Commit
•
f10985f
1
Parent(s):
77a05f7
End of training
Browse files- all_results.json +13 -0
- eval_results.json +8 -0
- runs/Oct26_04-05-49_4c56d22e99c3/events.out.tfevents.1729923978.4c56d22e99c3.214.1 +3 -0
- train_results.json +8 -0
- trainer_state.json +1182 -0
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.995776630689817,
|
| 3 |
+
"eval_accuracy": 0.834983498349835,
|
| 4 |
+
"eval_loss": 0.6400949358940125,
|
| 5 |
+
"eval_runtime": 121.5182,
|
| 6 |
+
"eval_samples_per_second": 62.336,
|
| 7 |
+
"eval_steps_per_second": 1.95,
|
| 8 |
+
"total_flos": 1.584209316447959e+19,
|
| 9 |
+
"train_loss": 1.1699181137825911,
|
| 10 |
+
"train_runtime": 8274.5103,
|
| 11 |
+
"train_samples_per_second": 24.717,
|
| 12 |
+
"train_steps_per_second": 0.193
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.995776630689817,
|
| 3 |
+
"eval_accuracy": 0.834983498349835,
|
| 4 |
+
"eval_loss": 0.6400949358940125,
|
| 5 |
+
"eval_runtime": 121.5182,
|
| 6 |
+
"eval_samples_per_second": 62.336,
|
| 7 |
+
"eval_steps_per_second": 1.95
|
| 8 |
+
}
|
runs/Oct26_04-05-49_4c56d22e99c3/events.out.tfevents.1729923978.4c56d22e99c3.214.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f36dbd3d84a6965d70511f0887ea0848067d8d3a6d0b9530eac04b0cc7211425
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.995776630689817,
|
| 3 |
+
"total_flos": 1.584209316447959e+19,
|
| 4 |
+
"train_loss": 1.1699181137825911,
|
| 5 |
+
"train_runtime": 8274.5103,
|
| 6 |
+
"train_samples_per_second": 24.717,
|
| 7 |
+
"train_steps_per_second": 0.193
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1182 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.834983498349835,
|
| 3 |
+
"best_model_checkpoint": "vit-base-patch16-224-finetuned-food101/checkpoint-1596",
|
| 4 |
+
"epoch": 2.995776630689817,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1596,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.018770530267480056,
|
| 13 |
+
"grad_norm": 2.9933698177337646,
|
| 14 |
+
"learning_rate": 3.125e-06,
|
| 15 |
+
"loss": 4.739,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.03754106053496011,
|
| 20 |
+
"grad_norm": 3.0962929725646973,
|
| 21 |
+
"learning_rate": 6.25e-06,
|
| 22 |
+
"loss": 4.7159,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05631159080244017,
|
| 27 |
+
"grad_norm": 3.2671728134155273,
|
| 28 |
+
"learning_rate": 9.375000000000001e-06,
|
| 29 |
+
"loss": 4.6674,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07508212106992022,
|
| 34 |
+
"grad_norm": 3.073523759841919,
|
| 35 |
+
"learning_rate": 1.25e-05,
|
| 36 |
+
"loss": 4.6293,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09385265133740028,
|
| 41 |
+
"grad_norm": 3.1302390098571777,
|
| 42 |
+
"learning_rate": 1.5625e-05,
|
| 43 |
+
"loss": 4.5709,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.11262318160488034,
|
| 48 |
+
"grad_norm": 2.9043614864349365,
|
| 49 |
+
"learning_rate": 1.8750000000000002e-05,
|
| 50 |
+
"loss": 4.4634,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.1313937118723604,
|
| 55 |
+
"grad_norm": 3.4175333976745605,
|
| 56 |
+
"learning_rate": 2.1875e-05,
|
| 57 |
+
"loss": 4.3879,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.15016424213984045,
|
| 62 |
+
"grad_norm": 3.323888063430786,
|
| 63 |
+
"learning_rate": 2.5e-05,
|
| 64 |
+
"loss": 4.2587,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.1689347724073205,
|
| 69 |
+
"grad_norm": 3.0213475227355957,
|
| 70 |
+
"learning_rate": 2.8125000000000003e-05,
|
| 71 |
+
"loss": 4.0885,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.18770530267480057,
|
| 76 |
+
"grad_norm": 3.02178955078125,
|
| 77 |
+
"learning_rate": 3.125e-05,
|
| 78 |
+
"loss": 3.9028,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.2064758329422806,
|
| 83 |
+
"grad_norm": 3.3495538234710693,
|
| 84 |
+
"learning_rate": 3.4375e-05,
|
| 85 |
+
"loss": 3.7353,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.22524636320976069,
|
| 90 |
+
"grad_norm": 3.1515848636627197,
|
| 91 |
+
"learning_rate": 3.7500000000000003e-05,
|
| 92 |
+
"loss": 3.4697,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.24401689347724073,
|
| 97 |
+
"grad_norm": 3.284679889678955,
|
| 98 |
+
"learning_rate": 4.0625000000000005e-05,
|
| 99 |
+
"loss": 3.2815,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.2627874237447208,
|
| 104 |
+
"grad_norm": 3.226576805114746,
|
| 105 |
+
"learning_rate": 4.375e-05,
|
| 106 |
+
"loss": 2.9895,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.28155795401220085,
|
| 111 |
+
"grad_norm": 3.181521415710449,
|
| 112 |
+
"learning_rate": 4.6875e-05,
|
| 113 |
+
"loss": 2.7421,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.3003284842796809,
|
| 118 |
+
"grad_norm": 3.1563026905059814,
|
| 119 |
+
"learning_rate": 5e-05,
|
| 120 |
+
"loss": 2.587,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.31909901454716094,
|
| 125 |
+
"grad_norm": 3.2322466373443604,
|
| 126 |
+
"learning_rate": 4.965181058495822e-05,
|
| 127 |
+
"loss": 2.3671,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.337869544814641,
|
| 132 |
+
"grad_norm": 3.2258055210113525,
|
| 133 |
+
"learning_rate": 4.930362116991643e-05,
|
| 134 |
+
"loss": 2.1425,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3566400750821211,
|
| 139 |
+
"grad_norm": 3.141265392303467,
|
| 140 |
+
"learning_rate": 4.895543175487465e-05,
|
| 141 |
+
"loss": 2.0185,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.37541060534960113,
|
| 146 |
+
"grad_norm": 3.157705307006836,
|
| 147 |
+
"learning_rate": 4.860724233983287e-05,
|
| 148 |
+
"loss": 1.9035,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.3941811356170812,
|
| 153 |
+
"grad_norm": 3.300532102584839,
|
| 154 |
+
"learning_rate": 4.825905292479109e-05,
|
| 155 |
+
"loss": 1.8364,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.4129516658845612,
|
| 160 |
+
"grad_norm": 3.4612457752227783,
|
| 161 |
+
"learning_rate": 4.79108635097493e-05,
|
| 162 |
+
"loss": 1.693,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.43172219615204127,
|
| 167 |
+
"grad_norm": 3.306436777114868,
|
| 168 |
+
"learning_rate": 4.756267409470752e-05,
|
| 169 |
+
"loss": 1.5683,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.45049272641952137,
|
| 174 |
+
"grad_norm": 3.522141695022583,
|
| 175 |
+
"learning_rate": 4.721448467966574e-05,
|
| 176 |
+
"loss": 1.5139,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.4692632566870014,
|
| 181 |
+
"grad_norm": 3.511997938156128,
|
| 182 |
+
"learning_rate": 4.686629526462396e-05,
|
| 183 |
+
"loss": 1.4348,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.48803378695448146,
|
| 188 |
+
"grad_norm": 3.343367576599121,
|
| 189 |
+
"learning_rate": 4.6518105849582176e-05,
|
| 190 |
+
"loss": 1.3879,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.5068043172219615,
|
| 195 |
+
"grad_norm": 3.2004270553588867,
|
| 196 |
+
"learning_rate": 4.6169916434540394e-05,
|
| 197 |
+
"loss": 1.3787,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5255748474894416,
|
| 202 |
+
"grad_norm": 3.113529682159424,
|
| 203 |
+
"learning_rate": 4.582172701949861e-05,
|
| 204 |
+
"loss": 1.3503,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.5443453777569216,
|
| 209 |
+
"grad_norm": 3.2955970764160156,
|
| 210 |
+
"learning_rate": 4.547353760445683e-05,
|
| 211 |
+
"loss": 1.3139,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.5631159080244017,
|
| 216 |
+
"grad_norm": 3.1911182403564453,
|
| 217 |
+
"learning_rate": 4.5125348189415044e-05,
|
| 218 |
+
"loss": 1.2995,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.5818864382918817,
|
| 223 |
+
"grad_norm": 3.416416645050049,
|
| 224 |
+
"learning_rate": 4.477715877437326e-05,
|
| 225 |
+
"loss": 1.2014,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.6006569685593618,
|
| 230 |
+
"grad_norm": 4.6215410232543945,
|
| 231 |
+
"learning_rate": 4.442896935933148e-05,
|
| 232 |
+
"loss": 1.1872,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.6194274988268419,
|
| 237 |
+
"grad_norm": 3.202340841293335,
|
| 238 |
+
"learning_rate": 4.40807799442897e-05,
|
| 239 |
+
"loss": 1.1448,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.6381980290943219,
|
| 244 |
+
"grad_norm": 3.1017255783081055,
|
| 245 |
+
"learning_rate": 4.373259052924791e-05,
|
| 246 |
+
"loss": 1.2542,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.656968559361802,
|
| 251 |
+
"grad_norm": 3.4146809577941895,
|
| 252 |
+
"learning_rate": 4.338440111420613e-05,
|
| 253 |
+
"loss": 1.143,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.675739089629282,
|
| 258 |
+
"grad_norm": 3.4860126972198486,
|
| 259 |
+
"learning_rate": 4.303621169916435e-05,
|
| 260 |
+
"loss": 1.1936,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.6945096198967621,
|
| 265 |
+
"grad_norm": 3.000502109527588,
|
| 266 |
+
"learning_rate": 4.268802228412256e-05,
|
| 267 |
+
"loss": 1.0896,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.7132801501642422,
|
| 272 |
+
"grad_norm": 3.3682737350463867,
|
| 273 |
+
"learning_rate": 4.233983286908078e-05,
|
| 274 |
+
"loss": 0.9863,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.7320506804317222,
|
| 279 |
+
"grad_norm": 2.903545379638672,
|
| 280 |
+
"learning_rate": 4.1991643454039e-05,
|
| 281 |
+
"loss": 1.0206,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.7508212106992023,
|
| 286 |
+
"grad_norm": 3.5237250328063965,
|
| 287 |
+
"learning_rate": 4.164345403899722e-05,
|
| 288 |
+
"loss": 1.0655,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.7695917409666823,
|
| 293 |
+
"grad_norm": 3.5761749744415283,
|
| 294 |
+
"learning_rate": 4.129526462395543e-05,
|
| 295 |
+
"loss": 1.0016,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.7883622712341624,
|
| 300 |
+
"grad_norm": 3.3111650943756104,
|
| 301 |
+
"learning_rate": 4.094707520891365e-05,
|
| 302 |
+
"loss": 1.0225,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.8071328015016425,
|
| 307 |
+
"grad_norm": 2.9664759635925293,
|
| 308 |
+
"learning_rate": 4.0598885793871866e-05,
|
| 309 |
+
"loss": 1.0624,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.8259033317691225,
|
| 314 |
+
"grad_norm": 3.750908136367798,
|
| 315 |
+
"learning_rate": 4.0250696378830085e-05,
|
| 316 |
+
"loss": 1.0223,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8446738620366026,
|
| 321 |
+
"grad_norm": 3.176248788833618,
|
| 322 |
+
"learning_rate": 3.9902506963788303e-05,
|
| 323 |
+
"loss": 0.9834,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.8634443923040825,
|
| 328 |
+
"grad_norm": 3.428868293762207,
|
| 329 |
+
"learning_rate": 3.955431754874652e-05,
|
| 330 |
+
"loss": 1.0338,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.8822149225715626,
|
| 335 |
+
"grad_norm": 3.6257553100585938,
|
| 336 |
+
"learning_rate": 3.920612813370474e-05,
|
| 337 |
+
"loss": 1.0156,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.9009854528390427,
|
| 342 |
+
"grad_norm": 3.195758104324341,
|
| 343 |
+
"learning_rate": 3.885793871866296e-05,
|
| 344 |
+
"loss": 1.0171,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.9197559831065227,
|
| 349 |
+
"grad_norm": 2.6419994831085205,
|
| 350 |
+
"learning_rate": 3.850974930362117e-05,
|
| 351 |
+
"loss": 0.9735,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.9385265133740028,
|
| 356 |
+
"grad_norm": 3.462597608566284,
|
| 357 |
+
"learning_rate": 3.816155988857939e-05,
|
| 358 |
+
"loss": 1.0049,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.9572970436414828,
|
| 363 |
+
"grad_norm": 2.8139870166778564,
|
| 364 |
+
"learning_rate": 3.781337047353761e-05,
|
| 365 |
+
"loss": 0.9696,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.9760675739089629,
|
| 370 |
+
"grad_norm": 3.027811288833618,
|
| 371 |
+
"learning_rate": 3.746518105849583e-05,
|
| 372 |
+
"loss": 0.9242,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.994838104176443,
|
| 377 |
+
"grad_norm": 3.6760294437408447,
|
| 378 |
+
"learning_rate": 3.711699164345404e-05,
|
| 379 |
+
"loss": 0.912,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.998592210229939,
|
| 384 |
+
"eval_accuracy": 0.7968316831683169,
|
| 385 |
+
"eval_loss": 0.8397366404533386,
|
| 386 |
+
"eval_runtime": 120.4032,
|
| 387 |
+
"eval_samples_per_second": 62.914,
|
| 388 |
+
"eval_steps_per_second": 1.968,
|
| 389 |
+
"step": 532
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.013608634443923,
|
| 393 |
+
"grad_norm": 2.9862356185913086,
|
| 394 |
+
"learning_rate": 3.676880222841226e-05,
|
| 395 |
+
"loss": 0.8398,
|
| 396 |
+
"step": 540
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.0323791647114031,
|
| 400 |
+
"grad_norm": 3.0735647678375244,
|
| 401 |
+
"learning_rate": 3.642061281337048e-05,
|
| 402 |
+
"loss": 0.8379,
|
| 403 |
+
"step": 550
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.0511496949788832,
|
| 407 |
+
"grad_norm": 3.5392494201660156,
|
| 408 |
+
"learning_rate": 3.607242339832869e-05,
|
| 409 |
+
"loss": 0.8264,
|
| 410 |
+
"step": 560
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.069920225246363,
|
| 414 |
+
"grad_norm": 3.511627435684204,
|
| 415 |
+
"learning_rate": 3.572423398328691e-05,
|
| 416 |
+
"loss": 0.8204,
|
| 417 |
+
"step": 570
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.0886907555138432,
|
| 421 |
+
"grad_norm": 3.393699884414673,
|
| 422 |
+
"learning_rate": 3.5376044568245126e-05,
|
| 423 |
+
"loss": 0.8963,
|
| 424 |
+
"step": 580
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.1074612857813233,
|
| 428 |
+
"grad_norm": 4.332097053527832,
|
| 429 |
+
"learning_rate": 3.5027855153203345e-05,
|
| 430 |
+
"loss": 0.8623,
|
| 431 |
+
"step": 590
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.1262318160488034,
|
| 435 |
+
"grad_norm": 3.1262927055358887,
|
| 436 |
+
"learning_rate": 3.4679665738161556e-05,
|
| 437 |
+
"loss": 0.8025,
|
| 438 |
+
"step": 600
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.1450023463162835,
|
| 442 |
+
"grad_norm": 3.1130638122558594,
|
| 443 |
+
"learning_rate": 3.4331476323119775e-05,
|
| 444 |
+
"loss": 0.8388,
|
| 445 |
+
"step": 610
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.1637728765837636,
|
| 449 |
+
"grad_norm": 3.509665012359619,
|
| 450 |
+
"learning_rate": 3.3983286908077994e-05,
|
| 451 |
+
"loss": 0.8306,
|
| 452 |
+
"step": 620
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.1825434068512435,
|
| 456 |
+
"grad_norm": 3.595564842224121,
|
| 457 |
+
"learning_rate": 3.363509749303621e-05,
|
| 458 |
+
"loss": 0.8241,
|
| 459 |
+
"step": 630
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.2013139371187236,
|
| 463 |
+
"grad_norm": 2.9736037254333496,
|
| 464 |
+
"learning_rate": 3.328690807799443e-05,
|
| 465 |
+
"loss": 0.8333,
|
| 466 |
+
"step": 640
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.2200844673862037,
|
| 470 |
+
"grad_norm": 3.5477538108825684,
|
| 471 |
+
"learning_rate": 3.293871866295265e-05,
|
| 472 |
+
"loss": 0.84,
|
| 473 |
+
"step": 650
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.2388549976536838,
|
| 477 |
+
"grad_norm": 3.617574453353882,
|
| 478 |
+
"learning_rate": 3.259052924791087e-05,
|
| 479 |
+
"loss": 0.7652,
|
| 480 |
+
"step": 660
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.2576255279211637,
|
| 484 |
+
"grad_norm": 3.621431589126587,
|
| 485 |
+
"learning_rate": 3.224233983286909e-05,
|
| 486 |
+
"loss": 0.7634,
|
| 487 |
+
"step": 670
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.2763960581886438,
|
| 491 |
+
"grad_norm": 3.5540874004364014,
|
| 492 |
+
"learning_rate": 3.18941504178273e-05,
|
| 493 |
+
"loss": 0.7951,
|
| 494 |
+
"step": 680
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.2951665884561239,
|
| 498 |
+
"grad_norm": 3.1876299381256104,
|
| 499 |
+
"learning_rate": 3.154596100278552e-05,
|
| 500 |
+
"loss": 0.787,
|
| 501 |
+
"step": 690
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.313937118723604,
|
| 505 |
+
"grad_norm": 3.8441293239593506,
|
| 506 |
+
"learning_rate": 3.1197771587743737e-05,
|
| 507 |
+
"loss": 0.7799,
|
| 508 |
+
"step": 700
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.332707648991084,
|
| 512 |
+
"grad_norm": 3.2278223037719727,
|
| 513 |
+
"learning_rate": 3.0849582172701955e-05,
|
| 514 |
+
"loss": 0.845,
|
| 515 |
+
"step": 710
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.3514781792585642,
|
| 519 |
+
"grad_norm": 3.1881773471832275,
|
| 520 |
+
"learning_rate": 3.050139275766017e-05,
|
| 521 |
+
"loss": 0.7415,
|
| 522 |
+
"step": 720
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.370248709526044,
|
| 526 |
+
"grad_norm": 3.34291672706604,
|
| 527 |
+
"learning_rate": 3.0153203342618386e-05,
|
| 528 |
+
"loss": 0.7424,
|
| 529 |
+
"step": 730
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.3890192397935242,
|
| 533 |
+
"grad_norm": 5.416168689727783,
|
| 534 |
+
"learning_rate": 2.98050139275766e-05,
|
| 535 |
+
"loss": 0.8025,
|
| 536 |
+
"step": 740
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.4077897700610043,
|
| 540 |
+
"grad_norm": 3.111943244934082,
|
| 541 |
+
"learning_rate": 2.945682451253482e-05,
|
| 542 |
+
"loss": 0.8537,
|
| 543 |
+
"step": 750
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.4265603003284844,
|
| 547 |
+
"grad_norm": 3.118394374847412,
|
| 548 |
+
"learning_rate": 2.9108635097493035e-05,
|
| 549 |
+
"loss": 0.7796,
|
| 550 |
+
"step": 760
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.4453308305959642,
|
| 554 |
+
"grad_norm": 3.8577425479888916,
|
| 555 |
+
"learning_rate": 2.8760445682451254e-05,
|
| 556 |
+
"loss": 0.7926,
|
| 557 |
+
"step": 770
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.4641013608634443,
|
| 561 |
+
"grad_norm": 4.177425384521484,
|
| 562 |
+
"learning_rate": 2.841225626740947e-05,
|
| 563 |
+
"loss": 0.7863,
|
| 564 |
+
"step": 780
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.4828718911309244,
|
| 568 |
+
"grad_norm": 3.465301275253296,
|
| 569 |
+
"learning_rate": 2.8064066852367688e-05,
|
| 570 |
+
"loss": 0.7687,
|
| 571 |
+
"step": 790
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.5016424213984045,
|
| 575 |
+
"grad_norm": 3.2810630798339844,
|
| 576 |
+
"learning_rate": 2.7715877437325903e-05,
|
| 577 |
+
"loss": 0.8167,
|
| 578 |
+
"step": 800
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.5204129516658846,
|
| 582 |
+
"grad_norm": 3.05501127243042,
|
| 583 |
+
"learning_rate": 2.736768802228412e-05,
|
| 584 |
+
"loss": 0.7299,
|
| 585 |
+
"step": 810
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.5391834819333647,
|
| 589 |
+
"grad_norm": 2.9829261302948,
|
| 590 |
+
"learning_rate": 2.7019498607242337e-05,
|
| 591 |
+
"loss": 0.7677,
|
| 592 |
+
"step": 820
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.5579540122008448,
|
| 596 |
+
"grad_norm": 3.561396360397339,
|
| 597 |
+
"learning_rate": 2.6671309192200562e-05,
|
| 598 |
+
"loss": 0.7813,
|
| 599 |
+
"step": 830
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.5767245424683247,
|
| 603 |
+
"grad_norm": 2.8982436656951904,
|
| 604 |
+
"learning_rate": 2.6323119777158778e-05,
|
| 605 |
+
"loss": 0.731,
|
| 606 |
+
"step": 840
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 1.5954950727358048,
|
| 610 |
+
"grad_norm": 3.406423330307007,
|
| 611 |
+
"learning_rate": 2.5974930362116996e-05,
|
| 612 |
+
"loss": 0.73,
|
| 613 |
+
"step": 850
|
| 614 |
+
},
|
| 615 |
+
{
|
| 616 |
+
"epoch": 1.6142656030032847,
|
| 617 |
+
"grad_norm": 3.0049679279327393,
|
| 618 |
+
"learning_rate": 2.562674094707521e-05,
|
| 619 |
+
"loss": 0.7826,
|
| 620 |
+
"step": 860
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 1.6330361332707648,
|
| 624 |
+
"grad_norm": 3.500458002090454,
|
| 625 |
+
"learning_rate": 2.527855153203343e-05,
|
| 626 |
+
"loss": 0.7499,
|
| 627 |
+
"step": 870
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 1.651806663538245,
|
| 631 |
+
"grad_norm": 3.0891637802124023,
|
| 632 |
+
"learning_rate": 2.4930362116991646e-05,
|
| 633 |
+
"loss": 0.7587,
|
| 634 |
+
"step": 880
|
| 635 |
+
},
|
| 636 |
+
{
|
| 637 |
+
"epoch": 1.670577193805725,
|
| 638 |
+
"grad_norm": 3.149742841720581,
|
| 639 |
+
"learning_rate": 2.4582172701949864e-05,
|
| 640 |
+
"loss": 0.7374,
|
| 641 |
+
"step": 890
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 1.689347724073205,
|
| 645 |
+
"grad_norm": 3.1518661975860596,
|
| 646 |
+
"learning_rate": 2.423398328690808e-05,
|
| 647 |
+
"loss": 0.7893,
|
| 648 |
+
"step": 900
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 1.7081182543406852,
|
| 652 |
+
"grad_norm": 3.08024525642395,
|
| 653 |
+
"learning_rate": 2.3885793871866298e-05,
|
| 654 |
+
"loss": 0.7853,
|
| 655 |
+
"step": 910
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 1.7268887846081653,
|
| 659 |
+
"grad_norm": 3.158390522003174,
|
| 660 |
+
"learning_rate": 2.3537604456824514e-05,
|
| 661 |
+
"loss": 0.7694,
|
| 662 |
+
"step": 920
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 1.7456593148756452,
|
| 666 |
+
"grad_norm": 3.0718069076538086,
|
| 667 |
+
"learning_rate": 2.318941504178273e-05,
|
| 668 |
+
"loss": 0.7843,
|
| 669 |
+
"step": 930
|
| 670 |
+
},
|
| 671 |
+
{
|
| 672 |
+
"epoch": 1.7644298451431253,
|
| 673 |
+
"grad_norm": 3.4767684936523438,
|
| 674 |
+
"learning_rate": 2.2841225626740948e-05,
|
| 675 |
+
"loss": 0.7483,
|
| 676 |
+
"step": 940
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 1.7832003754106054,
|
| 680 |
+
"grad_norm": 3.350541353225708,
|
| 681 |
+
"learning_rate": 2.2493036211699163e-05,
|
| 682 |
+
"loss": 0.7457,
|
| 683 |
+
"step": 950
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 1.8019709056780853,
|
| 687 |
+
"grad_norm": 3.4785940647125244,
|
| 688 |
+
"learning_rate": 2.214484679665738e-05,
|
| 689 |
+
"loss": 0.7586,
|
| 690 |
+
"step": 960
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 1.8207414359455654,
|
| 694 |
+
"grad_norm": 3.245809316635132,
|
| 695 |
+
"learning_rate": 2.17966573816156e-05,
|
| 696 |
+
"loss": 0.7035,
|
| 697 |
+
"step": 970
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 1.8395119662130455,
|
| 701 |
+
"grad_norm": 3.292048692703247,
|
| 702 |
+
"learning_rate": 2.144846796657382e-05,
|
| 703 |
+
"loss": 0.7366,
|
| 704 |
+
"step": 980
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 1.8582824964805256,
|
| 708 |
+
"grad_norm": 4.539200305938721,
|
| 709 |
+
"learning_rate": 2.1100278551532034e-05,
|
| 710 |
+
"loss": 0.7377,
|
| 711 |
+
"step": 990
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 1.8770530267480057,
|
| 715 |
+
"grad_norm": 2.6051247119903564,
|
| 716 |
+
"learning_rate": 2.0752089136490253e-05,
|
| 717 |
+
"loss": 0.8442,
|
| 718 |
+
"step": 1000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 1.8958235570154858,
|
| 722 |
+
"grad_norm": 3.7407584190368652,
|
| 723 |
+
"learning_rate": 2.0403899721448468e-05,
|
| 724 |
+
"loss": 0.732,
|
| 725 |
+
"step": 1010
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 1.9145940872829659,
|
| 729 |
+
"grad_norm": 3.855624198913574,
|
| 730 |
+
"learning_rate": 2.0055710306406687e-05,
|
| 731 |
+
"loss": 0.8147,
|
| 732 |
+
"step": 1020
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 1.9333646175504458,
|
| 736 |
+
"grad_norm": 3.7936253547668457,
|
| 737 |
+
"learning_rate": 1.9707520891364902e-05,
|
| 738 |
+
"loss": 0.6852,
|
| 739 |
+
"step": 1030
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.9521351478179259,
|
| 743 |
+
"grad_norm": 3.4030821323394775,
|
| 744 |
+
"learning_rate": 1.935933147632312e-05,
|
| 745 |
+
"loss": 0.7311,
|
| 746 |
+
"step": 1040
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.970905678085406,
|
| 750 |
+
"grad_norm": 3.5526273250579834,
|
| 751 |
+
"learning_rate": 1.9011142061281336e-05,
|
| 752 |
+
"loss": 0.7846,
|
| 753 |
+
"step": 1050
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 1.9896762083528858,
|
| 757 |
+
"grad_norm": 3.1862409114837646,
|
| 758 |
+
"learning_rate": 1.8662952646239558e-05,
|
| 759 |
+
"loss": 0.7233,
|
| 760 |
+
"step": 1060
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 1.999061473486626,
|
| 764 |
+
"eval_accuracy": 0.8294389438943894,
|
| 765 |
+
"eval_loss": 0.6780887842178345,
|
| 766 |
+
"eval_runtime": 125.2435,
|
| 767 |
+
"eval_samples_per_second": 60.482,
|
| 768 |
+
"eval_steps_per_second": 1.892,
|
| 769 |
+
"step": 1065
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 2.008446738620366,
|
| 773 |
+
"grad_norm": 3.521763563156128,
|
| 774 |
+
"learning_rate": 1.8314763231197773e-05,
|
| 775 |
+
"loss": 0.6907,
|
| 776 |
+
"step": 1070
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 2.027217268887846,
|
| 780 |
+
"grad_norm": 3.0328750610351562,
|
| 781 |
+
"learning_rate": 1.7966573816155992e-05,
|
| 782 |
+
"loss": 0.6829,
|
| 783 |
+
"step": 1080
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 2.045987799155326,
|
| 787 |
+
"grad_norm": 4.274998188018799,
|
| 788 |
+
"learning_rate": 1.7618384401114207e-05,
|
| 789 |
+
"loss": 0.635,
|
| 790 |
+
"step": 1090
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 2.0647583294228062,
|
| 794 |
+
"grad_norm": 3.0744423866271973,
|
| 795 |
+
"learning_rate": 1.7270194986072426e-05,
|
| 796 |
+
"loss": 0.7163,
|
| 797 |
+
"step": 1100
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 2.0835288596902863,
|
| 801 |
+
"grad_norm": 3.0927815437316895,
|
| 802 |
+
"learning_rate": 1.692200557103064e-05,
|
| 803 |
+
"loss": 0.6817,
|
| 804 |
+
"step": 1110
|
| 805 |
+
},
|
| 806 |
+
{
|
| 807 |
+
"epoch": 2.1022993899577664,
|
| 808 |
+
"grad_norm": 2.5418238639831543,
|
| 809 |
+
"learning_rate": 1.6573816155988857e-05,
|
| 810 |
+
"loss": 0.6235,
|
| 811 |
+
"step": 1120
|
| 812 |
+
},
|
| 813 |
+
{
|
| 814 |
+
"epoch": 2.1210699202252465,
|
| 815 |
+
"grad_norm": 3.1327621936798096,
|
| 816 |
+
"learning_rate": 1.6225626740947075e-05,
|
| 817 |
+
"loss": 0.6351,
|
| 818 |
+
"step": 1130
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 2.139840450492726,
|
| 822 |
+
"grad_norm": 4.186623573303223,
|
| 823 |
+
"learning_rate": 1.587743732590529e-05,
|
| 824 |
+
"loss": 0.6383,
|
| 825 |
+
"step": 1140
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 2.1586109807602063,
|
| 829 |
+
"grad_norm": 3.007977247238159,
|
| 830 |
+
"learning_rate": 1.552924791086351e-05,
|
| 831 |
+
"loss": 0.6994,
|
| 832 |
+
"step": 1150
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 2.1773815110276864,
|
| 836 |
+
"grad_norm": 3.232109308242798,
|
| 837 |
+
"learning_rate": 1.518105849582173e-05,
|
| 838 |
+
"loss": 0.6615,
|
| 839 |
+
"step": 1160
|
| 840 |
+
},
|
| 841 |
+
{
|
| 842 |
+
"epoch": 2.1961520412951665,
|
| 843 |
+
"grad_norm": 2.8354580402374268,
|
| 844 |
+
"learning_rate": 1.4832869080779947e-05,
|
| 845 |
+
"loss": 0.6686,
|
| 846 |
+
"step": 1170
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 2.2149225715626466,
|
| 850 |
+
"grad_norm": 3.6194567680358887,
|
| 851 |
+
"learning_rate": 1.4484679665738164e-05,
|
| 852 |
+
"loss": 0.6649,
|
| 853 |
+
"step": 1180
|
| 854 |
+
},
|
| 855 |
+
{
|
| 856 |
+
"epoch": 2.2336931018301267,
|
| 857 |
+
"grad_norm": 4.279024600982666,
|
| 858 |
+
"learning_rate": 1.413649025069638e-05,
|
| 859 |
+
"loss": 0.6529,
|
| 860 |
+
"step": 1190
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 2.252463632097607,
|
| 864 |
+
"grad_norm": 3.035931348800659,
|
| 865 |
+
"learning_rate": 1.3788300835654596e-05,
|
| 866 |
+
"loss": 0.6796,
|
| 867 |
+
"step": 1200
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 2.271234162365087,
|
| 871 |
+
"grad_norm": 3.2257440090179443,
|
| 872 |
+
"learning_rate": 1.3440111420612813e-05,
|
| 873 |
+
"loss": 0.6501,
|
| 874 |
+
"step": 1210
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 2.290004692632567,
|
| 878 |
+
"grad_norm": 3.176237106323242,
|
| 879 |
+
"learning_rate": 1.309192200557103e-05,
|
| 880 |
+
"loss": 0.7252,
|
| 881 |
+
"step": 1220
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 2.308775222900047,
|
| 885 |
+
"grad_norm": 3.575956344604492,
|
| 886 |
+
"learning_rate": 1.2743732590529247e-05,
|
| 887 |
+
"loss": 0.6798,
|
| 888 |
+
"step": 1230
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 2.327545753167527,
|
| 892 |
+
"grad_norm": 3.1033012866973877,
|
| 893 |
+
"learning_rate": 1.2395543175487466e-05,
|
| 894 |
+
"loss": 0.6291,
|
| 895 |
+
"step": 1240
|
| 896 |
+
},
|
| 897 |
+
{
|
| 898 |
+
"epoch": 2.346316283435007,
|
| 899 |
+
"grad_norm": 3.797513961791992,
|
| 900 |
+
"learning_rate": 1.2047353760445683e-05,
|
| 901 |
+
"loss": 0.6908,
|
| 902 |
+
"step": 1250
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 2.365086813702487,
|
| 906 |
+
"grad_norm": 3.399435043334961,
|
| 907 |
+
"learning_rate": 1.16991643454039e-05,
|
| 908 |
+
"loss": 0.5915,
|
| 909 |
+
"step": 1260
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 2.383857343969967,
|
| 913 |
+
"grad_norm": 2.9869837760925293,
|
| 914 |
+
"learning_rate": 1.1350974930362116e-05,
|
| 915 |
+
"loss": 0.6665,
|
| 916 |
+
"step": 1270
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 2.402627874237447,
|
| 920 |
+
"grad_norm": 3.171600580215454,
|
| 921 |
+
"learning_rate": 1.1002785515320335e-05,
|
| 922 |
+
"loss": 0.6202,
|
| 923 |
+
"step": 1280
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 2.4213984045049273,
|
| 927 |
+
"grad_norm": 2.0792176723480225,
|
| 928 |
+
"learning_rate": 1.0654596100278552e-05,
|
| 929 |
+
"loss": 0.5773,
|
| 930 |
+
"step": 1290
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 2.4401689347724074,
|
| 934 |
+
"grad_norm": 3.339883804321289,
|
| 935 |
+
"learning_rate": 1.0306406685236769e-05,
|
| 936 |
+
"loss": 0.6435,
|
| 937 |
+
"step": 1300
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 2.4589394650398875,
|
| 941 |
+
"grad_norm": 3.2008817195892334,
|
| 942 |
+
"learning_rate": 9.958217270194986e-06,
|
| 943 |
+
"loss": 0.6448,
|
| 944 |
+
"step": 1310
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 2.4777099953073676,
|
| 948 |
+
"grad_norm": 2.828279733657837,
|
| 949 |
+
"learning_rate": 9.610027855153205e-06,
|
| 950 |
+
"loss": 0.6121,
|
| 951 |
+
"step": 1320
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 2.4964805255748477,
|
| 955 |
+
"grad_norm": 2.920414924621582,
|
| 956 |
+
"learning_rate": 9.261838440111422e-06,
|
| 957 |
+
"loss": 0.6569,
|
| 958 |
+
"step": 1330
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 2.5152510558423273,
|
| 962 |
+
"grad_norm": 3.4083826541900635,
|
| 963 |
+
"learning_rate": 8.913649025069639e-06,
|
| 964 |
+
"loss": 0.6819,
|
| 965 |
+
"step": 1340
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 2.534021586109808,
|
| 969 |
+
"grad_norm": 2.8921279907226562,
|
| 970 |
+
"learning_rate": 8.565459610027856e-06,
|
| 971 |
+
"loss": 0.6257,
|
| 972 |
+
"step": 1350
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 2.5527921163772875,
|
| 976 |
+
"grad_norm": 3.33734130859375,
|
| 977 |
+
"learning_rate": 8.217270194986073e-06,
|
| 978 |
+
"loss": 0.6439,
|
| 979 |
+
"step": 1360
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 2.5715626466447676,
|
| 983 |
+
"grad_norm": 4.184520244598389,
|
| 984 |
+
"learning_rate": 7.869080779944291e-06,
|
| 985 |
+
"loss": 0.6742,
|
| 986 |
+
"step": 1370
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 2.5903331769122477,
|
| 990 |
+
"grad_norm": 3.5494306087493896,
|
| 991 |
+
"learning_rate": 7.5208913649025075e-06,
|
| 992 |
+
"loss": 0.6485,
|
| 993 |
+
"step": 1380
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 2.609103707179728,
|
| 997 |
+
"grad_norm": 2.791755437850952,
|
| 998 |
+
"learning_rate": 7.1727019498607245e-06,
|
| 999 |
+
"loss": 0.6061,
|
| 1000 |
+
"step": 1390
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 2.627874237447208,
|
| 1004 |
+
"grad_norm": 3.616508960723877,
|
| 1005 |
+
"learning_rate": 6.8245125348189415e-06,
|
| 1006 |
+
"loss": 0.7124,
|
| 1007 |
+
"step": 1400
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 2.646644767714688,
|
| 1011 |
+
"grad_norm": 3.5052719116210938,
|
| 1012 |
+
"learning_rate": 6.4763231197771585e-06,
|
| 1013 |
+
"loss": 0.6545,
|
| 1014 |
+
"step": 1410
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 2.665415297982168,
|
| 1018 |
+
"grad_norm": 3.463571786880493,
|
| 1019 |
+
"learning_rate": 6.128133704735376e-06,
|
| 1020 |
+
"loss": 0.5905,
|
| 1021 |
+
"step": 1420
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 2.684185828249648,
|
| 1025 |
+
"grad_norm": 3.2912559509277344,
|
| 1026 |
+
"learning_rate": 5.779944289693594e-06,
|
| 1027 |
+
"loss": 0.5468,
|
| 1028 |
+
"step": 1430
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 2.7029563585171283,
|
| 1032 |
+
"grad_norm": 3.0410854816436768,
|
| 1033 |
+
"learning_rate": 5.43175487465181e-06,
|
| 1034 |
+
"loss": 0.6286,
|
| 1035 |
+
"step": 1440
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 2.721726888784608,
|
| 1039 |
+
"grad_norm": 3.565953016281128,
|
| 1040 |
+
"learning_rate": 5.083565459610028e-06,
|
| 1041 |
+
"loss": 0.6567,
|
| 1042 |
+
"step": 1450
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 2.740497419052088,
|
| 1046 |
+
"grad_norm": 3.0948688983917236,
|
| 1047 |
+
"learning_rate": 4.735376044568245e-06,
|
| 1048 |
+
"loss": 0.644,
|
| 1049 |
+
"step": 1460
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 2.759267949319568,
|
| 1053 |
+
"grad_norm": 3.734225034713745,
|
| 1054 |
+
"learning_rate": 4.387186629526462e-06,
|
| 1055 |
+
"loss": 0.65,
|
| 1056 |
+
"step": 1470
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 2.7780384795870483,
|
| 1060 |
+
"grad_norm": 3.245823860168457,
|
| 1061 |
+
"learning_rate": 4.03899721448468e-06,
|
| 1062 |
+
"loss": 0.613,
|
| 1063 |
+
"step": 1480
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 2.7968090098545284,
|
| 1067 |
+
"grad_norm": 2.6476266384124756,
|
| 1068 |
+
"learning_rate": 3.690807799442897e-06,
|
| 1069 |
+
"loss": 0.6562,
|
| 1070 |
+
"step": 1490
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 2.8155795401220085,
|
| 1074 |
+
"grad_norm": 3.1951568126678467,
|
| 1075 |
+
"learning_rate": 3.3426183844011147e-06,
|
| 1076 |
+
"loss": 0.5897,
|
| 1077 |
+
"step": 1500
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 2.8343500703894886,
|
| 1081 |
+
"grad_norm": 3.754561424255371,
|
| 1082 |
+
"learning_rate": 2.9944289693593313e-06,
|
| 1083 |
+
"loss": 0.5686,
|
| 1084 |
+
"step": 1510
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 2.8531206006569687,
|
| 1088 |
+
"grad_norm": 2.6758432388305664,
|
| 1089 |
+
"learning_rate": 2.6462395543175487e-06,
|
| 1090 |
+
"loss": 0.6791,
|
| 1091 |
+
"step": 1520
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 2.871891130924449,
|
| 1095 |
+
"grad_norm": 2.6981005668640137,
|
| 1096 |
+
"learning_rate": 2.298050139275766e-06,
|
| 1097 |
+
"loss": 0.5544,
|
| 1098 |
+
"step": 1530
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 2.8906616611919285,
|
| 1102 |
+
"grad_norm": 3.503272294998169,
|
| 1103 |
+
"learning_rate": 1.9498607242339835e-06,
|
| 1104 |
+
"loss": 0.6921,
|
| 1105 |
+
"step": 1540
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 2.9094321914594086,
|
| 1109 |
+
"grad_norm": 3.2283263206481934,
|
| 1110 |
+
"learning_rate": 1.6016713091922007e-06,
|
| 1111 |
+
"loss": 0.6307,
|
| 1112 |
+
"step": 1550
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 2.9282027217268887,
|
| 1116 |
+
"grad_norm": 3.2601311206817627,
|
| 1117 |
+
"learning_rate": 1.253481894150418e-06,
|
| 1118 |
+
"loss": 0.653,
|
| 1119 |
+
"step": 1560
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 2.9469732519943688,
|
| 1123 |
+
"grad_norm": 3.5451338291168213,
|
| 1124 |
+
"learning_rate": 9.052924791086352e-07,
|
| 1125 |
+
"loss": 0.5953,
|
| 1126 |
+
"step": 1570
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 2.965743782261849,
|
| 1130 |
+
"grad_norm": 3.605889081954956,
|
| 1131 |
+
"learning_rate": 5.571030640668524e-07,
|
| 1132 |
+
"loss": 0.6338,
|
| 1133 |
+
"step": 1580
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 2.984514312529329,
|
| 1137 |
+
"grad_norm": 4.055668830871582,
|
| 1138 |
+
"learning_rate": 2.0891364902506967e-07,
|
| 1139 |
+
"loss": 0.6047,
|
| 1140 |
+
"step": 1590
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 2.995776630689817,
|
| 1144 |
+
"eval_accuracy": 0.834983498349835,
|
| 1145 |
+
"eval_loss": 0.6400949358940125,
|
| 1146 |
+
"eval_runtime": 126.916,
|
| 1147 |
+
"eval_samples_per_second": 59.685,
|
| 1148 |
+
"eval_steps_per_second": 1.867,
|
| 1149 |
+
"step": 1596
|
| 1150 |
+
},
|
| 1151 |
+
{
|
| 1152 |
+
"epoch": 2.995776630689817,
|
| 1153 |
+
"step": 1596,
|
| 1154 |
+
"total_flos": 1.584209316447959e+19,
|
| 1155 |
+
"train_loss": 1.1699181137825911,
|
| 1156 |
+
"train_runtime": 8274.5103,
|
| 1157 |
+
"train_samples_per_second": 24.717,
|
| 1158 |
+
"train_steps_per_second": 0.193
|
| 1159 |
+
}
|
| 1160 |
+
],
|
| 1161 |
+
"logging_steps": 10,
|
| 1162 |
+
"max_steps": 1596,
|
| 1163 |
+
"num_input_tokens_seen": 0,
|
| 1164 |
+
"num_train_epochs": 3,
|
| 1165 |
+
"save_steps": 500,
|
| 1166 |
+
"stateful_callbacks": {
|
| 1167 |
+
"TrainerControl": {
|
| 1168 |
+
"args": {
|
| 1169 |
+
"should_epoch_stop": false,
|
| 1170 |
+
"should_evaluate": false,
|
| 1171 |
+
"should_log": false,
|
| 1172 |
+
"should_save": true,
|
| 1173 |
+
"should_training_stop": true
|
| 1174 |
+
},
|
| 1175 |
+
"attributes": {}
|
| 1176 |
+
}
|
| 1177 |
+
},
|
| 1178 |
+
"total_flos": 1.584209316447959e+19,
|
| 1179 |
+
"train_batch_size": 32,
|
| 1180 |
+
"trial_name": null,
|
| 1181 |
+
"trial_params": null
|
| 1182 |
+
}
|