Training in progress, step 500

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:22f946dd187c71c743ed8a24c383b7621e8b8662ec78fbff6ed001870698a454
+oid sha256:b7ae6dd26bc1c419aab07240966e39546c64989855d6d985e7c22cc08ac52479
 size 598898116

run-3/checkpoint-500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2f4eff725814b13dc937fa55afed34b3497c98d1e48b5dc28cc14726451a72eb
+oid sha256:b7ae6dd26bc1c419aab07240966e39546c64989855d6d985e7c22cc08ac52479
 size 598898116

run-3/checkpoint-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:2f600a4eb072af1b2680d52a56cbd5f7424e22b03cfdb04ef085dfba10999f1f
+oid sha256:95d0e493985ac4f5d54096e2fce84dba3e7b34b3ac185eb1f2e3153c0dbc8f30
 size 1197884026

run-3/checkpoint-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d73e26459c92e195a7738eb8b675e4359a54bd79fe9532dd9d1f04483172c359
+oid sha256:9e1264523e958cf7990dc5f42d876cc12129475c4603804cf66868aaf25c2c24
 size 1064

run-3/checkpoint-500/trainer_state.json

@@ -11,50 +11,50 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.6764516129032258,
-      "eval_loss": 3.1047208309173584,
-      "eval_runtime": 20.0866,
-      "eval_samples_per_second": 154.332,
-      "eval_steps_per_second": 1.643,
+      "eval_accuracy": 0.7574193548387097,
+      "eval_loss": 2.6135873794555664,
+      "eval_runtime": 22.8142,
+      "eval_samples_per_second": 135.88,
+      "eval_steps_per_second": 1.446,
       "step": 159
     },
     {
       "epoch": 1.2578616352201257,
-      "grad_norm": 6.007691383361816,
-      "learning_rate": 1.49937106918239e-05,
-      "loss": 4.485,
+      "grad_norm": 4.528892517089844,
+      "learning_rate": 1.749685534591195e-05,
+      "loss": 3.9724,
       "step": 200
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.892258064516129,
-      "eval_loss": 1.8396068811416626,
-      "eval_runtime": 20.1816,
-      "eval_samples_per_second": 153.605,
-      "eval_steps_per_second": 1.635,
+      "eval_accuracy": 0.92,
+      "eval_loss": 1.4474385976791382,
+      "eval_runtime": 23.449,
+      "eval_samples_per_second": 132.202,
+      "eval_steps_per_second": 1.407,
       "step": 318
     },
     {
       "epoch": 2.5157232704402515,
-      "grad_norm": 6.994930267333984,
-      "learning_rate": 9.962264150943397e-06,
-      "loss": 2.021,
+      "grad_norm": 4.671586036682129,
+      "learning_rate": 1.4981132075471698e-05,
+      "loss": 1.573,
       "step": 400
     },
     {
       "epoch": 3.0,
-      "eval_accuracy": 0.9319354838709677,
-      "eval_loss": 1.3435319662094116,
-      "eval_runtime": 20.2565,
-      "eval_samples_per_second": 153.037,
-      "eval_steps_per_second": 1.629,
+      "eval_accuracy": 0.9429032258064516,
+      "eval_loss": 1.0164119005203247,
+      "eval_runtime": 22.8814,
+      "eval_samples_per_second": 135.481,
+      "eval_steps_per_second": 1.442,
       "step": 477
     }
   ],
   "logging_steps": 200,
-  "max_steps": 795,
+  "max_steps": 1590,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 5,
+  "num_train_epochs": 10,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -72,8 +72,8 @@
   "train_batch_size": 96,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.7221593654978639,
-    "num_train_epochs": 5,
-    "temperature": 12
+    "alpha": 0.44422800356103254,
+    "num_train_epochs": 10,
+    "temperature": 13
   }
 }

run-3/checkpoint-500/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:17adff4f3634e2a8d622b645ea126e53ba766395bed524f0e523cb41d5464456
+oid sha256:9db705550b987db55c9bb3aff5b3ff0fef828763e0672ac943e35e25b8f5e4e9
 size 5368

runs/Sep28_10-21-38_bb2384aee55a/events.out.tfevents.1759058880.bb2384aee55a.76.4

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3fbf21fe3734393b1c0633723a186f1ac72f5c30fa23a977ee63caf9c544fba3
+size 14502

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:59adeb0d3a55bdc7384874777e892e5d0cc98df3d3789609437ba56985f062cb
+oid sha256:9db705550b987db55c9bb3aff5b3ff0fef828763e0672ac943e35e25b8f5e4e9
 size 5368