968,
+      "eval_runtime": 54.8881,
+      "eval_samples_per_second": 91.277,
+      "eval_steps_per_second": 11.423,
       "step": 7000
     },
     {
-      "epoch": 128.0,
+      "epoch": 0.33,
       "learning_rate": 2.2211111111111114e-07,
-      "loss": 1.0085,
+      "loss": 1.1111,
       "step": 8000
     },
     {
-      "epoch": 128.0,
-      "eval_accuracy": 0.3534930139720559,
-      "eval_loss": 1.1201248168945312,
-      "eval_runtime": 54.0094,
-      "eval_samples_per_second": 92.762,
-      "eval_steps_per_second": 11.609,
+      "epoch": 0.33,
+      "eval_accuracy": 0.35708582834331337,
+      "eval_loss": 1.094658613204956,
+      "eval_runtime": 54.8233,
+      "eval_samples_per_second": 91.384,
+      "eval_steps_per_second": 11.437,
       "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 sha256:6c09d25f832c6c38e287ec7a7fe1b191faf603815bb3b91e32f16768a2d7d3ff
+oid sha256:a9a52921b15d7eeb53b158a9c0f943d88288ccd9232004774ffdf3a600a048d5
 size 4091