Training in progress, epoch 2

logs/events.out.tfevents.1709879794.6c1c016b0e3e.7263.20

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6e90bed77f9956ae21af7b4d0df26ce52cfe2fd253e88fc6636545811d0691bf
-size 5314
+oid sha256:c01f2e4cfe02a5019440b479d6844751da5b56d81608a9cfa06a0f5018b76476
+size 5984

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6aefede8e46ad5f22b888b8d497f4ec544fbae23a53d6e3df4ece10a06cbc53a
+oid sha256:4220a3eaf50ed5d3bbe14f3e4ea153f2c35d8f8c68e5d905cc4cdbbd981a1649
 size 17549312

run-20/checkpoint-96/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b1ab43cf677527e634592493ebdb4831e4c6e94b58dd754e1fbd5d5cf076b583
+oid sha256:4220a3eaf50ed5d3bbe14f3e4ea153f2c35d8f8c68e5d905cc4cdbbd981a1649
 size 17549312

run-20/checkpoint-96/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ed93744f5af312347e7bc8659ac15341e5e8ed12c37db920a692fb6fe55a6a21
+oid sha256:41f6ac2df2e05780ee4d2b58a440d2f5ea2010074970e164db7982bc857f97d0
 size 35122746

run-20/checkpoint-96/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:89409273acede6541260d3b0a38479cbc50d35519137f6acee5c9ffe76c364f2
+oid sha256:c94e801b174bf6c83c3f244b0bbc8c347b6365d33c2615ab0127a12a74c7a137
 size 1064

run-20/checkpoint-96/trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.7954990215264188,
+  "best_metric": 0.8170254403131115,
   "best_model_checkpoint": "tiny-bert-sst2-distilled/run-20/checkpoint-96",
   "epoch": 2.0,
   "eval_steps": 500,
@@ -10,56 +10,55 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "grad_norm": 1.460697889328003,
-      "learning_rate": 0.00025858582035056794,
-      "loss": 0.5513,
+      "grad_norm": 1.930127739906311,
+      "learning_rate": 0.00038924925342110863,
+      "loss": 0.5828,
       "step": 48
     },
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.773972602739726,
-      "eval_f1": 0.8053917438921651,
-      "eval_loss": 0.46848487854003906,
-      "eval_precision": 0.7071005917159763,
-      "eval_recall": 0.9354207436399217,
-      "eval_runtime": 28.0493,
-      "eval_samples_per_second": 36.436,
-      "eval_steps_per_second": 1.141,
+      "eval_accuracy": 0.7553816046966731,
+      "eval_f1": 0.7967479674796748,
+      "eval_loss": 0.4942302703857422,
+      "eval_precision": 0.6815020862308763,
+      "eval_recall": 0.958904109589041,
+      "eval_runtime": 27.5665,
+      "eval_samples_per_second": 37.074,
+      "eval_steps_per_second": 0.58,
       "step": 48
     },
     {
       "epoch": 2.0,
-      "grad_norm": 1.8203953504562378,
-      "learning_rate": 0.0,
-      "loss": 0.4771,
+      "grad_norm": 4.047781944274902,
+      "learning_rate": 0.0002594995022807391,
+      "loss": 0.4792,
       "step": 96
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.7954990215264188,
-      "eval_f1": 0.8184187662901825,
-      "eval_loss": 0.4446437954902649,
-      "eval_precision": 0.7359375,
-      "eval_recall": 0.9217221135029354,
-      "eval_runtime": 28.0332,
-      "eval_samples_per_second": 36.457,
-      "eval_steps_per_second": 1.142,
+      "eval_accuracy": 0.8170254403131115,
+      "eval_f1": 0.8270120259019426,
+      "eval_loss": 0.4411511719226837,
+      "eval_precision": 0.7842105263157895,
+      "eval_recall": 0.8747553816046967,
+      "eval_runtime": 27.2029,
+      "eval_samples_per_second": 37.57,
+      "eval_steps_per_second": 0.588,
       "step": 96
     }
   ],
   "logging_steps": 500,
-  "max_steps": 96,
+  "max_steps": 192,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 2,
+  "num_train_epochs": 4,
   "save_steps": 500,
   "total_flos": 471390394560.0,
   "train_batch_size": 64,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.8864140050782419,
-    "learning_rate": 0.0005171716407011359,
-    "num_train_epochs": 2,
-    "per_device_train_batch_size": 64,
-    "temperature": 26
+    "alpha": 0.9873470402997029,
+    "learning_rate": 0.0005189990045614782,
+    "num_train_epochs": 4,
+    "temperature": 29
   }
 }

run-20/checkpoint-96/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7a119276862b85d8c3e2b12e238cf425c26b0c4b7f12496442d4b30b53df88dd
+oid sha256:86ccdfa005aaa3037171e63b21106648822be6f97ddb8788a88937d2ac56456c
 size 4920