Training in progress, step 500

model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:17028b3e2170ffc8bc9ff89fc5f3cbd55d16101a92c963b6bcb44354f47ba633
+oid sha256:02dab28ae74faec556db50ebb3baf42a4ae9494207b13078539464f114ed0d4a
 size 598898116

run-2/checkpoint-500/model.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:95694a452ae9580a30c6f6be6e487d9e914d35551fcb83551d913999ac22ceaf
+oid sha256:02dab28ae74faec556db50ebb3baf42a4ae9494207b13078539464f114ed0d4a
 size 598898116

run-2/checkpoint-500/optimizer.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b57685a3502807808e441dd0b900238ad7b244a30b29b8bf9d01e36445cac736
+oid sha256:1d0557d4c96698905e1e0ed19104c669866ae3544a52a39ff0f0ddc6af425e19
 size 1197886411

run-2/checkpoint-500/scheduler.pt

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:b5aea2a4c12ecebbc4232d88b1bb48a10a0c81d431fc8be5878e5d1a32c56802
+oid sha256:9a09f9c36dde0a0a453a032a4c54a8e4cc9521c107df7189c063a10d22468170
 size 1465

run-2/checkpoint-500/trainer_state.json

@@ -11,25 +11,25 @@
   "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
     }
   ],
   "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": {
@@ -47,8 +47,8 @@
   "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-500/training_args.bin

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

training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:a5c5aa2e0f6144ff85d55658e39492889cffe974c1ae7f2b9e963705c16ca39d
+oid sha256:58684234268900eb5aa11de51a0372f0fe632929e930e5abd57754c9e23a4eaf
 size 5905