End of training
Browse files- all_results.json +13 -0
- eval_results.json +8 -0
- runs/Jan21_12-29-04_1b75f888b432/events.out.tfevents.1737467813.1b75f888b432.2972.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +2054 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.997333333333334,
|
| 3 |
+
"eval_accuracy": 1.0,
|
| 4 |
+
"eval_loss": 2.338912281629746e-06,
|
| 5 |
+
"eval_runtime": 40.3723,
|
| 6 |
+
"eval_samples_per_second": 198.156,
|
| 7 |
+
"eval_steps_per_second": 6.192,
|
| 8 |
+
"total_flos": 8.94051665811918e+18,
|
| 9 |
+
"train_loss": 0.011076973860487225,
|
| 10 |
+
"train_runtime": 5179.8163,
|
| 11 |
+
"train_samples_per_second": 69.501,
|
| 12 |
+
"train_steps_per_second": 0.542
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.997333333333334,
|
| 3 |
+
"eval_accuracy": 1.0,
|
| 4 |
+
"eval_loss": 2.338912281629746e-06,
|
| 5 |
+
"eval_runtime": 40.3723,
|
| 6 |
+
"eval_samples_per_second": 198.156,
|
| 7 |
+
"eval_steps_per_second": 6.192
|
| 8 |
+
}
|
runs/Jan21_12-29-04_1b75f888b432/events.out.tfevents.1737467813.1b75f888b432.2972.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d6ceb62b05c2ceaa221b569ce71e01104c6dd33f1af490311dee31ed9c75e6c7
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.997333333333334,
|
| 3 |
+
"total_flos": 8.94051665811918e+18,
|
| 4 |
+
"train_loss": 0.011076973860487225,
|
| 5 |
+
"train_runtime": 5179.8163,
|
| 6 |
+
"train_samples_per_second": 69.501,
|
| 7 |
+
"train_steps_per_second": 0.542
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2054 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.0,
|
| 3 |
+
"best_model_checkpoint": "/content/drive/MyDrive/Colab Notebooks/16_label_check_point/checkpoint-563",
|
| 4 |
+
"epoch": 4.997333333333334,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2810,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.017777777777777778,
|
| 13 |
+
"grad_norm": 10.457254409790039,
|
| 14 |
+
"learning_rate": 1.7793594306049826e-06,
|
| 15 |
+
"loss": 11.2334,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.035555555555555556,
|
| 20 |
+
"grad_norm": 6.316349506378174,
|
| 21 |
+
"learning_rate": 3.558718861209965e-06,
|
| 22 |
+
"loss": 11.1807,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05333333333333334,
|
| 27 |
+
"grad_norm": 5.264202117919922,
|
| 28 |
+
"learning_rate": 5.338078291814947e-06,
|
| 29 |
+
"loss": 11.1463,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07111111111111111,
|
| 34 |
+
"grad_norm": 5.205317497253418,
|
| 35 |
+
"learning_rate": 7.11743772241993e-06,
|
| 36 |
+
"loss": 11.0929,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.08888888888888889,
|
| 41 |
+
"grad_norm": 4.696351528167725,
|
| 42 |
+
"learning_rate": 8.896797153024912e-06,
|
| 43 |
+
"loss": 11.1015,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.10666666666666667,
|
| 48 |
+
"grad_norm": 5.9699320793151855,
|
| 49 |
+
"learning_rate": 1.0676156583629894e-05,
|
| 50 |
+
"loss": 11.0795,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12444444444444444,
|
| 55 |
+
"grad_norm": 7.235191822052002,
|
| 56 |
+
"learning_rate": 1.2455516014234877e-05,
|
| 57 |
+
"loss": 11.046,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14222222222222222,
|
| 62 |
+
"grad_norm": 14.865583419799805,
|
| 63 |
+
"learning_rate": 1.423487544483986e-05,
|
| 64 |
+
"loss": 11.0113,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.16,
|
| 69 |
+
"grad_norm": 17.412696838378906,
|
| 70 |
+
"learning_rate": 1.601423487544484e-05,
|
| 71 |
+
"loss": 10.8502,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.17777777777777778,
|
| 76 |
+
"grad_norm": 20.41786003112793,
|
| 77 |
+
"learning_rate": 1.7793594306049825e-05,
|
| 78 |
+
"loss": 10.5185,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.19555555555555557,
|
| 83 |
+
"grad_norm": 24.58249282836914,
|
| 84 |
+
"learning_rate": 1.9572953736654805e-05,
|
| 85 |
+
"loss": 9.8301,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.21333333333333335,
|
| 90 |
+
"grad_norm": 43.27064895629883,
|
| 91 |
+
"learning_rate": 2.135231316725979e-05,
|
| 92 |
+
"loss": 8.8636,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.2311111111111111,
|
| 97 |
+
"grad_norm": 46.2359733581543,
|
| 98 |
+
"learning_rate": 2.313167259786477e-05,
|
| 99 |
+
"loss": 7.4813,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.24888888888888888,
|
| 104 |
+
"grad_norm": 71.36530303955078,
|
| 105 |
+
"learning_rate": 2.4911032028469753e-05,
|
| 106 |
+
"loss": 6.0609,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.26666666666666666,
|
| 111 |
+
"grad_norm": 61.219024658203125,
|
| 112 |
+
"learning_rate": 2.669039145907473e-05,
|
| 113 |
+
"loss": 4.5934,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.28444444444444444,
|
| 118 |
+
"grad_norm": 52.64271545410156,
|
| 119 |
+
"learning_rate": 2.846975088967972e-05,
|
| 120 |
+
"loss": 3.3049,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3022222222222222,
|
| 125 |
+
"grad_norm": 40.045623779296875,
|
| 126 |
+
"learning_rate": 3.02491103202847e-05,
|
| 127 |
+
"loss": 2.0759,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.32,
|
| 132 |
+
"grad_norm": 32.63826370239258,
|
| 133 |
+
"learning_rate": 3.202846975088968e-05,
|
| 134 |
+
"loss": 1.2791,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3377777777777778,
|
| 139 |
+
"grad_norm": 39.16119384765625,
|
| 140 |
+
"learning_rate": 3.380782918149467e-05,
|
| 141 |
+
"loss": 0.9052,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.35555555555555557,
|
| 146 |
+
"grad_norm": 36.06990051269531,
|
| 147 |
+
"learning_rate": 3.558718861209965e-05,
|
| 148 |
+
"loss": 0.6508,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.37333333333333335,
|
| 153 |
+
"grad_norm": 22.00220489501953,
|
| 154 |
+
"learning_rate": 3.736654804270463e-05,
|
| 155 |
+
"loss": 0.6293,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.39111111111111113,
|
| 160 |
+
"grad_norm": 27.334341049194336,
|
| 161 |
+
"learning_rate": 3.914590747330961e-05,
|
| 162 |
+
"loss": 0.5774,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.4088888888888889,
|
| 167 |
+
"grad_norm": 21.130746841430664,
|
| 168 |
+
"learning_rate": 4.09252669039146e-05,
|
| 169 |
+
"loss": 0.4318,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.4266666666666667,
|
| 174 |
+
"grad_norm": 21.37102508544922,
|
| 175 |
+
"learning_rate": 4.270462633451958e-05,
|
| 176 |
+
"loss": 0.414,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4444444444444444,
|
| 181 |
+
"grad_norm": 86.8355712890625,
|
| 182 |
+
"learning_rate": 4.448398576512456e-05,
|
| 183 |
+
"loss": 0.4338,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.4622222222222222,
|
| 188 |
+
"grad_norm": 9.731348037719727,
|
| 189 |
+
"learning_rate": 4.626334519572954e-05,
|
| 190 |
+
"loss": 0.3978,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.48,
|
| 195 |
+
"grad_norm": 23.59229278564453,
|
| 196 |
+
"learning_rate": 4.8042704626334526e-05,
|
| 197 |
+
"loss": 0.2971,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.49777777777777776,
|
| 202 |
+
"grad_norm": 10.458292961120605,
|
| 203 |
+
"learning_rate": 4.9822064056939506e-05,
|
| 204 |
+
"loss": 0.3015,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5155555555555555,
|
| 209 |
+
"grad_norm": 13.044571876525879,
|
| 210 |
+
"learning_rate": 4.9822064056939506e-05,
|
| 211 |
+
"loss": 0.2013,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5333333333333333,
|
| 216 |
+
"grad_norm": 30.592700958251953,
|
| 217 |
+
"learning_rate": 4.962435745353895e-05,
|
| 218 |
+
"loss": 0.182,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5511111111111111,
|
| 223 |
+
"grad_norm": 26.582555770874023,
|
| 224 |
+
"learning_rate": 4.9426650850138396e-05,
|
| 225 |
+
"loss": 0.2276,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5688888888888889,
|
| 230 |
+
"grad_norm": 20.704526901245117,
|
| 231 |
+
"learning_rate": 4.9228944246737844e-05,
|
| 232 |
+
"loss": 0.1836,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5866666666666667,
|
| 237 |
+
"grad_norm": 29.400476455688477,
|
| 238 |
+
"learning_rate": 4.903123764333729e-05,
|
| 239 |
+
"loss": 0.2921,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6044444444444445,
|
| 244 |
+
"grad_norm": 23.031789779663086,
|
| 245 |
+
"learning_rate": 4.8833531039936733e-05,
|
| 246 |
+
"loss": 0.0848,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.6222222222222222,
|
| 251 |
+
"grad_norm": 0.9449447393417358,
|
| 252 |
+
"learning_rate": 4.863582443653618e-05,
|
| 253 |
+
"loss": 0.0741,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.64,
|
| 258 |
+
"grad_norm": 20.334348678588867,
|
| 259 |
+
"learning_rate": 4.843811783313563e-05,
|
| 260 |
+
"loss": 0.1798,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6577777777777778,
|
| 265 |
+
"grad_norm": 13.533489227294922,
|
| 266 |
+
"learning_rate": 4.824041122973508e-05,
|
| 267 |
+
"loss": 0.1007,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.6755555555555556,
|
| 272 |
+
"grad_norm": 1.5982263088226318,
|
| 273 |
+
"learning_rate": 4.8042704626334526e-05,
|
| 274 |
+
"loss": 0.1016,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6933333333333334,
|
| 279 |
+
"grad_norm": 2.323336362838745,
|
| 280 |
+
"learning_rate": 4.784499802293397e-05,
|
| 281 |
+
"loss": 0.1131,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7111111111111111,
|
| 286 |
+
"grad_norm": 24.458276748657227,
|
| 287 |
+
"learning_rate": 4.7647291419533415e-05,
|
| 288 |
+
"loss": 0.069,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7288888888888889,
|
| 293 |
+
"grad_norm": 29.530794143676758,
|
| 294 |
+
"learning_rate": 4.7449584816132864e-05,
|
| 295 |
+
"loss": 0.2415,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7466666666666667,
|
| 300 |
+
"grad_norm": 2.0636749267578125,
|
| 301 |
+
"learning_rate": 4.725187821273231e-05,
|
| 302 |
+
"loss": 0.1173,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.7644444444444445,
|
| 307 |
+
"grad_norm": 2.225900888442993,
|
| 308 |
+
"learning_rate": 4.705417160933175e-05,
|
| 309 |
+
"loss": 0.0918,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7822222222222223,
|
| 314 |
+
"grad_norm": 7.136375904083252,
|
| 315 |
+
"learning_rate": 4.68564650059312e-05,
|
| 316 |
+
"loss": 0.1589,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8,
|
| 321 |
+
"grad_norm": 2.6402971744537354,
|
| 322 |
+
"learning_rate": 4.665875840253064e-05,
|
| 323 |
+
"loss": 0.2547,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8177777777777778,
|
| 328 |
+
"grad_norm": 12.275321960449219,
|
| 329 |
+
"learning_rate": 4.64610517991301e-05,
|
| 330 |
+
"loss": 0.1082,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8355555555555556,
|
| 335 |
+
"grad_norm": 0.6636475920677185,
|
| 336 |
+
"learning_rate": 4.626334519572954e-05,
|
| 337 |
+
"loss": 0.0986,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.8533333333333334,
|
| 342 |
+
"grad_norm": 27.784332275390625,
|
| 343 |
+
"learning_rate": 4.606563859232899e-05,
|
| 344 |
+
"loss": 0.1056,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.8711111111111111,
|
| 349 |
+
"grad_norm": 22.047527313232422,
|
| 350 |
+
"learning_rate": 4.586793198892843e-05,
|
| 351 |
+
"loss": 0.1654,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.8888888888888888,
|
| 356 |
+
"grad_norm": 15.613493919372559,
|
| 357 |
+
"learning_rate": 4.5670225385527876e-05,
|
| 358 |
+
"loss": 0.1517,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9066666666666666,
|
| 363 |
+
"grad_norm": 20.039813995361328,
|
| 364 |
+
"learning_rate": 4.5472518782127324e-05,
|
| 365 |
+
"loss": 0.1283,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9244444444444444,
|
| 370 |
+
"grad_norm": 21.469423294067383,
|
| 371 |
+
"learning_rate": 4.527481217872677e-05,
|
| 372 |
+
"loss": 0.0707,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.9422222222222222,
|
| 377 |
+
"grad_norm": 38.251953125,
|
| 378 |
+
"learning_rate": 4.5077105575326214e-05,
|
| 379 |
+
"loss": 0.1116,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.96,
|
| 384 |
+
"grad_norm": 12.317658424377441,
|
| 385 |
+
"learning_rate": 4.487939897192566e-05,
|
| 386 |
+
"loss": 0.138,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.9777777777777777,
|
| 391 |
+
"grad_norm": 5.918329238891602,
|
| 392 |
+
"learning_rate": 4.468169236852511e-05,
|
| 393 |
+
"loss": 0.1053,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.9955555555555555,
|
| 398 |
+
"grad_norm": 12.31584358215332,
|
| 399 |
+
"learning_rate": 4.448398576512456e-05,
|
| 400 |
+
"loss": 0.1354,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.0,
|
| 405 |
+
"eval_accuracy": 1.0,
|
| 406 |
+
"eval_loss": 8.097552927210927e-05,
|
| 407 |
+
"eval_runtime": 74.3955,
|
| 408 |
+
"eval_samples_per_second": 107.533,
|
| 409 |
+
"eval_steps_per_second": 3.36,
|
| 410 |
+
"step": 563
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.0124444444444445,
|
| 414 |
+
"grad_norm": 10.894996643066406,
|
| 415 |
+
"learning_rate": 4.4286279161724006e-05,
|
| 416 |
+
"loss": 0.1134,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.0302222222222222,
|
| 421 |
+
"grad_norm": 16.037940979003906,
|
| 422 |
+
"learning_rate": 4.408857255832345e-05,
|
| 423 |
+
"loss": 0.0917,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.048,
|
| 428 |
+
"grad_norm": 0.164012148976326,
|
| 429 |
+
"learning_rate": 4.3890865954922896e-05,
|
| 430 |
+
"loss": 0.0816,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.0657777777777777,
|
| 435 |
+
"grad_norm": 0.05601898953318596,
|
| 436 |
+
"learning_rate": 4.3693159351522344e-05,
|
| 437 |
+
"loss": 0.0778,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.0835555555555556,
|
| 442 |
+
"grad_norm": 9.240290641784668,
|
| 443 |
+
"learning_rate": 4.349545274812179e-05,
|
| 444 |
+
"loss": 0.0893,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.1013333333333333,
|
| 449 |
+
"grad_norm": 8.448566436767578,
|
| 450 |
+
"learning_rate": 4.3297746144721233e-05,
|
| 451 |
+
"loss": 0.0905,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.1191111111111112,
|
| 456 |
+
"grad_norm": 1.9537012577056885,
|
| 457 |
+
"learning_rate": 4.310003954132068e-05,
|
| 458 |
+
"loss": 0.1561,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.1368888888888888,
|
| 463 |
+
"grad_norm": 25.13317108154297,
|
| 464 |
+
"learning_rate": 4.290233293792013e-05,
|
| 465 |
+
"loss": 0.0856,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.1546666666666667,
|
| 470 |
+
"grad_norm": 1.3837047815322876,
|
| 471 |
+
"learning_rate": 4.270462633451958e-05,
|
| 472 |
+
"loss": 0.0962,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.1724444444444444,
|
| 477 |
+
"grad_norm": 46.560367584228516,
|
| 478 |
+
"learning_rate": 4.250691973111902e-05,
|
| 479 |
+
"loss": 0.1009,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.1902222222222223,
|
| 484 |
+
"grad_norm": 42.40678787231445,
|
| 485 |
+
"learning_rate": 4.230921312771847e-05,
|
| 486 |
+
"loss": 0.1027,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.208,
|
| 491 |
+
"grad_norm": 1.2823681831359863,
|
| 492 |
+
"learning_rate": 4.211150652431791e-05,
|
| 493 |
+
"loss": 0.0609,
|
| 494 |
+
"step": 680
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.2257777777777779,
|
| 498 |
+
"grad_norm": 3.9979896545410156,
|
| 499 |
+
"learning_rate": 4.1913799920917364e-05,
|
| 500 |
+
"loss": 0.0884,
|
| 501 |
+
"step": 690
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.2435555555555555,
|
| 505 |
+
"grad_norm": 23.4843692779541,
|
| 506 |
+
"learning_rate": 4.1716093317516805e-05,
|
| 507 |
+
"loss": 0.049,
|
| 508 |
+
"step": 700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.2613333333333334,
|
| 512 |
+
"grad_norm": 0.9539620876312256,
|
| 513 |
+
"learning_rate": 4.151838671411625e-05,
|
| 514 |
+
"loss": 0.0481,
|
| 515 |
+
"step": 710
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.279111111111111,
|
| 519 |
+
"grad_norm": 0.23285511136054993,
|
| 520 |
+
"learning_rate": 4.13206801107157e-05,
|
| 521 |
+
"loss": 0.0388,
|
| 522 |
+
"step": 720
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.2968888888888888,
|
| 526 |
+
"grad_norm": 24.404285430908203,
|
| 527 |
+
"learning_rate": 4.112297350731515e-05,
|
| 528 |
+
"loss": 0.0425,
|
| 529 |
+
"step": 730
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.3146666666666667,
|
| 533 |
+
"grad_norm": 3.143155813217163,
|
| 534 |
+
"learning_rate": 4.09252669039146e-05,
|
| 535 |
+
"loss": 0.0313,
|
| 536 |
+
"step": 740
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.3324444444444445,
|
| 540 |
+
"grad_norm": 0.3580031991004944,
|
| 541 |
+
"learning_rate": 4.072756030051404e-05,
|
| 542 |
+
"loss": 0.0466,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.3502222222222222,
|
| 547 |
+
"grad_norm": 39.96261978149414,
|
| 548 |
+
"learning_rate": 4.052985369711349e-05,
|
| 549 |
+
"loss": 0.0706,
|
| 550 |
+
"step": 760
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.3679999999999999,
|
| 554 |
+
"grad_norm": 22.012971878051758,
|
| 555 |
+
"learning_rate": 4.033214709371293e-05,
|
| 556 |
+
"loss": 0.1249,
|
| 557 |
+
"step": 770
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.3857777777777778,
|
| 561 |
+
"grad_norm": 0.20229700207710266,
|
| 562 |
+
"learning_rate": 4.013444049031238e-05,
|
| 563 |
+
"loss": 0.1089,
|
| 564 |
+
"step": 780
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.4035555555555557,
|
| 568 |
+
"grad_norm": 36.518348693847656,
|
| 569 |
+
"learning_rate": 3.9936733886911825e-05,
|
| 570 |
+
"loss": 0.0514,
|
| 571 |
+
"step": 790
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.4213333333333333,
|
| 575 |
+
"grad_norm": 0.011868222616612911,
|
| 576 |
+
"learning_rate": 3.973902728351127e-05,
|
| 577 |
+
"loss": 0.0298,
|
| 578 |
+
"step": 800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.439111111111111,
|
| 582 |
+
"grad_norm": 23.455787658691406,
|
| 583 |
+
"learning_rate": 3.9541320680110714e-05,
|
| 584 |
+
"loss": 0.0365,
|
| 585 |
+
"step": 810
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.456888888888889,
|
| 589 |
+
"grad_norm": 0.5454981923103333,
|
| 590 |
+
"learning_rate": 3.934361407671016e-05,
|
| 591 |
+
"loss": 0.0603,
|
| 592 |
+
"step": 820
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.4746666666666668,
|
| 596 |
+
"grad_norm": 0.2658223509788513,
|
| 597 |
+
"learning_rate": 3.914590747330961e-05,
|
| 598 |
+
"loss": 0.0364,
|
| 599 |
+
"step": 830
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.4924444444444445,
|
| 603 |
+
"grad_norm": 32.6451301574707,
|
| 604 |
+
"learning_rate": 3.894820086990906e-05,
|
| 605 |
+
"loss": 0.1105,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.5102222222222221,
|
| 610 |
+
"grad_norm": 0.37322714924812317,
|
| 611 |
+
"learning_rate": 3.87504942665085e-05,
|
| 612 |
+
"loss": 0.0102,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.528,
|
| 617 |
+
"grad_norm": 1.2302775382995605,
|
| 618 |
+
"learning_rate": 3.855278766310795e-05,
|
| 619 |
+
"loss": 0.093,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.545777777777778,
|
| 624 |
+
"grad_norm": 0.11981203407049179,
|
| 625 |
+
"learning_rate": 3.8355081059707396e-05,
|
| 626 |
+
"loss": 0.0721,
|
| 627 |
+
"step": 870
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.5635555555555556,
|
| 631 |
+
"grad_norm": 0.09180541336536407,
|
| 632 |
+
"learning_rate": 3.8157374456306844e-05,
|
| 633 |
+
"loss": 0.0408,
|
| 634 |
+
"step": 880
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.5813333333333333,
|
| 638 |
+
"grad_norm": 29.872051239013672,
|
| 639 |
+
"learning_rate": 3.7959667852906285e-05,
|
| 640 |
+
"loss": 0.0393,
|
| 641 |
+
"step": 890
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.5991111111111111,
|
| 645 |
+
"grad_norm": 1.0710923671722412,
|
| 646 |
+
"learning_rate": 3.7761961249505734e-05,
|
| 647 |
+
"loss": 0.0023,
|
| 648 |
+
"step": 900
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.616888888888889,
|
| 652 |
+
"grad_norm": 20.386173248291016,
|
| 653 |
+
"learning_rate": 3.756425464610518e-05,
|
| 654 |
+
"loss": 0.0605,
|
| 655 |
+
"step": 910
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.6346666666666667,
|
| 659 |
+
"grad_norm": 0.03605956584215164,
|
| 660 |
+
"learning_rate": 3.736654804270463e-05,
|
| 661 |
+
"loss": 0.0059,
|
| 662 |
+
"step": 920
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.6524444444444444,
|
| 666 |
+
"grad_norm": 0.38812369108200073,
|
| 667 |
+
"learning_rate": 3.716884143930408e-05,
|
| 668 |
+
"loss": 0.0819,
|
| 669 |
+
"step": 930
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.6702222222222223,
|
| 673 |
+
"grad_norm": 27.319766998291016,
|
| 674 |
+
"learning_rate": 3.697113483590352e-05,
|
| 675 |
+
"loss": 0.0201,
|
| 676 |
+
"step": 940
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.688,
|
| 680 |
+
"grad_norm": 20.58792495727539,
|
| 681 |
+
"learning_rate": 3.677342823250297e-05,
|
| 682 |
+
"loss": 0.0678,
|
| 683 |
+
"step": 950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.7057777777777776,
|
| 687 |
+
"grad_norm": 0.33605310320854187,
|
| 688 |
+
"learning_rate": 3.6575721629102416e-05,
|
| 689 |
+
"loss": 0.0027,
|
| 690 |
+
"step": 960
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.7235555555555555,
|
| 694 |
+
"grad_norm": 0.13025720417499542,
|
| 695 |
+
"learning_rate": 3.6378015025701864e-05,
|
| 696 |
+
"loss": 0.0029,
|
| 697 |
+
"step": 970
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.7413333333333334,
|
| 701 |
+
"grad_norm": 31.07040023803711,
|
| 702 |
+
"learning_rate": 3.6180308422301305e-05,
|
| 703 |
+
"loss": 0.0626,
|
| 704 |
+
"step": 980
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.759111111111111,
|
| 708 |
+
"grad_norm": 0.055677346885204315,
|
| 709 |
+
"learning_rate": 3.598260181890075e-05,
|
| 710 |
+
"loss": 0.0079,
|
| 711 |
+
"step": 990
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.7768888888888887,
|
| 715 |
+
"grad_norm": 0.012918527238070965,
|
| 716 |
+
"learning_rate": 3.5784895215500194e-05,
|
| 717 |
+
"loss": 0.0308,
|
| 718 |
+
"step": 1000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.7946666666666666,
|
| 722 |
+
"grad_norm": 12.896405220031738,
|
| 723 |
+
"learning_rate": 3.558718861209965e-05,
|
| 724 |
+
"loss": 0.0551,
|
| 725 |
+
"step": 1010
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.8124444444444445,
|
| 729 |
+
"grad_norm": 0.0037423851899802685,
|
| 730 |
+
"learning_rate": 3.538948200869909e-05,
|
| 731 |
+
"loss": 0.0437,
|
| 732 |
+
"step": 1020
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.8302222222222222,
|
| 736 |
+
"grad_norm": 0.2329370528459549,
|
| 737 |
+
"learning_rate": 3.519177540529854e-05,
|
| 738 |
+
"loss": 0.0281,
|
| 739 |
+
"step": 1030
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.8479999999999999,
|
| 743 |
+
"grad_norm": 0.027233602479100227,
|
| 744 |
+
"learning_rate": 3.499406880189798e-05,
|
| 745 |
+
"loss": 0.0136,
|
| 746 |
+
"step": 1040
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.8657777777777778,
|
| 750 |
+
"grad_norm": 0.1924924999475479,
|
| 751 |
+
"learning_rate": 3.4796362198497435e-05,
|
| 752 |
+
"loss": 0.056,
|
| 753 |
+
"step": 1050
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.8835555555555556,
|
| 757 |
+
"grad_norm": 0.02651727944612503,
|
| 758 |
+
"learning_rate": 3.4598655595096876e-05,
|
| 759 |
+
"loss": 0.0647,
|
| 760 |
+
"step": 1060
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.9013333333333333,
|
| 764 |
+
"grad_norm": 38.13447570800781,
|
| 765 |
+
"learning_rate": 3.4400948991696325e-05,
|
| 766 |
+
"loss": 0.0216,
|
| 767 |
+
"step": 1070
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.919111111111111,
|
| 771 |
+
"grad_norm": 11.278813362121582,
|
| 772 |
+
"learning_rate": 3.420324238829577e-05,
|
| 773 |
+
"loss": 0.0395,
|
| 774 |
+
"step": 1080
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.9368888888888889,
|
| 778 |
+
"grad_norm": 0.5866456031799316,
|
| 779 |
+
"learning_rate": 3.4005535784895214e-05,
|
| 780 |
+
"loss": 0.0433,
|
| 781 |
+
"step": 1090
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.9546666666666668,
|
| 785 |
+
"grad_norm": 2.6308796405792236,
|
| 786 |
+
"learning_rate": 3.380782918149467e-05,
|
| 787 |
+
"loss": 0.0118,
|
| 788 |
+
"step": 1100
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.9724444444444444,
|
| 792 |
+
"grad_norm": 0.1729055494070053,
|
| 793 |
+
"learning_rate": 3.361012257809411e-05,
|
| 794 |
+
"loss": 0.0251,
|
| 795 |
+
"step": 1110
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.9902222222222221,
|
| 799 |
+
"grad_norm": 10.363531112670898,
|
| 800 |
+
"learning_rate": 3.341241597469356e-05,
|
| 801 |
+
"loss": 0.0656,
|
| 802 |
+
"step": 1120
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 2.0,
|
| 806 |
+
"eval_accuracy": 1.0,
|
| 807 |
+
"eval_loss": 2.5387274945387617e-05,
|
| 808 |
+
"eval_runtime": 75.7395,
|
| 809 |
+
"eval_samples_per_second": 105.625,
|
| 810 |
+
"eval_steps_per_second": 3.301,
|
| 811 |
+
"step": 1126
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 2.010666666666667,
|
| 815 |
+
"grad_norm": 29.318574905395508,
|
| 816 |
+
"learning_rate": 3.3214709371293e-05,
|
| 817 |
+
"loss": 0.2308,
|
| 818 |
+
"step": 1130
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.0284444444444443,
|
| 822 |
+
"grad_norm": 67.69547271728516,
|
| 823 |
+
"learning_rate": 3.301700276789245e-05,
|
| 824 |
+
"loss": 0.0604,
|
| 825 |
+
"step": 1140
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 2.046222222222222,
|
| 829 |
+
"grad_norm": 0.18736213445663452,
|
| 830 |
+
"learning_rate": 3.2819296164491896e-05,
|
| 831 |
+
"loss": 0.0546,
|
| 832 |
+
"step": 1150
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.064,
|
| 836 |
+
"grad_norm": 15.664798736572266,
|
| 837 |
+
"learning_rate": 3.2621589561091344e-05,
|
| 838 |
+
"loss": 0.0304,
|
| 839 |
+
"step": 1160
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 2.081777777777778,
|
| 843 |
+
"grad_norm": 0.13788799941539764,
|
| 844 |
+
"learning_rate": 3.2423882957690785e-05,
|
| 845 |
+
"loss": 0.0029,
|
| 846 |
+
"step": 1170
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.0995555555555554,
|
| 850 |
+
"grad_norm": 4.969343185424805,
|
| 851 |
+
"learning_rate": 3.2226176354290234e-05,
|
| 852 |
+
"loss": 0.0099,
|
| 853 |
+
"step": 1180
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 2.1173333333333333,
|
| 857 |
+
"grad_norm": 1.4300692081451416,
|
| 858 |
+
"learning_rate": 3.202846975088968e-05,
|
| 859 |
+
"loss": 0.0181,
|
| 860 |
+
"step": 1190
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.135111111111111,
|
| 864 |
+
"grad_norm": 0.019833851605653763,
|
| 865 |
+
"learning_rate": 3.183076314748913e-05,
|
| 866 |
+
"loss": 0.0107,
|
| 867 |
+
"step": 1200
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 2.152888888888889,
|
| 871 |
+
"grad_norm": 0.01474601961672306,
|
| 872 |
+
"learning_rate": 3.163305654408857e-05,
|
| 873 |
+
"loss": 0.0169,
|
| 874 |
+
"step": 1210
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 2.1706666666666665,
|
| 878 |
+
"grad_norm": 0.004971742630004883,
|
| 879 |
+
"learning_rate": 3.143534994068802e-05,
|
| 880 |
+
"loss": 0.0254,
|
| 881 |
+
"step": 1220
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 2.1884444444444444,
|
| 885 |
+
"grad_norm": 0.42502257227897644,
|
| 886 |
+
"learning_rate": 3.123764333728747e-05,
|
| 887 |
+
"loss": 0.0623,
|
| 888 |
+
"step": 1230
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 2.2062222222222223,
|
| 892 |
+
"grad_norm": 30.118955612182617,
|
| 893 |
+
"learning_rate": 3.1039936733886916e-05,
|
| 894 |
+
"loss": 0.0456,
|
| 895 |
+
"step": 1240
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 2.224,
|
| 899 |
+
"grad_norm": 7.1990132331848145,
|
| 900 |
+
"learning_rate": 3.0842230130486364e-05,
|
| 901 |
+
"loss": 0.0468,
|
| 902 |
+
"step": 1250
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.2417777777777776,
|
| 906 |
+
"grad_norm": 0.021625256165862083,
|
| 907 |
+
"learning_rate": 3.0644523527085805e-05,
|
| 908 |
+
"loss": 0.0375,
|
| 909 |
+
"step": 1260
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 2.2595555555555555,
|
| 913 |
+
"grad_norm": 0.08051316440105438,
|
| 914 |
+
"learning_rate": 3.044681692368525e-05,
|
| 915 |
+
"loss": 0.0097,
|
| 916 |
+
"step": 1270
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 2.2773333333333334,
|
| 920 |
+
"grad_norm": 38.26923751831055,
|
| 921 |
+
"learning_rate": 3.02491103202847e-05,
|
| 922 |
+
"loss": 0.0247,
|
| 923 |
+
"step": 1280
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.295111111111111,
|
| 927 |
+
"grad_norm": 1.5090163946151733,
|
| 928 |
+
"learning_rate": 3.0051403716884146e-05,
|
| 929 |
+
"loss": 0.0207,
|
| 930 |
+
"step": 1290
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 2.3128888888888888,
|
| 934 |
+
"grad_norm": 14.290279388427734,
|
| 935 |
+
"learning_rate": 2.985369711348359e-05,
|
| 936 |
+
"loss": 0.0041,
|
| 937 |
+
"step": 1300
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 2.3306666666666667,
|
| 941 |
+
"grad_norm": 0.025663571432232857,
|
| 942 |
+
"learning_rate": 2.9655990510083035e-05,
|
| 943 |
+
"loss": 0.0013,
|
| 944 |
+
"step": 1310
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.3484444444444446,
|
| 948 |
+
"grad_norm": 13.073821067810059,
|
| 949 |
+
"learning_rate": 2.9458283906682484e-05,
|
| 950 |
+
"loss": 0.006,
|
| 951 |
+
"step": 1320
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 2.3662222222222224,
|
| 955 |
+
"grad_norm": 0.03257250785827637,
|
| 956 |
+
"learning_rate": 2.9260577303281932e-05,
|
| 957 |
+
"loss": 0.0383,
|
| 958 |
+
"step": 1330
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 2.384,
|
| 962 |
+
"grad_norm": 2.0085232257843018,
|
| 963 |
+
"learning_rate": 2.906287069988138e-05,
|
| 964 |
+
"loss": 0.0552,
|
| 965 |
+
"step": 1340
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 2.401777777777778,
|
| 969 |
+
"grad_norm": 19.939329147338867,
|
| 970 |
+
"learning_rate": 2.8865164096480825e-05,
|
| 971 |
+
"loss": 0.0714,
|
| 972 |
+
"step": 1350
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 2.4195555555555557,
|
| 976 |
+
"grad_norm": 0.028250480070710182,
|
| 977 |
+
"learning_rate": 2.866745749308027e-05,
|
| 978 |
+
"loss": 0.0227,
|
| 979 |
+
"step": 1360
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 2.437333333333333,
|
| 983 |
+
"grad_norm": 0.6903110146522522,
|
| 984 |
+
"learning_rate": 2.846975088967972e-05,
|
| 985 |
+
"loss": 0.0085,
|
| 986 |
+
"step": 1370
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.455111111111111,
|
| 990 |
+
"grad_norm": 0.22341494262218475,
|
| 991 |
+
"learning_rate": 2.8272044286279166e-05,
|
| 992 |
+
"loss": 0.0093,
|
| 993 |
+
"step": 1380
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 2.472888888888889,
|
| 997 |
+
"grad_norm": 0.015194721519947052,
|
| 998 |
+
"learning_rate": 2.807433768287861e-05,
|
| 999 |
+
"loss": 0.0413,
|
| 1000 |
+
"step": 1390
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.490666666666667,
|
| 1004 |
+
"grad_norm": 0.035975806415081024,
|
| 1005 |
+
"learning_rate": 2.7876631079478055e-05,
|
| 1006 |
+
"loss": 0.0013,
|
| 1007 |
+
"step": 1400
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 2.5084444444444447,
|
| 1011 |
+
"grad_norm": 0.32504796981811523,
|
| 1012 |
+
"learning_rate": 2.76789244760775e-05,
|
| 1013 |
+
"loss": 0.0547,
|
| 1014 |
+
"step": 1410
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.526222222222222,
|
| 1018 |
+
"grad_norm": 0.34928593039512634,
|
| 1019 |
+
"learning_rate": 2.748121787267695e-05,
|
| 1020 |
+
"loss": 0.0556,
|
| 1021 |
+
"step": 1420
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 2.544,
|
| 1025 |
+
"grad_norm": 0.045603252947330475,
|
| 1026 |
+
"learning_rate": 2.7283511269276396e-05,
|
| 1027 |
+
"loss": 0.0156,
|
| 1028 |
+
"step": 1430
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.561777777777778,
|
| 1032 |
+
"grad_norm": 11.350424766540527,
|
| 1033 |
+
"learning_rate": 2.708580466587584e-05,
|
| 1034 |
+
"loss": 0.011,
|
| 1035 |
+
"step": 1440
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 2.5795555555555554,
|
| 1039 |
+
"grad_norm": 0.07788264751434326,
|
| 1040 |
+
"learning_rate": 2.6888098062475286e-05,
|
| 1041 |
+
"loss": 0.0607,
|
| 1042 |
+
"step": 1450
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 2.5973333333333333,
|
| 1046 |
+
"grad_norm": 0.06617221236228943,
|
| 1047 |
+
"learning_rate": 2.669039145907473e-05,
|
| 1048 |
+
"loss": 0.0102,
|
| 1049 |
+
"step": 1460
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 2.615111111111111,
|
| 1053 |
+
"grad_norm": 34.64754867553711,
|
| 1054 |
+
"learning_rate": 2.6492684855674182e-05,
|
| 1055 |
+
"loss": 0.0268,
|
| 1056 |
+
"step": 1470
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.632888888888889,
|
| 1060 |
+
"grad_norm": 9.72877311706543,
|
| 1061 |
+
"learning_rate": 2.6294978252273626e-05,
|
| 1062 |
+
"loss": 0.0051,
|
| 1063 |
+
"step": 1480
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.6506666666666665,
|
| 1067 |
+
"grad_norm": 21.619274139404297,
|
| 1068 |
+
"learning_rate": 2.609727164887307e-05,
|
| 1069 |
+
"loss": 0.0092,
|
| 1070 |
+
"step": 1490
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.6684444444444444,
|
| 1074 |
+
"grad_norm": 4.081634521484375,
|
| 1075 |
+
"learning_rate": 2.589956504547252e-05,
|
| 1076 |
+
"loss": 0.0369,
|
| 1077 |
+
"step": 1500
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.6862222222222223,
|
| 1081 |
+
"grad_norm": 0.009345272555947304,
|
| 1082 |
+
"learning_rate": 2.5701858442071967e-05,
|
| 1083 |
+
"loss": 0.018,
|
| 1084 |
+
"step": 1510
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.7039999999999997,
|
| 1088 |
+
"grad_norm": 73.03565979003906,
|
| 1089 |
+
"learning_rate": 2.5504151838671416e-05,
|
| 1090 |
+
"loss": 0.0945,
|
| 1091 |
+
"step": 1520
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.7217777777777776,
|
| 1095 |
+
"grad_norm": 1.0712828636169434,
|
| 1096 |
+
"learning_rate": 2.530644523527086e-05,
|
| 1097 |
+
"loss": 0.0156,
|
| 1098 |
+
"step": 1530
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.7395555555555555,
|
| 1102 |
+
"grad_norm": 0.023015221580863,
|
| 1103 |
+
"learning_rate": 2.5108738631870305e-05,
|
| 1104 |
+
"loss": 0.0408,
|
| 1105 |
+
"step": 1540
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.7573333333333334,
|
| 1109 |
+
"grad_norm": 3.0739543437957764,
|
| 1110 |
+
"learning_rate": 2.4911032028469753e-05,
|
| 1111 |
+
"loss": 0.0384,
|
| 1112 |
+
"step": 1550
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.7751111111111113,
|
| 1116 |
+
"grad_norm": 0.017695285379886627,
|
| 1117 |
+
"learning_rate": 2.4713325425069198e-05,
|
| 1118 |
+
"loss": 0.034,
|
| 1119 |
+
"step": 1560
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.7928888888888888,
|
| 1123 |
+
"grad_norm": 0.013055549003183842,
|
| 1124 |
+
"learning_rate": 2.4515618821668646e-05,
|
| 1125 |
+
"loss": 0.074,
|
| 1126 |
+
"step": 1570
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.8106666666666666,
|
| 1130 |
+
"grad_norm": 6.4839582443237305,
|
| 1131 |
+
"learning_rate": 2.431791221826809e-05,
|
| 1132 |
+
"loss": 0.0016,
|
| 1133 |
+
"step": 1580
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 2.8284444444444445,
|
| 1137 |
+
"grad_norm": 0.17747992277145386,
|
| 1138 |
+
"learning_rate": 2.412020561486754e-05,
|
| 1139 |
+
"loss": 0.0579,
|
| 1140 |
+
"step": 1590
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.846222222222222,
|
| 1144 |
+
"grad_norm": 0.07140109688043594,
|
| 1145 |
+
"learning_rate": 2.3922499011466984e-05,
|
| 1146 |
+
"loss": 0.0303,
|
| 1147 |
+
"step": 1600
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 2.864,
|
| 1151 |
+
"grad_norm": 0.0027039784472435713,
|
| 1152 |
+
"learning_rate": 2.3724792408066432e-05,
|
| 1153 |
+
"loss": 0.0337,
|
| 1154 |
+
"step": 1610
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.8817777777777778,
|
| 1158 |
+
"grad_norm": 0.015552740544080734,
|
| 1159 |
+
"learning_rate": 2.3527085804665877e-05,
|
| 1160 |
+
"loss": 0.0131,
|
| 1161 |
+
"step": 1620
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 2.8995555555555557,
|
| 1165 |
+
"grad_norm": 0.014052975922822952,
|
| 1166 |
+
"learning_rate": 2.332937920126532e-05,
|
| 1167 |
+
"loss": 0.0434,
|
| 1168 |
+
"step": 1630
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 2.9173333333333336,
|
| 1172 |
+
"grad_norm": 0.7165421843528748,
|
| 1173 |
+
"learning_rate": 2.313167259786477e-05,
|
| 1174 |
+
"loss": 0.0069,
|
| 1175 |
+
"step": 1640
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 2.935111111111111,
|
| 1179 |
+
"grad_norm": 0.020382430404424667,
|
| 1180 |
+
"learning_rate": 2.2933965994464214e-05,
|
| 1181 |
+
"loss": 0.0419,
|
| 1182 |
+
"step": 1650
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.952888888888889,
|
| 1186 |
+
"grad_norm": 0.07127852737903595,
|
| 1187 |
+
"learning_rate": 2.2736259391063662e-05,
|
| 1188 |
+
"loss": 0.0303,
|
| 1189 |
+
"step": 1660
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 2.970666666666667,
|
| 1193 |
+
"grad_norm": 2.5529978275299072,
|
| 1194 |
+
"learning_rate": 2.2538552787663107e-05,
|
| 1195 |
+
"loss": 0.0063,
|
| 1196 |
+
"step": 1670
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.9884444444444442,
|
| 1200 |
+
"grad_norm": 0.0018354392377659678,
|
| 1201 |
+
"learning_rate": 2.2340846184262555e-05,
|
| 1202 |
+
"loss": 0.0323,
|
| 1203 |
+
"step": 1680
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.999111111111111,
|
| 1207 |
+
"eval_accuracy": 1.0,
|
| 1208 |
+
"eval_loss": 6.3951174524845555e-06,
|
| 1209 |
+
"eval_runtime": 39.8564,
|
| 1210 |
+
"eval_samples_per_second": 200.721,
|
| 1211 |
+
"eval_steps_per_second": 6.273,
|
| 1212 |
+
"step": 1686
|
| 1213 |
+
},
|
| 1214 |
+
{
|
| 1215 |
+
"epoch": 3.007111111111111,
|
| 1216 |
+
"grad_norm": 0.0571288987994194,
|
| 1217 |
+
"learning_rate": 2.2143139580862003e-05,
|
| 1218 |
+
"loss": 0.0435,
|
| 1219 |
+
"step": 1690
|
| 1220 |
+
},
|
| 1221 |
+
{
|
| 1222 |
+
"epoch": 3.024888888888889,
|
| 1223 |
+
"grad_norm": 0.015790535137057304,
|
| 1224 |
+
"learning_rate": 2.1945432977461448e-05,
|
| 1225 |
+
"loss": 0.0058,
|
| 1226 |
+
"step": 1700
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 3.042666666666667,
|
| 1230 |
+
"grad_norm": 0.9516779184341431,
|
| 1231 |
+
"learning_rate": 2.1747726374060896e-05,
|
| 1232 |
+
"loss": 0.0109,
|
| 1233 |
+
"step": 1710
|
| 1234 |
+
},
|
| 1235 |
+
{
|
| 1236 |
+
"epoch": 3.0604444444444443,
|
| 1237 |
+
"grad_norm": 0.05595998093485832,
|
| 1238 |
+
"learning_rate": 2.155001977066034e-05,
|
| 1239 |
+
"loss": 0.01,
|
| 1240 |
+
"step": 1720
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 3.078222222222222,
|
| 1244 |
+
"grad_norm": 0.07043913751840591,
|
| 1245 |
+
"learning_rate": 2.135231316725979e-05,
|
| 1246 |
+
"loss": 0.0017,
|
| 1247 |
+
"step": 1730
|
| 1248 |
+
},
|
| 1249 |
+
{
|
| 1250 |
+
"epoch": 3.096,
|
| 1251 |
+
"grad_norm": 0.00207032123580575,
|
| 1252 |
+
"learning_rate": 2.1154606563859234e-05,
|
| 1253 |
+
"loss": 0.0026,
|
| 1254 |
+
"step": 1740
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 3.113777777777778,
|
| 1258 |
+
"grad_norm": 0.0049515170976519585,
|
| 1259 |
+
"learning_rate": 2.0956899960458682e-05,
|
| 1260 |
+
"loss": 0.0049,
|
| 1261 |
+
"step": 1750
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 3.1315555555555554,
|
| 1265 |
+
"grad_norm": 0.08053428679704666,
|
| 1266 |
+
"learning_rate": 2.0759193357058127e-05,
|
| 1267 |
+
"loss": 0.0169,
|
| 1268 |
+
"step": 1760
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 3.1493333333333333,
|
| 1272 |
+
"grad_norm": 0.0009357993258163333,
|
| 1273 |
+
"learning_rate": 2.0561486753657575e-05,
|
| 1274 |
+
"loss": 0.0124,
|
| 1275 |
+
"step": 1770
|
| 1276 |
+
},
|
| 1277 |
+
{
|
| 1278 |
+
"epoch": 3.167111111111111,
|
| 1279 |
+
"grad_norm": 0.014298039488494396,
|
| 1280 |
+
"learning_rate": 2.036378015025702e-05,
|
| 1281 |
+
"loss": 0.0007,
|
| 1282 |
+
"step": 1780
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 3.1848888888888887,
|
| 1286 |
+
"grad_norm": 1.3868434429168701,
|
| 1287 |
+
"learning_rate": 2.0166073546856464e-05,
|
| 1288 |
+
"loss": 0.0122,
|
| 1289 |
+
"step": 1790
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 3.2026666666666666,
|
| 1293 |
+
"grad_norm": 0.0033609354868531227,
|
| 1294 |
+
"learning_rate": 1.9968366943455912e-05,
|
| 1295 |
+
"loss": 0.0097,
|
| 1296 |
+
"step": 1800
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 3.2204444444444444,
|
| 1300 |
+
"grad_norm": 10.693577766418457,
|
| 1301 |
+
"learning_rate": 1.9770660340055357e-05,
|
| 1302 |
+
"loss": 0.0637,
|
| 1303 |
+
"step": 1810
|
| 1304 |
+
},
|
| 1305 |
+
{
|
| 1306 |
+
"epoch": 3.2382222222222223,
|
| 1307 |
+
"grad_norm": 0.006601002067327499,
|
| 1308 |
+
"learning_rate": 1.9572953736654805e-05,
|
| 1309 |
+
"loss": 0.0718,
|
| 1310 |
+
"step": 1820
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 3.2560000000000002,
|
| 1314 |
+
"grad_norm": 0.03210904076695442,
|
| 1315 |
+
"learning_rate": 1.937524713325425e-05,
|
| 1316 |
+
"loss": 0.0014,
|
| 1317 |
+
"step": 1830
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 3.2737777777777777,
|
| 1321 |
+
"grad_norm": 0.007109949365258217,
|
| 1322 |
+
"learning_rate": 1.9177540529853698e-05,
|
| 1323 |
+
"loss": 0.0065,
|
| 1324 |
+
"step": 1840
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 3.2915555555555556,
|
| 1328 |
+
"grad_norm": 0.0019361014710739255,
|
| 1329 |
+
"learning_rate": 1.8979833926453143e-05,
|
| 1330 |
+
"loss": 0.0008,
|
| 1331 |
+
"step": 1850
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 3.3093333333333335,
|
| 1335 |
+
"grad_norm": 0.08277800679206848,
|
| 1336 |
+
"learning_rate": 1.878212732305259e-05,
|
| 1337 |
+
"loss": 0.0004,
|
| 1338 |
+
"step": 1860
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 3.327111111111111,
|
| 1342 |
+
"grad_norm": 0.003757915459573269,
|
| 1343 |
+
"learning_rate": 1.858442071965204e-05,
|
| 1344 |
+
"loss": 0.0074,
|
| 1345 |
+
"step": 1870
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 3.344888888888889,
|
| 1349 |
+
"grad_norm": 13.0957670211792,
|
| 1350 |
+
"learning_rate": 1.8386714116251484e-05,
|
| 1351 |
+
"loss": 0.0177,
|
| 1352 |
+
"step": 1880
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 3.3626666666666667,
|
| 1356 |
+
"grad_norm": 0.07693179696798325,
|
| 1357 |
+
"learning_rate": 1.8189007512850932e-05,
|
| 1358 |
+
"loss": 0.0093,
|
| 1359 |
+
"step": 1890
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 3.3804444444444446,
|
| 1363 |
+
"grad_norm": 0.007269900757819414,
|
| 1364 |
+
"learning_rate": 1.7991300909450377e-05,
|
| 1365 |
+
"loss": 0.0004,
|
| 1366 |
+
"step": 1900
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 3.398222222222222,
|
| 1370 |
+
"grad_norm": 0.149241641163826,
|
| 1371 |
+
"learning_rate": 1.7793594306049825e-05,
|
| 1372 |
+
"loss": 0.0008,
|
| 1373 |
+
"step": 1910
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 3.416,
|
| 1377 |
+
"grad_norm": 0.13814906775951385,
|
| 1378 |
+
"learning_rate": 1.759588770264927e-05,
|
| 1379 |
+
"loss": 0.0228,
|
| 1380 |
+
"step": 1920
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 3.433777777777778,
|
| 1384 |
+
"grad_norm": 0.012578528374433517,
|
| 1385 |
+
"learning_rate": 1.7398181099248718e-05,
|
| 1386 |
+
"loss": 0.0031,
|
| 1387 |
+
"step": 1930
|
| 1388 |
+
},
|
| 1389 |
+
{
|
| 1390 |
+
"epoch": 3.4515555555555557,
|
| 1391 |
+
"grad_norm": 0.002241666428744793,
|
| 1392 |
+
"learning_rate": 1.7200474495848162e-05,
|
| 1393 |
+
"loss": 0.0018,
|
| 1394 |
+
"step": 1940
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 3.469333333333333,
|
| 1398 |
+
"grad_norm": 0.005206266883760691,
|
| 1399 |
+
"learning_rate": 1.7002767892447607e-05,
|
| 1400 |
+
"loss": 0.0469,
|
| 1401 |
+
"step": 1950
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 3.487111111111111,
|
| 1405 |
+
"grad_norm": 11.583967208862305,
|
| 1406 |
+
"learning_rate": 1.6805061289047055e-05,
|
| 1407 |
+
"loss": 0.0691,
|
| 1408 |
+
"step": 1960
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 3.504888888888889,
|
| 1412 |
+
"grad_norm": 0.006195690017193556,
|
| 1413 |
+
"learning_rate": 1.66073546856465e-05,
|
| 1414 |
+
"loss": 0.0199,
|
| 1415 |
+
"step": 1970
|
| 1416 |
+
},
|
| 1417 |
+
{
|
| 1418 |
+
"epoch": 3.522666666666667,
|
| 1419 |
+
"grad_norm": 8.7911958694458,
|
| 1420 |
+
"learning_rate": 1.6409648082245948e-05,
|
| 1421 |
+
"loss": 0.0023,
|
| 1422 |
+
"step": 1980
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 3.5404444444444443,
|
| 1426 |
+
"grad_norm": 0.002020699204877019,
|
| 1427 |
+
"learning_rate": 1.6211941478845393e-05,
|
| 1428 |
+
"loss": 0.0022,
|
| 1429 |
+
"step": 1990
|
| 1430 |
+
},
|
| 1431 |
+
{
|
| 1432 |
+
"epoch": 3.558222222222222,
|
| 1433 |
+
"grad_norm": 0.006164974998682737,
|
| 1434 |
+
"learning_rate": 1.601423487544484e-05,
|
| 1435 |
+
"loss": 0.0004,
|
| 1436 |
+
"step": 2000
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 3.576,
|
| 1440 |
+
"grad_norm": 0.004664299543946981,
|
| 1441 |
+
"learning_rate": 1.5816528272044286e-05,
|
| 1442 |
+
"loss": 0.0013,
|
| 1443 |
+
"step": 2010
|
| 1444 |
+
},
|
| 1445 |
+
{
|
| 1446 |
+
"epoch": 3.5937777777777775,
|
| 1447 |
+
"grad_norm": 0.47931966185569763,
|
| 1448 |
+
"learning_rate": 1.5618821668643734e-05,
|
| 1449 |
+
"loss": 0.0003,
|
| 1450 |
+
"step": 2020
|
| 1451 |
+
},
|
| 1452 |
+
{
|
| 1453 |
+
"epoch": 3.6115555555555554,
|
| 1454 |
+
"grad_norm": 43.603816986083984,
|
| 1455 |
+
"learning_rate": 1.5421115065243182e-05,
|
| 1456 |
+
"loss": 0.0085,
|
| 1457 |
+
"step": 2030
|
| 1458 |
+
},
|
| 1459 |
+
{
|
| 1460 |
+
"epoch": 3.6293333333333333,
|
| 1461 |
+
"grad_norm": 0.0044282907620072365,
|
| 1462 |
+
"learning_rate": 1.5223408461842625e-05,
|
| 1463 |
+
"loss": 0.0589,
|
| 1464 |
+
"step": 2040
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 3.647111111111111,
|
| 1468 |
+
"grad_norm": 0.001094396342523396,
|
| 1469 |
+
"learning_rate": 1.5025701858442073e-05,
|
| 1470 |
+
"loss": 0.0006,
|
| 1471 |
+
"step": 2050
|
| 1472 |
+
},
|
| 1473 |
+
{
|
| 1474 |
+
"epoch": 3.664888888888889,
|
| 1475 |
+
"grad_norm": 0.008125217631459236,
|
| 1476 |
+
"learning_rate": 1.4827995255041518e-05,
|
| 1477 |
+
"loss": 0.0002,
|
| 1478 |
+
"step": 2060
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 3.6826666666666665,
|
| 1482 |
+
"grad_norm": 19.899005889892578,
|
| 1483 |
+
"learning_rate": 1.4630288651640966e-05,
|
| 1484 |
+
"loss": 0.0241,
|
| 1485 |
+
"step": 2070
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 3.7004444444444444,
|
| 1489 |
+
"grad_norm": 0.19454504549503326,
|
| 1490 |
+
"learning_rate": 1.4432582048240412e-05,
|
| 1491 |
+
"loss": 0.0005,
|
| 1492 |
+
"step": 2080
|
| 1493 |
+
},
|
| 1494 |
+
{
|
| 1495 |
+
"epoch": 3.7182222222222223,
|
| 1496 |
+
"grad_norm": 0.003989567514508963,
|
| 1497 |
+
"learning_rate": 1.423487544483986e-05,
|
| 1498 |
+
"loss": 0.0045,
|
| 1499 |
+
"step": 2090
|
| 1500 |
+
},
|
| 1501 |
+
{
|
| 1502 |
+
"epoch": 3.7359999999999998,
|
| 1503 |
+
"grad_norm": 0.0012507745996117592,
|
| 1504 |
+
"learning_rate": 1.4037168841439305e-05,
|
| 1505 |
+
"loss": 0.0011,
|
| 1506 |
+
"step": 2100
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 3.7537777777777777,
|
| 1510 |
+
"grad_norm": 0.033174123615026474,
|
| 1511 |
+
"learning_rate": 1.383946223803875e-05,
|
| 1512 |
+
"loss": 0.0005,
|
| 1513 |
+
"step": 2110
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 3.7715555555555556,
|
| 1517 |
+
"grad_norm": 0.04268620163202286,
|
| 1518 |
+
"learning_rate": 1.3641755634638198e-05,
|
| 1519 |
+
"loss": 0.0111,
|
| 1520 |
+
"step": 2120
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 3.7893333333333334,
|
| 1524 |
+
"grad_norm": 0.041019320487976074,
|
| 1525 |
+
"learning_rate": 1.3444049031237643e-05,
|
| 1526 |
+
"loss": 0.0018,
|
| 1527 |
+
"step": 2130
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 3.8071111111111113,
|
| 1531 |
+
"grad_norm": 0.1453588902950287,
|
| 1532 |
+
"learning_rate": 1.3246342427837091e-05,
|
| 1533 |
+
"loss": 0.0276,
|
| 1534 |
+
"step": 2140
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 3.824888888888889,
|
| 1538 |
+
"grad_norm": 0.0027844184078276157,
|
| 1539 |
+
"learning_rate": 1.3048635824436536e-05,
|
| 1540 |
+
"loss": 0.0018,
|
| 1541 |
+
"step": 2150
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 3.8426666666666667,
|
| 1545 |
+
"grad_norm": 0.0006637629121541977,
|
| 1546 |
+
"learning_rate": 1.2850929221035984e-05,
|
| 1547 |
+
"loss": 0.015,
|
| 1548 |
+
"step": 2160
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 3.8604444444444446,
|
| 1552 |
+
"grad_norm": 0.00038110537570901215,
|
| 1553 |
+
"learning_rate": 1.265322261763543e-05,
|
| 1554 |
+
"loss": 0.0017,
|
| 1555 |
+
"step": 2170
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 3.878222222222222,
|
| 1559 |
+
"grad_norm": 0.0005764598026871681,
|
| 1560 |
+
"learning_rate": 1.2455516014234877e-05,
|
| 1561 |
+
"loss": 0.0443,
|
| 1562 |
+
"step": 2180
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 3.896,
|
| 1566 |
+
"grad_norm": 6.875983238220215,
|
| 1567 |
+
"learning_rate": 1.2257809410834323e-05,
|
| 1568 |
+
"loss": 0.0357,
|
| 1569 |
+
"step": 2190
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 3.913777777777778,
|
| 1573 |
+
"grad_norm": 51.81392288208008,
|
| 1574 |
+
"learning_rate": 1.206010280743377e-05,
|
| 1575 |
+
"loss": 0.074,
|
| 1576 |
+
"step": 2200
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 3.9315555555555557,
|
| 1580 |
+
"grad_norm": 0.0016795358387753367,
|
| 1581 |
+
"learning_rate": 1.1862396204033216e-05,
|
| 1582 |
+
"loss": 0.0046,
|
| 1583 |
+
"step": 2210
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 3.9493333333333336,
|
| 1587 |
+
"grad_norm": 1.101694107055664,
|
| 1588 |
+
"learning_rate": 1.166468960063266e-05,
|
| 1589 |
+
"loss": 0.0004,
|
| 1590 |
+
"step": 2220
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 3.967111111111111,
|
| 1594 |
+
"grad_norm": 0.03407168760895729,
|
| 1595 |
+
"learning_rate": 1.1466982997232107e-05,
|
| 1596 |
+
"loss": 0.0004,
|
| 1597 |
+
"step": 2230
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 3.984888888888889,
|
| 1601 |
+
"grad_norm": 0.012767530046403408,
|
| 1602 |
+
"learning_rate": 1.1269276393831553e-05,
|
| 1603 |
+
"loss": 0.0426,
|
| 1604 |
+
"step": 2240
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 4.0,
|
| 1608 |
+
"eval_accuracy": 1.0,
|
| 1609 |
+
"eval_loss": 7.458427717210725e-06,
|
| 1610 |
+
"eval_runtime": 41.6071,
|
| 1611 |
+
"eval_samples_per_second": 192.275,
|
| 1612 |
+
"eval_steps_per_second": 6.009,
|
| 1613 |
+
"step": 2249
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 4.001777777777778,
|
| 1617 |
+
"grad_norm": 0.0036902178544551134,
|
| 1618 |
+
"learning_rate": 1.1071569790431002e-05,
|
| 1619 |
+
"loss": 0.0003,
|
| 1620 |
+
"step": 2250
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 4.019555555555556,
|
| 1624 |
+
"grad_norm": 0.00998405460268259,
|
| 1625 |
+
"learning_rate": 1.0873863187030448e-05,
|
| 1626 |
+
"loss": 0.002,
|
| 1627 |
+
"step": 2260
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 4.037333333333334,
|
| 1631 |
+
"grad_norm": 23.953229904174805,
|
| 1632 |
+
"learning_rate": 1.0676156583629894e-05,
|
| 1633 |
+
"loss": 0.0187,
|
| 1634 |
+
"step": 2270
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 4.0551111111111116,
|
| 1638 |
+
"grad_norm": 0.008150537498295307,
|
| 1639 |
+
"learning_rate": 1.0478449980229341e-05,
|
| 1640 |
+
"loss": 0.0153,
|
| 1641 |
+
"step": 2280
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 4.072888888888889,
|
| 1645 |
+
"grad_norm": 0.5894471406936646,
|
| 1646 |
+
"learning_rate": 1.0280743376828787e-05,
|
| 1647 |
+
"loss": 0.0002,
|
| 1648 |
+
"step": 2290
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 4.0906666666666665,
|
| 1652 |
+
"grad_norm": 0.07007890194654465,
|
| 1653 |
+
"learning_rate": 1.0083036773428232e-05,
|
| 1654 |
+
"loss": 0.0367,
|
| 1655 |
+
"step": 2300
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 4.108444444444444,
|
| 1659 |
+
"grad_norm": 0.07020383328199387,
|
| 1660 |
+
"learning_rate": 9.885330170027678e-06,
|
| 1661 |
+
"loss": 0.0017,
|
| 1662 |
+
"step": 2310
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 4.126222222222222,
|
| 1666 |
+
"grad_norm": 0.0013433824060484767,
|
| 1667 |
+
"learning_rate": 9.687623566627125e-06,
|
| 1668 |
+
"loss": 0.0176,
|
| 1669 |
+
"step": 2320
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 4.144,
|
| 1673 |
+
"grad_norm": 0.0036678831093013287,
|
| 1674 |
+
"learning_rate": 9.489916963226571e-06,
|
| 1675 |
+
"loss": 0.001,
|
| 1676 |
+
"step": 2330
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 4.161777777777778,
|
| 1680 |
+
"grad_norm": 20.646207809448242,
|
| 1681 |
+
"learning_rate": 9.29221035982602e-06,
|
| 1682 |
+
"loss": 0.0071,
|
| 1683 |
+
"step": 2340
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 4.179555555555556,
|
| 1687 |
+
"grad_norm": 0.004499041475355625,
|
| 1688 |
+
"learning_rate": 9.094503756425466e-06,
|
| 1689 |
+
"loss": 0.0004,
|
| 1690 |
+
"step": 2350
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 4.197333333333333,
|
| 1694 |
+
"grad_norm": 0.0007168107549659908,
|
| 1695 |
+
"learning_rate": 8.896797153024912e-06,
|
| 1696 |
+
"loss": 0.0025,
|
| 1697 |
+
"step": 2360
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 4.215111111111111,
|
| 1701 |
+
"grad_norm": 0.015021364204585552,
|
| 1702 |
+
"learning_rate": 8.699090549624359e-06,
|
| 1703 |
+
"loss": 0.0277,
|
| 1704 |
+
"step": 2370
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 4.232888888888889,
|
| 1708 |
+
"grad_norm": 0.006119410507380962,
|
| 1709 |
+
"learning_rate": 8.501383946223804e-06,
|
| 1710 |
+
"loss": 0.0009,
|
| 1711 |
+
"step": 2380
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 4.250666666666667,
|
| 1715 |
+
"grad_norm": 0.0018322835676372051,
|
| 1716 |
+
"learning_rate": 8.30367734282325e-06,
|
| 1717 |
+
"loss": 0.0009,
|
| 1718 |
+
"step": 2390
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 4.2684444444444445,
|
| 1722 |
+
"grad_norm": 0.0025883447378873825,
|
| 1723 |
+
"learning_rate": 8.105970739422696e-06,
|
| 1724 |
+
"loss": 0.0179,
|
| 1725 |
+
"step": 2400
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 4.286222222222222,
|
| 1729 |
+
"grad_norm": 0.010295086540281773,
|
| 1730 |
+
"learning_rate": 7.908264136022143e-06,
|
| 1731 |
+
"loss": 0.0002,
|
| 1732 |
+
"step": 2410
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 4.304,
|
| 1736 |
+
"grad_norm": 0.27159199118614197,
|
| 1737 |
+
"learning_rate": 7.710557532621591e-06,
|
| 1738 |
+
"loss": 0.0003,
|
| 1739 |
+
"step": 2420
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 4.321777777777778,
|
| 1743 |
+
"grad_norm": 0.014537914656102657,
|
| 1744 |
+
"learning_rate": 7.5128509292210365e-06,
|
| 1745 |
+
"loss": 0.0108,
|
| 1746 |
+
"step": 2430
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 4.339555555555555,
|
| 1750 |
+
"grad_norm": 0.001482433988712728,
|
| 1751 |
+
"learning_rate": 7.315144325820483e-06,
|
| 1752 |
+
"loss": 0.0003,
|
| 1753 |
+
"step": 2440
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 4.357333333333333,
|
| 1757 |
+
"grad_norm": 0.0015277402708306909,
|
| 1758 |
+
"learning_rate": 7.11743772241993e-06,
|
| 1759 |
+
"loss": 0.0496,
|
| 1760 |
+
"step": 2450
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 4.375111111111111,
|
| 1764 |
+
"grad_norm": 0.005141290370374918,
|
| 1765 |
+
"learning_rate": 6.919731119019375e-06,
|
| 1766 |
+
"loss": 0.0001,
|
| 1767 |
+
"step": 2460
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 4.392888888888889,
|
| 1771 |
+
"grad_norm": 27.9423770904541,
|
| 1772 |
+
"learning_rate": 6.722024515618821e-06,
|
| 1773 |
+
"loss": 0.0393,
|
| 1774 |
+
"step": 2470
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 4.410666666666667,
|
| 1778 |
+
"grad_norm": 0.010187560692429543,
|
| 1779 |
+
"learning_rate": 6.524317912218268e-06,
|
| 1780 |
+
"loss": 0.0001,
|
| 1781 |
+
"step": 2480
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 4.428444444444445,
|
| 1785 |
+
"grad_norm": 0.002554529346525669,
|
| 1786 |
+
"learning_rate": 6.326611308817715e-06,
|
| 1787 |
+
"loss": 0.0031,
|
| 1788 |
+
"step": 2490
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 4.4462222222222225,
|
| 1792 |
+
"grad_norm": 4.20240592956543,
|
| 1793 |
+
"learning_rate": 6.1289047054171615e-06,
|
| 1794 |
+
"loss": 0.0016,
|
| 1795 |
+
"step": 2500
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 4.464,
|
| 1799 |
+
"grad_norm": 0.013741197995841503,
|
| 1800 |
+
"learning_rate": 5.931198102016608e-06,
|
| 1801 |
+
"loss": 0.0003,
|
| 1802 |
+
"step": 2510
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 4.481777777777777,
|
| 1806 |
+
"grad_norm": 0.043951794505119324,
|
| 1807 |
+
"learning_rate": 5.7334914986160535e-06,
|
| 1808 |
+
"loss": 0.0125,
|
| 1809 |
+
"step": 2520
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 4.499555555555555,
|
| 1813 |
+
"grad_norm": 0.11376281827688217,
|
| 1814 |
+
"learning_rate": 5.535784895215501e-06,
|
| 1815 |
+
"loss": 0.003,
|
| 1816 |
+
"step": 2530
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 4.517333333333333,
|
| 1820 |
+
"grad_norm": 0.05734412372112274,
|
| 1821 |
+
"learning_rate": 5.338078291814947e-06,
|
| 1822 |
+
"loss": 0.0009,
|
| 1823 |
+
"step": 2540
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 4.535111111111111,
|
| 1827 |
+
"grad_norm": 0.0010632964549586177,
|
| 1828 |
+
"learning_rate": 5.140371688414394e-06,
|
| 1829 |
+
"loss": 0.0196,
|
| 1830 |
+
"step": 2550
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 4.552888888888889,
|
| 1834 |
+
"grad_norm": 0.0036729658022522926,
|
| 1835 |
+
"learning_rate": 4.942665085013839e-06,
|
| 1836 |
+
"loss": 0.0071,
|
| 1837 |
+
"step": 2560
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 4.570666666666667,
|
| 1841 |
+
"grad_norm": 5.985267162322998,
|
| 1842 |
+
"learning_rate": 4.744958481613286e-06,
|
| 1843 |
+
"loss": 0.0311,
|
| 1844 |
+
"step": 2570
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 4.588444444444445,
|
| 1848 |
+
"grad_norm": 0.996809720993042,
|
| 1849 |
+
"learning_rate": 4.547251878212733e-06,
|
| 1850 |
+
"loss": 0.0012,
|
| 1851 |
+
"step": 2580
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 4.606222222222222,
|
| 1855 |
+
"grad_norm": 0.11869648844003677,
|
| 1856 |
+
"learning_rate": 4.349545274812179e-06,
|
| 1857 |
+
"loss": 0.0003,
|
| 1858 |
+
"step": 2590
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 4.624,
|
| 1862 |
+
"grad_norm": 0.006143218372017145,
|
| 1863 |
+
"learning_rate": 4.151838671411625e-06,
|
| 1864 |
+
"loss": 0.0013,
|
| 1865 |
+
"step": 2600
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 4.641777777777778,
|
| 1869 |
+
"grad_norm": 0.024631284177303314,
|
| 1870 |
+
"learning_rate": 3.954132068011071e-06,
|
| 1871 |
+
"loss": 0.0388,
|
| 1872 |
+
"step": 2610
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 4.6595555555555555,
|
| 1876 |
+
"grad_norm": 0.0017836794722825289,
|
| 1877 |
+
"learning_rate": 3.7564254646105183e-06,
|
| 1878 |
+
"loss": 0.0015,
|
| 1879 |
+
"step": 2620
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 4.677333333333333,
|
| 1883 |
+
"grad_norm": 0.003801500890403986,
|
| 1884 |
+
"learning_rate": 3.558718861209965e-06,
|
| 1885 |
+
"loss": 0.0003,
|
| 1886 |
+
"step": 2630
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 4.695111111111111,
|
| 1890 |
+
"grad_norm": 0.004178278613835573,
|
| 1891 |
+
"learning_rate": 3.3610122578094107e-06,
|
| 1892 |
+
"loss": 0.001,
|
| 1893 |
+
"step": 2640
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 4.712888888888889,
|
| 1897 |
+
"grad_norm": 0.0044832993298769,
|
| 1898 |
+
"learning_rate": 3.1633056544088575e-06,
|
| 1899 |
+
"loss": 0.0194,
|
| 1900 |
+
"step": 2650
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 4.730666666666667,
|
| 1904 |
+
"grad_norm": 0.0029173328075557947,
|
| 1905 |
+
"learning_rate": 2.965599051008304e-06,
|
| 1906 |
+
"loss": 0.0302,
|
| 1907 |
+
"step": 2660
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 4.748444444444445,
|
| 1911 |
+
"grad_norm": 0.0005038917297497392,
|
| 1912 |
+
"learning_rate": 2.7678924476077504e-06,
|
| 1913 |
+
"loss": 0.011,
|
| 1914 |
+
"step": 2670
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 4.766222222222222,
|
| 1918 |
+
"grad_norm": 0.01968969963490963,
|
| 1919 |
+
"learning_rate": 2.570185844207197e-06,
|
| 1920 |
+
"loss": 0.0002,
|
| 1921 |
+
"step": 2680
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 4.784,
|
| 1925 |
+
"grad_norm": 0.02507755346596241,
|
| 1926 |
+
"learning_rate": 2.372479240806643e-06,
|
| 1927 |
+
"loss": 0.0012,
|
| 1928 |
+
"step": 2690
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 4.801777777777778,
|
| 1932 |
+
"grad_norm": 6.288967609405518,
|
| 1933 |
+
"learning_rate": 2.1747726374060897e-06,
|
| 1934 |
+
"loss": 0.0195,
|
| 1935 |
+
"step": 2700
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 4.819555555555556,
|
| 1939 |
+
"grad_norm": 0.19547367095947266,
|
| 1940 |
+
"learning_rate": 1.9770660340055357e-06,
|
| 1941 |
+
"loss": 0.0034,
|
| 1942 |
+
"step": 2710
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 4.8373333333333335,
|
| 1946 |
+
"grad_norm": 0.007160472217947245,
|
| 1947 |
+
"learning_rate": 1.7793594306049826e-06,
|
| 1948 |
+
"loss": 0.0008,
|
| 1949 |
+
"step": 2720
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 4.855111111111111,
|
| 1953 |
+
"grad_norm": 0.1027381643652916,
|
| 1954 |
+
"learning_rate": 1.5816528272044288e-06,
|
| 1955 |
+
"loss": 0.0006,
|
| 1956 |
+
"step": 2730
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 4.872888888888889,
|
| 1960 |
+
"grad_norm": 0.6163949966430664,
|
| 1961 |
+
"learning_rate": 1.3839462238038752e-06,
|
| 1962 |
+
"loss": 0.0001,
|
| 1963 |
+
"step": 2740
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 4.890666666666666,
|
| 1967 |
+
"grad_norm": 0.0065285759046673775,
|
| 1968 |
+
"learning_rate": 1.1862396204033214e-06,
|
| 1969 |
+
"loss": 0.0,
|
| 1970 |
+
"step": 2750
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 4.908444444444444,
|
| 1974 |
+
"grad_norm": 2.0664429664611816,
|
| 1975 |
+
"learning_rate": 9.885330170027678e-07,
|
| 1976 |
+
"loss": 0.0033,
|
| 1977 |
+
"step": 2760
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 4.926222222222222,
|
| 1981 |
+
"grad_norm": 0.04097575694322586,
|
| 1982 |
+
"learning_rate": 7.908264136022144e-07,
|
| 1983 |
+
"loss": 0.001,
|
| 1984 |
+
"step": 2770
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 4.944,
|
| 1988 |
+
"grad_norm": 0.0015862607397139072,
|
| 1989 |
+
"learning_rate": 5.931198102016607e-07,
|
| 1990 |
+
"loss": 0.0001,
|
| 1991 |
+
"step": 2780
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 4.961777777777778,
|
| 1995 |
+
"grad_norm": 0.0021847274620085955,
|
| 1996 |
+
"learning_rate": 3.954132068011072e-07,
|
| 1997 |
+
"loss": 0.0006,
|
| 1998 |
+
"step": 2790
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 4.979555555555556,
|
| 2002 |
+
"grad_norm": 0.00527564063668251,
|
| 2003 |
+
"learning_rate": 1.977066034005536e-07,
|
| 2004 |
+
"loss": 0.0243,
|
| 2005 |
+
"step": 2800
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 4.997333333333334,
|
| 2009 |
+
"grad_norm": 0.0007434898870997131,
|
| 2010 |
+
"learning_rate": 0.0,
|
| 2011 |
+
"loss": 0.0525,
|
| 2012 |
+
"step": 2810
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 4.997333333333334,
|
| 2016 |
+
"eval_accuracy": 1.0,
|
| 2017 |
+
"eval_loss": 2.338912281629746e-06,
|
| 2018 |
+
"eval_runtime": 40.6863,
|
| 2019 |
+
"eval_samples_per_second": 196.626,
|
| 2020 |
+
"eval_steps_per_second": 6.145,
|
| 2021 |
+
"step": 2810
|
| 2022 |
+
},
|
| 2023 |
+
{
|
| 2024 |
+
"epoch": 4.997333333333334,
|
| 2025 |
+
"step": 2810,
|
| 2026 |
+
"total_flos": 8.94051665811918e+18,
|
| 2027 |
+
"train_loss": 0.011076973860487225,
|
| 2028 |
+
"train_runtime": 5179.8163,
|
| 2029 |
+
"train_samples_per_second": 69.501,
|
| 2030 |
+
"train_steps_per_second": 0.542
|
| 2031 |
+
}
|
| 2032 |
+
],
|
| 2033 |
+
"logging_steps": 10,
|
| 2034 |
+
"max_steps": 2810,
|
| 2035 |
+
"num_input_tokens_seen": 0,
|
| 2036 |
+
"num_train_epochs": 5,
|
| 2037 |
+
"save_steps": 500,
|
| 2038 |
+
"stateful_callbacks": {
|
| 2039 |
+
"TrainerControl": {
|
| 2040 |
+
"args": {
|
| 2041 |
+
"should_epoch_stop": false,
|
| 2042 |
+
"should_evaluate": false,
|
| 2043 |
+
"should_log": false,
|
| 2044 |
+
"should_save": true,
|
| 2045 |
+
"should_training_stop": true
|
| 2046 |
+
},
|
| 2047 |
+
"attributes": {}
|
| 2048 |
+
}
|
| 2049 |
+
},
|
| 2050 |
+
"total_flos": 8.94051665811918e+18,
|
| 2051 |
+
"train_batch_size": 32,
|
| 2052 |
+
"trial_name": null,
|
| 2053 |
+
"trial_params": null
|
| 2054 |
+
}
|