Training in progress, step 100, checkpoint

checkpoint-100/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:96df8dd678de7c9d23cf5f8dda9133a83dd98dfb460ea95a4fb2dc22d8752365
+oid sha256:656ed54fe0a8b84aa66db81f903d3ddbb3aca7edcab5c1769c3d44f87fe90939
 size 1474893317

checkpoint-100/pytorch_model.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:4141ef29c5d1a3d0269840b8d7cb6b4cbd2f3c34a45521eb10df3e8f3d1ce399
+oid sha256:057eda2c682bf9d9c903eb92f7c38cd743ca463e0c19fef70ca36d331551a8b1
 size 737457141

checkpoint-100/rng_state.pth

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:5ee42ba47149d52ce0f1b0e713f48f980d6f585d6f1d401fc0f38d3db1416b62
+oid sha256:8e9624411eb0f1fd660261f3e4d9bcc1fe5110e59f1a1401372c693dafb1d75e
 size 14575

checkpoint-100/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6722ec8c45928bf7d3dc083d89fb6e8c1e5f98313ddbf4c3c80ba8ab0cf5d2ef
+oid sha256:20769b67f2ef261d19043e9e784bd3f0a87b8b69b311de1d28c2468463bc32ba
 size 627

checkpoint-100/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.6272028088569641,
+  "best_metric": 0.5696164965629578,
   "best_model_checkpoint": "/kaggle/output/checkpoint-100",
   "epoch": 0.2066115702479339,
   "eval_steps": 50,
@@ -10,34 +10,46 @@
   "log_history": [
     {
       "epoch": 0.0,
-      "learning_rate": 1.6666666666666667e-06,
-      "loss": 0.6979,
+      "learning_rate": 0.0,
+      "loss": 0.7246,
       "step": 1
     },
+    {
+      "epoch": 0.1,
+      "learning_rate": 7.333333333333332e-05,
+      "loss": 0.5982,
+      "step": 50
+    },
     {
       "epoch": 0.1,
       "eval_accuracy": 0.75,
-      "eval_loss": 0.6723740696907043,
-      "eval_runtime": 28.903,
-      "eval_samples_per_second": 40.965,
-      "eval_steps_per_second": 20.482,
+      "eval_loss": 0.6296697854995728,
+      "eval_runtime": 30.5612,
+      "eval_samples_per_second": 38.742,
+      "eval_steps_per_second": 19.371,
       "step": 50
     },
+    {
+      "epoch": 0.21,
+      "learning_rate": 0.00015,
+      "loss": 0.5505,
+      "step": 100
+    },
     {
       "epoch": 0.21,
       "eval_accuracy": 0.75,
-      "eval_loss": 0.6272028088569641,
-      "eval_runtime": 28.756,
-      "eval_samples_per_second": 41.174,
-      "eval_steps_per_second": 20.587,
+      "eval_loss": 0.5696164965629578,
+      "eval_runtime": 30.5816,
+      "eval_samples_per_second": 38.716,
+      "eval_steps_per_second": 19.358,
       "step": 100
     }
   ],
-  "logging_steps": 1000,
+  "logging_steps": 50,
   "max_steps": 1000,
   "num_train_epochs": 3,
   "save_steps": 100,
-  "total_flos": 322260294311640.0,
+  "total_flos": 348571871706840.0,
   "trial_name": null,
   "trial_params": null
 }

checkpoint-100/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fbe5e88ca49040823bdb83dc035b8a4470c58ae105b0482f624b7d48f5e08cbe
+oid sha256:cd0372335b8d69875fb40a5865f6914bab9ff539640d01b439e8f707b1e5a43d
 size 4027