End of training
Browse files- all_results.json +13 -0
- eval_results.json +8 -0
- train_results.json +8 -0
- trainer_state.json +2125 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.9989842559674962,
|
| 3 |
+
"eval_accuracy": 0.9820952380952381,
|
| 4 |
+
"eval_loss": 0.062325455248355865,
|
| 5 |
+
"eval_runtime": 746.7763,
|
| 6 |
+
"eval_samples_per_second": 28.121,
|
| 7 |
+
"eval_steps_per_second": 0.88,
|
| 8 |
+
"total_flos": 1.3566159569165903e+20,
|
| 9 |
+
"train_loss": 0.14182275843612224,
|
| 10 |
+
"train_runtime": 17938.7006,
|
| 11 |
+
"train_samples_per_second": 21.072,
|
| 12 |
+
"train_steps_per_second": 0.165
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.9989842559674962,
|
| 3 |
+
"eval_accuracy": 0.9820952380952381,
|
| 4 |
+
"eval_loss": 0.062325455248355865,
|
| 5 |
+
"eval_runtime": 746.7763,
|
| 6 |
+
"eval_samples_per_second": 28.121,
|
| 7 |
+
"eval_steps_per_second": 0.88
|
| 8 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.9989842559674962,
|
| 3 |
+
"total_flos": 1.3566159569165903e+20,
|
| 4 |
+
"train_loss": 0.14182275843612224,
|
| 5 |
+
"train_runtime": 17938.7006,
|
| 6 |
+
"train_samples_per_second": 21.072,
|
| 7 |
+
"train_steps_per_second": 0.165
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2125 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9820952380952381,
|
| 3 |
+
"best_model_checkpoint": "dinov2-large-cluster-finetune-linear-probe-trueface/checkpoint-2952",
|
| 4 |
+
"epoch": 1.9989842559674962,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2952,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.006771626883358727,
|
| 13 |
+
"grad_norm": 5.265126705169678,
|
| 14 |
+
"learning_rate": 1.6891891891891894e-06,
|
| 15 |
+
"loss": 0.8799,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.013543253766717453,
|
| 20 |
+
"grad_norm": 6.357407093048096,
|
| 21 |
+
"learning_rate": 3.3783783783783788e-06,
|
| 22 |
+
"loss": 0.8361,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.02031488065007618,
|
| 27 |
+
"grad_norm": 4.737895965576172,
|
| 28 |
+
"learning_rate": 5.067567567567568e-06,
|
| 29 |
+
"loss": 0.8673,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.027086507533434907,
|
| 34 |
+
"grad_norm": 5.784124374389648,
|
| 35 |
+
"learning_rate": 6.7567567567567575e-06,
|
| 36 |
+
"loss": 0.87,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.03385813441679363,
|
| 41 |
+
"grad_norm": 6.456064701080322,
|
| 42 |
+
"learning_rate": 8.445945945945946e-06,
|
| 43 |
+
"loss": 0.8188,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.04062976130015236,
|
| 48 |
+
"grad_norm": 6.111750602722168,
|
| 49 |
+
"learning_rate": 1.0135135135135136e-05,
|
| 50 |
+
"loss": 0.8241,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.04740138818351109,
|
| 55 |
+
"grad_norm": 4.581135272979736,
|
| 56 |
+
"learning_rate": 1.1824324324324325e-05,
|
| 57 |
+
"loss": 0.8163,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.05417301506686981,
|
| 62 |
+
"grad_norm": 4.400207996368408,
|
| 63 |
+
"learning_rate": 1.3513513513513515e-05,
|
| 64 |
+
"loss": 0.8066,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.06094464195022854,
|
| 69 |
+
"grad_norm": 4.919775009155273,
|
| 70 |
+
"learning_rate": 1.5202702702702704e-05,
|
| 71 |
+
"loss": 0.7875,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.06771626883358726,
|
| 76 |
+
"grad_norm": 4.747180461883545,
|
| 77 |
+
"learning_rate": 1.6891891891891892e-05,
|
| 78 |
+
"loss": 0.7627,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.074487895716946,
|
| 83 |
+
"grad_norm": 4.32789421081543,
|
| 84 |
+
"learning_rate": 1.8581081081081082e-05,
|
| 85 |
+
"loss": 0.7603,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.08125952260030472,
|
| 90 |
+
"grad_norm": 4.564222812652588,
|
| 91 |
+
"learning_rate": 2.0270270270270273e-05,
|
| 92 |
+
"loss": 0.7323,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.08803114948366345,
|
| 97 |
+
"grad_norm": 4.640108108520508,
|
| 98 |
+
"learning_rate": 2.195945945945946e-05,
|
| 99 |
+
"loss": 0.7282,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.09480277636702218,
|
| 104 |
+
"grad_norm": 4.069850921630859,
|
| 105 |
+
"learning_rate": 2.364864864864865e-05,
|
| 106 |
+
"loss": 0.6938,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.1015744032503809,
|
| 111 |
+
"grad_norm": 4.110805988311768,
|
| 112 |
+
"learning_rate": 2.533783783783784e-05,
|
| 113 |
+
"loss": 0.6607,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.10834603013373963,
|
| 118 |
+
"grad_norm": 4.257925510406494,
|
| 119 |
+
"learning_rate": 2.702702702702703e-05,
|
| 120 |
+
"loss": 0.6171,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.11511765701709836,
|
| 125 |
+
"grad_norm": 3.9571423530578613,
|
| 126 |
+
"learning_rate": 2.8716216216216217e-05,
|
| 127 |
+
"loss": 0.6084,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.12188928390045708,
|
| 132 |
+
"grad_norm": 3.1258833408355713,
|
| 133 |
+
"learning_rate": 3.0405405405405407e-05,
|
| 134 |
+
"loss": 0.5765,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.1286609107838158,
|
| 139 |
+
"grad_norm": 3.235241651535034,
|
| 140 |
+
"learning_rate": 3.20945945945946e-05,
|
| 141 |
+
"loss": 0.5534,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.13543253766717453,
|
| 146 |
+
"grad_norm": 4.166315078735352,
|
| 147 |
+
"learning_rate": 3.3783783783783784e-05,
|
| 148 |
+
"loss": 0.5382,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.14220416455053325,
|
| 153 |
+
"grad_norm": 3.878629446029663,
|
| 154 |
+
"learning_rate": 3.547297297297297e-05,
|
| 155 |
+
"loss": 0.5036,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.148975791433892,
|
| 160 |
+
"grad_norm": 2.8281824588775635,
|
| 161 |
+
"learning_rate": 3.7162162162162165e-05,
|
| 162 |
+
"loss": 0.4804,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.15574741831725072,
|
| 167 |
+
"grad_norm": 2.894228219985962,
|
| 168 |
+
"learning_rate": 3.885135135135135e-05,
|
| 169 |
+
"loss": 0.4522,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.16251904520060945,
|
| 174 |
+
"grad_norm": 2.7571442127227783,
|
| 175 |
+
"learning_rate": 4.0540540540540545e-05,
|
| 176 |
+
"loss": 0.4496,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.16929067208396817,
|
| 181 |
+
"grad_norm": 2.831394672393799,
|
| 182 |
+
"learning_rate": 4.222972972972973e-05,
|
| 183 |
+
"loss": 0.412,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.1760622989673269,
|
| 188 |
+
"grad_norm": 2.329063892364502,
|
| 189 |
+
"learning_rate": 4.391891891891892e-05,
|
| 190 |
+
"loss": 0.386,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.1828339258506856,
|
| 195 |
+
"grad_norm": 2.811448812484741,
|
| 196 |
+
"learning_rate": 4.560810810810811e-05,
|
| 197 |
+
"loss": 0.3755,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.18960555273404436,
|
| 202 |
+
"grad_norm": 2.9765515327453613,
|
| 203 |
+
"learning_rate": 4.72972972972973e-05,
|
| 204 |
+
"loss": 0.3544,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.19637717961740309,
|
| 209 |
+
"grad_norm": 2.6609086990356445,
|
| 210 |
+
"learning_rate": 4.8986486486486486e-05,
|
| 211 |
+
"loss": 0.3317,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.2031488065007618,
|
| 216 |
+
"grad_norm": 4.433963298797607,
|
| 217 |
+
"learning_rate": 4.992469879518072e-05,
|
| 218 |
+
"loss": 0.3321,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.20992043338412053,
|
| 223 |
+
"grad_norm": 1.8786640167236328,
|
| 224 |
+
"learning_rate": 4.9736445783132535e-05,
|
| 225 |
+
"loss": 0.3046,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.21669206026747925,
|
| 230 |
+
"grad_norm": 2.7514021396636963,
|
| 231 |
+
"learning_rate": 4.954819277108434e-05,
|
| 232 |
+
"loss": 0.2974,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.22346368715083798,
|
| 237 |
+
"grad_norm": 3.1674134731292725,
|
| 238 |
+
"learning_rate": 4.9359939759036146e-05,
|
| 239 |
+
"loss": 0.2762,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.23023531403419673,
|
| 244 |
+
"grad_norm": 1.7846009731292725,
|
| 245 |
+
"learning_rate": 4.917168674698795e-05,
|
| 246 |
+
"loss": 0.2604,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.23700694091755545,
|
| 251 |
+
"grad_norm": 1.7402037382125854,
|
| 252 |
+
"learning_rate": 4.898343373493976e-05,
|
| 253 |
+
"loss": 0.263,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.24377856780091417,
|
| 258 |
+
"grad_norm": 1.480891466140747,
|
| 259 |
+
"learning_rate": 4.879518072289157e-05,
|
| 260 |
+
"loss": 0.25,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.2505501946842729,
|
| 265 |
+
"grad_norm": 2.047008514404297,
|
| 266 |
+
"learning_rate": 4.8606927710843376e-05,
|
| 267 |
+
"loss": 0.2281,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.2573218215676316,
|
| 272 |
+
"grad_norm": 1.3639116287231445,
|
| 273 |
+
"learning_rate": 4.841867469879519e-05,
|
| 274 |
+
"loss": 0.2264,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.26409344845099036,
|
| 279 |
+
"grad_norm": 1.8563684225082397,
|
| 280 |
+
"learning_rate": 4.823042168674699e-05,
|
| 281 |
+
"loss": 0.2245,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.27086507533434906,
|
| 286 |
+
"grad_norm": 1.3617446422576904,
|
| 287 |
+
"learning_rate": 4.804216867469879e-05,
|
| 288 |
+
"loss": 0.2177,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.2776367022177078,
|
| 293 |
+
"grad_norm": 1.7790151834487915,
|
| 294 |
+
"learning_rate": 4.7853915662650606e-05,
|
| 295 |
+
"loss": 0.2191,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2844083291010665,
|
| 300 |
+
"grad_norm": 1.2283886671066284,
|
| 301 |
+
"learning_rate": 4.766566265060241e-05,
|
| 302 |
+
"loss": 0.208,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.29117995598442525,
|
| 307 |
+
"grad_norm": 1.2149569988250732,
|
| 308 |
+
"learning_rate": 4.7477409638554224e-05,
|
| 309 |
+
"loss": 0.1974,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.297951582867784,
|
| 314 |
+
"grad_norm": 1.4879294633865356,
|
| 315 |
+
"learning_rate": 4.728915662650602e-05,
|
| 316 |
+
"loss": 0.1905,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.3047232097511427,
|
| 321 |
+
"grad_norm": 1.4393724203109741,
|
| 322 |
+
"learning_rate": 4.710090361445783e-05,
|
| 323 |
+
"loss": 0.194,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.31149483663450145,
|
| 328 |
+
"grad_norm": 1.5028496980667114,
|
| 329 |
+
"learning_rate": 4.691265060240964e-05,
|
| 330 |
+
"loss": 0.1831,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.31826646351786014,
|
| 335 |
+
"grad_norm": 1.403731346130371,
|
| 336 |
+
"learning_rate": 4.672439759036145e-05,
|
| 337 |
+
"loss": 0.193,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.3250380904012189,
|
| 342 |
+
"grad_norm": 1.6740349531173706,
|
| 343 |
+
"learning_rate": 4.653614457831326e-05,
|
| 344 |
+
"loss": 0.1724,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.33180971728457764,
|
| 349 |
+
"grad_norm": 0.9081584215164185,
|
| 350 |
+
"learning_rate": 4.634789156626506e-05,
|
| 351 |
+
"loss": 0.1593,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.33858134416793634,
|
| 356 |
+
"grad_norm": 1.0645991563796997,
|
| 357 |
+
"learning_rate": 4.615963855421687e-05,
|
| 358 |
+
"loss": 0.1611,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.3453529710512951,
|
| 363 |
+
"grad_norm": 2.582812786102295,
|
| 364 |
+
"learning_rate": 4.597138554216868e-05,
|
| 365 |
+
"loss": 0.1558,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.3521245979346538,
|
| 370 |
+
"grad_norm": 1.2780678272247314,
|
| 371 |
+
"learning_rate": 4.578313253012048e-05,
|
| 372 |
+
"loss": 0.1426,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.35889622481801253,
|
| 377 |
+
"grad_norm": 2.387918710708618,
|
| 378 |
+
"learning_rate": 4.5594879518072295e-05,
|
| 379 |
+
"loss": 0.1573,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.3656678517013712,
|
| 384 |
+
"grad_norm": 1.4807089567184448,
|
| 385 |
+
"learning_rate": 4.5406626506024094e-05,
|
| 386 |
+
"loss": 0.1618,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.37243947858473,
|
| 391 |
+
"grad_norm": 1.0865875482559204,
|
| 392 |
+
"learning_rate": 4.5218373493975907e-05,
|
| 393 |
+
"loss": 0.1439,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.3792111054680887,
|
| 398 |
+
"grad_norm": 0.9812231063842773,
|
| 399 |
+
"learning_rate": 4.503012048192771e-05,
|
| 400 |
+
"loss": 0.1422,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.3859827323514474,
|
| 405 |
+
"grad_norm": 0.9900213479995728,
|
| 406 |
+
"learning_rate": 4.4841867469879525e-05,
|
| 407 |
+
"loss": 0.1308,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.39275435923480617,
|
| 412 |
+
"grad_norm": 1.0598454475402832,
|
| 413 |
+
"learning_rate": 4.465361445783133e-05,
|
| 414 |
+
"loss": 0.1565,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.39952598611816487,
|
| 419 |
+
"grad_norm": 1.130131483078003,
|
| 420 |
+
"learning_rate": 4.446536144578313e-05,
|
| 421 |
+
"loss": 0.1432,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.4062976130015236,
|
| 426 |
+
"grad_norm": 1.4709850549697876,
|
| 427 |
+
"learning_rate": 4.427710843373494e-05,
|
| 428 |
+
"loss": 0.1427,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.41306923988488237,
|
| 433 |
+
"grad_norm": 1.2057602405548096,
|
| 434 |
+
"learning_rate": 4.408885542168675e-05,
|
| 435 |
+
"loss": 0.1377,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.41984086676824106,
|
| 440 |
+
"grad_norm": 1.0241261720657349,
|
| 441 |
+
"learning_rate": 4.390060240963856e-05,
|
| 442 |
+
"loss": 0.1303,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.4266124936515998,
|
| 447 |
+
"grad_norm": 0.70640629529953,
|
| 448 |
+
"learning_rate": 4.3712349397590366e-05,
|
| 449 |
+
"loss": 0.1284,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.4333841205349585,
|
| 454 |
+
"grad_norm": 1.1402077674865723,
|
| 455 |
+
"learning_rate": 4.352409638554217e-05,
|
| 456 |
+
"loss": 0.1153,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.44015574741831726,
|
| 461 |
+
"grad_norm": 2.3130173683166504,
|
| 462 |
+
"learning_rate": 4.333584337349398e-05,
|
| 463 |
+
"loss": 0.1353,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.44692737430167595,
|
| 468 |
+
"grad_norm": 1.018059253692627,
|
| 469 |
+
"learning_rate": 4.3147590361445783e-05,
|
| 470 |
+
"loss": 0.1198,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.4536990011850347,
|
| 475 |
+
"grad_norm": 1.1822158098220825,
|
| 476 |
+
"learning_rate": 4.2959337349397596e-05,
|
| 477 |
+
"loss": 0.1224,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.46047062806839345,
|
| 482 |
+
"grad_norm": 0.8007648587226868,
|
| 483 |
+
"learning_rate": 4.27710843373494e-05,
|
| 484 |
+
"loss": 0.1156,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.46724225495175215,
|
| 489 |
+
"grad_norm": 2.0608577728271484,
|
| 490 |
+
"learning_rate": 4.258283132530121e-05,
|
| 491 |
+
"loss": 0.1355,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.4740138818351109,
|
| 496 |
+
"grad_norm": 0.8175686597824097,
|
| 497 |
+
"learning_rate": 4.239457831325301e-05,
|
| 498 |
+
"loss": 0.117,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.4807855087184696,
|
| 503 |
+
"grad_norm": 0.9864722490310669,
|
| 504 |
+
"learning_rate": 4.220632530120482e-05,
|
| 505 |
+
"loss": 0.1143,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.48755713560182834,
|
| 510 |
+
"grad_norm": 1.1170806884765625,
|
| 511 |
+
"learning_rate": 4.201807228915663e-05,
|
| 512 |
+
"loss": 0.1128,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.4943287624851871,
|
| 517 |
+
"grad_norm": 0.8426876664161682,
|
| 518 |
+
"learning_rate": 4.182981927710844e-05,
|
| 519 |
+
"loss": 0.1039,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.5011003893685458,
|
| 524 |
+
"grad_norm": 1.0189573764801025,
|
| 525 |
+
"learning_rate": 4.164156626506024e-05,
|
| 526 |
+
"loss": 0.1056,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.5078720162519045,
|
| 531 |
+
"grad_norm": 1.0524259805679321,
|
| 532 |
+
"learning_rate": 4.145331325301205e-05,
|
| 533 |
+
"loss": 0.106,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.5146436431352632,
|
| 538 |
+
"grad_norm": 0.8824974298477173,
|
| 539 |
+
"learning_rate": 4.126506024096386e-05,
|
| 540 |
+
"loss": 0.1077,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.521415270018622,
|
| 545 |
+
"grad_norm": 1.034146785736084,
|
| 546 |
+
"learning_rate": 4.107680722891567e-05,
|
| 547 |
+
"loss": 0.1147,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.5281868969019807,
|
| 552 |
+
"grad_norm": 1.1661494970321655,
|
| 553 |
+
"learning_rate": 4.088855421686747e-05,
|
| 554 |
+
"loss": 0.1159,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.5349585237853395,
|
| 559 |
+
"grad_norm": 1.9664863348007202,
|
| 560 |
+
"learning_rate": 4.070030120481928e-05,
|
| 561 |
+
"loss": 0.1155,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.5417301506686981,
|
| 566 |
+
"grad_norm": 0.9557604193687439,
|
| 567 |
+
"learning_rate": 4.0512048192771084e-05,
|
| 568 |
+
"loss": 0.1117,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.5485017775520569,
|
| 573 |
+
"grad_norm": 1.1883480548858643,
|
| 574 |
+
"learning_rate": 4.03237951807229e-05,
|
| 575 |
+
"loss": 0.1001,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.5552734044354156,
|
| 580 |
+
"grad_norm": 1.3493871688842773,
|
| 581 |
+
"learning_rate": 4.01355421686747e-05,
|
| 582 |
+
"loss": 0.1022,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.5620450313187744,
|
| 587 |
+
"grad_norm": 0.8323326110839844,
|
| 588 |
+
"learning_rate": 3.994728915662651e-05,
|
| 589 |
+
"loss": 0.1084,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.568816658202133,
|
| 594 |
+
"grad_norm": 1.3079185485839844,
|
| 595 |
+
"learning_rate": 3.9759036144578314e-05,
|
| 596 |
+
"loss": 0.1054,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.5755882850854918,
|
| 601 |
+
"grad_norm": 0.961869478225708,
|
| 602 |
+
"learning_rate": 3.957078313253012e-05,
|
| 603 |
+
"loss": 0.1027,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.5823599119688505,
|
| 608 |
+
"grad_norm": 0.7965933680534363,
|
| 609 |
+
"learning_rate": 3.938253012048193e-05,
|
| 610 |
+
"loss": 0.101,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.5891315388522093,
|
| 615 |
+
"grad_norm": 0.5749910473823547,
|
| 616 |
+
"learning_rate": 3.919427710843374e-05,
|
| 617 |
+
"loss": 0.0942,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.595903165735568,
|
| 622 |
+
"grad_norm": 1.314102292060852,
|
| 623 |
+
"learning_rate": 3.9006024096385544e-05,
|
| 624 |
+
"loss": 0.1062,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.6026747926189266,
|
| 629 |
+
"grad_norm": 0.9225144982337952,
|
| 630 |
+
"learning_rate": 3.881777108433735e-05,
|
| 631 |
+
"loss": 0.0895,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.6094464195022854,
|
| 636 |
+
"grad_norm": 1.0227304697036743,
|
| 637 |
+
"learning_rate": 3.8629518072289155e-05,
|
| 638 |
+
"loss": 0.1057,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.6162180463856441,
|
| 643 |
+
"grad_norm": 0.9698050618171692,
|
| 644 |
+
"learning_rate": 3.844126506024097e-05,
|
| 645 |
+
"loss": 0.0996,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.6229896732690029,
|
| 650 |
+
"grad_norm": 0.7296442985534668,
|
| 651 |
+
"learning_rate": 3.8253012048192774e-05,
|
| 652 |
+
"loss": 0.1035,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.6297613001523616,
|
| 657 |
+
"grad_norm": 0.943133533000946,
|
| 658 |
+
"learning_rate": 3.806475903614458e-05,
|
| 659 |
+
"loss": 0.0892,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.6365329270357203,
|
| 664 |
+
"grad_norm": 1.618379831314087,
|
| 665 |
+
"learning_rate": 3.7876506024096385e-05,
|
| 666 |
+
"loss": 0.0828,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.643304553919079,
|
| 671 |
+
"grad_norm": 0.8589970469474792,
|
| 672 |
+
"learning_rate": 3.76882530120482e-05,
|
| 673 |
+
"loss": 0.1033,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.6500761808024378,
|
| 678 |
+
"grad_norm": 0.7783709764480591,
|
| 679 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 680 |
+
"loss": 0.0943,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.6568478076857965,
|
| 685 |
+
"grad_norm": 1.0342212915420532,
|
| 686 |
+
"learning_rate": 3.731174698795181e-05,
|
| 687 |
+
"loss": 0.0979,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.6636194345691553,
|
| 692 |
+
"grad_norm": 0.815461277961731,
|
| 693 |
+
"learning_rate": 3.7123493975903615e-05,
|
| 694 |
+
"loss": 0.092,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.6703910614525139,
|
| 699 |
+
"grad_norm": 0.7246832251548767,
|
| 700 |
+
"learning_rate": 3.693524096385542e-05,
|
| 701 |
+
"loss": 0.0843,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.6771626883358727,
|
| 706 |
+
"grad_norm": 0.9605166912078857,
|
| 707 |
+
"learning_rate": 3.674698795180723e-05,
|
| 708 |
+
"loss": 0.0835,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.6839343152192314,
|
| 713 |
+
"grad_norm": 1.1440260410308838,
|
| 714 |
+
"learning_rate": 3.655873493975904e-05,
|
| 715 |
+
"loss": 0.0886,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.6907059421025902,
|
| 720 |
+
"grad_norm": 1.0243327617645264,
|
| 721 |
+
"learning_rate": 3.6370481927710845e-05,
|
| 722 |
+
"loss": 0.089,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.6974775689859489,
|
| 727 |
+
"grad_norm": 0.8508220314979553,
|
| 728 |
+
"learning_rate": 3.618222891566265e-05,
|
| 729 |
+
"loss": 0.0795,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.7042491958693076,
|
| 734 |
+
"grad_norm": 0.852389931678772,
|
| 735 |
+
"learning_rate": 3.5993975903614456e-05,
|
| 736 |
+
"loss": 0.091,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.7110208227526663,
|
| 741 |
+
"grad_norm": 0.9550991058349609,
|
| 742 |
+
"learning_rate": 3.580572289156627e-05,
|
| 743 |
+
"loss": 0.0856,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.7177924496360251,
|
| 748 |
+
"grad_norm": 1.009476661682129,
|
| 749 |
+
"learning_rate": 3.5617469879518075e-05,
|
| 750 |
+
"loss": 0.084,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.7245640765193838,
|
| 755 |
+
"grad_norm": 0.8816896080970764,
|
| 756 |
+
"learning_rate": 3.542921686746988e-05,
|
| 757 |
+
"loss": 0.0873,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.7313357034027425,
|
| 762 |
+
"grad_norm": 0.8172402381896973,
|
| 763 |
+
"learning_rate": 3.5240963855421686e-05,
|
| 764 |
+
"loss": 0.0785,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.7381073302861012,
|
| 769 |
+
"grad_norm": 1.3699508905410767,
|
| 770 |
+
"learning_rate": 3.505271084337349e-05,
|
| 771 |
+
"loss": 0.0785,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.74487895716946,
|
| 776 |
+
"grad_norm": 0.6807621121406555,
|
| 777 |
+
"learning_rate": 3.4864457831325304e-05,
|
| 778 |
+
"loss": 0.0797,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.7516505840528187,
|
| 783 |
+
"grad_norm": 0.8369671106338501,
|
| 784 |
+
"learning_rate": 3.467620481927711e-05,
|
| 785 |
+
"loss": 0.0845,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.7584222109361775,
|
| 790 |
+
"grad_norm": 0.7025638818740845,
|
| 791 |
+
"learning_rate": 3.4487951807228916e-05,
|
| 792 |
+
"loss": 0.0699,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.7651938378195361,
|
| 797 |
+
"grad_norm": 0.6446508765220642,
|
| 798 |
+
"learning_rate": 3.429969879518072e-05,
|
| 799 |
+
"loss": 0.0879,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.7719654647028948,
|
| 804 |
+
"grad_norm": 1.6889147758483887,
|
| 805 |
+
"learning_rate": 3.4111445783132534e-05,
|
| 806 |
+
"loss": 0.0966,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.7787370915862536,
|
| 811 |
+
"grad_norm": 0.7355449199676514,
|
| 812 |
+
"learning_rate": 3.392319277108434e-05,
|
| 813 |
+
"loss": 0.0824,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.7855087184696123,
|
| 818 |
+
"grad_norm": 0.5761105418205261,
|
| 819 |
+
"learning_rate": 3.3734939759036146e-05,
|
| 820 |
+
"loss": 0.0759,
|
| 821 |
+
"step": 1160
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.7922803453529711,
|
| 825 |
+
"grad_norm": 0.7694604396820068,
|
| 826 |
+
"learning_rate": 3.354668674698795e-05,
|
| 827 |
+
"loss": 0.0839,
|
| 828 |
+
"step": 1170
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.7990519722363297,
|
| 832 |
+
"grad_norm": 0.8113671541213989,
|
| 833 |
+
"learning_rate": 3.335843373493976e-05,
|
| 834 |
+
"loss": 0.0799,
|
| 835 |
+
"step": 1180
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.8058235991196885,
|
| 839 |
+
"grad_norm": 0.6517748832702637,
|
| 840 |
+
"learning_rate": 3.317018072289157e-05,
|
| 841 |
+
"loss": 0.078,
|
| 842 |
+
"step": 1190
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.8125952260030472,
|
| 846 |
+
"grad_norm": 0.8274058103561401,
|
| 847 |
+
"learning_rate": 3.2981927710843376e-05,
|
| 848 |
+
"loss": 0.0789,
|
| 849 |
+
"step": 1200
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.819366852886406,
|
| 853 |
+
"grad_norm": 0.7922660112380981,
|
| 854 |
+
"learning_rate": 3.279367469879519e-05,
|
| 855 |
+
"loss": 0.0764,
|
| 856 |
+
"step": 1210
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.8261384797697647,
|
| 860 |
+
"grad_norm": 1.2404780387878418,
|
| 861 |
+
"learning_rate": 3.260542168674699e-05,
|
| 862 |
+
"loss": 0.0731,
|
| 863 |
+
"step": 1220
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.8329101066531234,
|
| 867 |
+
"grad_norm": 0.6905914545059204,
|
| 868 |
+
"learning_rate": 3.241716867469879e-05,
|
| 869 |
+
"loss": 0.0704,
|
| 870 |
+
"step": 1230
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.8396817335364821,
|
| 874 |
+
"grad_norm": 0.9580501914024353,
|
| 875 |
+
"learning_rate": 3.2228915662650605e-05,
|
| 876 |
+
"loss": 0.0806,
|
| 877 |
+
"step": 1240
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.8464533604198409,
|
| 881 |
+
"grad_norm": 0.4647589325904846,
|
| 882 |
+
"learning_rate": 3.204066265060241e-05,
|
| 883 |
+
"loss": 0.0749,
|
| 884 |
+
"step": 1250
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.8532249873031996,
|
| 888 |
+
"grad_norm": 0.7317982316017151,
|
| 889 |
+
"learning_rate": 3.1852409638554224e-05,
|
| 890 |
+
"loss": 0.0739,
|
| 891 |
+
"step": 1260
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.8599966141865584,
|
| 895 |
+
"grad_norm": 0.7131925821304321,
|
| 896 |
+
"learning_rate": 3.166415662650602e-05,
|
| 897 |
+
"loss": 0.0742,
|
| 898 |
+
"step": 1270
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.866768241069917,
|
| 902 |
+
"grad_norm": 0.42962631583213806,
|
| 903 |
+
"learning_rate": 3.147590361445783e-05,
|
| 904 |
+
"loss": 0.0595,
|
| 905 |
+
"step": 1280
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.8735398679532758,
|
| 909 |
+
"grad_norm": 1.0177807807922363,
|
| 910 |
+
"learning_rate": 3.128765060240964e-05,
|
| 911 |
+
"loss": 0.0815,
|
| 912 |
+
"step": 1290
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.8803114948366345,
|
| 916 |
+
"grad_norm": 0.7908213138580322,
|
| 917 |
+
"learning_rate": 3.1099397590361447e-05,
|
| 918 |
+
"loss": 0.0758,
|
| 919 |
+
"step": 1300
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.8870831217199933,
|
| 923 |
+
"grad_norm": 0.7449843883514404,
|
| 924 |
+
"learning_rate": 3.091114457831326e-05,
|
| 925 |
+
"loss": 0.0789,
|
| 926 |
+
"step": 1310
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.8938547486033519,
|
| 930 |
+
"grad_norm": 0.8616693615913391,
|
| 931 |
+
"learning_rate": 3.072289156626506e-05,
|
| 932 |
+
"loss": 0.0653,
|
| 933 |
+
"step": 1320
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.9006263754867107,
|
| 937 |
+
"grad_norm": 1.0630964040756226,
|
| 938 |
+
"learning_rate": 3.053463855421687e-05,
|
| 939 |
+
"loss": 0.081,
|
| 940 |
+
"step": 1330
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.9073980023700694,
|
| 944 |
+
"grad_norm": 0.7529353499412537,
|
| 945 |
+
"learning_rate": 3.0346385542168676e-05,
|
| 946 |
+
"loss": 0.0753,
|
| 947 |
+
"step": 1340
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.9141696292534282,
|
| 951 |
+
"grad_norm": 0.7995216846466064,
|
| 952 |
+
"learning_rate": 3.0158132530120482e-05,
|
| 953 |
+
"loss": 0.066,
|
| 954 |
+
"step": 1350
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.9209412561367869,
|
| 958 |
+
"grad_norm": 0.3852786719799042,
|
| 959 |
+
"learning_rate": 2.996987951807229e-05,
|
| 960 |
+
"loss": 0.07,
|
| 961 |
+
"step": 1360
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.9277128830201455,
|
| 965 |
+
"grad_norm": 0.6470819711685181,
|
| 966 |
+
"learning_rate": 2.9781626506024097e-05,
|
| 967 |
+
"loss": 0.0815,
|
| 968 |
+
"step": 1370
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.9344845099035043,
|
| 972 |
+
"grad_norm": 0.9942120313644409,
|
| 973 |
+
"learning_rate": 2.9593373493975906e-05,
|
| 974 |
+
"loss": 0.0749,
|
| 975 |
+
"step": 1380
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.941256136786863,
|
| 979 |
+
"grad_norm": 1.0045307874679565,
|
| 980 |
+
"learning_rate": 2.9405120481927712e-05,
|
| 981 |
+
"loss": 0.0797,
|
| 982 |
+
"step": 1390
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.9480277636702218,
|
| 986 |
+
"grad_norm": 0.5135801434516907,
|
| 987 |
+
"learning_rate": 2.921686746987952e-05,
|
| 988 |
+
"loss": 0.0647,
|
| 989 |
+
"step": 1400
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.9547993905535805,
|
| 993 |
+
"grad_norm": 1.0146623849868774,
|
| 994 |
+
"learning_rate": 2.9028614457831327e-05,
|
| 995 |
+
"loss": 0.078,
|
| 996 |
+
"step": 1410
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.9615710174369392,
|
| 1000 |
+
"grad_norm": 0.730907142162323,
|
| 1001 |
+
"learning_rate": 2.8840361445783133e-05,
|
| 1002 |
+
"loss": 0.0671,
|
| 1003 |
+
"step": 1420
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.9683426443202979,
|
| 1007 |
+
"grad_norm": 1.0035794973373413,
|
| 1008 |
+
"learning_rate": 2.8652108433734942e-05,
|
| 1009 |
+
"loss": 0.0804,
|
| 1010 |
+
"step": 1430
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.9751142712036567,
|
| 1014 |
+
"grad_norm": 0.6765666604042053,
|
| 1015 |
+
"learning_rate": 2.8463855421686748e-05,
|
| 1016 |
+
"loss": 0.0691,
|
| 1017 |
+
"step": 1440
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.9818858980870154,
|
| 1021 |
+
"grad_norm": 0.8572028279304504,
|
| 1022 |
+
"learning_rate": 2.8275602409638557e-05,
|
| 1023 |
+
"loss": 0.0716,
|
| 1024 |
+
"step": 1450
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.9886575249703742,
|
| 1028 |
+
"grad_norm": 0.4426514208316803,
|
| 1029 |
+
"learning_rate": 2.8087349397590362e-05,
|
| 1030 |
+
"loss": 0.0709,
|
| 1031 |
+
"step": 1460
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9954291518537328,
|
| 1035 |
+
"grad_norm": 0.5503798723220825,
|
| 1036 |
+
"learning_rate": 2.789909638554217e-05,
|
| 1037 |
+
"loss": 0.0709,
|
| 1038 |
+
"step": 1470
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.9994921279837481,
|
| 1042 |
+
"eval_accuracy": 0.9789523809523809,
|
| 1043 |
+
"eval_loss": 0.07530121505260468,
|
| 1044 |
+
"eval_runtime": 1028.8877,
|
| 1045 |
+
"eval_samples_per_second": 20.41,
|
| 1046 |
+
"eval_steps_per_second": 0.639,
|
| 1047 |
+
"step": 1476
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 1.0022007787370917,
|
| 1051 |
+
"grad_norm": 0.808159589767456,
|
| 1052 |
+
"learning_rate": 2.7710843373493977e-05,
|
| 1053 |
+
"loss": 0.0653,
|
| 1054 |
+
"step": 1480
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 1.0089724056204503,
|
| 1058 |
+
"grad_norm": 1.711690902709961,
|
| 1059 |
+
"learning_rate": 2.7522590361445783e-05,
|
| 1060 |
+
"loss": 0.0712,
|
| 1061 |
+
"step": 1490
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.015744032503809,
|
| 1065 |
+
"grad_norm": 0.584488570690155,
|
| 1066 |
+
"learning_rate": 2.7334337349397592e-05,
|
| 1067 |
+
"loss": 0.0748,
|
| 1068 |
+
"step": 1500
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.0225156593871678,
|
| 1072 |
+
"grad_norm": 0.6118115186691284,
|
| 1073 |
+
"learning_rate": 2.7146084337349398e-05,
|
| 1074 |
+
"loss": 0.0663,
|
| 1075 |
+
"step": 1510
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 1.0292872862705265,
|
| 1079 |
+
"grad_norm": 0.7841724753379822,
|
| 1080 |
+
"learning_rate": 2.6957831325301207e-05,
|
| 1081 |
+
"loss": 0.0664,
|
| 1082 |
+
"step": 1520
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 1.0360589131538853,
|
| 1086 |
+
"grad_norm": 0.6936156153678894,
|
| 1087 |
+
"learning_rate": 2.6769578313253013e-05,
|
| 1088 |
+
"loss": 0.0642,
|
| 1089 |
+
"step": 1530
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 1.042830540037244,
|
| 1093 |
+
"grad_norm": 0.7428813576698303,
|
| 1094 |
+
"learning_rate": 2.658132530120482e-05,
|
| 1095 |
+
"loss": 0.0694,
|
| 1096 |
+
"step": 1540
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 1.0496021669206026,
|
| 1100 |
+
"grad_norm": 1.0296515226364136,
|
| 1101 |
+
"learning_rate": 2.6393072289156628e-05,
|
| 1102 |
+
"loss": 0.0754,
|
| 1103 |
+
"step": 1550
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.0563737938039615,
|
| 1107 |
+
"grad_norm": 0.6250594258308411,
|
| 1108 |
+
"learning_rate": 2.6204819277108434e-05,
|
| 1109 |
+
"loss": 0.0646,
|
| 1110 |
+
"step": 1560
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 1.06314542068732,
|
| 1114 |
+
"grad_norm": 1.5104742050170898,
|
| 1115 |
+
"learning_rate": 2.6016566265060243e-05,
|
| 1116 |
+
"loss": 0.073,
|
| 1117 |
+
"step": 1570
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 1.069917047570679,
|
| 1121 |
+
"grad_norm": 0.8500051498413086,
|
| 1122 |
+
"learning_rate": 2.582831325301205e-05,
|
| 1123 |
+
"loss": 0.0677,
|
| 1124 |
+
"step": 1580
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 1.0766886744540376,
|
| 1128 |
+
"grad_norm": 1.0728169679641724,
|
| 1129 |
+
"learning_rate": 2.5640060240963858e-05,
|
| 1130 |
+
"loss": 0.0692,
|
| 1131 |
+
"step": 1590
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 1.0834603013373962,
|
| 1135 |
+
"grad_norm": 0.7029580473899841,
|
| 1136 |
+
"learning_rate": 2.5451807228915663e-05,
|
| 1137 |
+
"loss": 0.0681,
|
| 1138 |
+
"step": 1600
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 1.090231928220755,
|
| 1142 |
+
"grad_norm": 0.7941862344741821,
|
| 1143 |
+
"learning_rate": 2.526355421686747e-05,
|
| 1144 |
+
"loss": 0.0727,
|
| 1145 |
+
"step": 1610
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 1.0970035551041137,
|
| 1149 |
+
"grad_norm": 0.8318243026733398,
|
| 1150 |
+
"learning_rate": 2.5075301204819278e-05,
|
| 1151 |
+
"loss": 0.0764,
|
| 1152 |
+
"step": 1620
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 1.1037751819874724,
|
| 1156 |
+
"grad_norm": 0.9039504528045654,
|
| 1157 |
+
"learning_rate": 2.4887048192771087e-05,
|
| 1158 |
+
"loss": 0.0817,
|
| 1159 |
+
"step": 1630
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 1.1105468088708312,
|
| 1163 |
+
"grad_norm": 0.8218241930007935,
|
| 1164 |
+
"learning_rate": 2.4698795180722893e-05,
|
| 1165 |
+
"loss": 0.0705,
|
| 1166 |
+
"step": 1640
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.1173184357541899,
|
| 1170 |
+
"grad_norm": 0.5053806900978088,
|
| 1171 |
+
"learning_rate": 2.45105421686747e-05,
|
| 1172 |
+
"loss": 0.0738,
|
| 1173 |
+
"step": 1650
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 1.1240900626375487,
|
| 1177 |
+
"grad_norm": 0.5837478637695312,
|
| 1178 |
+
"learning_rate": 2.4322289156626508e-05,
|
| 1179 |
+
"loss": 0.0623,
|
| 1180 |
+
"step": 1660
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 1.1308616895209074,
|
| 1184 |
+
"grad_norm": 0.6927766799926758,
|
| 1185 |
+
"learning_rate": 2.4134036144578314e-05,
|
| 1186 |
+
"loss": 0.0692,
|
| 1187 |
+
"step": 1670
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 1.137633316404266,
|
| 1191 |
+
"grad_norm": 0.8067636489868164,
|
| 1192 |
+
"learning_rate": 2.3945783132530123e-05,
|
| 1193 |
+
"loss": 0.0716,
|
| 1194 |
+
"step": 1680
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 1.1444049432876249,
|
| 1198 |
+
"grad_norm": 0.8656367063522339,
|
| 1199 |
+
"learning_rate": 2.375753012048193e-05,
|
| 1200 |
+
"loss": 0.0648,
|
| 1201 |
+
"step": 1690
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 1.1511765701709835,
|
| 1205 |
+
"grad_norm": 0.5790455341339111,
|
| 1206 |
+
"learning_rate": 2.3569277108433734e-05,
|
| 1207 |
+
"loss": 0.0682,
|
| 1208 |
+
"step": 1700
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 1.1579481970543424,
|
| 1212 |
+
"grad_norm": 0.9087624549865723,
|
| 1213 |
+
"learning_rate": 2.3381024096385544e-05,
|
| 1214 |
+
"loss": 0.0606,
|
| 1215 |
+
"step": 1710
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 1.164719823937701,
|
| 1219 |
+
"grad_norm": 0.6014874577522278,
|
| 1220 |
+
"learning_rate": 2.319277108433735e-05,
|
| 1221 |
+
"loss": 0.078,
|
| 1222 |
+
"step": 1720
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 1.1714914508210597,
|
| 1226 |
+
"grad_norm": 0.6091852784156799,
|
| 1227 |
+
"learning_rate": 2.300451807228916e-05,
|
| 1228 |
+
"loss": 0.0647,
|
| 1229 |
+
"step": 1730
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 1.1782630777044185,
|
| 1233 |
+
"grad_norm": 0.581322431564331,
|
| 1234 |
+
"learning_rate": 2.2816265060240964e-05,
|
| 1235 |
+
"loss": 0.0712,
|
| 1236 |
+
"step": 1740
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 1.1850347045877772,
|
| 1240 |
+
"grad_norm": 0.4964611530303955,
|
| 1241 |
+
"learning_rate": 2.2628012048192773e-05,
|
| 1242 |
+
"loss": 0.0524,
|
| 1243 |
+
"step": 1750
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 1.191806331471136,
|
| 1247 |
+
"grad_norm": 0.8920990228652954,
|
| 1248 |
+
"learning_rate": 2.243975903614458e-05,
|
| 1249 |
+
"loss": 0.0719,
|
| 1250 |
+
"step": 1760
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.1985779583544947,
|
| 1254 |
+
"grad_norm": 0.6709417104721069,
|
| 1255 |
+
"learning_rate": 2.2251506024096385e-05,
|
| 1256 |
+
"loss": 0.0695,
|
| 1257 |
+
"step": 1770
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 1.2053495852378533,
|
| 1261 |
+
"grad_norm": 0.7828611731529236,
|
| 1262 |
+
"learning_rate": 2.2063253012048194e-05,
|
| 1263 |
+
"loss": 0.0742,
|
| 1264 |
+
"step": 1780
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 1.2121212121212122,
|
| 1268 |
+
"grad_norm": 1.2962734699249268,
|
| 1269 |
+
"learning_rate": 2.1875e-05,
|
| 1270 |
+
"loss": 0.0743,
|
| 1271 |
+
"step": 1790
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.2188928390045708,
|
| 1275 |
+
"grad_norm": 0.5535785555839539,
|
| 1276 |
+
"learning_rate": 2.168674698795181e-05,
|
| 1277 |
+
"loss": 0.064,
|
| 1278 |
+
"step": 1800
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 1.2256644658879297,
|
| 1282 |
+
"grad_norm": 1.0534149408340454,
|
| 1283 |
+
"learning_rate": 2.1498493975903615e-05,
|
| 1284 |
+
"loss": 0.0594,
|
| 1285 |
+
"step": 1810
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 1.2324360927712883,
|
| 1289 |
+
"grad_norm": 0.8949538469314575,
|
| 1290 |
+
"learning_rate": 2.1310240963855424e-05,
|
| 1291 |
+
"loss": 0.0607,
|
| 1292 |
+
"step": 1820
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 1.239207719654647,
|
| 1296 |
+
"grad_norm": 0.7287763357162476,
|
| 1297 |
+
"learning_rate": 2.112198795180723e-05,
|
| 1298 |
+
"loss": 0.076,
|
| 1299 |
+
"step": 1830
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 1.2459793465380058,
|
| 1303 |
+
"grad_norm": 1.2560794353485107,
|
| 1304 |
+
"learning_rate": 2.0933734939759035e-05,
|
| 1305 |
+
"loss": 0.0662,
|
| 1306 |
+
"step": 1840
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 1.2527509734213644,
|
| 1310 |
+
"grad_norm": 0.4771580398082733,
|
| 1311 |
+
"learning_rate": 2.0745481927710844e-05,
|
| 1312 |
+
"loss": 0.0541,
|
| 1313 |
+
"step": 1850
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 1.2595226003047233,
|
| 1317 |
+
"grad_norm": 0.9189515113830566,
|
| 1318 |
+
"learning_rate": 2.055722891566265e-05,
|
| 1319 |
+
"loss": 0.058,
|
| 1320 |
+
"step": 1860
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 1.266294227188082,
|
| 1324 |
+
"grad_norm": 1.0164196491241455,
|
| 1325 |
+
"learning_rate": 2.036897590361446e-05,
|
| 1326 |
+
"loss": 0.0604,
|
| 1327 |
+
"step": 1870
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 1.2730658540714406,
|
| 1331 |
+
"grad_norm": 0.8721578121185303,
|
| 1332 |
+
"learning_rate": 2.018072289156627e-05,
|
| 1333 |
+
"loss": 0.0731,
|
| 1334 |
+
"step": 1880
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 1.2798374809547994,
|
| 1338 |
+
"grad_norm": 0.8412027955055237,
|
| 1339 |
+
"learning_rate": 1.999246987951807e-05,
|
| 1340 |
+
"loss": 0.0636,
|
| 1341 |
+
"step": 1890
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 1.286609107838158,
|
| 1345 |
+
"grad_norm": 0.784599244594574,
|
| 1346 |
+
"learning_rate": 1.980421686746988e-05,
|
| 1347 |
+
"loss": 0.0636,
|
| 1348 |
+
"step": 1900
|
| 1349 |
+
},
|
| 1350 |
+
{
|
| 1351 |
+
"epoch": 1.293380734721517,
|
| 1352 |
+
"grad_norm": 0.5514250993728638,
|
| 1353 |
+
"learning_rate": 1.9615963855421686e-05,
|
| 1354 |
+
"loss": 0.0677,
|
| 1355 |
+
"step": 1910
|
| 1356 |
+
},
|
| 1357 |
+
{
|
| 1358 |
+
"epoch": 1.3001523616048756,
|
| 1359 |
+
"grad_norm": 0.6008131504058838,
|
| 1360 |
+
"learning_rate": 1.9427710843373495e-05,
|
| 1361 |
+
"loss": 0.0637,
|
| 1362 |
+
"step": 1920
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 1.3069239884882342,
|
| 1366 |
+
"grad_norm": 0.9827722311019897,
|
| 1367 |
+
"learning_rate": 1.9239457831325304e-05,
|
| 1368 |
+
"loss": 0.0581,
|
| 1369 |
+
"step": 1930
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 1.313695615371593,
|
| 1373 |
+
"grad_norm": 0.699483335018158,
|
| 1374 |
+
"learning_rate": 1.905120481927711e-05,
|
| 1375 |
+
"loss": 0.0651,
|
| 1376 |
+
"step": 1940
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 1.3204672422549517,
|
| 1380 |
+
"grad_norm": 0.7085596919059753,
|
| 1381 |
+
"learning_rate": 1.8862951807228916e-05,
|
| 1382 |
+
"loss": 0.0589,
|
| 1383 |
+
"step": 1950
|
| 1384 |
+
},
|
| 1385 |
+
{
|
| 1386 |
+
"epoch": 1.3272388691383106,
|
| 1387 |
+
"grad_norm": 1.0991511344909668,
|
| 1388 |
+
"learning_rate": 1.867469879518072e-05,
|
| 1389 |
+
"loss": 0.0699,
|
| 1390 |
+
"step": 1960
|
| 1391 |
+
},
|
| 1392 |
+
{
|
| 1393 |
+
"epoch": 1.3340104960216692,
|
| 1394 |
+
"grad_norm": 0.8005927801132202,
|
| 1395 |
+
"learning_rate": 1.848644578313253e-05,
|
| 1396 |
+
"loss": 0.0673,
|
| 1397 |
+
"step": 1970
|
| 1398 |
+
},
|
| 1399 |
+
{
|
| 1400 |
+
"epoch": 1.3407821229050279,
|
| 1401 |
+
"grad_norm": 0.7139153480529785,
|
| 1402 |
+
"learning_rate": 1.829819277108434e-05,
|
| 1403 |
+
"loss": 0.0571,
|
| 1404 |
+
"step": 1980
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 1.3475537497883867,
|
| 1408 |
+
"grad_norm": 0.4111141562461853,
|
| 1409 |
+
"learning_rate": 1.8109939759036145e-05,
|
| 1410 |
+
"loss": 0.0587,
|
| 1411 |
+
"step": 1990
|
| 1412 |
+
},
|
| 1413 |
+
{
|
| 1414 |
+
"epoch": 1.3543253766717454,
|
| 1415 |
+
"grad_norm": 0.5993856191635132,
|
| 1416 |
+
"learning_rate": 1.7921686746987955e-05,
|
| 1417 |
+
"loss": 0.0649,
|
| 1418 |
+
"step": 2000
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 1.3610970035551042,
|
| 1422 |
+
"grad_norm": 0.9376055598258972,
|
| 1423 |
+
"learning_rate": 1.773343373493976e-05,
|
| 1424 |
+
"loss": 0.0672,
|
| 1425 |
+
"step": 2010
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"epoch": 1.3678686304384629,
|
| 1429 |
+
"grad_norm": 0.4629497230052948,
|
| 1430 |
+
"learning_rate": 1.7545180722891566e-05,
|
| 1431 |
+
"loss": 0.0517,
|
| 1432 |
+
"step": 2020
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 1.3746402573218215,
|
| 1436 |
+
"grad_norm": 0.4839510917663574,
|
| 1437 |
+
"learning_rate": 1.7356927710843375e-05,
|
| 1438 |
+
"loss": 0.0524,
|
| 1439 |
+
"step": 2030
|
| 1440 |
+
},
|
| 1441 |
+
{
|
| 1442 |
+
"epoch": 1.3814118842051804,
|
| 1443 |
+
"grad_norm": 0.5999952554702759,
|
| 1444 |
+
"learning_rate": 1.716867469879518e-05,
|
| 1445 |
+
"loss": 0.056,
|
| 1446 |
+
"step": 2040
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 1.388183511088539,
|
| 1450 |
+
"grad_norm": 0.5760485529899597,
|
| 1451 |
+
"learning_rate": 1.698042168674699e-05,
|
| 1452 |
+
"loss": 0.0663,
|
| 1453 |
+
"step": 2050
|
| 1454 |
+
},
|
| 1455 |
+
{
|
| 1456 |
+
"epoch": 1.3949551379718979,
|
| 1457 |
+
"grad_norm": 0.7293563485145569,
|
| 1458 |
+
"learning_rate": 1.6792168674698796e-05,
|
| 1459 |
+
"loss": 0.0649,
|
| 1460 |
+
"step": 2060
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 1.4017267648552565,
|
| 1464 |
+
"grad_norm": 0.6858052611351013,
|
| 1465 |
+
"learning_rate": 1.6603915662650605e-05,
|
| 1466 |
+
"loss": 0.0545,
|
| 1467 |
+
"step": 2070
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 1.4084983917386151,
|
| 1471 |
+
"grad_norm": 0.9105307459831238,
|
| 1472 |
+
"learning_rate": 1.641566265060241e-05,
|
| 1473 |
+
"loss": 0.0616,
|
| 1474 |
+
"step": 2080
|
| 1475 |
+
},
|
| 1476 |
+
{
|
| 1477 |
+
"epoch": 1.415270018621974,
|
| 1478 |
+
"grad_norm": 0.4213721752166748,
|
| 1479 |
+
"learning_rate": 1.6227409638554216e-05,
|
| 1480 |
+
"loss": 0.0647,
|
| 1481 |
+
"step": 2090
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 1.4220416455053326,
|
| 1485 |
+
"grad_norm": 0.5692495703697205,
|
| 1486 |
+
"learning_rate": 1.6039156626506026e-05,
|
| 1487 |
+
"loss": 0.0602,
|
| 1488 |
+
"step": 2100
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 1.4288132723886915,
|
| 1492 |
+
"grad_norm": 0.70749431848526,
|
| 1493 |
+
"learning_rate": 1.585090361445783e-05,
|
| 1494 |
+
"loss": 0.0638,
|
| 1495 |
+
"step": 2110
|
| 1496 |
+
},
|
| 1497 |
+
{
|
| 1498 |
+
"epoch": 1.4355848992720501,
|
| 1499 |
+
"grad_norm": 0.7523058652877808,
|
| 1500 |
+
"learning_rate": 1.566265060240964e-05,
|
| 1501 |
+
"loss": 0.0565,
|
| 1502 |
+
"step": 2120
|
| 1503 |
+
},
|
| 1504 |
+
{
|
| 1505 |
+
"epoch": 1.4423565261554088,
|
| 1506 |
+
"grad_norm": 0.5956985354423523,
|
| 1507 |
+
"learning_rate": 1.5474397590361446e-05,
|
| 1508 |
+
"loss": 0.0584,
|
| 1509 |
+
"step": 2130
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 1.4491281530387676,
|
| 1513 |
+
"grad_norm": 0.7276691794395447,
|
| 1514 |
+
"learning_rate": 1.5286144578313255e-05,
|
| 1515 |
+
"loss": 0.0589,
|
| 1516 |
+
"step": 2140
|
| 1517 |
+
},
|
| 1518 |
+
{
|
| 1519 |
+
"epoch": 1.4558997799221263,
|
| 1520 |
+
"grad_norm": 0.9040044546127319,
|
| 1521 |
+
"learning_rate": 1.509789156626506e-05,
|
| 1522 |
+
"loss": 0.0635,
|
| 1523 |
+
"step": 2150
|
| 1524 |
+
},
|
| 1525 |
+
{
|
| 1526 |
+
"epoch": 1.4626714068054851,
|
| 1527 |
+
"grad_norm": 0.7457456588745117,
|
| 1528 |
+
"learning_rate": 1.4909638554216867e-05,
|
| 1529 |
+
"loss": 0.0654,
|
| 1530 |
+
"step": 2160
|
| 1531 |
+
},
|
| 1532 |
+
{
|
| 1533 |
+
"epoch": 1.4694430336888438,
|
| 1534 |
+
"grad_norm": 0.7019338011741638,
|
| 1535 |
+
"learning_rate": 1.4721385542168676e-05,
|
| 1536 |
+
"loss": 0.0641,
|
| 1537 |
+
"step": 2170
|
| 1538 |
+
},
|
| 1539 |
+
{
|
| 1540 |
+
"epoch": 1.4762146605722024,
|
| 1541 |
+
"grad_norm": 0.8061505556106567,
|
| 1542 |
+
"learning_rate": 1.4533132530120484e-05,
|
| 1543 |
+
"loss": 0.0573,
|
| 1544 |
+
"step": 2180
|
| 1545 |
+
},
|
| 1546 |
+
{
|
| 1547 |
+
"epoch": 1.4829862874555613,
|
| 1548 |
+
"grad_norm": 0.6644711494445801,
|
| 1549 |
+
"learning_rate": 1.4344879518072291e-05,
|
| 1550 |
+
"loss": 0.0654,
|
| 1551 |
+
"step": 2190
|
| 1552 |
+
},
|
| 1553 |
+
{
|
| 1554 |
+
"epoch": 1.48975791433892,
|
| 1555 |
+
"grad_norm": 0.8262456059455872,
|
| 1556 |
+
"learning_rate": 1.4156626506024098e-05,
|
| 1557 |
+
"loss": 0.0642,
|
| 1558 |
+
"step": 2200
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 1.4965295412222788,
|
| 1562 |
+
"grad_norm": 0.9707741737365723,
|
| 1563 |
+
"learning_rate": 1.3968373493975902e-05,
|
| 1564 |
+
"loss": 0.0629,
|
| 1565 |
+
"step": 2210
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 1.5033011681056374,
|
| 1569 |
+
"grad_norm": 1.0274876356124878,
|
| 1570 |
+
"learning_rate": 1.378012048192771e-05,
|
| 1571 |
+
"loss": 0.0584,
|
| 1572 |
+
"step": 2220
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 1.510072794988996,
|
| 1576 |
+
"grad_norm": 0.5561116933822632,
|
| 1577 |
+
"learning_rate": 1.3591867469879519e-05,
|
| 1578 |
+
"loss": 0.0703,
|
| 1579 |
+
"step": 2230
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 1.5168444218723547,
|
| 1583 |
+
"grad_norm": 0.7350441217422485,
|
| 1584 |
+
"learning_rate": 1.3403614457831327e-05,
|
| 1585 |
+
"loss": 0.0679,
|
| 1586 |
+
"step": 2240
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 1.5236160487557135,
|
| 1590 |
+
"grad_norm": 1.2632744312286377,
|
| 1591 |
+
"learning_rate": 1.3215361445783134e-05,
|
| 1592 |
+
"loss": 0.0572,
|
| 1593 |
+
"step": 2250
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 1.5303876756390724,
|
| 1597 |
+
"grad_norm": 0.970457911491394,
|
| 1598 |
+
"learning_rate": 1.3027108433734941e-05,
|
| 1599 |
+
"loss": 0.057,
|
| 1600 |
+
"step": 2260
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 1.537159302522431,
|
| 1604 |
+
"grad_norm": 0.8355304598808289,
|
| 1605 |
+
"learning_rate": 1.2838855421686745e-05,
|
| 1606 |
+
"loss": 0.0543,
|
| 1607 |
+
"step": 2270
|
| 1608 |
+
},
|
| 1609 |
+
{
|
| 1610 |
+
"epoch": 1.5439309294057897,
|
| 1611 |
+
"grad_norm": 0.61915522813797,
|
| 1612 |
+
"learning_rate": 1.2650602409638555e-05,
|
| 1613 |
+
"loss": 0.0623,
|
| 1614 |
+
"step": 2280
|
| 1615 |
+
},
|
| 1616 |
+
{
|
| 1617 |
+
"epoch": 1.5507025562891483,
|
| 1618 |
+
"grad_norm": 0.5436218976974487,
|
| 1619 |
+
"learning_rate": 1.2462349397590362e-05,
|
| 1620 |
+
"loss": 0.0621,
|
| 1621 |
+
"step": 2290
|
| 1622 |
+
},
|
| 1623 |
+
{
|
| 1624 |
+
"epoch": 1.5574741831725072,
|
| 1625 |
+
"grad_norm": 0.8874384164810181,
|
| 1626 |
+
"learning_rate": 1.227409638554217e-05,
|
| 1627 |
+
"loss": 0.0598,
|
| 1628 |
+
"step": 2300
|
| 1629 |
+
},
|
| 1630 |
+
{
|
| 1631 |
+
"epoch": 1.564245810055866,
|
| 1632 |
+
"grad_norm": 0.6226460933685303,
|
| 1633 |
+
"learning_rate": 1.2085843373493977e-05,
|
| 1634 |
+
"loss": 0.0549,
|
| 1635 |
+
"step": 2310
|
| 1636 |
+
},
|
| 1637 |
+
{
|
| 1638 |
+
"epoch": 1.5710174369392247,
|
| 1639 |
+
"grad_norm": 0.648560643196106,
|
| 1640 |
+
"learning_rate": 1.1897590361445783e-05,
|
| 1641 |
+
"loss": 0.0581,
|
| 1642 |
+
"step": 2320
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"epoch": 1.5777890638225833,
|
| 1646 |
+
"grad_norm": 0.6028856039047241,
|
| 1647 |
+
"learning_rate": 1.170933734939759e-05,
|
| 1648 |
+
"loss": 0.0628,
|
| 1649 |
+
"step": 2330
|
| 1650 |
+
},
|
| 1651 |
+
{
|
| 1652 |
+
"epoch": 1.584560690705942,
|
| 1653 |
+
"grad_norm": 0.5027485489845276,
|
| 1654 |
+
"learning_rate": 1.1521084337349398e-05,
|
| 1655 |
+
"loss": 0.0655,
|
| 1656 |
+
"step": 2340
|
| 1657 |
+
},
|
| 1658 |
+
{
|
| 1659 |
+
"epoch": 1.5913323175893008,
|
| 1660 |
+
"grad_norm": 1.5245540142059326,
|
| 1661 |
+
"learning_rate": 1.1332831325301205e-05,
|
| 1662 |
+
"loss": 0.0637,
|
| 1663 |
+
"step": 2350
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 1.5981039444726597,
|
| 1667 |
+
"grad_norm": 0.7036380767822266,
|
| 1668 |
+
"learning_rate": 1.1144578313253013e-05,
|
| 1669 |
+
"loss": 0.0582,
|
| 1670 |
+
"step": 2360
|
| 1671 |
+
},
|
| 1672 |
+
{
|
| 1673 |
+
"epoch": 1.6048755713560183,
|
| 1674 |
+
"grad_norm": 0.7167279124259949,
|
| 1675 |
+
"learning_rate": 1.095632530120482e-05,
|
| 1676 |
+
"loss": 0.0542,
|
| 1677 |
+
"step": 2370
|
| 1678 |
+
},
|
| 1679 |
+
{
|
| 1680 |
+
"epoch": 1.611647198239377,
|
| 1681 |
+
"grad_norm": 0.9925076365470886,
|
| 1682 |
+
"learning_rate": 1.0768072289156627e-05,
|
| 1683 |
+
"loss": 0.0521,
|
| 1684 |
+
"step": 2380
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"epoch": 1.6184188251227356,
|
| 1688 |
+
"grad_norm": 0.8511892557144165,
|
| 1689 |
+
"learning_rate": 1.0579819277108433e-05,
|
| 1690 |
+
"loss": 0.0603,
|
| 1691 |
+
"step": 2390
|
| 1692 |
+
},
|
| 1693 |
+
{
|
| 1694 |
+
"epoch": 1.6251904520060945,
|
| 1695 |
+
"grad_norm": 1.3457633256912231,
|
| 1696 |
+
"learning_rate": 1.0391566265060242e-05,
|
| 1697 |
+
"loss": 0.0654,
|
| 1698 |
+
"step": 2400
|
| 1699 |
+
},
|
| 1700 |
+
{
|
| 1701 |
+
"epoch": 1.6319620788894533,
|
| 1702 |
+
"grad_norm": 0.8031491041183472,
|
| 1703 |
+
"learning_rate": 1.020331325301205e-05,
|
| 1704 |
+
"loss": 0.0526,
|
| 1705 |
+
"step": 2410
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 1.638733705772812,
|
| 1709 |
+
"grad_norm": 0.6882662773132324,
|
| 1710 |
+
"learning_rate": 1.0015060240963856e-05,
|
| 1711 |
+
"loss": 0.0585,
|
| 1712 |
+
"step": 2420
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 1.6455053326561706,
|
| 1716 |
+
"grad_norm": 0.5608332753181458,
|
| 1717 |
+
"learning_rate": 9.826807228915663e-06,
|
| 1718 |
+
"loss": 0.0613,
|
| 1719 |
+
"step": 2430
|
| 1720 |
+
},
|
| 1721 |
+
{
|
| 1722 |
+
"epoch": 1.6522769595395292,
|
| 1723 |
+
"grad_norm": 0.512611985206604,
|
| 1724 |
+
"learning_rate": 9.63855421686747e-06,
|
| 1725 |
+
"loss": 0.0535,
|
| 1726 |
+
"step": 2440
|
| 1727 |
+
},
|
| 1728 |
+
{
|
| 1729 |
+
"epoch": 1.659048586422888,
|
| 1730 |
+
"grad_norm": 0.48724594712257385,
|
| 1731 |
+
"learning_rate": 9.450301204819278e-06,
|
| 1732 |
+
"loss": 0.0554,
|
| 1733 |
+
"step": 2450
|
| 1734 |
+
},
|
| 1735 |
+
{
|
| 1736 |
+
"epoch": 1.665820213306247,
|
| 1737 |
+
"grad_norm": 0.7341836094856262,
|
| 1738 |
+
"learning_rate": 9.262048192771085e-06,
|
| 1739 |
+
"loss": 0.0598,
|
| 1740 |
+
"step": 2460
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 1.6725918401896056,
|
| 1744 |
+
"grad_norm": 0.5964226722717285,
|
| 1745 |
+
"learning_rate": 9.073795180722893e-06,
|
| 1746 |
+
"loss": 0.0742,
|
| 1747 |
+
"step": 2470
|
| 1748 |
+
},
|
| 1749 |
+
{
|
| 1750 |
+
"epoch": 1.6793634670729642,
|
| 1751 |
+
"grad_norm": 1.4178098440170288,
|
| 1752 |
+
"learning_rate": 8.885542168674699e-06,
|
| 1753 |
+
"loss": 0.0716,
|
| 1754 |
+
"step": 2480
|
| 1755 |
+
},
|
| 1756 |
+
{
|
| 1757 |
+
"epoch": 1.6861350939563229,
|
| 1758 |
+
"grad_norm": 0.7655882239341736,
|
| 1759 |
+
"learning_rate": 8.697289156626506e-06,
|
| 1760 |
+
"loss": 0.0509,
|
| 1761 |
+
"step": 2490
|
| 1762 |
+
},
|
| 1763 |
+
{
|
| 1764 |
+
"epoch": 1.6929067208396817,
|
| 1765 |
+
"grad_norm": 0.7825894355773926,
|
| 1766 |
+
"learning_rate": 8.509036144578313e-06,
|
| 1767 |
+
"loss": 0.056,
|
| 1768 |
+
"step": 2500
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 1.6996783477230406,
|
| 1772 |
+
"grad_norm": 0.7740017771720886,
|
| 1773 |
+
"learning_rate": 8.320783132530121e-06,
|
| 1774 |
+
"loss": 0.0669,
|
| 1775 |
+
"step": 2510
|
| 1776 |
+
},
|
| 1777 |
+
{
|
| 1778 |
+
"epoch": 1.7064499746063992,
|
| 1779 |
+
"grad_norm": 0.4478660523891449,
|
| 1780 |
+
"learning_rate": 8.132530120481928e-06,
|
| 1781 |
+
"loss": 0.0711,
|
| 1782 |
+
"step": 2520
|
| 1783 |
+
},
|
| 1784 |
+
{
|
| 1785 |
+
"epoch": 1.7132216014897579,
|
| 1786 |
+
"grad_norm": 0.9198179244995117,
|
| 1787 |
+
"learning_rate": 7.944277108433736e-06,
|
| 1788 |
+
"loss": 0.0534,
|
| 1789 |
+
"step": 2530
|
| 1790 |
+
},
|
| 1791 |
+
{
|
| 1792 |
+
"epoch": 1.7199932283731165,
|
| 1793 |
+
"grad_norm": 0.5813512802124023,
|
| 1794 |
+
"learning_rate": 7.756024096385543e-06,
|
| 1795 |
+
"loss": 0.0612,
|
| 1796 |
+
"step": 2540
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 1.7267648552564754,
|
| 1800 |
+
"grad_norm": 0.6739583611488342,
|
| 1801 |
+
"learning_rate": 7.56777108433735e-06,
|
| 1802 |
+
"loss": 0.0617,
|
| 1803 |
+
"step": 2550
|
| 1804 |
+
},
|
| 1805 |
+
{
|
| 1806 |
+
"epoch": 1.7335364821398342,
|
| 1807 |
+
"grad_norm": 0.6534674167633057,
|
| 1808 |
+
"learning_rate": 7.379518072289157e-06,
|
| 1809 |
+
"loss": 0.0639,
|
| 1810 |
+
"step": 2560
|
| 1811 |
+
},
|
| 1812 |
+
{
|
| 1813 |
+
"epoch": 1.7403081090231929,
|
| 1814 |
+
"grad_norm": 0.6965113878250122,
|
| 1815 |
+
"learning_rate": 7.191265060240965e-06,
|
| 1816 |
+
"loss": 0.0543,
|
| 1817 |
+
"step": 2570
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 1.7470797359065515,
|
| 1821 |
+
"grad_norm": 0.7045726776123047,
|
| 1822 |
+
"learning_rate": 7.003012048192771e-06,
|
| 1823 |
+
"loss": 0.0557,
|
| 1824 |
+
"step": 2580
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 1.7538513627899102,
|
| 1828 |
+
"grad_norm": 0.7541308403015137,
|
| 1829 |
+
"learning_rate": 6.814759036144579e-06,
|
| 1830 |
+
"loss": 0.0609,
|
| 1831 |
+
"step": 2590
|
| 1832 |
+
},
|
| 1833 |
+
{
|
| 1834 |
+
"epoch": 1.760622989673269,
|
| 1835 |
+
"grad_norm": 1.2629750967025757,
|
| 1836 |
+
"learning_rate": 6.626506024096386e-06,
|
| 1837 |
+
"loss": 0.055,
|
| 1838 |
+
"step": 2600
|
| 1839 |
+
},
|
| 1840 |
+
{
|
| 1841 |
+
"epoch": 1.7673946165566279,
|
| 1842 |
+
"grad_norm": 0.8643785119056702,
|
| 1843 |
+
"learning_rate": 6.438253012048193e-06,
|
| 1844 |
+
"loss": 0.0675,
|
| 1845 |
+
"step": 2610
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 1.7741662434399865,
|
| 1849 |
+
"grad_norm": 0.7467519640922546,
|
| 1850 |
+
"learning_rate": 6.25e-06,
|
| 1851 |
+
"loss": 0.0641,
|
| 1852 |
+
"step": 2620
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 1.7809378703233452,
|
| 1856 |
+
"grad_norm": 0.4777075946331024,
|
| 1857 |
+
"learning_rate": 6.061746987951807e-06,
|
| 1858 |
+
"loss": 0.0605,
|
| 1859 |
+
"step": 2630
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 1.7877094972067038,
|
| 1863 |
+
"grad_norm": 0.6084995865821838,
|
| 1864 |
+
"learning_rate": 5.873493975903615e-06,
|
| 1865 |
+
"loss": 0.0545,
|
| 1866 |
+
"step": 2640
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 1.7944811240900627,
|
| 1870 |
+
"grad_norm": 0.9602014422416687,
|
| 1871 |
+
"learning_rate": 5.685240963855422e-06,
|
| 1872 |
+
"loss": 0.0484,
|
| 1873 |
+
"step": 2650
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 1.8012527509734215,
|
| 1877 |
+
"grad_norm": 0.7709717154502869,
|
| 1878 |
+
"learning_rate": 5.496987951807229e-06,
|
| 1879 |
+
"loss": 0.0637,
|
| 1880 |
+
"step": 2660
|
| 1881 |
+
},
|
| 1882 |
+
{
|
| 1883 |
+
"epoch": 1.8080243778567802,
|
| 1884 |
+
"grad_norm": 0.422376424074173,
|
| 1885 |
+
"learning_rate": 5.308734939759037e-06,
|
| 1886 |
+
"loss": 0.0659,
|
| 1887 |
+
"step": 2670
|
| 1888 |
+
},
|
| 1889 |
+
{
|
| 1890 |
+
"epoch": 1.8147960047401388,
|
| 1891 |
+
"grad_norm": 0.4001966118812561,
|
| 1892 |
+
"learning_rate": 5.120481927710843e-06,
|
| 1893 |
+
"loss": 0.0609,
|
| 1894 |
+
"step": 2680
|
| 1895 |
+
},
|
| 1896 |
+
{
|
| 1897 |
+
"epoch": 1.8215676316234974,
|
| 1898 |
+
"grad_norm": 0.4686823785305023,
|
| 1899 |
+
"learning_rate": 4.932228915662651e-06,
|
| 1900 |
+
"loss": 0.0588,
|
| 1901 |
+
"step": 2690
|
| 1902 |
+
},
|
| 1903 |
+
{
|
| 1904 |
+
"epoch": 1.8283392585068563,
|
| 1905 |
+
"grad_norm": 0.4594326615333557,
|
| 1906 |
+
"learning_rate": 4.743975903614458e-06,
|
| 1907 |
+
"loss": 0.0558,
|
| 1908 |
+
"step": 2700
|
| 1909 |
+
},
|
| 1910 |
+
{
|
| 1911 |
+
"epoch": 1.8351108853902152,
|
| 1912 |
+
"grad_norm": 0.7824903130531311,
|
| 1913 |
+
"learning_rate": 4.555722891566265e-06,
|
| 1914 |
+
"loss": 0.0536,
|
| 1915 |
+
"step": 2710
|
| 1916 |
+
},
|
| 1917 |
+
{
|
| 1918 |
+
"epoch": 1.8418825122735738,
|
| 1919 |
+
"grad_norm": 0.6944392323493958,
|
| 1920 |
+
"learning_rate": 4.367469879518073e-06,
|
| 1921 |
+
"loss": 0.0653,
|
| 1922 |
+
"step": 2720
|
| 1923 |
+
},
|
| 1924 |
+
{
|
| 1925 |
+
"epoch": 1.8486541391569324,
|
| 1926 |
+
"grad_norm": 0.665273129940033,
|
| 1927 |
+
"learning_rate": 4.17921686746988e-06,
|
| 1928 |
+
"loss": 0.0544,
|
| 1929 |
+
"step": 2730
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 1.855425766040291,
|
| 1933 |
+
"grad_norm": 0.6628099083900452,
|
| 1934 |
+
"learning_rate": 3.990963855421686e-06,
|
| 1935 |
+
"loss": 0.056,
|
| 1936 |
+
"step": 2740
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 1.86219739292365,
|
| 1940 |
+
"grad_norm": 0.79542475938797,
|
| 1941 |
+
"learning_rate": 3.802710843373494e-06,
|
| 1942 |
+
"loss": 0.0549,
|
| 1943 |
+
"step": 2750
|
| 1944 |
+
},
|
| 1945 |
+
{
|
| 1946 |
+
"epoch": 1.8689690198070086,
|
| 1947 |
+
"grad_norm": 0.7736166715621948,
|
| 1948 |
+
"learning_rate": 3.614457831325301e-06,
|
| 1949 |
+
"loss": 0.0606,
|
| 1950 |
+
"step": 2760
|
| 1951 |
+
},
|
| 1952 |
+
{
|
| 1953 |
+
"epoch": 1.8757406466903674,
|
| 1954 |
+
"grad_norm": 1.0499205589294434,
|
| 1955 |
+
"learning_rate": 3.426204819277109e-06,
|
| 1956 |
+
"loss": 0.0656,
|
| 1957 |
+
"step": 2770
|
| 1958 |
+
},
|
| 1959 |
+
{
|
| 1960 |
+
"epoch": 1.882512273573726,
|
| 1961 |
+
"grad_norm": 0.831328809261322,
|
| 1962 |
+
"learning_rate": 3.2379518072289157e-06,
|
| 1963 |
+
"loss": 0.0575,
|
| 1964 |
+
"step": 2780
|
| 1965 |
+
},
|
| 1966 |
+
{
|
| 1967 |
+
"epoch": 1.8892839004570847,
|
| 1968 |
+
"grad_norm": 0.7741392254829407,
|
| 1969 |
+
"learning_rate": 3.049698795180723e-06,
|
| 1970 |
+
"loss": 0.0493,
|
| 1971 |
+
"step": 2790
|
| 1972 |
+
},
|
| 1973 |
+
{
|
| 1974 |
+
"epoch": 1.8960555273404436,
|
| 1975 |
+
"grad_norm": 0.5156965851783752,
|
| 1976 |
+
"learning_rate": 2.86144578313253e-06,
|
| 1977 |
+
"loss": 0.0574,
|
| 1978 |
+
"step": 2800
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 1.9028271542238022,
|
| 1982 |
+
"grad_norm": 0.7923727035522461,
|
| 1983 |
+
"learning_rate": 2.6731927710843376e-06,
|
| 1984 |
+
"loss": 0.0496,
|
| 1985 |
+
"step": 2810
|
| 1986 |
+
},
|
| 1987 |
+
{
|
| 1988 |
+
"epoch": 1.909598781107161,
|
| 1989 |
+
"grad_norm": 0.6281595230102539,
|
| 1990 |
+
"learning_rate": 2.4849397590361446e-06,
|
| 1991 |
+
"loss": 0.0564,
|
| 1992 |
+
"step": 2820
|
| 1993 |
+
},
|
| 1994 |
+
{
|
| 1995 |
+
"epoch": 1.9163704079905197,
|
| 1996 |
+
"grad_norm": 0.475424200296402,
|
| 1997 |
+
"learning_rate": 2.296686746987952e-06,
|
| 1998 |
+
"loss": 0.0518,
|
| 1999 |
+
"step": 2830
|
| 2000 |
+
},
|
| 2001 |
+
{
|
| 2002 |
+
"epoch": 1.9231420348738784,
|
| 2003 |
+
"grad_norm": 0.808529794216156,
|
| 2004 |
+
"learning_rate": 2.108433734939759e-06,
|
| 2005 |
+
"loss": 0.0519,
|
| 2006 |
+
"step": 2840
|
| 2007 |
+
},
|
| 2008 |
+
{
|
| 2009 |
+
"epoch": 1.9299136617572372,
|
| 2010 |
+
"grad_norm": 0.9661712050437927,
|
| 2011 |
+
"learning_rate": 1.920180722891566e-06,
|
| 2012 |
+
"loss": 0.057,
|
| 2013 |
+
"step": 2850
|
| 2014 |
+
},
|
| 2015 |
+
{
|
| 2016 |
+
"epoch": 1.9366852886405959,
|
| 2017 |
+
"grad_norm": 0.7859961986541748,
|
| 2018 |
+
"learning_rate": 1.7319277108433736e-06,
|
| 2019 |
+
"loss": 0.0597,
|
| 2020 |
+
"step": 2860
|
| 2021 |
+
},
|
| 2022 |
+
{
|
| 2023 |
+
"epoch": 1.9434569155239547,
|
| 2024 |
+
"grad_norm": 0.8376522660255432,
|
| 2025 |
+
"learning_rate": 1.5436746987951808e-06,
|
| 2026 |
+
"loss": 0.0562,
|
| 2027 |
+
"step": 2870
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 1.9502285424073134,
|
| 2031 |
+
"grad_norm": 1.249200463294983,
|
| 2032 |
+
"learning_rate": 1.355421686746988e-06,
|
| 2033 |
+
"loss": 0.0582,
|
| 2034 |
+
"step": 2880
|
| 2035 |
+
},
|
| 2036 |
+
{
|
| 2037 |
+
"epoch": 1.957000169290672,
|
| 2038 |
+
"grad_norm": 0.6086472272872925,
|
| 2039 |
+
"learning_rate": 1.167168674698795e-06,
|
| 2040 |
+
"loss": 0.0575,
|
| 2041 |
+
"step": 2890
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 1.9637717961740309,
|
| 2045 |
+
"grad_norm": 0.5852237939834595,
|
| 2046 |
+
"learning_rate": 9.789156626506025e-07,
|
| 2047 |
+
"loss": 0.0523,
|
| 2048 |
+
"step": 2900
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 1.9705434230573895,
|
| 2052 |
+
"grad_norm": 0.5044909119606018,
|
| 2053 |
+
"learning_rate": 7.906626506024097e-07,
|
| 2054 |
+
"loss": 0.0599,
|
| 2055 |
+
"step": 2910
|
| 2056 |
+
},
|
| 2057 |
+
{
|
| 2058 |
+
"epoch": 1.9773150499407484,
|
| 2059 |
+
"grad_norm": 0.5792267918586731,
|
| 2060 |
+
"learning_rate": 6.024096385542169e-07,
|
| 2061 |
+
"loss": 0.0514,
|
| 2062 |
+
"step": 2920
|
| 2063 |
+
},
|
| 2064 |
+
{
|
| 2065 |
+
"epoch": 1.984086676824107,
|
| 2066 |
+
"grad_norm": 0.6914946436882019,
|
| 2067 |
+
"learning_rate": 4.1415662650602414e-07,
|
| 2068 |
+
"loss": 0.0582,
|
| 2069 |
+
"step": 2930
|
| 2070 |
+
},
|
| 2071 |
+
{
|
| 2072 |
+
"epoch": 1.9908583037074656,
|
| 2073 |
+
"grad_norm": 1.0799570083618164,
|
| 2074 |
+
"learning_rate": 2.2590361445783133e-07,
|
| 2075 |
+
"loss": 0.0467,
|
| 2076 |
+
"step": 2940
|
| 2077 |
+
},
|
| 2078 |
+
{
|
| 2079 |
+
"epoch": 1.9976299305908245,
|
| 2080 |
+
"grad_norm": 0.7630107998847961,
|
| 2081 |
+
"learning_rate": 3.7650602409638556e-08,
|
| 2082 |
+
"loss": 0.0634,
|
| 2083 |
+
"step": 2950
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 1.9989842559674962,
|
| 2087 |
+
"eval_accuracy": 0.9820952380952381,
|
| 2088 |
+
"eval_loss": 0.062325455248355865,
|
| 2089 |
+
"eval_runtime": 751.2071,
|
| 2090 |
+
"eval_samples_per_second": 27.955,
|
| 2091 |
+
"eval_steps_per_second": 0.875,
|
| 2092 |
+
"step": 2952
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 1.9989842559674962,
|
| 2096 |
+
"step": 2952,
|
| 2097 |
+
"total_flos": 1.3566159569165903e+20,
|
| 2098 |
+
"train_loss": 0.14182275843612224,
|
| 2099 |
+
"train_runtime": 17938.7006,
|
| 2100 |
+
"train_samples_per_second": 21.072,
|
| 2101 |
+
"train_steps_per_second": 0.165
|
| 2102 |
+
}
|
| 2103 |
+
],
|
| 2104 |
+
"logging_steps": 10,
|
| 2105 |
+
"max_steps": 2952,
|
| 2106 |
+
"num_input_tokens_seen": 0,
|
| 2107 |
+
"num_train_epochs": 2,
|
| 2108 |
+
"save_steps": 500,
|
| 2109 |
+
"stateful_callbacks": {
|
| 2110 |
+
"TrainerControl": {
|
| 2111 |
+
"args": {
|
| 2112 |
+
"should_epoch_stop": false,
|
| 2113 |
+
"should_evaluate": false,
|
| 2114 |
+
"should_log": false,
|
| 2115 |
+
"should_save": true,
|
| 2116 |
+
"should_training_stop": true
|
| 2117 |
+
},
|
| 2118 |
+
"attributes": {}
|
| 2119 |
+
}
|
| 2120 |
+
},
|
| 2121 |
+
"total_flos": 1.3566159569165903e+20,
|
| 2122 |
+
"train_batch_size": 32,
|
| 2123 |
+
"trial_name": null,
|
| 2124 |
+
"trial_params": null
|
| 2125 |
+
}
|