Training in progress, step 1000

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:02dab28ae74faec556db50ebb3baf42a4ae9494207b13078539464f114ed0d4a
+oid sha256:da708200d7bc0aee3ca398d01b91bfdebb5e2b1eadb96197c64b765419de7cad
 size 598898116

run-2/checkpoint-1000/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8193fee56dc53365e927c2d9fd3ea0bc4cc3a1e83f2ea8b008b1691e9f019cba
+oid sha256:da708200d7bc0aee3ca398d01b91bfdebb5e2b1eadb96197c64b765419de7cad
 size 598898116

run-2/checkpoint-1000/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a365afa7456a10f06515312a47741b6d1acd07735039211c717ca52d97a8d0ef
+oid sha256:5887173a33336117b385484b05ad0c47a48dbddae5aa18cf6507a5f96281ec8c
 size 1197886411

run-2/checkpoint-1000/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:57715a5dc14c56466f04a233671161ef1dbc32c0cd25ac501a889debf2bff9af
+oid sha256:2113d291798b628dfd227e15753b04d5ba59b745b7635da61797987791bd2209
 size 1465

run-2/checkpoint-1000/trainer_state.json

@@ -11,41 +11,41 @@
   "log_history": [
     {
       "epoch": 1.0,
-      "eval_accuracy": 0.9070967741935484,
-      "eval_loss": 1.3142679929733276,
-      "eval_runtime": 17.875,
-      "eval_samples_per_second": 173.427,
-      "eval_steps_per_second": 5.427,
+      "eval_accuracy": 0.8709677419354839,
+      "eval_loss": 2.2650985717773438,
+      "eval_runtime": 17.8499,
+      "eval_samples_per_second": 173.671,
+      "eval_steps_per_second": 5.434,
       "step": 477
     },
     {
       "epoch": 1.0482180293501049,
-      "grad_norm": 6.521761417388916,
-      "learning_rate": 1.4769392033542977e-05,
-      "loss": 2.7869,
+      "grad_norm": 14.103622436523438,
+      "learning_rate": 1.302585604472397e-05,
+      "loss": 6.1632,
       "step": 500
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.9435483870967742,
-      "eval_loss": 0.8224020004272461,
-      "eval_runtime": 17.7944,
-      "eval_samples_per_second": 174.212,
-      "eval_steps_per_second": 5.451,
+      "eval_accuracy": 0.9187096774193548,
+      "eval_loss": 1.4211522340774536,
+      "eval_runtime": 24.025,
+      "eval_samples_per_second": 129.032,
+      "eval_steps_per_second": 4.037,
       "step": 954
     },
     {
       "epoch": 2.0964360587002098,
-      "grad_norm": 3.0208613872528076,
-      "learning_rate": 9.528301886792455e-06,
-      "loss": 0.9064,
+      "grad_norm": 6.999321937561035,
+      "learning_rate": 6.037735849056604e-06,
+      "loss": 1.5867,
       "step": 1000
     }
   ],
   "logging_steps": 500,
-  "max_steps": 1908,
+  "max_steps": 1431,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 4,
+  "num_train_epochs": 3,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -63,8 +63,8 @@
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": {
-    "alpha": 0.856440788642103,
-    "num_train_epochs": 4,
-    "temperature": 8.405670661191898
+    "alpha": 0.8329290526522004,
+    "num_train_epochs": 3,
+    "temperature": 4.156866262168391
   }
 }

run-2/checkpoint-1000/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:3e06f1bb4bc4cbbd72bb74742f00db632f86b3856c71e606f51ec5790460bfd7
+oid sha256:58684234268900eb5aa11de51a0372f0fe632929e930e5abd57754c9e23a4eaf
 size 5905