Commit
•
3e41e34
1
Parent(s):
e9c1c29
End of training
Browse files- all_results.json +13 -0
- eval_results.json +8 -0
- runs/Feb22_02-39-49_6e171dfc349c/events.out.tfevents.1708570540.6e171dfc349c.4642.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +1181 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.9992091735863978,
|
| 4 |
+
"eval_loss": 0.0019393304828554392,
|
| 5 |
+
"eval_runtime": 27.0155,
|
| 6 |
+
"eval_samples_per_second": 93.613,
|
| 7 |
+
"eval_steps_per_second": 11.734,
|
| 8 |
+
"total_flos": 3.915825219204563e+18,
|
| 9 |
+
"train_loss": 0.0232889701421461,
|
| 10 |
+
"train_runtime": 841.0378,
|
| 11 |
+
"train_samples_per_second": 60.122,
|
| 12 |
+
"train_steps_per_second": 1.879
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"eval_accuracy": 0.9992091735863978,
|
| 4 |
+
"eval_loss": 0.0019393304828554392,
|
| 5 |
+
"eval_runtime": 27.0155,
|
| 6 |
+
"eval_samples_per_second": 93.613,
|
| 7 |
+
"eval_steps_per_second": 11.734
|
| 8 |
+
}
|
runs/Feb22_02-39-49_6e171dfc349c/events.out.tfevents.1708570540.6e171dfc349c.4642.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b4e65edf0e1565b596d4bce6a135f5d4f032b39c6087cfa04312b0425e6a355
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 3.915825219204563e+18,
|
| 4 |
+
"train_loss": 0.0232889701421461,
|
| 5 |
+
"train_runtime": 841.0378,
|
| 6 |
+
"train_samples_per_second": 60.122,
|
| 7 |
+
"train_steps_per_second": 1.879
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1181 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.9992091735863978,
|
| 3 |
+
"best_model_checkpoint": "vit-base-patch16-224-clothes-filter/checkpoint-1265",
|
| 4 |
+
"epoch": 4.996047430830039,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1580,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.03,
|
| 13 |
+
"grad_norm": 4.780308246612549,
|
| 14 |
+
"learning_rate": 3.1645569620253167e-06,
|
| 15 |
+
"loss": 0.6282,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.06,
|
| 20 |
+
"grad_norm": 4.528079986572266,
|
| 21 |
+
"learning_rate": 6.329113924050633e-06,
|
| 22 |
+
"loss": 0.5794,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.09,
|
| 27 |
+
"grad_norm": 3.32427978515625,
|
| 28 |
+
"learning_rate": 9.49367088607595e-06,
|
| 29 |
+
"loss": 0.4682,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.13,
|
| 34 |
+
"grad_norm": 3.10469388961792,
|
| 35 |
+
"learning_rate": 1.2658227848101267e-05,
|
| 36 |
+
"loss": 0.3436,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.16,
|
| 41 |
+
"grad_norm": 1.8978112936019897,
|
| 42 |
+
"learning_rate": 1.5822784810126583e-05,
|
| 43 |
+
"loss": 0.2182,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.19,
|
| 48 |
+
"grad_norm": 1.4250308275222778,
|
| 49 |
+
"learning_rate": 1.89873417721519e-05,
|
| 50 |
+
"loss": 0.1309,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.22,
|
| 55 |
+
"grad_norm": 1.0968660116195679,
|
| 56 |
+
"learning_rate": 2.2151898734177217e-05,
|
| 57 |
+
"loss": 0.0826,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.25,
|
| 62 |
+
"grad_norm": 2.002328872680664,
|
| 63 |
+
"learning_rate": 2.5316455696202533e-05,
|
| 64 |
+
"loss": 0.0477,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.28,
|
| 69 |
+
"grad_norm": 0.20249253511428833,
|
| 70 |
+
"learning_rate": 2.848101265822785e-05,
|
| 71 |
+
"loss": 0.0207,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.32,
|
| 76 |
+
"grad_norm": 0.15333689749240875,
|
| 77 |
+
"learning_rate": 3.1645569620253167e-05,
|
| 78 |
+
"loss": 0.0164,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.35,
|
| 83 |
+
"grad_norm": 0.08237770944833755,
|
| 84 |
+
"learning_rate": 3.4810126582278487e-05,
|
| 85 |
+
"loss": 0.0419,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.38,
|
| 90 |
+
"grad_norm": 0.6087208986282349,
|
| 91 |
+
"learning_rate": 3.79746835443038e-05,
|
| 92 |
+
"loss": 0.0081,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.41,
|
| 97 |
+
"grad_norm": 0.059435680508613586,
|
| 98 |
+
"learning_rate": 4.113924050632912e-05,
|
| 99 |
+
"loss": 0.007,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.44,
|
| 104 |
+
"grad_norm": 0.08203254640102386,
|
| 105 |
+
"learning_rate": 4.430379746835443e-05,
|
| 106 |
+
"loss": 0.022,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.47,
|
| 111 |
+
"grad_norm": 0.059913892298936844,
|
| 112 |
+
"learning_rate": 4.7468354430379746e-05,
|
| 113 |
+
"loss": 0.02,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.51,
|
| 118 |
+
"grad_norm": 0.04219400882720947,
|
| 119 |
+
"learning_rate": 4.9929676511955e-05,
|
| 120 |
+
"loss": 0.0241,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.54,
|
| 125 |
+
"grad_norm": 0.4119158089160919,
|
| 126 |
+
"learning_rate": 4.957805907172996e-05,
|
| 127 |
+
"loss": 0.011,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.57,
|
| 132 |
+
"grad_norm": 0.08845602720975876,
|
| 133 |
+
"learning_rate": 4.9226441631504925e-05,
|
| 134 |
+
"loss": 0.012,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.6,
|
| 139 |
+
"grad_norm": 0.022793015465140343,
|
| 140 |
+
"learning_rate": 4.887482419127989e-05,
|
| 141 |
+
"loss": 0.0044,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.63,
|
| 146 |
+
"grad_norm": 0.1266743242740631,
|
| 147 |
+
"learning_rate": 4.852320675105486e-05,
|
| 148 |
+
"loss": 0.0026,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.66,
|
| 153 |
+
"grad_norm": 0.2308553010225296,
|
| 154 |
+
"learning_rate": 4.817158931082982e-05,
|
| 155 |
+
"loss": 0.0088,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.7,
|
| 160 |
+
"grad_norm": 0.2685852348804474,
|
| 161 |
+
"learning_rate": 4.7819971870604783e-05,
|
| 162 |
+
"loss": 0.0191,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.73,
|
| 167 |
+
"grad_norm": 0.012726900167763233,
|
| 168 |
+
"learning_rate": 4.7468354430379746e-05,
|
| 169 |
+
"loss": 0.0084,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.76,
|
| 174 |
+
"grad_norm": 0.30946388840675354,
|
| 175 |
+
"learning_rate": 4.7116736990154716e-05,
|
| 176 |
+
"loss": 0.0242,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.79,
|
| 181 |
+
"grad_norm": 0.03251117840409279,
|
| 182 |
+
"learning_rate": 4.676511954992968e-05,
|
| 183 |
+
"loss": 0.0028,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.82,
|
| 188 |
+
"grad_norm": 0.622480034828186,
|
| 189 |
+
"learning_rate": 4.641350210970464e-05,
|
| 190 |
+
"loss": 0.0123,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.85,
|
| 195 |
+
"grad_norm": 0.013131936080753803,
|
| 196 |
+
"learning_rate": 4.606188466947961e-05,
|
| 197 |
+
"loss": 0.0013,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.89,
|
| 202 |
+
"grad_norm": 0.042751457542181015,
|
| 203 |
+
"learning_rate": 4.5710267229254575e-05,
|
| 204 |
+
"loss": 0.0403,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.92,
|
| 209 |
+
"grad_norm": 0.023626720532774925,
|
| 210 |
+
"learning_rate": 4.535864978902954e-05,
|
| 211 |
+
"loss": 0.0376,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.95,
|
| 216 |
+
"grad_norm": 0.03286024183034897,
|
| 217 |
+
"learning_rate": 4.50070323488045e-05,
|
| 218 |
+
"loss": 0.0095,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.98,
|
| 223 |
+
"grad_norm": 0.19246149063110352,
|
| 224 |
+
"learning_rate": 4.465541490857947e-05,
|
| 225 |
+
"loss": 0.0462,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.0,
|
| 230 |
+
"eval_accuracy": 0.9936733886911823,
|
| 231 |
+
"eval_loss": 0.016152312979102135,
|
| 232 |
+
"eval_runtime": 16.8828,
|
| 233 |
+
"eval_samples_per_second": 149.798,
|
| 234 |
+
"eval_steps_per_second": 18.777,
|
| 235 |
+
"step": 316
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.01,
|
| 239 |
+
"grad_norm": 4.946556568145752,
|
| 240 |
+
"learning_rate": 4.430379746835443e-05,
|
| 241 |
+
"loss": 0.02,
|
| 242 |
+
"step": 320
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.04,
|
| 246 |
+
"grad_norm": 1.085129737854004,
|
| 247 |
+
"learning_rate": 4.3952180028129396e-05,
|
| 248 |
+
"loss": 0.0067,
|
| 249 |
+
"step": 330
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.08,
|
| 253 |
+
"grad_norm": 0.026508348062634468,
|
| 254 |
+
"learning_rate": 4.3600562587904366e-05,
|
| 255 |
+
"loss": 0.0151,
|
| 256 |
+
"step": 340
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 1.11,
|
| 260 |
+
"grad_norm": 0.13778889179229736,
|
| 261 |
+
"learning_rate": 4.324894514767933e-05,
|
| 262 |
+
"loss": 0.007,
|
| 263 |
+
"step": 350
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.14,
|
| 267 |
+
"grad_norm": 3.9214696884155273,
|
| 268 |
+
"learning_rate": 4.289732770745429e-05,
|
| 269 |
+
"loss": 0.0384,
|
| 270 |
+
"step": 360
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 1.17,
|
| 274 |
+
"grad_norm": 0.047678280621767044,
|
| 275 |
+
"learning_rate": 4.2545710267229255e-05,
|
| 276 |
+
"loss": 0.001,
|
| 277 |
+
"step": 370
|
| 278 |
+
},
|
| 279 |
+
{
|
| 280 |
+
"epoch": 1.2,
|
| 281 |
+
"grad_norm": 1.579052448272705,
|
| 282 |
+
"learning_rate": 4.2194092827004224e-05,
|
| 283 |
+
"loss": 0.0254,
|
| 284 |
+
"step": 380
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.23,
|
| 288 |
+
"grad_norm": 0.04058700054883957,
|
| 289 |
+
"learning_rate": 4.184247538677919e-05,
|
| 290 |
+
"loss": 0.0037,
|
| 291 |
+
"step": 390
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.26,
|
| 295 |
+
"grad_norm": 0.02530195750296116,
|
| 296 |
+
"learning_rate": 4.149085794655415e-05,
|
| 297 |
+
"loss": 0.0018,
|
| 298 |
+
"step": 400
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 1.3,
|
| 302 |
+
"grad_norm": 2.1537139415740967,
|
| 303 |
+
"learning_rate": 4.113924050632912e-05,
|
| 304 |
+
"loss": 0.0046,
|
| 305 |
+
"step": 410
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.33,
|
| 309 |
+
"grad_norm": 0.4535873234272003,
|
| 310 |
+
"learning_rate": 4.078762306610408e-05,
|
| 311 |
+
"loss": 0.006,
|
| 312 |
+
"step": 420
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.36,
|
| 316 |
+
"grad_norm": 0.050164151936769485,
|
| 317 |
+
"learning_rate": 4.0436005625879046e-05,
|
| 318 |
+
"loss": 0.0055,
|
| 319 |
+
"step": 430
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.39,
|
| 323 |
+
"grad_norm": 0.015650559216737747,
|
| 324 |
+
"learning_rate": 4.008438818565401e-05,
|
| 325 |
+
"loss": 0.0159,
|
| 326 |
+
"step": 440
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.42,
|
| 330 |
+
"grad_norm": 0.02410263940691948,
|
| 331 |
+
"learning_rate": 3.973277074542898e-05,
|
| 332 |
+
"loss": 0.0008,
|
| 333 |
+
"step": 450
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.45,
|
| 337 |
+
"grad_norm": 0.005263881757855415,
|
| 338 |
+
"learning_rate": 3.938115330520394e-05,
|
| 339 |
+
"loss": 0.004,
|
| 340 |
+
"step": 460
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.49,
|
| 344 |
+
"grad_norm": 0.005594250746071339,
|
| 345 |
+
"learning_rate": 3.9029535864978904e-05,
|
| 346 |
+
"loss": 0.0004,
|
| 347 |
+
"step": 470
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.52,
|
| 351 |
+
"grad_norm": 0.008686939254403114,
|
| 352 |
+
"learning_rate": 3.867791842475387e-05,
|
| 353 |
+
"loss": 0.0014,
|
| 354 |
+
"step": 480
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.55,
|
| 358 |
+
"grad_norm": 0.00374673493206501,
|
| 359 |
+
"learning_rate": 3.832630098452884e-05,
|
| 360 |
+
"loss": 0.0086,
|
| 361 |
+
"step": 490
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.58,
|
| 365 |
+
"grad_norm": 0.004833437968045473,
|
| 366 |
+
"learning_rate": 3.79746835443038e-05,
|
| 367 |
+
"loss": 0.0136,
|
| 368 |
+
"step": 500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.61,
|
| 372 |
+
"grad_norm": 2.200552463531494,
|
| 373 |
+
"learning_rate": 3.762306610407876e-05,
|
| 374 |
+
"loss": 0.0156,
|
| 375 |
+
"step": 510
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.64,
|
| 379 |
+
"grad_norm": 0.1714821308851242,
|
| 380 |
+
"learning_rate": 3.727144866385373e-05,
|
| 381 |
+
"loss": 0.0006,
|
| 382 |
+
"step": 520
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.68,
|
| 386 |
+
"grad_norm": 0.009503379464149475,
|
| 387 |
+
"learning_rate": 3.6919831223628695e-05,
|
| 388 |
+
"loss": 0.0142,
|
| 389 |
+
"step": 530
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.71,
|
| 393 |
+
"grad_norm": 0.004495191853493452,
|
| 394 |
+
"learning_rate": 3.656821378340366e-05,
|
| 395 |
+
"loss": 0.0119,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.74,
|
| 400 |
+
"grad_norm": 0.2889450192451477,
|
| 401 |
+
"learning_rate": 3.621659634317862e-05,
|
| 402 |
+
"loss": 0.0021,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.77,
|
| 407 |
+
"grad_norm": 0.04732218757271767,
|
| 408 |
+
"learning_rate": 3.586497890295359e-05,
|
| 409 |
+
"loss": 0.005,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.8,
|
| 414 |
+
"grad_norm": 0.710472047328949,
|
| 415 |
+
"learning_rate": 3.551336146272855e-05,
|
| 416 |
+
"loss": 0.0009,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.83,
|
| 421 |
+
"grad_norm": 0.025769829750061035,
|
| 422 |
+
"learning_rate": 3.516174402250352e-05,
|
| 423 |
+
"loss": 0.0013,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.87,
|
| 428 |
+
"grad_norm": 5.155148983001709,
|
| 429 |
+
"learning_rate": 3.4810126582278487e-05,
|
| 430 |
+
"loss": 0.016,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.9,
|
| 435 |
+
"grad_norm": 0.009949293918907642,
|
| 436 |
+
"learning_rate": 3.445850914205345e-05,
|
| 437 |
+
"loss": 0.0005,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.93,
|
| 442 |
+
"grad_norm": 0.004877807106822729,
|
| 443 |
+
"learning_rate": 3.410689170182841e-05,
|
| 444 |
+
"loss": 0.006,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.96,
|
| 449 |
+
"grad_norm": 3.116375684738159,
|
| 450 |
+
"learning_rate": 3.3755274261603375e-05,
|
| 451 |
+
"loss": 0.0076,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.99,
|
| 456 |
+
"grad_norm": 0.0024900627322494984,
|
| 457 |
+
"learning_rate": 3.3403656821378345e-05,
|
| 458 |
+
"loss": 0.0108,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.0,
|
| 463 |
+
"eval_accuracy": 0.9984183471727955,
|
| 464 |
+
"eval_loss": 0.004202807787805796,
|
| 465 |
+
"eval_runtime": 17.1708,
|
| 466 |
+
"eval_samples_per_second": 147.285,
|
| 467 |
+
"eval_steps_per_second": 18.462,
|
| 468 |
+
"step": 632
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 2.02,
|
| 472 |
+
"grad_norm": 0.027852624654769897,
|
| 473 |
+
"learning_rate": 3.305203938115331e-05,
|
| 474 |
+
"loss": 0.0126,
|
| 475 |
+
"step": 640
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 2.06,
|
| 479 |
+
"grad_norm": 0.00895751640200615,
|
| 480 |
+
"learning_rate": 3.270042194092827e-05,
|
| 481 |
+
"loss": 0.0247,
|
| 482 |
+
"step": 650
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.09,
|
| 486 |
+
"grad_norm": 0.005895175039768219,
|
| 487 |
+
"learning_rate": 3.234880450070324e-05,
|
| 488 |
+
"loss": 0.0031,
|
| 489 |
+
"step": 660
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.12,
|
| 493 |
+
"grad_norm": 0.0067118131555616856,
|
| 494 |
+
"learning_rate": 3.1997187060478204e-05,
|
| 495 |
+
"loss": 0.0123,
|
| 496 |
+
"step": 670
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 2.15,
|
| 500 |
+
"grad_norm": 3.036545991897583,
|
| 501 |
+
"learning_rate": 3.1645569620253167e-05,
|
| 502 |
+
"loss": 0.0139,
|
| 503 |
+
"step": 680
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 2.18,
|
| 507 |
+
"grad_norm": 0.013006511144340038,
|
| 508 |
+
"learning_rate": 3.129395218002813e-05,
|
| 509 |
+
"loss": 0.0007,
|
| 510 |
+
"step": 690
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 2.21,
|
| 514 |
+
"grad_norm": 0.011808563023805618,
|
| 515 |
+
"learning_rate": 3.09423347398031e-05,
|
| 516 |
+
"loss": 0.001,
|
| 517 |
+
"step": 700
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 2.25,
|
| 521 |
+
"grad_norm": 0.0029538481030613184,
|
| 522 |
+
"learning_rate": 3.059071729957806e-05,
|
| 523 |
+
"loss": 0.004,
|
| 524 |
+
"step": 710
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.28,
|
| 528 |
+
"grad_norm": 0.0031332324724644423,
|
| 529 |
+
"learning_rate": 3.0239099859353025e-05,
|
| 530 |
+
"loss": 0.0006,
|
| 531 |
+
"step": 720
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.31,
|
| 535 |
+
"grad_norm": 0.07193127274513245,
|
| 536 |
+
"learning_rate": 2.9887482419127988e-05,
|
| 537 |
+
"loss": 0.0134,
|
| 538 |
+
"step": 730
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 2.34,
|
| 542 |
+
"grad_norm": 0.009125332348048687,
|
| 543 |
+
"learning_rate": 2.9535864978902954e-05,
|
| 544 |
+
"loss": 0.0009,
|
| 545 |
+
"step": 740
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 2.37,
|
| 549 |
+
"grad_norm": 0.03563140332698822,
|
| 550 |
+
"learning_rate": 2.9184247538677924e-05,
|
| 551 |
+
"loss": 0.008,
|
| 552 |
+
"step": 750
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 2.4,
|
| 556 |
+
"grad_norm": 0.4033479690551758,
|
| 557 |
+
"learning_rate": 2.8832630098452884e-05,
|
| 558 |
+
"loss": 0.0006,
|
| 559 |
+
"step": 760
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 2.43,
|
| 563 |
+
"grad_norm": 0.005182890221476555,
|
| 564 |
+
"learning_rate": 2.848101265822785e-05,
|
| 565 |
+
"loss": 0.0003,
|
| 566 |
+
"step": 770
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.47,
|
| 570 |
+
"grad_norm": 0.06785815209150314,
|
| 571 |
+
"learning_rate": 2.8129395218002813e-05,
|
| 572 |
+
"loss": 0.0005,
|
| 573 |
+
"step": 780
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.5,
|
| 577 |
+
"grad_norm": 0.00515472237020731,
|
| 578 |
+
"learning_rate": 2.777777777777778e-05,
|
| 579 |
+
"loss": 0.0116,
|
| 580 |
+
"step": 790
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 2.53,
|
| 584 |
+
"grad_norm": 0.06535543501377106,
|
| 585 |
+
"learning_rate": 2.7426160337552742e-05,
|
| 586 |
+
"loss": 0.0002,
|
| 587 |
+
"step": 800
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 2.56,
|
| 591 |
+
"grad_norm": 0.2995426654815674,
|
| 592 |
+
"learning_rate": 2.707454289732771e-05,
|
| 593 |
+
"loss": 0.014,
|
| 594 |
+
"step": 810
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 2.59,
|
| 598 |
+
"grad_norm": 0.011195150204002857,
|
| 599 |
+
"learning_rate": 2.672292545710267e-05,
|
| 600 |
+
"loss": 0.0257,
|
| 601 |
+
"step": 820
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 2.62,
|
| 605 |
+
"grad_norm": 0.00852285698056221,
|
| 606 |
+
"learning_rate": 2.6371308016877638e-05,
|
| 607 |
+
"loss": 0.0249,
|
| 608 |
+
"step": 830
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.66,
|
| 612 |
+
"grad_norm": 0.3155508041381836,
|
| 613 |
+
"learning_rate": 2.6019690576652604e-05,
|
| 614 |
+
"loss": 0.0008,
|
| 615 |
+
"step": 840
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.69,
|
| 619 |
+
"grad_norm": 0.00613983441144228,
|
| 620 |
+
"learning_rate": 2.5668073136427567e-05,
|
| 621 |
+
"loss": 0.0024,
|
| 622 |
+
"step": 850
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 2.72,
|
| 626 |
+
"grad_norm": 0.18053768575191498,
|
| 627 |
+
"learning_rate": 2.5316455696202533e-05,
|
| 628 |
+
"loss": 0.0021,
|
| 629 |
+
"step": 860
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 2.75,
|
| 633 |
+
"grad_norm": 0.3204176127910614,
|
| 634 |
+
"learning_rate": 2.49648382559775e-05,
|
| 635 |
+
"loss": 0.0087,
|
| 636 |
+
"step": 870
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 2.78,
|
| 640 |
+
"grad_norm": 0.019261712208390236,
|
| 641 |
+
"learning_rate": 2.4613220815752462e-05,
|
| 642 |
+
"loss": 0.0004,
|
| 643 |
+
"step": 880
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 2.81,
|
| 647 |
+
"grad_norm": 0.04258158802986145,
|
| 648 |
+
"learning_rate": 2.426160337552743e-05,
|
| 649 |
+
"loss": 0.0157,
|
| 650 |
+
"step": 890
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.85,
|
| 654 |
+
"grad_norm": 0.8219675421714783,
|
| 655 |
+
"learning_rate": 2.3909985935302392e-05,
|
| 656 |
+
"loss": 0.0018,
|
| 657 |
+
"step": 900
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.88,
|
| 661 |
+
"grad_norm": 0.007423636503517628,
|
| 662 |
+
"learning_rate": 2.3558368495077358e-05,
|
| 663 |
+
"loss": 0.0004,
|
| 664 |
+
"step": 910
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 2.91,
|
| 668 |
+
"grad_norm": 0.039796918630599976,
|
| 669 |
+
"learning_rate": 2.320675105485232e-05,
|
| 670 |
+
"loss": 0.0005,
|
| 671 |
+
"step": 920
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 2.94,
|
| 675 |
+
"grad_norm": 0.007585118990391493,
|
| 676 |
+
"learning_rate": 2.2855133614627287e-05,
|
| 677 |
+
"loss": 0.0408,
|
| 678 |
+
"step": 930
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 2.97,
|
| 682 |
+
"grad_norm": 0.004360842052847147,
|
| 683 |
+
"learning_rate": 2.250351617440225e-05,
|
| 684 |
+
"loss": 0.0051,
|
| 685 |
+
"step": 940
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 3.0,
|
| 689 |
+
"eval_accuracy": 0.9972321075523922,
|
| 690 |
+
"eval_loss": 0.0071137435734272,
|
| 691 |
+
"eval_runtime": 16.9966,
|
| 692 |
+
"eval_samples_per_second": 148.794,
|
| 693 |
+
"eval_steps_per_second": 18.651,
|
| 694 |
+
"step": 948
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 3.0,
|
| 698 |
+
"grad_norm": 0.012101028114557266,
|
| 699 |
+
"learning_rate": 2.2151898734177217e-05,
|
| 700 |
+
"loss": 0.0003,
|
| 701 |
+
"step": 950
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 3.04,
|
| 705 |
+
"grad_norm": 0.021550431847572327,
|
| 706 |
+
"learning_rate": 2.1800281293952183e-05,
|
| 707 |
+
"loss": 0.0036,
|
| 708 |
+
"step": 960
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 3.07,
|
| 712 |
+
"grad_norm": 0.003491973737254739,
|
| 713 |
+
"learning_rate": 2.1448663853727146e-05,
|
| 714 |
+
"loss": 0.0002,
|
| 715 |
+
"step": 970
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 3.1,
|
| 719 |
+
"grad_norm": 0.0050508989952504635,
|
| 720 |
+
"learning_rate": 2.1097046413502112e-05,
|
| 721 |
+
"loss": 0.0007,
|
| 722 |
+
"step": 980
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 3.13,
|
| 726 |
+
"grad_norm": 0.004054068587720394,
|
| 727 |
+
"learning_rate": 2.0745428973277075e-05,
|
| 728 |
+
"loss": 0.0003,
|
| 729 |
+
"step": 990
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 3.16,
|
| 733 |
+
"grad_norm": 0.0022561594378203154,
|
| 734 |
+
"learning_rate": 2.039381153305204e-05,
|
| 735 |
+
"loss": 0.0007,
|
| 736 |
+
"step": 1000
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 3.19,
|
| 740 |
+
"grad_norm": 0.016699934378266335,
|
| 741 |
+
"learning_rate": 2.0042194092827004e-05,
|
| 742 |
+
"loss": 0.0052,
|
| 743 |
+
"step": 1010
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 3.23,
|
| 747 |
+
"grad_norm": 0.016947178170084953,
|
| 748 |
+
"learning_rate": 1.969057665260197e-05,
|
| 749 |
+
"loss": 0.0027,
|
| 750 |
+
"step": 1020
|
| 751 |
+
},
|
| 752 |
+
{
|
| 753 |
+
"epoch": 3.26,
|
| 754 |
+
"grad_norm": 0.0021668283734470606,
|
| 755 |
+
"learning_rate": 1.9338959212376934e-05,
|
| 756 |
+
"loss": 0.0085,
|
| 757 |
+
"step": 1030
|
| 758 |
+
},
|
| 759 |
+
{
|
| 760 |
+
"epoch": 3.29,
|
| 761 |
+
"grad_norm": 0.0019585881382226944,
|
| 762 |
+
"learning_rate": 1.89873417721519e-05,
|
| 763 |
+
"loss": 0.0003,
|
| 764 |
+
"step": 1040
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 3.32,
|
| 768 |
+
"grad_norm": 0.02020511031150818,
|
| 769 |
+
"learning_rate": 1.8635724331926866e-05,
|
| 770 |
+
"loss": 0.0084,
|
| 771 |
+
"step": 1050
|
| 772 |
+
},
|
| 773 |
+
{
|
| 774 |
+
"epoch": 3.35,
|
| 775 |
+
"grad_norm": 0.002657042583450675,
|
| 776 |
+
"learning_rate": 1.828410689170183e-05,
|
| 777 |
+
"loss": 0.0003,
|
| 778 |
+
"step": 1060
|
| 779 |
+
},
|
| 780 |
+
{
|
| 781 |
+
"epoch": 3.38,
|
| 782 |
+
"grad_norm": 0.002752606989815831,
|
| 783 |
+
"learning_rate": 1.7932489451476795e-05,
|
| 784 |
+
"loss": 0.0278,
|
| 785 |
+
"step": 1070
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 3.42,
|
| 789 |
+
"grad_norm": 0.4717991352081299,
|
| 790 |
+
"learning_rate": 1.758087201125176e-05,
|
| 791 |
+
"loss": 0.0004,
|
| 792 |
+
"step": 1080
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 3.45,
|
| 796 |
+
"grad_norm": 0.01203097589313984,
|
| 797 |
+
"learning_rate": 1.7229254571026725e-05,
|
| 798 |
+
"loss": 0.0271,
|
| 799 |
+
"step": 1090
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 3.48,
|
| 803 |
+
"grad_norm": 0.001858622650615871,
|
| 804 |
+
"learning_rate": 1.6877637130801688e-05,
|
| 805 |
+
"loss": 0.0341,
|
| 806 |
+
"step": 1100
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 3.51,
|
| 810 |
+
"grad_norm": 0.021344369277358055,
|
| 811 |
+
"learning_rate": 1.6526019690576654e-05,
|
| 812 |
+
"loss": 0.0198,
|
| 813 |
+
"step": 1110
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 3.54,
|
| 817 |
+
"grad_norm": 0.05389473959803581,
|
| 818 |
+
"learning_rate": 1.617440225035162e-05,
|
| 819 |
+
"loss": 0.0022,
|
| 820 |
+
"step": 1120
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 3.57,
|
| 824 |
+
"grad_norm": 0.0023719954770058393,
|
| 825 |
+
"learning_rate": 1.5822784810126583e-05,
|
| 826 |
+
"loss": 0.0069,
|
| 827 |
+
"step": 1130
|
| 828 |
+
},
|
| 829 |
+
{
|
| 830 |
+
"epoch": 3.6,
|
| 831 |
+
"grad_norm": 0.0039686476811766624,
|
| 832 |
+
"learning_rate": 1.547116736990155e-05,
|
| 833 |
+
"loss": 0.0007,
|
| 834 |
+
"step": 1140
|
| 835 |
+
},
|
| 836 |
+
{
|
| 837 |
+
"epoch": 3.64,
|
| 838 |
+
"grad_norm": 0.004471825901418924,
|
| 839 |
+
"learning_rate": 1.5119549929676513e-05,
|
| 840 |
+
"loss": 0.0003,
|
| 841 |
+
"step": 1150
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 3.67,
|
| 845 |
+
"grad_norm": 0.0025274273939430714,
|
| 846 |
+
"learning_rate": 1.4767932489451477e-05,
|
| 847 |
+
"loss": 0.0004,
|
| 848 |
+
"step": 1160
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 3.7,
|
| 852 |
+
"grad_norm": 0.005175269674509764,
|
| 853 |
+
"learning_rate": 1.4416315049226442e-05,
|
| 854 |
+
"loss": 0.0037,
|
| 855 |
+
"step": 1170
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"epoch": 3.73,
|
| 859 |
+
"grad_norm": 0.4415951073169708,
|
| 860 |
+
"learning_rate": 1.4064697609001406e-05,
|
| 861 |
+
"loss": 0.0022,
|
| 862 |
+
"step": 1180
|
| 863 |
+
},
|
| 864 |
+
{
|
| 865 |
+
"epoch": 3.76,
|
| 866 |
+
"grad_norm": 0.02205347828567028,
|
| 867 |
+
"learning_rate": 1.3713080168776371e-05,
|
| 868 |
+
"loss": 0.0049,
|
| 869 |
+
"step": 1190
|
| 870 |
+
},
|
| 871 |
+
{
|
| 872 |
+
"epoch": 3.79,
|
| 873 |
+
"grad_norm": 0.0073618339374661446,
|
| 874 |
+
"learning_rate": 1.3361462728551336e-05,
|
| 875 |
+
"loss": 0.0005,
|
| 876 |
+
"step": 1200
|
| 877 |
+
},
|
| 878 |
+
{
|
| 879 |
+
"epoch": 3.83,
|
| 880 |
+
"grad_norm": 0.01835988275706768,
|
| 881 |
+
"learning_rate": 1.3009845288326302e-05,
|
| 882 |
+
"loss": 0.0002,
|
| 883 |
+
"step": 1210
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 3.86,
|
| 887 |
+
"grad_norm": 0.006863409653306007,
|
| 888 |
+
"learning_rate": 1.2658227848101267e-05,
|
| 889 |
+
"loss": 0.0012,
|
| 890 |
+
"step": 1220
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 3.89,
|
| 894 |
+
"grad_norm": 0.009447610937058926,
|
| 895 |
+
"learning_rate": 1.2306610407876231e-05,
|
| 896 |
+
"loss": 0.0008,
|
| 897 |
+
"step": 1230
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 3.92,
|
| 901 |
+
"grad_norm": 0.0040683564729988575,
|
| 902 |
+
"learning_rate": 1.1954992967651196e-05,
|
| 903 |
+
"loss": 0.0003,
|
| 904 |
+
"step": 1240
|
| 905 |
+
},
|
| 906 |
+
{
|
| 907 |
+
"epoch": 3.95,
|
| 908 |
+
"grad_norm": 0.0015517990104854107,
|
| 909 |
+
"learning_rate": 1.160337552742616e-05,
|
| 910 |
+
"loss": 0.0004,
|
| 911 |
+
"step": 1250
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 3.98,
|
| 915 |
+
"grad_norm": 0.0038061882369220257,
|
| 916 |
+
"learning_rate": 1.1251758087201125e-05,
|
| 917 |
+
"loss": 0.0002,
|
| 918 |
+
"step": 1260
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 4.0,
|
| 922 |
+
"eval_accuracy": 0.9992091735863978,
|
| 923 |
+
"eval_loss": 0.0019393304828554392,
|
| 924 |
+
"eval_runtime": 17.2093,
|
| 925 |
+
"eval_samples_per_second": 146.955,
|
| 926 |
+
"eval_steps_per_second": 18.42,
|
| 927 |
+
"step": 1265
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 4.02,
|
| 931 |
+
"grad_norm": 0.021453727036714554,
|
| 932 |
+
"learning_rate": 1.0900140646976091e-05,
|
| 933 |
+
"loss": 0.0005,
|
| 934 |
+
"step": 1270
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 4.05,
|
| 938 |
+
"grad_norm": 0.003984913229942322,
|
| 939 |
+
"learning_rate": 1.0548523206751056e-05,
|
| 940 |
+
"loss": 0.0005,
|
| 941 |
+
"step": 1280
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 4.08,
|
| 945 |
+
"grad_norm": 0.00517477560788393,
|
| 946 |
+
"learning_rate": 1.019690576652602e-05,
|
| 947 |
+
"loss": 0.001,
|
| 948 |
+
"step": 1290
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 4.11,
|
| 952 |
+
"grad_norm": 0.002068225760012865,
|
| 953 |
+
"learning_rate": 9.845288326300985e-06,
|
| 954 |
+
"loss": 0.0002,
|
| 955 |
+
"step": 1300
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 4.14,
|
| 959 |
+
"grad_norm": 0.00740728760138154,
|
| 960 |
+
"learning_rate": 9.49367088607595e-06,
|
| 961 |
+
"loss": 0.0002,
|
| 962 |
+
"step": 1310
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 4.17,
|
| 966 |
+
"grad_norm": 0.0014793493319302797,
|
| 967 |
+
"learning_rate": 9.142053445850915e-06,
|
| 968 |
+
"loss": 0.0017,
|
| 969 |
+
"step": 1320
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 4.21,
|
| 973 |
+
"grad_norm": 0.0028481709305197,
|
| 974 |
+
"learning_rate": 8.79043600562588e-06,
|
| 975 |
+
"loss": 0.0011,
|
| 976 |
+
"step": 1330
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 4.24,
|
| 980 |
+
"grad_norm": 0.06655070930719376,
|
| 981 |
+
"learning_rate": 8.438818565400844e-06,
|
| 982 |
+
"loss": 0.0009,
|
| 983 |
+
"step": 1340
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 4.27,
|
| 987 |
+
"grad_norm": 0.012430869042873383,
|
| 988 |
+
"learning_rate": 8.08720112517581e-06,
|
| 989 |
+
"loss": 0.0014,
|
| 990 |
+
"step": 1350
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 4.3,
|
| 994 |
+
"grad_norm": 0.0027373009361326694,
|
| 995 |
+
"learning_rate": 7.735583684950775e-06,
|
| 996 |
+
"loss": 0.0067,
|
| 997 |
+
"step": 1360
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 4.33,
|
| 1001 |
+
"grad_norm": 0.0010446230880916119,
|
| 1002 |
+
"learning_rate": 7.3839662447257386e-06,
|
| 1003 |
+
"loss": 0.0001,
|
| 1004 |
+
"step": 1370
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 4.36,
|
| 1008 |
+
"grad_norm": 0.0015856565441936255,
|
| 1009 |
+
"learning_rate": 7.032348804500703e-06,
|
| 1010 |
+
"loss": 0.0002,
|
| 1011 |
+
"step": 1380
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 4.4,
|
| 1015 |
+
"grad_norm": 0.0012825512094423175,
|
| 1016 |
+
"learning_rate": 6.680731364275668e-06,
|
| 1017 |
+
"loss": 0.0003,
|
| 1018 |
+
"step": 1390
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 4.43,
|
| 1022 |
+
"grad_norm": 0.0009867001790553331,
|
| 1023 |
+
"learning_rate": 6.329113924050633e-06,
|
| 1024 |
+
"loss": 0.0002,
|
| 1025 |
+
"step": 1400
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 4.46,
|
| 1029 |
+
"grad_norm": 0.0017112062778323889,
|
| 1030 |
+
"learning_rate": 5.977496483825598e-06,
|
| 1031 |
+
"loss": 0.0003,
|
| 1032 |
+
"step": 1410
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 4.49,
|
| 1036 |
+
"grad_norm": 0.0032221453730016947,
|
| 1037 |
+
"learning_rate": 5.6258790436005626e-06,
|
| 1038 |
+
"loss": 0.0002,
|
| 1039 |
+
"step": 1420
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 4.52,
|
| 1043 |
+
"grad_norm": 0.003395942272618413,
|
| 1044 |
+
"learning_rate": 5.274261603375528e-06,
|
| 1045 |
+
"loss": 0.0001,
|
| 1046 |
+
"step": 1430
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 4.55,
|
| 1050 |
+
"grad_norm": 0.0007977127097547054,
|
| 1051 |
+
"learning_rate": 4.922644163150493e-06,
|
| 1052 |
+
"loss": 0.0132,
|
| 1053 |
+
"step": 1440
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 4.58,
|
| 1057 |
+
"grad_norm": 0.0035839290358126163,
|
| 1058 |
+
"learning_rate": 4.571026722925457e-06,
|
| 1059 |
+
"loss": 0.0001,
|
| 1060 |
+
"step": 1450
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 4.62,
|
| 1064 |
+
"grad_norm": 2.0375587940216064,
|
| 1065 |
+
"learning_rate": 4.219409282700422e-06,
|
| 1066 |
+
"loss": 0.0166,
|
| 1067 |
+
"step": 1460
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 4.65,
|
| 1071 |
+
"grad_norm": 0.005914665758609772,
|
| 1072 |
+
"learning_rate": 3.867791842475387e-06,
|
| 1073 |
+
"loss": 0.0001,
|
| 1074 |
+
"step": 1470
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 4.68,
|
| 1078 |
+
"grad_norm": 0.6946305632591248,
|
| 1079 |
+
"learning_rate": 3.5161744022503516e-06,
|
| 1080 |
+
"loss": 0.0011,
|
| 1081 |
+
"step": 1480
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 4.71,
|
| 1085 |
+
"grad_norm": 0.003537875832989812,
|
| 1086 |
+
"learning_rate": 3.1645569620253167e-06,
|
| 1087 |
+
"loss": 0.0129,
|
| 1088 |
+
"step": 1490
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 4.74,
|
| 1092 |
+
"grad_norm": 0.01320402417331934,
|
| 1093 |
+
"learning_rate": 2.8129395218002813e-06,
|
| 1094 |
+
"loss": 0.0092,
|
| 1095 |
+
"step": 1500
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 4.77,
|
| 1099 |
+
"grad_norm": 0.10570605099201202,
|
| 1100 |
+
"learning_rate": 2.4613220815752463e-06,
|
| 1101 |
+
"loss": 0.0008,
|
| 1102 |
+
"step": 1510
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 4.81,
|
| 1106 |
+
"grad_norm": 0.0033983599860221148,
|
| 1107 |
+
"learning_rate": 2.109704641350211e-06,
|
| 1108 |
+
"loss": 0.0002,
|
| 1109 |
+
"step": 1520
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 4.84,
|
| 1113 |
+
"grad_norm": 0.0025147399865090847,
|
| 1114 |
+
"learning_rate": 1.7580872011251758e-06,
|
| 1115 |
+
"loss": 0.0001,
|
| 1116 |
+
"step": 1530
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 4.87,
|
| 1120 |
+
"grad_norm": 0.004903899040073156,
|
| 1121 |
+
"learning_rate": 1.4064697609001406e-06,
|
| 1122 |
+
"loss": 0.016,
|
| 1123 |
+
"step": 1540
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 4.9,
|
| 1127 |
+
"grad_norm": 0.002249601762741804,
|
| 1128 |
+
"learning_rate": 1.0548523206751055e-06,
|
| 1129 |
+
"loss": 0.0024,
|
| 1130 |
+
"step": 1550
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 4.93,
|
| 1134 |
+
"grad_norm": 0.005820867605507374,
|
| 1135 |
+
"learning_rate": 7.032348804500703e-07,
|
| 1136 |
+
"loss": 0.0003,
|
| 1137 |
+
"step": 1560
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 4.96,
|
| 1141 |
+
"grad_norm": 0.46763402223587036,
|
| 1142 |
+
"learning_rate": 3.5161744022503516e-07,
|
| 1143 |
+
"loss": 0.0006,
|
| 1144 |
+
"step": 1570
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 5.0,
|
| 1148 |
+
"grad_norm": 1.628798246383667,
|
| 1149 |
+
"learning_rate": 0.0,
|
| 1150 |
+
"loss": 0.0019,
|
| 1151 |
+
"step": 1580
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 5.0,
|
| 1155 |
+
"eval_accuracy": 0.9992091735863978,
|
| 1156 |
+
"eval_loss": 0.0018690477591007948,
|
| 1157 |
+
"eval_runtime": 17.1455,
|
| 1158 |
+
"eval_samples_per_second": 147.502,
|
| 1159 |
+
"eval_steps_per_second": 18.489,
|
| 1160 |
+
"step": 1580
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 5.0,
|
| 1164 |
+
"step": 1580,
|
| 1165 |
+
"total_flos": 3.915825219204563e+18,
|
| 1166 |
+
"train_loss": 0.0232889701421461,
|
| 1167 |
+
"train_runtime": 841.0378,
|
| 1168 |
+
"train_samples_per_second": 60.122,
|
| 1169 |
+
"train_steps_per_second": 1.879
|
| 1170 |
+
}
|
| 1171 |
+
],
|
| 1172 |
+
"logging_steps": 10,
|
| 1173 |
+
"max_steps": 1580,
|
| 1174 |
+
"num_input_tokens_seen": 0,
|
| 1175 |
+
"num_train_epochs": 5,
|
| 1176 |
+
"save_steps": 500,
|
| 1177 |
+
"total_flos": 3.915825219204563e+18,
|
| 1178 |
+
"train_batch_size": 8,
|
| 1179 |
+
"trial_name": null,
|
| 1180 |
+
"trial_params": null
|
| 1181 |
+
}
|