Training in progress, epoch 2

logs/events.out.tfevents.1719310313.852b1e905a9a.223.5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:40dd7e9b5670b46dc7a6663a517ea1b9da12a2014e95a3a84bcb900af0806e36
-size 5429
+oid sha256:d9bfa6b7eb660bf1c9c59b52d24d804e5112bc82c9910e223008034395fbfed8
+size 5963

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3ae20a202c55d96dbc7a99eb68b3c49d5f2889c97c0f7b7581a4369c38393b73
+oid sha256:dc0fee5c27b13a0a9c632d2aac5f70e2a09379e669796176adfed370c09bec74
 size 17549312

run-0/checkpoint-1054/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:9ffba682ed7f803f9b0e6756d1d7262c0a8ff7725ee17d25a4400fb904738bf0
+oid sha256:dc0fee5c27b13a0a9c632d2aac5f70e2a09379e669796176adfed370c09bec74
 size 17549312

run-0/checkpoint-1054/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7cc2a85d4ea840d3fcc66aea49ec835a62a0e8cec385ee7e3944eeb287a7e1d9
+oid sha256:180bafd9040bf7f35eca41800b8bc8fea98cfd11101725956e8708ac27fa2b2f
 size 35123898

run-0/checkpoint-1054/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ecce5e800ae20802be0ec9e25ab4e60fcfb940623cf20ae609f503abadbbe480
+oid sha256:100464d33f6556fd1f8f558519b1030d76a9ca5f1f9e07ec8e190d748723f68d
 size 1064

run-0/checkpoint-1054/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.7786697247706422,
+  "best_metric": 0.7809633027522935,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-1054",
   "epoch": 2.0,
   "eval_steps": 500,
@@ -10,41 +10,41 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 5.95961856842041,
-      "learning_rate": 9.55389368279823e-06,
-      "loss": 1.5369,
+      "grad_norm": 12.377176284790039,
+      "learning_rate": 1.5980854192157344e-05,
+      "loss": 3.0218,
       "step": 527
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.7339449541284404,
-      "eval_loss": 1.2773902416229248,
-      "eval_runtime": 2.5957,
-      "eval_samples_per_second": 335.939,
-      "eval_steps_per_second": 2.697,
+      "eval_accuracy": 0.7672018348623854,
+      "eval_loss": 2.191861391067505,
+      "eval_runtime": 2.5906,
+      "eval_samples_per_second": 336.603,
+      "eval_steps_per_second": 2.702,
       "step": 527
     },
     {
       "epoch": 2.0,
-      "grad_norm": 15.252978324890137,
-      "learning_rate": 8.492349940265094e-06,
-      "loss": 1.2159,
+      "grad_norm": Infinity,
+      "learning_rate": 1.2796813034327135e-05,
+      "loss": 2.0797,
       "step": 1054
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7786697247706422,
-      "eval_loss": 1.022659182548523,
-      "eval_runtime": 2.5741,
-      "eval_samples_per_second": 338.753,
-      "eval_steps_per_second": 2.719,
+      "eval_accuracy": 0.7809633027522935,
+      "eval_loss": 1.8230279684066772,
+      "eval_runtime": 2.5719,
+      "eval_samples_per_second": 339.043,
+      "eval_steps_per_second": 2.722,
       "step": 1054
     }
   ],
   "logging_steps": 500,
-  "max_steps": 5270,
+  "max_steps": 3162,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 6,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -62,9 +62,9 @@
   "train_batch_size": 128,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.27608826195592573,
-    "learning_rate": 1.0615437425331367e-05,
-    "num_train_epochs": 10,
-    "temperature": 2
+    "alpha": 0.275210688972374,
+    "learning_rate": 1.9177025030588814e-05,
+    "num_train_epochs": 6,
+    "temperature": 6
   }
 }

run-0/checkpoint-1054/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:c8b6a60f7b85b38fa45cddf1a417ee51250fe5822237403416bf2406ff2cdb84
+oid sha256:15e7b4feae857373f91378a3d0efc15d0ec396bfba71e74c965086843aa6acf4
 size 5176