End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.9882,
|
| 4 |
+
"eval_loss": 0.03509841486811638,
|
| 5 |
+
"eval_runtime": 66.4143,
|
| 6 |
+
"eval_samples_per_second": 75.285,
|
| 7 |
+
"eval_steps_per_second": 4.713,
|
| 8 |
+
"total_flos": 1.045459878817923e+19,
|
| 9 |
+
"train_loss": 0.37951253058504525,
|
| 10 |
+
"train_runtime": 4954.8349,
|
| 11 |
+
"train_samples_per_second": 27.246,
|
| 12 |
+
"train_steps_per_second": 0.426
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_accuracy": 0.9882,
|
| 4 |
+
"eval_loss": 0.03509841486811638,
|
| 5 |
+
"eval_runtime": 66.4143,
|
| 6 |
+
"eval_samples_per_second": 75.285,
|
| 7 |
+
"eval_steps_per_second": 4.713
|
| 8 |
+
}
|
runs/Mar22_14-42-17_5aac9b83cbd0/events.out.tfevents.1711123650.5aac9b83cbd0.15185.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2b8bd3cc06060ec79d95414b12e5d5c3f67ce7b71d8aeba0cc5fbacbed3960e3
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.045459878817923e+19,
|
| 4 |
+
"train_loss": 0.37951253058504525,
|
| 5 |
+
"train_runtime": 4954.8349,
|
| 6 |
+
"train_samples_per_second": 27.246,
|
| 7 |
+
"train_steps_per_second": 0.426
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1527 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9882,
|
| 3 |
+
"best_model_checkpoint": "beit-base-patch16-224-finetuned-eurosat/checkpoint-2109",
|
| 4 |
+
"epoch": 2.998933522929257,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 2109,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.01,
|
| 13 |
+
"grad_norm": 42.045841217041016,
|
| 14 |
+
"learning_rate": 2.3696682464454976e-06,
|
| 15 |
+
"loss": 2.6733,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03,
|
| 20 |
+
"grad_norm": 19.79863166809082,
|
| 21 |
+
"learning_rate": 4.739336492890995e-06,
|
| 22 |
+
"loss": 2.5329,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04,
|
| 27 |
+
"grad_norm": 21.456212997436523,
|
| 28 |
+
"learning_rate": 7.109004739336493e-06,
|
| 29 |
+
"loss": 2.3177,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06,
|
| 34 |
+
"grad_norm": 24.893077850341797,
|
| 35 |
+
"learning_rate": 9.47867298578199e-06,
|
| 36 |
+
"loss": 2.0632,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.07,
|
| 41 |
+
"grad_norm": 25.683191299438477,
|
| 42 |
+
"learning_rate": 1.184834123222749e-05,
|
| 43 |
+
"loss": 1.7585,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09,
|
| 48 |
+
"grad_norm": 27.140174865722656,
|
| 49 |
+
"learning_rate": 1.4218009478672985e-05,
|
| 50 |
+
"loss": 1.4911,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.1,
|
| 55 |
+
"grad_norm": 17.97708511352539,
|
| 56 |
+
"learning_rate": 1.6587677725118483e-05,
|
| 57 |
+
"loss": 1.2294,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.11,
|
| 62 |
+
"grad_norm": 36.129730224609375,
|
| 63 |
+
"learning_rate": 1.895734597156398e-05,
|
| 64 |
+
"loss": 1.0185,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.13,
|
| 69 |
+
"grad_norm": 22.884685516357422,
|
| 70 |
+
"learning_rate": 2.132701421800948e-05,
|
| 71 |
+
"loss": 0.8923,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.14,
|
| 76 |
+
"grad_norm": 19.778934478759766,
|
| 77 |
+
"learning_rate": 2.369668246445498e-05,
|
| 78 |
+
"loss": 0.7784,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.16,
|
| 83 |
+
"grad_norm": 20.47595977783203,
|
| 84 |
+
"learning_rate": 2.6066350710900477e-05,
|
| 85 |
+
"loss": 0.6687,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.17,
|
| 90 |
+
"grad_norm": 17.096622467041016,
|
| 91 |
+
"learning_rate": 2.843601895734597e-05,
|
| 92 |
+
"loss": 0.6549,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.18,
|
| 97 |
+
"grad_norm": 18.601505279541016,
|
| 98 |
+
"learning_rate": 3.080568720379147e-05,
|
| 99 |
+
"loss": 0.6253,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.2,
|
| 104 |
+
"grad_norm": 15.65523624420166,
|
| 105 |
+
"learning_rate": 3.3175355450236966e-05,
|
| 106 |
+
"loss": 0.6118,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.21,
|
| 111 |
+
"grad_norm": 15.892389297485352,
|
| 112 |
+
"learning_rate": 3.554502369668247e-05,
|
| 113 |
+
"loss": 0.6227,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.23,
|
| 118 |
+
"grad_norm": 12.784531593322754,
|
| 119 |
+
"learning_rate": 3.791469194312796e-05,
|
| 120 |
+
"loss": 0.4946,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.24,
|
| 125 |
+
"grad_norm": 13.44602108001709,
|
| 126 |
+
"learning_rate": 4.028436018957346e-05,
|
| 127 |
+
"loss": 0.6287,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.26,
|
| 132 |
+
"grad_norm": 17.538671493530273,
|
| 133 |
+
"learning_rate": 4.265402843601896e-05,
|
| 134 |
+
"loss": 0.5645,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.27,
|
| 139 |
+
"grad_norm": 11.633671760559082,
|
| 140 |
+
"learning_rate": 4.502369668246446e-05,
|
| 141 |
+
"loss": 0.5169,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.28,
|
| 146 |
+
"grad_norm": 13.391677856445312,
|
| 147 |
+
"learning_rate": 4.739336492890996e-05,
|
| 148 |
+
"loss": 0.6072,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.3,
|
| 153 |
+
"grad_norm": 8.77043342590332,
|
| 154 |
+
"learning_rate": 4.976303317535545e-05,
|
| 155 |
+
"loss": 0.4768,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.31,
|
| 160 |
+
"grad_norm": 11.799147605895996,
|
| 161 |
+
"learning_rate": 4.976290832455216e-05,
|
| 162 |
+
"loss": 0.4924,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.33,
|
| 167 |
+
"grad_norm": 8.894740104675293,
|
| 168 |
+
"learning_rate": 4.949947312961012e-05,
|
| 169 |
+
"loss": 0.4062,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.34,
|
| 174 |
+
"grad_norm": 11.966766357421875,
|
| 175 |
+
"learning_rate": 4.923603793466807e-05,
|
| 176 |
+
"loss": 0.5118,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.36,
|
| 181 |
+
"grad_norm": 12.86320686340332,
|
| 182 |
+
"learning_rate": 4.8972602739726034e-05,
|
| 183 |
+
"loss": 0.5241,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.37,
|
| 188 |
+
"grad_norm": 8.136141777038574,
|
| 189 |
+
"learning_rate": 4.8709167544783986e-05,
|
| 190 |
+
"loss": 0.4875,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.38,
|
| 195 |
+
"grad_norm": 12.6289644241333,
|
| 196 |
+
"learning_rate": 4.8445732349841945e-05,
|
| 197 |
+
"loss": 0.482,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.4,
|
| 202 |
+
"grad_norm": 11.868968963623047,
|
| 203 |
+
"learning_rate": 4.8182297154899896e-05,
|
| 204 |
+
"loss": 0.5488,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.41,
|
| 209 |
+
"grad_norm": 11.955994606018066,
|
| 210 |
+
"learning_rate": 4.791886195995785e-05,
|
| 211 |
+
"loss": 0.4242,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.43,
|
| 216 |
+
"grad_norm": 10.719066619873047,
|
| 217 |
+
"learning_rate": 4.7655426765015806e-05,
|
| 218 |
+
"loss": 0.4832,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.44,
|
| 223 |
+
"grad_norm": 9.737288475036621,
|
| 224 |
+
"learning_rate": 4.7391991570073765e-05,
|
| 225 |
+
"loss": 0.4619,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.46,
|
| 230 |
+
"grad_norm": 9.085290908813477,
|
| 231 |
+
"learning_rate": 4.712855637513172e-05,
|
| 232 |
+
"loss": 0.4583,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.47,
|
| 237 |
+
"grad_norm": 5.268682956695557,
|
| 238 |
+
"learning_rate": 4.6865121180189675e-05,
|
| 239 |
+
"loss": 0.4926,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.48,
|
| 244 |
+
"grad_norm": 8.957120895385742,
|
| 245 |
+
"learning_rate": 4.6601685985247633e-05,
|
| 246 |
+
"loss": 0.4615,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5,
|
| 251 |
+
"grad_norm": 5.488365650177002,
|
| 252 |
+
"learning_rate": 4.6338250790305585e-05,
|
| 253 |
+
"loss": 0.4487,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.51,
|
| 258 |
+
"grad_norm": 9.455901145935059,
|
| 259 |
+
"learning_rate": 4.6074815595363544e-05,
|
| 260 |
+
"loss": 0.4429,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.53,
|
| 265 |
+
"grad_norm": 10.715672492980957,
|
| 266 |
+
"learning_rate": 4.58113804004215e-05,
|
| 267 |
+
"loss": 0.4603,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.54,
|
| 272 |
+
"grad_norm": 10.67427921295166,
|
| 273 |
+
"learning_rate": 4.5547945205479454e-05,
|
| 274 |
+
"loss": 0.3811,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.55,
|
| 279 |
+
"grad_norm": 7.144228935241699,
|
| 280 |
+
"learning_rate": 4.528451001053741e-05,
|
| 281 |
+
"loss": 0.4789,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.57,
|
| 286 |
+
"grad_norm": 10.277647972106934,
|
| 287 |
+
"learning_rate": 4.5021074815595364e-05,
|
| 288 |
+
"loss": 0.4265,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.58,
|
| 293 |
+
"grad_norm": 9.004936218261719,
|
| 294 |
+
"learning_rate": 4.4757639620653316e-05,
|
| 295 |
+
"loss": 0.468,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.6,
|
| 300 |
+
"grad_norm": 8.17403507232666,
|
| 301 |
+
"learning_rate": 4.449420442571128e-05,
|
| 302 |
+
"loss": 0.3897,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.61,
|
| 307 |
+
"grad_norm": 7.23596715927124,
|
| 308 |
+
"learning_rate": 4.423076923076923e-05,
|
| 309 |
+
"loss": 0.3734,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.63,
|
| 314 |
+
"grad_norm": 8.713665962219238,
|
| 315 |
+
"learning_rate": 4.396733403582719e-05,
|
| 316 |
+
"loss": 0.4557,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.64,
|
| 321 |
+
"grad_norm": 9.222480773925781,
|
| 322 |
+
"learning_rate": 4.370389884088514e-05,
|
| 323 |
+
"loss": 0.3975,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.65,
|
| 328 |
+
"grad_norm": 8.174620628356934,
|
| 329 |
+
"learning_rate": 4.34404636459431e-05,
|
| 330 |
+
"loss": 0.4742,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.67,
|
| 335 |
+
"grad_norm": 10.548660278320312,
|
| 336 |
+
"learning_rate": 4.317702845100105e-05,
|
| 337 |
+
"loss": 0.4072,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.68,
|
| 342 |
+
"grad_norm": 10.924005508422852,
|
| 343 |
+
"learning_rate": 4.291359325605901e-05,
|
| 344 |
+
"loss": 0.3547,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.7,
|
| 349 |
+
"grad_norm": 7.682478427886963,
|
| 350 |
+
"learning_rate": 4.265015806111697e-05,
|
| 351 |
+
"loss": 0.3249,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.71,
|
| 356 |
+
"grad_norm": 9.409101486206055,
|
| 357 |
+
"learning_rate": 4.238672286617492e-05,
|
| 358 |
+
"loss": 0.3852,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.73,
|
| 363 |
+
"grad_norm": 8.597813606262207,
|
| 364 |
+
"learning_rate": 4.212328767123288e-05,
|
| 365 |
+
"loss": 0.3998,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.74,
|
| 370 |
+
"grad_norm": 8.316597938537598,
|
| 371 |
+
"learning_rate": 4.185985247629083e-05,
|
| 372 |
+
"loss": 0.4554,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.75,
|
| 377 |
+
"grad_norm": 7.785365581512451,
|
| 378 |
+
"learning_rate": 4.159641728134879e-05,
|
| 379 |
+
"loss": 0.4193,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.77,
|
| 384 |
+
"grad_norm": 9.083528518676758,
|
| 385 |
+
"learning_rate": 4.133298208640675e-05,
|
| 386 |
+
"loss": 0.3247,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.78,
|
| 391 |
+
"grad_norm": 6.537896633148193,
|
| 392 |
+
"learning_rate": 4.10695468914647e-05,
|
| 393 |
+
"loss": 0.3701,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.8,
|
| 398 |
+
"grad_norm": 4.992751598358154,
|
| 399 |
+
"learning_rate": 4.080611169652266e-05,
|
| 400 |
+
"loss": 0.3539,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.81,
|
| 405 |
+
"grad_norm": 9.362491607666016,
|
| 406 |
+
"learning_rate": 4.054267650158061e-05,
|
| 407 |
+
"loss": 0.3521,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.82,
|
| 412 |
+
"grad_norm": 9.023208618164062,
|
| 413 |
+
"learning_rate": 4.027924130663857e-05,
|
| 414 |
+
"loss": 0.3028,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.84,
|
| 419 |
+
"grad_norm": 11.851508140563965,
|
| 420 |
+
"learning_rate": 4.001580611169653e-05,
|
| 421 |
+
"loss": 0.3854,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.85,
|
| 426 |
+
"grad_norm": 11.16539192199707,
|
| 427 |
+
"learning_rate": 3.975237091675448e-05,
|
| 428 |
+
"loss": 0.4021,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.87,
|
| 433 |
+
"grad_norm": 7.1588311195373535,
|
| 434 |
+
"learning_rate": 3.948893572181244e-05,
|
| 435 |
+
"loss": 0.3865,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.88,
|
| 440 |
+
"grad_norm": 8.23454761505127,
|
| 441 |
+
"learning_rate": 3.922550052687039e-05,
|
| 442 |
+
"loss": 0.3941,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.9,
|
| 447 |
+
"grad_norm": 9.590554237365723,
|
| 448 |
+
"learning_rate": 3.896206533192835e-05,
|
| 449 |
+
"loss": 0.3831,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.91,
|
| 454 |
+
"grad_norm": 7.980281829833984,
|
| 455 |
+
"learning_rate": 3.86986301369863e-05,
|
| 456 |
+
"loss": 0.4134,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.92,
|
| 461 |
+
"grad_norm": 6.231551647186279,
|
| 462 |
+
"learning_rate": 3.843519494204426e-05,
|
| 463 |
+
"loss": 0.3746,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.94,
|
| 468 |
+
"grad_norm": 8.723283767700195,
|
| 469 |
+
"learning_rate": 3.8171759747102217e-05,
|
| 470 |
+
"loss": 0.3349,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.95,
|
| 475 |
+
"grad_norm": 7.533834934234619,
|
| 476 |
+
"learning_rate": 3.790832455216017e-05,
|
| 477 |
+
"loss": 0.3229,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.97,
|
| 482 |
+
"grad_norm": 6.750338554382324,
|
| 483 |
+
"learning_rate": 3.764488935721813e-05,
|
| 484 |
+
"loss": 0.3163,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.98,
|
| 489 |
+
"grad_norm": 11.813076972961426,
|
| 490 |
+
"learning_rate": 3.738145416227608e-05,
|
| 491 |
+
"loss": 0.4089,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.0,
|
| 496 |
+
"grad_norm": 7.000370979309082,
|
| 497 |
+
"learning_rate": 3.711801896733404e-05,
|
| 498 |
+
"loss": 0.4327,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.0,
|
| 503 |
+
"eval_accuracy": 0.9774,
|
| 504 |
+
"eval_loss": 0.06401650607585907,
|
| 505 |
+
"eval_runtime": 65.8888,
|
| 506 |
+
"eval_samples_per_second": 75.885,
|
| 507 |
+
"eval_steps_per_second": 4.75,
|
| 508 |
+
"step": 703
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.01,
|
| 512 |
+
"grad_norm": 7.698484420776367,
|
| 513 |
+
"learning_rate": 3.6854583772391995e-05,
|
| 514 |
+
"loss": 0.359,
|
| 515 |
+
"step": 710
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.02,
|
| 519 |
+
"grad_norm": 10.548850059509277,
|
| 520 |
+
"learning_rate": 3.6591148577449954e-05,
|
| 521 |
+
"loss": 0.3162,
|
| 522 |
+
"step": 720
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.04,
|
| 526 |
+
"grad_norm": 10.149295806884766,
|
| 527 |
+
"learning_rate": 3.6327713382507905e-05,
|
| 528 |
+
"loss": 0.3265,
|
| 529 |
+
"step": 730
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.05,
|
| 533 |
+
"grad_norm": 6.814464092254639,
|
| 534 |
+
"learning_rate": 3.606427818756586e-05,
|
| 535 |
+
"loss": 0.3536,
|
| 536 |
+
"step": 740
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.07,
|
| 540 |
+
"grad_norm": 8.711482048034668,
|
| 541 |
+
"learning_rate": 3.5800842992623816e-05,
|
| 542 |
+
"loss": 0.379,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.08,
|
| 547 |
+
"grad_norm": 8.469487190246582,
|
| 548 |
+
"learning_rate": 3.5537407797681774e-05,
|
| 549 |
+
"loss": 0.3117,
|
| 550 |
+
"step": 760
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.09,
|
| 554 |
+
"grad_norm": 9.90764045715332,
|
| 555 |
+
"learning_rate": 3.527397260273973e-05,
|
| 556 |
+
"loss": 0.3177,
|
| 557 |
+
"step": 770
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.11,
|
| 561 |
+
"grad_norm": 7.835967063903809,
|
| 562 |
+
"learning_rate": 3.5010537407797684e-05,
|
| 563 |
+
"loss": 0.347,
|
| 564 |
+
"step": 780
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.12,
|
| 568 |
+
"grad_norm": 5.563517093658447,
|
| 569 |
+
"learning_rate": 3.4747102212855636e-05,
|
| 570 |
+
"loss": 0.3244,
|
| 571 |
+
"step": 790
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.14,
|
| 575 |
+
"grad_norm": 10.490474700927734,
|
| 576 |
+
"learning_rate": 3.4483667017913594e-05,
|
| 577 |
+
"loss": 0.3159,
|
| 578 |
+
"step": 800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.15,
|
| 582 |
+
"grad_norm": 7.063011169433594,
|
| 583 |
+
"learning_rate": 3.4220231822971546e-05,
|
| 584 |
+
"loss": 0.2711,
|
| 585 |
+
"step": 810
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.17,
|
| 589 |
+
"grad_norm": 9.191926002502441,
|
| 590 |
+
"learning_rate": 3.3956796628029505e-05,
|
| 591 |
+
"loss": 0.3441,
|
| 592 |
+
"step": 820
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.18,
|
| 596 |
+
"grad_norm": 5.757563591003418,
|
| 597 |
+
"learning_rate": 3.369336143308746e-05,
|
| 598 |
+
"loss": 0.3562,
|
| 599 |
+
"step": 830
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.19,
|
| 603 |
+
"grad_norm": 5.849490642547607,
|
| 604 |
+
"learning_rate": 3.342992623814542e-05,
|
| 605 |
+
"loss": 0.2797,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.21,
|
| 610 |
+
"grad_norm": 5.61791467666626,
|
| 611 |
+
"learning_rate": 3.316649104320337e-05,
|
| 612 |
+
"loss": 0.3362,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.22,
|
| 617 |
+
"grad_norm": 7.7571916580200195,
|
| 618 |
+
"learning_rate": 3.2903055848261325e-05,
|
| 619 |
+
"loss": 0.2682,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.24,
|
| 624 |
+
"grad_norm": 9.108525276184082,
|
| 625 |
+
"learning_rate": 3.2639620653319283e-05,
|
| 626 |
+
"loss": 0.3755,
|
| 627 |
+
"step": 870
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.25,
|
| 631 |
+
"grad_norm": 5.915694713592529,
|
| 632 |
+
"learning_rate": 3.237618545837724e-05,
|
| 633 |
+
"loss": 0.3061,
|
| 634 |
+
"step": 880
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.27,
|
| 638 |
+
"grad_norm": 7.744335651397705,
|
| 639 |
+
"learning_rate": 3.21127502634352e-05,
|
| 640 |
+
"loss": 0.367,
|
| 641 |
+
"step": 890
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.28,
|
| 645 |
+
"grad_norm": 9.851384162902832,
|
| 646 |
+
"learning_rate": 3.184931506849315e-05,
|
| 647 |
+
"loss": 0.3318,
|
| 648 |
+
"step": 900
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.29,
|
| 652 |
+
"grad_norm": 6.104685306549072,
|
| 653 |
+
"learning_rate": 3.1585879873551104e-05,
|
| 654 |
+
"loss": 0.3477,
|
| 655 |
+
"step": 910
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.31,
|
| 659 |
+
"grad_norm": 7.819393157958984,
|
| 660 |
+
"learning_rate": 3.132244467860906e-05,
|
| 661 |
+
"loss": 0.2994,
|
| 662 |
+
"step": 920
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.32,
|
| 666 |
+
"grad_norm": 9.686634063720703,
|
| 667 |
+
"learning_rate": 3.105900948366702e-05,
|
| 668 |
+
"loss": 0.2885,
|
| 669 |
+
"step": 930
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.34,
|
| 673 |
+
"grad_norm": 4.374415397644043,
|
| 674 |
+
"learning_rate": 3.079557428872498e-05,
|
| 675 |
+
"loss": 0.3053,
|
| 676 |
+
"step": 940
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.35,
|
| 680 |
+
"grad_norm": 9.000449180603027,
|
| 681 |
+
"learning_rate": 3.053213909378293e-05,
|
| 682 |
+
"loss": 0.2896,
|
| 683 |
+
"step": 950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.37,
|
| 687 |
+
"grad_norm": 6.736356258392334,
|
| 688 |
+
"learning_rate": 3.026870389884089e-05,
|
| 689 |
+
"loss": 0.2897,
|
| 690 |
+
"step": 960
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.38,
|
| 694 |
+
"grad_norm": 6.521255016326904,
|
| 695 |
+
"learning_rate": 3.000526870389884e-05,
|
| 696 |
+
"loss": 0.3365,
|
| 697 |
+
"step": 970
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.39,
|
| 701 |
+
"grad_norm": 7.526865482330322,
|
| 702 |
+
"learning_rate": 2.9741833508956796e-05,
|
| 703 |
+
"loss": 0.2817,
|
| 704 |
+
"step": 980
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.41,
|
| 708 |
+
"grad_norm": 3.9108359813690186,
|
| 709 |
+
"learning_rate": 2.9478398314014755e-05,
|
| 710 |
+
"loss": 0.3002,
|
| 711 |
+
"step": 990
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.42,
|
| 715 |
+
"grad_norm": 9.110422134399414,
|
| 716 |
+
"learning_rate": 2.921496311907271e-05,
|
| 717 |
+
"loss": 0.3055,
|
| 718 |
+
"step": 1000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.44,
|
| 722 |
+
"grad_norm": 5.4320573806762695,
|
| 723 |
+
"learning_rate": 2.8951527924130668e-05,
|
| 724 |
+
"loss": 0.3126,
|
| 725 |
+
"step": 1010
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.45,
|
| 729 |
+
"grad_norm": 8.311667442321777,
|
| 730 |
+
"learning_rate": 2.868809272918862e-05,
|
| 731 |
+
"loss": 0.3177,
|
| 732 |
+
"step": 1020
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.46,
|
| 736 |
+
"grad_norm": 8.694578170776367,
|
| 737 |
+
"learning_rate": 2.842465753424658e-05,
|
| 738 |
+
"loss": 0.2458,
|
| 739 |
+
"step": 1030
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.48,
|
| 743 |
+
"grad_norm": 6.780071258544922,
|
| 744 |
+
"learning_rate": 2.8161222339304533e-05,
|
| 745 |
+
"loss": 0.2171,
|
| 746 |
+
"step": 1040
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.49,
|
| 750 |
+
"grad_norm": 8.7374906539917,
|
| 751 |
+
"learning_rate": 2.7897787144362485e-05,
|
| 752 |
+
"loss": 0.2914,
|
| 753 |
+
"step": 1050
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.51,
|
| 757 |
+
"grad_norm": 8.430846214294434,
|
| 758 |
+
"learning_rate": 2.7634351949420444e-05,
|
| 759 |
+
"loss": 0.2981,
|
| 760 |
+
"step": 1060
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.52,
|
| 764 |
+
"grad_norm": 6.208935737609863,
|
| 765 |
+
"learning_rate": 2.73709167544784e-05,
|
| 766 |
+
"loss": 0.2859,
|
| 767 |
+
"step": 1070
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 1.54,
|
| 771 |
+
"grad_norm": 6.287723541259766,
|
| 772 |
+
"learning_rate": 2.7107481559536357e-05,
|
| 773 |
+
"loss": 0.2559,
|
| 774 |
+
"step": 1080
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.55,
|
| 778 |
+
"grad_norm": 9.158524513244629,
|
| 779 |
+
"learning_rate": 2.6844046364594312e-05,
|
| 780 |
+
"loss": 0.2544,
|
| 781 |
+
"step": 1090
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 1.56,
|
| 785 |
+
"grad_norm": 8.676222801208496,
|
| 786 |
+
"learning_rate": 2.6580611169652264e-05,
|
| 787 |
+
"loss": 0.2648,
|
| 788 |
+
"step": 1100
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 1.58,
|
| 792 |
+
"grad_norm": 5.815190315246582,
|
| 793 |
+
"learning_rate": 2.6317175974710222e-05,
|
| 794 |
+
"loss": 0.2794,
|
| 795 |
+
"step": 1110
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 1.59,
|
| 799 |
+
"grad_norm": 7.133479595184326,
|
| 800 |
+
"learning_rate": 2.6053740779768177e-05,
|
| 801 |
+
"loss": 0.3008,
|
| 802 |
+
"step": 1120
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 1.61,
|
| 806 |
+
"grad_norm": 9.465476989746094,
|
| 807 |
+
"learning_rate": 2.5790305584826136e-05,
|
| 808 |
+
"loss": 0.3032,
|
| 809 |
+
"step": 1130
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 1.62,
|
| 813 |
+
"grad_norm": 5.298926830291748,
|
| 814 |
+
"learning_rate": 2.5526870389884088e-05,
|
| 815 |
+
"loss": 0.2616,
|
| 816 |
+
"step": 1140
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 1.64,
|
| 820 |
+
"grad_norm": 3.1998720169067383,
|
| 821 |
+
"learning_rate": 2.5263435194942046e-05,
|
| 822 |
+
"loss": 0.3016,
|
| 823 |
+
"step": 1150
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 1.65,
|
| 827 |
+
"grad_norm": 9.716959953308105,
|
| 828 |
+
"learning_rate": 2.5e-05,
|
| 829 |
+
"loss": 0.2625,
|
| 830 |
+
"step": 1160
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.66,
|
| 834 |
+
"grad_norm": 9.040936470031738,
|
| 835 |
+
"learning_rate": 2.4736564805057956e-05,
|
| 836 |
+
"loss": 0.2634,
|
| 837 |
+
"step": 1170
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 1.68,
|
| 841 |
+
"grad_norm": 7.872269630432129,
|
| 842 |
+
"learning_rate": 2.4473129610115915e-05,
|
| 843 |
+
"loss": 0.2744,
|
| 844 |
+
"step": 1180
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 1.69,
|
| 848 |
+
"grad_norm": 4.666429042816162,
|
| 849 |
+
"learning_rate": 2.420969441517387e-05,
|
| 850 |
+
"loss": 0.2552,
|
| 851 |
+
"step": 1190
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.71,
|
| 855 |
+
"grad_norm": 5.423487663269043,
|
| 856 |
+
"learning_rate": 2.394625922023182e-05,
|
| 857 |
+
"loss": 0.3138,
|
| 858 |
+
"step": 1200
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 1.72,
|
| 862 |
+
"grad_norm": 5.613728046417236,
|
| 863 |
+
"learning_rate": 2.368282402528978e-05,
|
| 864 |
+
"loss": 0.3048,
|
| 865 |
+
"step": 1210
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.73,
|
| 869 |
+
"grad_norm": 5.959019660949707,
|
| 870 |
+
"learning_rate": 2.3419388830347735e-05,
|
| 871 |
+
"loss": 0.2784,
|
| 872 |
+
"step": 1220
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 1.75,
|
| 876 |
+
"grad_norm": 5.206528663635254,
|
| 877 |
+
"learning_rate": 2.315595363540569e-05,
|
| 878 |
+
"loss": 0.3245,
|
| 879 |
+
"step": 1230
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 1.76,
|
| 883 |
+
"grad_norm": 6.318498611450195,
|
| 884 |
+
"learning_rate": 2.289251844046365e-05,
|
| 885 |
+
"loss": 0.299,
|
| 886 |
+
"step": 1240
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 1.78,
|
| 890 |
+
"grad_norm": 5.1116437911987305,
|
| 891 |
+
"learning_rate": 2.2629083245521604e-05,
|
| 892 |
+
"loss": 0.2706,
|
| 893 |
+
"step": 1250
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 1.79,
|
| 897 |
+
"grad_norm": 9.792173385620117,
|
| 898 |
+
"learning_rate": 2.236564805057956e-05,
|
| 899 |
+
"loss": 0.2698,
|
| 900 |
+
"step": 1260
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 1.81,
|
| 904 |
+
"grad_norm": 5.214483737945557,
|
| 905 |
+
"learning_rate": 2.2102212855637514e-05,
|
| 906 |
+
"loss": 0.2968,
|
| 907 |
+
"step": 1270
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 1.82,
|
| 911 |
+
"grad_norm": 5.681826114654541,
|
| 912 |
+
"learning_rate": 2.183877766069547e-05,
|
| 913 |
+
"loss": 0.2893,
|
| 914 |
+
"step": 1280
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 1.83,
|
| 918 |
+
"grad_norm": 4.573737621307373,
|
| 919 |
+
"learning_rate": 2.1575342465753427e-05,
|
| 920 |
+
"loss": 0.2424,
|
| 921 |
+
"step": 1290
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.85,
|
| 925 |
+
"grad_norm": 5.404233932495117,
|
| 926 |
+
"learning_rate": 2.1311907270811383e-05,
|
| 927 |
+
"loss": 0.2956,
|
| 928 |
+
"step": 1300
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 1.86,
|
| 932 |
+
"grad_norm": 7.8233160972595215,
|
| 933 |
+
"learning_rate": 2.1048472075869338e-05,
|
| 934 |
+
"loss": 0.2992,
|
| 935 |
+
"step": 1310
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 1.88,
|
| 939 |
+
"grad_norm": 5.9538373947143555,
|
| 940 |
+
"learning_rate": 2.0785036880927293e-05,
|
| 941 |
+
"loss": 0.2843,
|
| 942 |
+
"step": 1320
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 1.89,
|
| 946 |
+
"grad_norm": 6.776533603668213,
|
| 947 |
+
"learning_rate": 2.0521601685985248e-05,
|
| 948 |
+
"loss": 0.2255,
|
| 949 |
+
"step": 1330
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 1.91,
|
| 953 |
+
"grad_norm": 4.279127597808838,
|
| 954 |
+
"learning_rate": 2.0258166491043203e-05,
|
| 955 |
+
"loss": 0.2791,
|
| 956 |
+
"step": 1340
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.92,
|
| 960 |
+
"grad_norm": 9.389575958251953,
|
| 961 |
+
"learning_rate": 1.999473129610116e-05,
|
| 962 |
+
"loss": 0.3064,
|
| 963 |
+
"step": 1350
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 1.93,
|
| 967 |
+
"grad_norm": 6.343210697174072,
|
| 968 |
+
"learning_rate": 1.9731296101159116e-05,
|
| 969 |
+
"loss": 0.215,
|
| 970 |
+
"step": 1360
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 1.95,
|
| 974 |
+
"grad_norm": 6.596449375152588,
|
| 975 |
+
"learning_rate": 1.946786090621707e-05,
|
| 976 |
+
"loss": 0.2429,
|
| 977 |
+
"step": 1370
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 1.96,
|
| 981 |
+
"grad_norm": 6.299315929412842,
|
| 982 |
+
"learning_rate": 1.9204425711275027e-05,
|
| 983 |
+
"loss": 0.3099,
|
| 984 |
+
"step": 1380
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 1.98,
|
| 988 |
+
"grad_norm": 7.726513385772705,
|
| 989 |
+
"learning_rate": 1.894099051633298e-05,
|
| 990 |
+
"loss": 0.2435,
|
| 991 |
+
"step": 1390
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 1.99,
|
| 995 |
+
"grad_norm": 7.626183986663818,
|
| 996 |
+
"learning_rate": 1.8677555321390937e-05,
|
| 997 |
+
"loss": 0.2814,
|
| 998 |
+
"step": 1400
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 2.0,
|
| 1002 |
+
"eval_accuracy": 0.9834,
|
| 1003 |
+
"eval_loss": 0.048424676060676575,
|
| 1004 |
+
"eval_runtime": 66.0101,
|
| 1005 |
+
"eval_samples_per_second": 75.746,
|
| 1006 |
+
"eval_steps_per_second": 4.742,
|
| 1007 |
+
"step": 1406
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 2.0,
|
| 1011 |
+
"grad_norm": 5.147545337677002,
|
| 1012 |
+
"learning_rate": 1.8414120126448895e-05,
|
| 1013 |
+
"loss": 0.2343,
|
| 1014 |
+
"step": 1410
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.02,
|
| 1018 |
+
"grad_norm": 3.9744009971618652,
|
| 1019 |
+
"learning_rate": 1.815068493150685e-05,
|
| 1020 |
+
"loss": 0.2542,
|
| 1021 |
+
"step": 1420
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 2.03,
|
| 1025 |
+
"grad_norm": 7.23543643951416,
|
| 1026 |
+
"learning_rate": 1.7887249736564805e-05,
|
| 1027 |
+
"loss": 0.2568,
|
| 1028 |
+
"step": 1430
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.05,
|
| 1032 |
+
"grad_norm": 4.5337395668029785,
|
| 1033 |
+
"learning_rate": 1.7623814541622764e-05,
|
| 1034 |
+
"loss": 0.2348,
|
| 1035 |
+
"step": 1440
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 2.06,
|
| 1039 |
+
"grad_norm": 7.225223541259766,
|
| 1040 |
+
"learning_rate": 1.7360379346680716e-05,
|
| 1041 |
+
"loss": 0.2861,
|
| 1042 |
+
"step": 1450
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 2.08,
|
| 1046 |
+
"grad_norm": 6.333385467529297,
|
| 1047 |
+
"learning_rate": 1.7096944151738674e-05,
|
| 1048 |
+
"loss": 0.2237,
|
| 1049 |
+
"step": 1460
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 2.09,
|
| 1053 |
+
"grad_norm": 9.622603416442871,
|
| 1054 |
+
"learning_rate": 1.683350895679663e-05,
|
| 1055 |
+
"loss": 0.2707,
|
| 1056 |
+
"step": 1470
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.1,
|
| 1060 |
+
"grad_norm": 3.6370978355407715,
|
| 1061 |
+
"learning_rate": 1.6570073761854584e-05,
|
| 1062 |
+
"loss": 0.2177,
|
| 1063 |
+
"step": 1480
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.12,
|
| 1067 |
+
"grad_norm": 5.092869758605957,
|
| 1068 |
+
"learning_rate": 1.630663856691254e-05,
|
| 1069 |
+
"loss": 0.174,
|
| 1070 |
+
"step": 1490
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.13,
|
| 1074 |
+
"grad_norm": 7.764402389526367,
|
| 1075 |
+
"learning_rate": 1.6043203371970498e-05,
|
| 1076 |
+
"loss": 0.2166,
|
| 1077 |
+
"step": 1500
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.15,
|
| 1081 |
+
"grad_norm": 5.907904624938965,
|
| 1082 |
+
"learning_rate": 1.577976817702845e-05,
|
| 1083 |
+
"loss": 0.219,
|
| 1084 |
+
"step": 1510
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.16,
|
| 1088 |
+
"grad_norm": 5.130086898803711,
|
| 1089 |
+
"learning_rate": 1.5516332982086408e-05,
|
| 1090 |
+
"loss": 0.1956,
|
| 1091 |
+
"step": 1520
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.18,
|
| 1095 |
+
"grad_norm": 7.971607685089111,
|
| 1096 |
+
"learning_rate": 1.5252897787144363e-05,
|
| 1097 |
+
"loss": 0.1871,
|
| 1098 |
+
"step": 1530
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.19,
|
| 1102 |
+
"grad_norm": 10.202670097351074,
|
| 1103 |
+
"learning_rate": 1.498946259220232e-05,
|
| 1104 |
+
"loss": 0.2444,
|
| 1105 |
+
"step": 1540
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.2,
|
| 1109 |
+
"grad_norm": 5.461312770843506,
|
| 1110 |
+
"learning_rate": 1.4726027397260275e-05,
|
| 1111 |
+
"loss": 0.2538,
|
| 1112 |
+
"step": 1550
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.22,
|
| 1116 |
+
"grad_norm": 8.133660316467285,
|
| 1117 |
+
"learning_rate": 1.4462592202318232e-05,
|
| 1118 |
+
"loss": 0.2324,
|
| 1119 |
+
"step": 1560
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.23,
|
| 1123 |
+
"grad_norm": 7.940431594848633,
|
| 1124 |
+
"learning_rate": 1.4199157007376185e-05,
|
| 1125 |
+
"loss": 0.2264,
|
| 1126 |
+
"step": 1570
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.25,
|
| 1130 |
+
"grad_norm": 6.688675403594971,
|
| 1131 |
+
"learning_rate": 1.3935721812434142e-05,
|
| 1132 |
+
"loss": 0.199,
|
| 1133 |
+
"step": 1580
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 2.26,
|
| 1137 |
+
"grad_norm": 7.409407138824463,
|
| 1138 |
+
"learning_rate": 1.3672286617492097e-05,
|
| 1139 |
+
"loss": 0.2311,
|
| 1140 |
+
"step": 1590
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.28,
|
| 1144 |
+
"grad_norm": 9.85133171081543,
|
| 1145 |
+
"learning_rate": 1.3408851422550054e-05,
|
| 1146 |
+
"loss": 0.2153,
|
| 1147 |
+
"step": 1600
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 2.29,
|
| 1151 |
+
"grad_norm": 6.340139865875244,
|
| 1152 |
+
"learning_rate": 1.3145416227608009e-05,
|
| 1153 |
+
"loss": 0.2717,
|
| 1154 |
+
"step": 1610
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 2.3,
|
| 1158 |
+
"grad_norm": 5.589156627655029,
|
| 1159 |
+
"learning_rate": 1.2881981032665966e-05,
|
| 1160 |
+
"loss": 0.2068,
|
| 1161 |
+
"step": 1620
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 2.32,
|
| 1165 |
+
"grad_norm": 8.83266830444336,
|
| 1166 |
+
"learning_rate": 1.2618545837723922e-05,
|
| 1167 |
+
"loss": 0.2459,
|
| 1168 |
+
"step": 1630
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 2.33,
|
| 1172 |
+
"grad_norm": 6.671909809112549,
|
| 1173 |
+
"learning_rate": 1.2355110642781877e-05,
|
| 1174 |
+
"loss": 0.2332,
|
| 1175 |
+
"step": 1640
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 2.35,
|
| 1179 |
+
"grad_norm": 8.771206855773926,
|
| 1180 |
+
"learning_rate": 1.209167544783983e-05,
|
| 1181 |
+
"loss": 0.2526,
|
| 1182 |
+
"step": 1650
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 2.36,
|
| 1186 |
+
"grad_norm": 9.015884399414062,
|
| 1187 |
+
"learning_rate": 1.1828240252897788e-05,
|
| 1188 |
+
"loss": 0.2239,
|
| 1189 |
+
"step": 1660
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 2.37,
|
| 1193 |
+
"grad_norm": 9.24109172821045,
|
| 1194 |
+
"learning_rate": 1.1564805057955744e-05,
|
| 1195 |
+
"loss": 0.2304,
|
| 1196 |
+
"step": 1670
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 2.39,
|
| 1200 |
+
"grad_norm": 6.8663330078125,
|
| 1201 |
+
"learning_rate": 1.1301369863013698e-05,
|
| 1202 |
+
"loss": 0.2107,
|
| 1203 |
+
"step": 1680
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 2.4,
|
| 1207 |
+
"grad_norm": 5.675321578979492,
|
| 1208 |
+
"learning_rate": 1.1037934668071655e-05,
|
| 1209 |
+
"loss": 0.2607,
|
| 1210 |
+
"step": 1690
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 2.42,
|
| 1214 |
+
"grad_norm": 6.900195598602295,
|
| 1215 |
+
"learning_rate": 1.0774499473129611e-05,
|
| 1216 |
+
"loss": 0.2323,
|
| 1217 |
+
"step": 1700
|
| 1218 |
+
},
|
| 1219 |
+
{
|
| 1220 |
+
"epoch": 2.43,
|
| 1221 |
+
"grad_norm": 11.515998840332031,
|
| 1222 |
+
"learning_rate": 1.0511064278187566e-05,
|
| 1223 |
+
"loss": 0.2704,
|
| 1224 |
+
"step": 1710
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 2.45,
|
| 1228 |
+
"grad_norm": 7.85768461227417,
|
| 1229 |
+
"learning_rate": 1.0247629083245521e-05,
|
| 1230 |
+
"loss": 0.2175,
|
| 1231 |
+
"step": 1720
|
| 1232 |
+
},
|
| 1233 |
+
{
|
| 1234 |
+
"epoch": 2.46,
|
| 1235 |
+
"grad_norm": 4.934202194213867,
|
| 1236 |
+
"learning_rate": 9.984193888303478e-06,
|
| 1237 |
+
"loss": 0.2069,
|
| 1238 |
+
"step": 1730
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 2.47,
|
| 1242 |
+
"grad_norm": 5.035346031188965,
|
| 1243 |
+
"learning_rate": 9.720758693361433e-06,
|
| 1244 |
+
"loss": 0.2145,
|
| 1245 |
+
"step": 1740
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 2.49,
|
| 1249 |
+
"grad_norm": 7.098711967468262,
|
| 1250 |
+
"learning_rate": 9.457323498419388e-06,
|
| 1251 |
+
"loss": 0.2642,
|
| 1252 |
+
"step": 1750
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 2.5,
|
| 1256 |
+
"grad_norm": 7.717360019683838,
|
| 1257 |
+
"learning_rate": 9.193888303477345e-06,
|
| 1258 |
+
"loss": 0.1801,
|
| 1259 |
+
"step": 1760
|
| 1260 |
+
},
|
| 1261 |
+
{
|
| 1262 |
+
"epoch": 2.52,
|
| 1263 |
+
"grad_norm": 7.103116035461426,
|
| 1264 |
+
"learning_rate": 8.930453108535302e-06,
|
| 1265 |
+
"loss": 0.212,
|
| 1266 |
+
"step": 1770
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 2.53,
|
| 1270 |
+
"grad_norm": 4.945562839508057,
|
| 1271 |
+
"learning_rate": 8.667017913593255e-06,
|
| 1272 |
+
"loss": 0.2158,
|
| 1273 |
+
"step": 1780
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 2.55,
|
| 1277 |
+
"grad_norm": 8.597467422485352,
|
| 1278 |
+
"learning_rate": 8.403582718651212e-06,
|
| 1279 |
+
"loss": 0.233,
|
| 1280 |
+
"step": 1790
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 2.56,
|
| 1284 |
+
"grad_norm": 3.4780356884002686,
|
| 1285 |
+
"learning_rate": 8.140147523709169e-06,
|
| 1286 |
+
"loss": 0.2139,
|
| 1287 |
+
"step": 1800
|
| 1288 |
+
},
|
| 1289 |
+
{
|
| 1290 |
+
"epoch": 2.57,
|
| 1291 |
+
"grad_norm": 8.698033332824707,
|
| 1292 |
+
"learning_rate": 7.876712328767124e-06,
|
| 1293 |
+
"loss": 0.2285,
|
| 1294 |
+
"step": 1810
|
| 1295 |
+
},
|
| 1296 |
+
{
|
| 1297 |
+
"epoch": 2.59,
|
| 1298 |
+
"grad_norm": 8.690217971801758,
|
| 1299 |
+
"learning_rate": 7.613277133825079e-06,
|
| 1300 |
+
"loss": 0.221,
|
| 1301 |
+
"step": 1820
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 2.6,
|
| 1305 |
+
"grad_norm": 6.253328800201416,
|
| 1306 |
+
"learning_rate": 7.349841938883036e-06,
|
| 1307 |
+
"loss": 0.2156,
|
| 1308 |
+
"step": 1830
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 2.62,
|
| 1312 |
+
"grad_norm": 6.390902042388916,
|
| 1313 |
+
"learning_rate": 7.08640674394099e-06,
|
| 1314 |
+
"loss": 0.2134,
|
| 1315 |
+
"step": 1840
|
| 1316 |
+
},
|
| 1317 |
+
{
|
| 1318 |
+
"epoch": 2.63,
|
| 1319 |
+
"grad_norm": 3.520571231842041,
|
| 1320 |
+
"learning_rate": 6.822971548998947e-06,
|
| 1321 |
+
"loss": 0.2177,
|
| 1322 |
+
"step": 1850
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 2.64,
|
| 1326 |
+
"grad_norm": 4.312754154205322,
|
| 1327 |
+
"learning_rate": 6.559536354056903e-06,
|
| 1328 |
+
"loss": 0.1913,
|
| 1329 |
+
"step": 1860
|
| 1330 |
+
},
|
| 1331 |
+
{
|
| 1332 |
+
"epoch": 2.66,
|
| 1333 |
+
"grad_norm": 4.5632758140563965,
|
| 1334 |
+
"learning_rate": 6.296101159114858e-06,
|
| 1335 |
+
"loss": 0.2014,
|
| 1336 |
+
"step": 1870
|
| 1337 |
+
},
|
| 1338 |
+
{
|
| 1339 |
+
"epoch": 2.67,
|
| 1340 |
+
"grad_norm": 8.447988510131836,
|
| 1341 |
+
"learning_rate": 6.032665964172814e-06,
|
| 1342 |
+
"loss": 0.2445,
|
| 1343 |
+
"step": 1880
|
| 1344 |
+
},
|
| 1345 |
+
{
|
| 1346 |
+
"epoch": 2.69,
|
| 1347 |
+
"grad_norm": 4.897846698760986,
|
| 1348 |
+
"learning_rate": 5.76923076923077e-06,
|
| 1349 |
+
"loss": 0.1689,
|
| 1350 |
+
"step": 1890
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 2.7,
|
| 1354 |
+
"grad_norm": 6.561771392822266,
|
| 1355 |
+
"learning_rate": 5.505795574288726e-06,
|
| 1356 |
+
"loss": 0.2089,
|
| 1357 |
+
"step": 1900
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 2.72,
|
| 1361 |
+
"grad_norm": 8.186767578125,
|
| 1362 |
+
"learning_rate": 5.242360379346681e-06,
|
| 1363 |
+
"loss": 0.2337,
|
| 1364 |
+
"step": 1910
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 2.73,
|
| 1368 |
+
"grad_norm": 3.8493165969848633,
|
| 1369 |
+
"learning_rate": 4.978925184404637e-06,
|
| 1370 |
+
"loss": 0.2202,
|
| 1371 |
+
"step": 1920
|
| 1372 |
+
},
|
| 1373 |
+
{
|
| 1374 |
+
"epoch": 2.74,
|
| 1375 |
+
"grad_norm": 4.461836814880371,
|
| 1376 |
+
"learning_rate": 4.715489989462593e-06,
|
| 1377 |
+
"loss": 0.2052,
|
| 1378 |
+
"step": 1930
|
| 1379 |
+
},
|
| 1380 |
+
{
|
| 1381 |
+
"epoch": 2.76,
|
| 1382 |
+
"grad_norm": 6.122828483581543,
|
| 1383 |
+
"learning_rate": 4.452054794520548e-06,
|
| 1384 |
+
"loss": 0.2447,
|
| 1385 |
+
"step": 1940
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 2.77,
|
| 1389 |
+
"grad_norm": 7.432437419891357,
|
| 1390 |
+
"learning_rate": 4.188619599578504e-06,
|
| 1391 |
+
"loss": 0.2148,
|
| 1392 |
+
"step": 1950
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 2.79,
|
| 1396 |
+
"grad_norm": 5.491920471191406,
|
| 1397 |
+
"learning_rate": 3.92518440463646e-06,
|
| 1398 |
+
"loss": 0.2425,
|
| 1399 |
+
"step": 1960
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"epoch": 2.8,
|
| 1403 |
+
"grad_norm": 5.547910690307617,
|
| 1404 |
+
"learning_rate": 3.661749209694415e-06,
|
| 1405 |
+
"loss": 0.2409,
|
| 1406 |
+
"step": 1970
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 2.82,
|
| 1410 |
+
"grad_norm": 7.146701335906982,
|
| 1411 |
+
"learning_rate": 3.398314014752371e-06,
|
| 1412 |
+
"loss": 0.2349,
|
| 1413 |
+
"step": 1980
|
| 1414 |
+
},
|
| 1415 |
+
{
|
| 1416 |
+
"epoch": 2.83,
|
| 1417 |
+
"grad_norm": 9.67960262298584,
|
| 1418 |
+
"learning_rate": 3.1348788198103265e-06,
|
| 1419 |
+
"loss": 0.2119,
|
| 1420 |
+
"step": 1990
|
| 1421 |
+
},
|
| 1422 |
+
{
|
| 1423 |
+
"epoch": 2.84,
|
| 1424 |
+
"grad_norm": 7.780117511749268,
|
| 1425 |
+
"learning_rate": 2.8714436248682825e-06,
|
| 1426 |
+
"loss": 0.225,
|
| 1427 |
+
"step": 2000
|
| 1428 |
+
},
|
| 1429 |
+
{
|
| 1430 |
+
"epoch": 2.86,
|
| 1431 |
+
"grad_norm": 7.887894153594971,
|
| 1432 |
+
"learning_rate": 2.6080084299262384e-06,
|
| 1433 |
+
"loss": 0.1839,
|
| 1434 |
+
"step": 2010
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 2.87,
|
| 1438 |
+
"grad_norm": 3.4736483097076416,
|
| 1439 |
+
"learning_rate": 2.3445732349841943e-06,
|
| 1440 |
+
"loss": 0.1897,
|
| 1441 |
+
"step": 2020
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 2.89,
|
| 1445 |
+
"grad_norm": 3.9504001140594482,
|
| 1446 |
+
"learning_rate": 2.08113804004215e-06,
|
| 1447 |
+
"loss": 0.1665,
|
| 1448 |
+
"step": 2030
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 2.9,
|
| 1452 |
+
"grad_norm": 3.5273189544677734,
|
| 1453 |
+
"learning_rate": 1.8177028451001056e-06,
|
| 1454 |
+
"loss": 0.1694,
|
| 1455 |
+
"step": 2040
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 2.92,
|
| 1459 |
+
"grad_norm": 6.938018321990967,
|
| 1460 |
+
"learning_rate": 1.554267650158061e-06,
|
| 1461 |
+
"loss": 0.2071,
|
| 1462 |
+
"step": 2050
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 2.93,
|
| 1466 |
+
"grad_norm": 3.9890809059143066,
|
| 1467 |
+
"learning_rate": 1.290832455216017e-06,
|
| 1468 |
+
"loss": 0.1837,
|
| 1469 |
+
"step": 2060
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 2.94,
|
| 1473 |
+
"grad_norm": 5.470897197723389,
|
| 1474 |
+
"learning_rate": 1.0273972602739725e-06,
|
| 1475 |
+
"loss": 0.1977,
|
| 1476 |
+
"step": 2070
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 2.96,
|
| 1480 |
+
"grad_norm": 7.521297931671143,
|
| 1481 |
+
"learning_rate": 7.639620653319284e-07,
|
| 1482 |
+
"loss": 0.2391,
|
| 1483 |
+
"step": 2080
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 2.97,
|
| 1487 |
+
"grad_norm": 5.917148590087891,
|
| 1488 |
+
"learning_rate": 5.00526870389884e-07,
|
| 1489 |
+
"loss": 0.1858,
|
| 1490 |
+
"step": 2090
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 2.99,
|
| 1494 |
+
"grad_norm": 5.207727909088135,
|
| 1495 |
+
"learning_rate": 2.3709167544783985e-07,
|
| 1496 |
+
"loss": 0.1655,
|
| 1497 |
+
"step": 2100
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 3.0,
|
| 1501 |
+
"eval_accuracy": 0.9882,
|
| 1502 |
+
"eval_loss": 0.03509841486811638,
|
| 1503 |
+
"eval_runtime": 65.8746,
|
| 1504 |
+
"eval_samples_per_second": 75.902,
|
| 1505 |
+
"eval_steps_per_second": 4.751,
|
| 1506 |
+
"step": 2109
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 3.0,
|
| 1510 |
+
"step": 2109,
|
| 1511 |
+
"total_flos": 1.045459878817923e+19,
|
| 1512 |
+
"train_loss": 0.37951253058504525,
|
| 1513 |
+
"train_runtime": 4954.8349,
|
| 1514 |
+
"train_samples_per_second": 27.246,
|
| 1515 |
+
"train_steps_per_second": 0.426
|
| 1516 |
+
}
|
| 1517 |
+
],
|
| 1518 |
+
"logging_steps": 10,
|
| 1519 |
+
"max_steps": 2109,
|
| 1520 |
+
"num_input_tokens_seen": 0,
|
| 1521 |
+
"num_train_epochs": 3,
|
| 1522 |
+
"save_steps": 500,
|
| 1523 |
+
"total_flos": 1.045459878817923e+19,
|
| 1524 |
+
"train_batch_size": 16,
|
| 1525 |
+
"trial_name": null,
|
| 1526 |
+
"trial_params": null
|
| 1527 |
+
}
|