Training in progress, step 2460, checkpoint

checkpoint-2460/trainer_state.json

@@ -9,36 +9,71 @@
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
+    {
+      "epoch": 0.0020325203252032522,
+      "grad_norm": 28.81853675842285,
+      "learning_rate": 0.0,
+      "loss": 10.3353,
+      "step": 1
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.4544554948806763,
+      "learning_rate": 9.82e-05,
+      "loss": 4.8617,
+      "step": 492
+    },
     {
       "epoch": 1.0,
       "eval_loss": 3.9908180236816406,
-      "eval_runtime": 66.5009,
-      "eval_samples_per_second": 504.745,
-      "eval_steps_per_second": 1.985,
+      "eval_runtime": 66.4665,
+      "eval_samples_per_second": 505.006,
+      "eval_steps_per_second": 1.986,
       "step": 492
     },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.4267009496688843,
+      "learning_rate": 7.529411764705883e-05,
+      "loss": 3.5249,
+      "step": 984
+    },
     {
       "epoch": 2.0,
       "eval_loss": 3.653867483139038,
-      "eval_runtime": 66.5302,
-      "eval_samples_per_second": 504.523,
-      "eval_steps_per_second": 1.984,
+      "eval_runtime": 66.5662,
+      "eval_samples_per_second": 504.25,
+      "eval_steps_per_second": 1.983,
       "step": 984
     },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.5028904676437378,
+      "learning_rate": 5.0127877237851665e-05,
+      "loss": 3.257,
+      "step": 1476
+    },
     {
       "epoch": 3.0,
       "eval_loss": 3.5320048332214355,
-      "eval_runtime": 66.5567,
-      "eval_samples_per_second": 504.322,
-      "eval_steps_per_second": 1.983,
+      "eval_runtime": 66.4925,
+      "eval_samples_per_second": 504.809,
+      "eval_steps_per_second": 1.985,
       "step": 1476
     },
+    {
+      "epoch": 4.0,
+      "grad_norm": 1.6736714839935303,
+      "learning_rate": 2.4961636828644502e-05,
+      "loss": 3.1208,
+      "step": 1968
+    },
     {
       "epoch": 4.0,
       "eval_loss": 3.473982334136963,
-      "eval_runtime": 66.5024,
-      "eval_samples_per_second": 504.734,
-      "eval_steps_per_second": 1.985,
+      "eval_runtime": 66.5329,
+      "eval_samples_per_second": 504.503,
+      "eval_steps_per_second": 1.984,
       "step": 1968
     }
   ],

checkpoint-2460/training_args.bin

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:7be473c73c39056482d0a43c66eaed18208f6dc8449be0dc16aabad059cd82cc
+oid sha256:3290c0b51af746ef3fa5659adc5fa025f21c3647c818c43be3ade238b5ea6a1e
 size 5969