Training in progress, step 800
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- model.safetensors +1 -1
- runs/Mar14_18-22-37_675d-3/events.out.tfevents.1710440608.675d-3.2467532.1 +3 -0
- runs/Mar14_19-57-09_675d-3/events.out.tfevents.1710444418.675d-3.2472008.0 +3 -0
- train_results.json +8 -0
- trainer_state.json +1216 -0
- training_args.bin +1 -1
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_cer": 1.0,
|
| 4 |
+
"eval_loss": 2.859034538269043,
|
| 5 |
+
"eval_runtime": 21.2144,
|
| 6 |
+
"eval_samples": 2642,
|
| 7 |
+
"eval_samples_per_second": 124.538,
|
| 8 |
+
"eval_steps_per_second": 1.98,
|
| 9 |
+
"eval_wer": 1.0,
|
| 10 |
+
"train_loss": 4.4063241723037905,
|
| 11 |
+
"train_runtime": 1105.5076,
|
| 12 |
+
"train_samples": 28538,
|
| 13 |
+
"train_samples_per_second": 77.443,
|
| 14 |
+
"train_steps_per_second": 0.152
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_cer": 1.0,
|
| 4 |
+
"eval_loss": 2.859034538269043,
|
| 5 |
+
"eval_runtime": 21.2144,
|
| 6 |
+
"eval_samples": 2642,
|
| 7 |
+
"eval_samples_per_second": 124.538,
|
| 8 |
+
"eval_steps_per_second": 1.98,
|
| 9 |
+
"eval_wer": 1.0
|
| 10 |
+
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1261933700
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:148ed659c84901fa64b4919c674545daef3b7392e18e54ebfcaa11cf8b8bc8be
|
| 3 |
size 1261933700
|
runs/Mar14_18-22-37_675d-3/events.out.tfevents.1710440608.675d-3.2467532.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:94e566c112849cccd13c009c602ea5159b02c0675cc64c514c2c1e2e0d78c309
|
| 3 |
+
size 453
|
runs/Mar14_19-57-09_675d-3/events.out.tfevents.1710444418.675d-3.2472008.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fe1a5e22956b2223c5157f4c8c485b8e996b983a56a4b7041b9732740cfa1181
|
| 3 |
+
size 177195
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 4.4063241723037905,
|
| 4 |
+
"train_runtime": 1105.5076,
|
| 5 |
+
"train_samples": 28538,
|
| 6 |
+
"train_samples_per_second": 77.443,
|
| 7 |
+
"train_steps_per_second": 0.152
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1216 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 168,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.02,
|
| 13 |
+
"grad_norm": Infinity,
|
| 14 |
+
"learning_rate": 0.0,
|
| 15 |
+
"loss": 8.9123,
|
| 16 |
+
"step": 1
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.04,
|
| 20 |
+
"grad_norm": 2.591931104660034,
|
| 21 |
+
"learning_rate": 6e-07,
|
| 22 |
+
"loss": 9.0614,
|
| 23 |
+
"step": 2
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.05,
|
| 27 |
+
"grad_norm": 2.5352702140808105,
|
| 28 |
+
"learning_rate": 1.2e-06,
|
| 29 |
+
"loss": 8.8832,
|
| 30 |
+
"step": 3
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.07,
|
| 34 |
+
"grad_norm": 2.5569372177124023,
|
| 35 |
+
"learning_rate": 1.8e-06,
|
| 36 |
+
"loss": 8.909,
|
| 37 |
+
"step": 4
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.09,
|
| 41 |
+
"grad_norm": 2.5344245433807373,
|
| 42 |
+
"learning_rate": 2.4e-06,
|
| 43 |
+
"loss": 8.8885,
|
| 44 |
+
"step": 5
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.11,
|
| 48 |
+
"grad_norm": 2.566680669784546,
|
| 49 |
+
"learning_rate": 2.9999999999999997e-06,
|
| 50 |
+
"loss": 8.9777,
|
| 51 |
+
"step": 6
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.12,
|
| 55 |
+
"grad_norm": 2.5253851413726807,
|
| 56 |
+
"learning_rate": 3.6e-06,
|
| 57 |
+
"loss": 8.9826,
|
| 58 |
+
"step": 7
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.14,
|
| 62 |
+
"grad_norm": 2.5269064903259277,
|
| 63 |
+
"learning_rate": 4.2e-06,
|
| 64 |
+
"loss": 8.7751,
|
| 65 |
+
"step": 8
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.16,
|
| 69 |
+
"grad_norm": 2.6110405921936035,
|
| 70 |
+
"learning_rate": 4.8e-06,
|
| 71 |
+
"loss": 9.0477,
|
| 72 |
+
"step": 9
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.18,
|
| 76 |
+
"grad_norm": 2.625199556350708,
|
| 77 |
+
"learning_rate": 5.399999999999999e-06,
|
| 78 |
+
"loss": 8.9635,
|
| 79 |
+
"step": 10
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.2,
|
| 83 |
+
"grad_norm": 2.6010096073150635,
|
| 84 |
+
"learning_rate": 5.999999999999999e-06,
|
| 85 |
+
"loss": 8.9304,
|
| 86 |
+
"step": 11
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.21,
|
| 90 |
+
"grad_norm": 2.617136001586914,
|
| 91 |
+
"learning_rate": 6.599999999999999e-06,
|
| 92 |
+
"loss": 8.9034,
|
| 93 |
+
"step": 12
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.23,
|
| 97 |
+
"grad_norm": 2.7402148246765137,
|
| 98 |
+
"learning_rate": 7.2e-06,
|
| 99 |
+
"loss": 9.0438,
|
| 100 |
+
"step": 13
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.25,
|
| 104 |
+
"grad_norm": 2.7554776668548584,
|
| 105 |
+
"learning_rate": 7.799999999999998e-06,
|
| 106 |
+
"loss": 8.8399,
|
| 107 |
+
"step": 14
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.27,
|
| 111 |
+
"grad_norm": 2.7558281421661377,
|
| 112 |
+
"learning_rate": 8.4e-06,
|
| 113 |
+
"loss": 8.6818,
|
| 114 |
+
"step": 15
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.29,
|
| 118 |
+
"grad_norm": 2.8014261722564697,
|
| 119 |
+
"learning_rate": 8.999999999999999e-06,
|
| 120 |
+
"loss": 8.7148,
|
| 121 |
+
"step": 16
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.3,
|
| 125 |
+
"grad_norm": 3.0471227169036865,
|
| 126 |
+
"learning_rate": 9.6e-06,
|
| 127 |
+
"loss": 8.6494,
|
| 128 |
+
"step": 17
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.32,
|
| 132 |
+
"grad_norm": 3.0272316932678223,
|
| 133 |
+
"learning_rate": 1.02e-05,
|
| 134 |
+
"loss": 8.5651,
|
| 135 |
+
"step": 18
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.34,
|
| 139 |
+
"grad_norm": 3.201422929763794,
|
| 140 |
+
"learning_rate": 1.0799999999999998e-05,
|
| 141 |
+
"loss": 8.8344,
|
| 142 |
+
"step": 19
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.36,
|
| 146 |
+
"grad_norm": 3.3028013706207275,
|
| 147 |
+
"learning_rate": 1.14e-05,
|
| 148 |
+
"loss": 8.5798,
|
| 149 |
+
"step": 20
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.38,
|
| 153 |
+
"grad_norm": 3.3581230640411377,
|
| 154 |
+
"learning_rate": 1.1999999999999999e-05,
|
| 155 |
+
"loss": 8.462,
|
| 156 |
+
"step": 21
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.39,
|
| 160 |
+
"grad_norm": 3.490736484527588,
|
| 161 |
+
"learning_rate": 1.26e-05,
|
| 162 |
+
"loss": 8.5277,
|
| 163 |
+
"step": 22
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.41,
|
| 167 |
+
"grad_norm": 3.643688678741455,
|
| 168 |
+
"learning_rate": 1.3199999999999997e-05,
|
| 169 |
+
"loss": 8.4209,
|
| 170 |
+
"step": 23
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.43,
|
| 174 |
+
"grad_norm": 3.639347791671753,
|
| 175 |
+
"learning_rate": 1.3799999999999998e-05,
|
| 176 |
+
"loss": 8.2214,
|
| 177 |
+
"step": 24
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.45,
|
| 181 |
+
"grad_norm": 4.0215983390808105,
|
| 182 |
+
"learning_rate": 1.44e-05,
|
| 183 |
+
"loss": 8.6764,
|
| 184 |
+
"step": 25
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.46,
|
| 188 |
+
"grad_norm": 4.044653415679932,
|
| 189 |
+
"learning_rate": 1.4999999999999999e-05,
|
| 190 |
+
"loss": 8.1195,
|
| 191 |
+
"step": 26
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.48,
|
| 195 |
+
"grad_norm": 4.363345146179199,
|
| 196 |
+
"learning_rate": 1.5599999999999996e-05,
|
| 197 |
+
"loss": 8.2267,
|
| 198 |
+
"step": 27
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.5,
|
| 202 |
+
"grad_norm": 4.63635778427124,
|
| 203 |
+
"learning_rate": 1.6199999999999997e-05,
|
| 204 |
+
"loss": 8.0302,
|
| 205 |
+
"step": 28
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.52,
|
| 209 |
+
"grad_norm": 4.918217658996582,
|
| 210 |
+
"learning_rate": 1.68e-05,
|
| 211 |
+
"loss": 8.0072,
|
| 212 |
+
"step": 29
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.54,
|
| 216 |
+
"grad_norm": 5.041306018829346,
|
| 217 |
+
"learning_rate": 1.74e-05,
|
| 218 |
+
"loss": 7.8004,
|
| 219 |
+
"step": 30
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.55,
|
| 223 |
+
"grad_norm": 5.159876823425293,
|
| 224 |
+
"learning_rate": 1.7999999999999997e-05,
|
| 225 |
+
"loss": 7.7674,
|
| 226 |
+
"step": 31
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.57,
|
| 230 |
+
"grad_norm": 5.359218120574951,
|
| 231 |
+
"learning_rate": 1.8599999999999998e-05,
|
| 232 |
+
"loss": 7.7172,
|
| 233 |
+
"step": 32
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.59,
|
| 237 |
+
"grad_norm": 5.789942741394043,
|
| 238 |
+
"learning_rate": 1.92e-05,
|
| 239 |
+
"loss": 7.5795,
|
| 240 |
+
"step": 33
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.61,
|
| 244 |
+
"grad_norm": 6.3346171379089355,
|
| 245 |
+
"learning_rate": 1.98e-05,
|
| 246 |
+
"loss": 7.5992,
|
| 247 |
+
"step": 34
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.62,
|
| 251 |
+
"grad_norm": 6.570509433746338,
|
| 252 |
+
"learning_rate": 2.04e-05,
|
| 253 |
+
"loss": 7.2858,
|
| 254 |
+
"step": 35
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.64,
|
| 258 |
+
"grad_norm": 7.087466716766357,
|
| 259 |
+
"learning_rate": 2.1e-05,
|
| 260 |
+
"loss": 7.2201,
|
| 261 |
+
"step": 36
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.66,
|
| 265 |
+
"grad_norm": 7.214906692504883,
|
| 266 |
+
"learning_rate": 2.1599999999999996e-05,
|
| 267 |
+
"loss": 6.9394,
|
| 268 |
+
"step": 37
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.68,
|
| 272 |
+
"grad_norm": Infinity,
|
| 273 |
+
"learning_rate": 2.1599999999999996e-05,
|
| 274 |
+
"loss": 6.9783,
|
| 275 |
+
"step": 38
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.7,
|
| 279 |
+
"grad_norm": 7.592892169952393,
|
| 280 |
+
"learning_rate": 2.2199999999999998e-05,
|
| 281 |
+
"loss": 6.6987,
|
| 282 |
+
"step": 39
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.71,
|
| 286 |
+
"grad_norm": 8.196120262145996,
|
| 287 |
+
"learning_rate": 2.28e-05,
|
| 288 |
+
"loss": 6.6235,
|
| 289 |
+
"step": 40
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.73,
|
| 293 |
+
"grad_norm": 8.517624855041504,
|
| 294 |
+
"learning_rate": 2.34e-05,
|
| 295 |
+
"loss": 6.3869,
|
| 296 |
+
"step": 41
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.75,
|
| 300 |
+
"grad_norm": 8.981674194335938,
|
| 301 |
+
"learning_rate": 2.3999999999999997e-05,
|
| 302 |
+
"loss": 6.2016,
|
| 303 |
+
"step": 42
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.77,
|
| 307 |
+
"grad_norm": 9.2571439743042,
|
| 308 |
+
"learning_rate": 2.4599999999999998e-05,
|
| 309 |
+
"loss": 5.927,
|
| 310 |
+
"step": 43
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.79,
|
| 314 |
+
"grad_norm": 9.87216567993164,
|
| 315 |
+
"learning_rate": 2.52e-05,
|
| 316 |
+
"loss": 5.9454,
|
| 317 |
+
"step": 44
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.8,
|
| 321 |
+
"grad_norm": 9.831586837768555,
|
| 322 |
+
"learning_rate": 2.5799999999999997e-05,
|
| 323 |
+
"loss": 5.5236,
|
| 324 |
+
"step": 45
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.82,
|
| 328 |
+
"grad_norm": 9.9969482421875,
|
| 329 |
+
"learning_rate": 2.6399999999999995e-05,
|
| 330 |
+
"loss": 5.3099,
|
| 331 |
+
"step": 46
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.84,
|
| 335 |
+
"grad_norm": 10.3759183883667,
|
| 336 |
+
"learning_rate": 2.6999999999999996e-05,
|
| 337 |
+
"loss": 5.167,
|
| 338 |
+
"step": 47
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.86,
|
| 342 |
+
"grad_norm": 10.099465370178223,
|
| 343 |
+
"learning_rate": 2.7599999999999997e-05,
|
| 344 |
+
"loss": 4.8973,
|
| 345 |
+
"step": 48
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.88,
|
| 349 |
+
"grad_norm": 10.04648494720459,
|
| 350 |
+
"learning_rate": 2.8199999999999998e-05,
|
| 351 |
+
"loss": 4.7279,
|
| 352 |
+
"step": 49
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.89,
|
| 356 |
+
"grad_norm": 10.015722274780273,
|
| 357 |
+
"learning_rate": 2.88e-05,
|
| 358 |
+
"loss": 4.6312,
|
| 359 |
+
"step": 50
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.91,
|
| 363 |
+
"grad_norm": 9.263956069946289,
|
| 364 |
+
"learning_rate": 2.94e-05,
|
| 365 |
+
"loss": 4.3213,
|
| 366 |
+
"step": 51
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.93,
|
| 370 |
+
"grad_norm": 9.205384254455566,
|
| 371 |
+
"learning_rate": 2.9999999999999997e-05,
|
| 372 |
+
"loss": 4.2187,
|
| 373 |
+
"step": 52
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.95,
|
| 377 |
+
"grad_norm": 8.679594993591309,
|
| 378 |
+
"learning_rate": 3.06e-05,
|
| 379 |
+
"loss": 4.0743,
|
| 380 |
+
"step": 53
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.96,
|
| 384 |
+
"grad_norm": 8.269742012023926,
|
| 385 |
+
"learning_rate": 3.119999999999999e-05,
|
| 386 |
+
"loss": 3.9656,
|
| 387 |
+
"step": 54
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.98,
|
| 391 |
+
"grad_norm": 7.4761128425598145,
|
| 392 |
+
"learning_rate": 3.1799999999999994e-05,
|
| 393 |
+
"loss": 3.832,
|
| 394 |
+
"step": 55
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 1.0,
|
| 398 |
+
"grad_norm": 6.980052471160889,
|
| 399 |
+
"learning_rate": 3.2399999999999995e-05,
|
| 400 |
+
"loss": 3.77,
|
| 401 |
+
"step": 56
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 1.02,
|
| 405 |
+
"grad_norm": 6.122258186340332,
|
| 406 |
+
"learning_rate": 3.2999999999999996e-05,
|
| 407 |
+
"loss": 3.6187,
|
| 408 |
+
"step": 57
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 1.04,
|
| 412 |
+
"grad_norm": 5.600069999694824,
|
| 413 |
+
"learning_rate": 3.36e-05,
|
| 414 |
+
"loss": 3.5517,
|
| 415 |
+
"step": 58
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 1.05,
|
| 419 |
+
"grad_norm": 5.075619220733643,
|
| 420 |
+
"learning_rate": 3.42e-05,
|
| 421 |
+
"loss": 3.4915,
|
| 422 |
+
"step": 59
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.07,
|
| 426 |
+
"grad_norm": 4.481492042541504,
|
| 427 |
+
"learning_rate": 3.48e-05,
|
| 428 |
+
"loss": 3.4368,
|
| 429 |
+
"step": 60
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.09,
|
| 433 |
+
"grad_norm": 3.752028465270996,
|
| 434 |
+
"learning_rate": 3.539999999999999e-05,
|
| 435 |
+
"loss": 3.3771,
|
| 436 |
+
"step": 61
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 1.11,
|
| 440 |
+
"grad_norm": 3.3141655921936035,
|
| 441 |
+
"learning_rate": 3.5999999999999994e-05,
|
| 442 |
+
"loss": 3.355,
|
| 443 |
+
"step": 62
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 1.12,
|
| 447 |
+
"grad_norm": 2.8507885932922363,
|
| 448 |
+
"learning_rate": 3.6599999999999995e-05,
|
| 449 |
+
"loss": 3.3178,
|
| 450 |
+
"step": 63
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 1.14,
|
| 454 |
+
"grad_norm": 2.23435640335083,
|
| 455 |
+
"learning_rate": 3.7199999999999996e-05,
|
| 456 |
+
"loss": 3.2465,
|
| 457 |
+
"step": 64
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.16,
|
| 461 |
+
"grad_norm": 1.9935156106948853,
|
| 462 |
+
"learning_rate": 3.78e-05,
|
| 463 |
+
"loss": 3.2263,
|
| 464 |
+
"step": 65
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.18,
|
| 468 |
+
"grad_norm": 1.7533106803894043,
|
| 469 |
+
"learning_rate": 3.84e-05,
|
| 470 |
+
"loss": 3.2015,
|
| 471 |
+
"step": 66
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 1.2,
|
| 475 |
+
"grad_norm": 1.662286400794983,
|
| 476 |
+
"learning_rate": 3.9e-05,
|
| 477 |
+
"loss": 3.1798,
|
| 478 |
+
"step": 67
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 1.21,
|
| 482 |
+
"grad_norm": 1.5815298557281494,
|
| 483 |
+
"learning_rate": 3.96e-05,
|
| 484 |
+
"loss": 3.163,
|
| 485 |
+
"step": 68
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 1.23,
|
| 489 |
+
"grad_norm": 1.6589211225509644,
|
| 490 |
+
"learning_rate": 4.02e-05,
|
| 491 |
+
"loss": 3.1624,
|
| 492 |
+
"step": 69
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 1.25,
|
| 496 |
+
"grad_norm": 1.4096239805221558,
|
| 497 |
+
"learning_rate": 4.08e-05,
|
| 498 |
+
"loss": 3.1033,
|
| 499 |
+
"step": 70
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 1.27,
|
| 503 |
+
"grad_norm": 1.174314022064209,
|
| 504 |
+
"learning_rate": 4.14e-05,
|
| 505 |
+
"loss": 3.0756,
|
| 506 |
+
"step": 71
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 1.29,
|
| 510 |
+
"grad_norm": 1.0445579290390015,
|
| 511 |
+
"learning_rate": 4.2e-05,
|
| 512 |
+
"loss": 3.0549,
|
| 513 |
+
"step": 72
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 1.3,
|
| 517 |
+
"grad_norm": 1.0027287006378174,
|
| 518 |
+
"learning_rate": 4.259999999999999e-05,
|
| 519 |
+
"loss": 3.0395,
|
| 520 |
+
"step": 73
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 1.32,
|
| 524 |
+
"grad_norm": 0.8988072276115417,
|
| 525 |
+
"learning_rate": 4.319999999999999e-05,
|
| 526 |
+
"loss": 3.0235,
|
| 527 |
+
"step": 74
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 1.34,
|
| 531 |
+
"grad_norm": 1.0056397914886475,
|
| 532 |
+
"learning_rate": 4.3799999999999994e-05,
|
| 533 |
+
"loss": 3.0391,
|
| 534 |
+
"step": 75
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 1.36,
|
| 538 |
+
"grad_norm": 0.8533895611763,
|
| 539 |
+
"learning_rate": 4.4399999999999995e-05,
|
| 540 |
+
"loss": 2.9842,
|
| 541 |
+
"step": 76
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 1.38,
|
| 545 |
+
"grad_norm": 0.7804594039916992,
|
| 546 |
+
"learning_rate": 4.4999999999999996e-05,
|
| 547 |
+
"loss": 2.9738,
|
| 548 |
+
"step": 77
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 1.39,
|
| 552 |
+
"grad_norm": 0.7258276343345642,
|
| 553 |
+
"learning_rate": 4.56e-05,
|
| 554 |
+
"loss": 2.9663,
|
| 555 |
+
"step": 78
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 1.41,
|
| 559 |
+
"grad_norm": 0.7681087255477905,
|
| 560 |
+
"learning_rate": 4.62e-05,
|
| 561 |
+
"loss": 2.9634,
|
| 562 |
+
"step": 79
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 1.43,
|
| 566 |
+
"grad_norm": 0.5595097541809082,
|
| 567 |
+
"learning_rate": 4.68e-05,
|
| 568 |
+
"loss": 2.9571,
|
| 569 |
+
"step": 80
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.45,
|
| 573 |
+
"grad_norm": 0.5609928965568542,
|
| 574 |
+
"learning_rate": 4.7399999999999993e-05,
|
| 575 |
+
"loss": 2.9781,
|
| 576 |
+
"step": 81
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.46,
|
| 580 |
+
"grad_norm": 0.46705162525177,
|
| 581 |
+
"learning_rate": 4.7999999999999994e-05,
|
| 582 |
+
"loss": 2.93,
|
| 583 |
+
"step": 82
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 1.48,
|
| 587 |
+
"grad_norm": 0.5322052836418152,
|
| 588 |
+
"learning_rate": 4.8599999999999995e-05,
|
| 589 |
+
"loss": 2.923,
|
| 590 |
+
"step": 83
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 1.5,
|
| 594 |
+
"grad_norm": 0.5382172465324402,
|
| 595 |
+
"learning_rate": 4.9199999999999997e-05,
|
| 596 |
+
"loss": 2.9224,
|
| 597 |
+
"step": 84
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 1.52,
|
| 601 |
+
"grad_norm": 0.4733451008796692,
|
| 602 |
+
"learning_rate": 4.98e-05,
|
| 603 |
+
"loss": 2.9204,
|
| 604 |
+
"step": 85
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 1.54,
|
| 608 |
+
"grad_norm": 0.3383927345275879,
|
| 609 |
+
"learning_rate": 5.04e-05,
|
| 610 |
+
"loss": 2.9134,
|
| 611 |
+
"step": 86
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 1.55,
|
| 615 |
+
"grad_norm": 0.3574801981449127,
|
| 616 |
+
"learning_rate": 5.1e-05,
|
| 617 |
+
"loss": 2.9322,
|
| 618 |
+
"step": 87
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.57,
|
| 622 |
+
"grad_norm": 0.46184679865837097,
|
| 623 |
+
"learning_rate": 5.1599999999999994e-05,
|
| 624 |
+
"loss": 2.9262,
|
| 625 |
+
"step": 88
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.59,
|
| 629 |
+
"grad_norm": 0.5523853898048401,
|
| 630 |
+
"learning_rate": 5.2199999999999995e-05,
|
| 631 |
+
"loss": 2.8996,
|
| 632 |
+
"step": 89
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.61,
|
| 636 |
+
"grad_norm": 0.5265095829963684,
|
| 637 |
+
"learning_rate": 5.279999999999999e-05,
|
| 638 |
+
"loss": 2.9001,
|
| 639 |
+
"step": 90
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.62,
|
| 643 |
+
"grad_norm": 0.283716082572937,
|
| 644 |
+
"learning_rate": 5.339999999999999e-05,
|
| 645 |
+
"loss": 2.8979,
|
| 646 |
+
"step": 91
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.64,
|
| 650 |
+
"grad_norm": 0.33560270071029663,
|
| 651 |
+
"learning_rate": 5.399999999999999e-05,
|
| 652 |
+
"loss": 2.895,
|
| 653 |
+
"step": 92
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.66,
|
| 657 |
+
"grad_norm": 0.2877383828163147,
|
| 658 |
+
"learning_rate": 5.459999999999999e-05,
|
| 659 |
+
"loss": 2.9077,
|
| 660 |
+
"step": 93
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.68,
|
| 664 |
+
"grad_norm": 0.4475316107273102,
|
| 665 |
+
"learning_rate": 5.519999999999999e-05,
|
| 666 |
+
"loss": 2.9159,
|
| 667 |
+
"step": 94
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.7,
|
| 671 |
+
"grad_norm": 0.3262443244457245,
|
| 672 |
+
"learning_rate": 5.5799999999999994e-05,
|
| 673 |
+
"loss": 2.8856,
|
| 674 |
+
"step": 95
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.71,
|
| 678 |
+
"grad_norm": 0.2758294939994812,
|
| 679 |
+
"learning_rate": 5.6399999999999995e-05,
|
| 680 |
+
"loss": 2.8894,
|
| 681 |
+
"step": 96
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.73,
|
| 685 |
+
"grad_norm": 0.22843773663043976,
|
| 686 |
+
"learning_rate": 5.6999999999999996e-05,
|
| 687 |
+
"loss": 2.8889,
|
| 688 |
+
"step": 97
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.75,
|
| 692 |
+
"grad_norm": 0.21555787324905396,
|
| 693 |
+
"learning_rate": 5.76e-05,
|
| 694 |
+
"loss": 2.8882,
|
| 695 |
+
"step": 98
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.77,
|
| 699 |
+
"grad_norm": 0.21384727954864502,
|
| 700 |
+
"learning_rate": 5.82e-05,
|
| 701 |
+
"loss": 2.8905,
|
| 702 |
+
"step": 99
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.79,
|
| 706 |
+
"grad_norm": 0.3471096158027649,
|
| 707 |
+
"learning_rate": 5.88e-05,
|
| 708 |
+
"loss": 2.9166,
|
| 709 |
+
"step": 100
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.79,
|
| 713 |
+
"eval_cer": 1.0,
|
| 714 |
+
"eval_loss": 2.920933485031128,
|
| 715 |
+
"eval_runtime": 27.1302,
|
| 716 |
+
"eval_samples_per_second": 97.382,
|
| 717 |
+
"eval_steps_per_second": 1.548,
|
| 718 |
+
"eval_wer": 1.0,
|
| 719 |
+
"step": 100
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 1.8,
|
| 723 |
+
"grad_norm": 0.6376880407333374,
|
| 724 |
+
"learning_rate": 5.94e-05,
|
| 725 |
+
"loss": 2.8808,
|
| 726 |
+
"step": 101
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 1.82,
|
| 730 |
+
"grad_norm": 0.3349604308605194,
|
| 731 |
+
"learning_rate": 5.9999999999999995e-05,
|
| 732 |
+
"loss": 2.8767,
|
| 733 |
+
"step": 102
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 1.84,
|
| 737 |
+
"grad_norm": 0.25555065274238586,
|
| 738 |
+
"learning_rate": 6.0599999999999996e-05,
|
| 739 |
+
"loss": 2.8732,
|
| 740 |
+
"step": 103
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 1.86,
|
| 744 |
+
"grad_norm": 0.29420363903045654,
|
| 745 |
+
"learning_rate": 6.12e-05,
|
| 746 |
+
"loss": 2.8786,
|
| 747 |
+
"step": 104
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 1.88,
|
| 751 |
+
"grad_norm": 0.305449515581131,
|
| 752 |
+
"learning_rate": 6.18e-05,
|
| 753 |
+
"loss": 2.8787,
|
| 754 |
+
"step": 105
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 1.89,
|
| 758 |
+
"grad_norm": 0.44796454906463623,
|
| 759 |
+
"learning_rate": 6.239999999999999e-05,
|
| 760 |
+
"loss": 2.9139,
|
| 761 |
+
"step": 106
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 1.91,
|
| 765 |
+
"grad_norm": 0.8851379156112671,
|
| 766 |
+
"learning_rate": 6.299999999999999e-05,
|
| 767 |
+
"loss": 2.8776,
|
| 768 |
+
"step": 107
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 1.93,
|
| 772 |
+
"grad_norm": 0.4452502131462097,
|
| 773 |
+
"learning_rate": 6.359999999999999e-05,
|
| 774 |
+
"loss": 2.873,
|
| 775 |
+
"step": 108
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 1.95,
|
| 779 |
+
"grad_norm": 0.3026847243309021,
|
| 780 |
+
"learning_rate": 6.419999999999999e-05,
|
| 781 |
+
"loss": 2.869,
|
| 782 |
+
"step": 109
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.96,
|
| 786 |
+
"grad_norm": 0.6372184157371521,
|
| 787 |
+
"learning_rate": 6.479999999999999e-05,
|
| 788 |
+
"loss": 2.8696,
|
| 789 |
+
"step": 110
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 1.98,
|
| 793 |
+
"grad_norm": 0.5308623313903809,
|
| 794 |
+
"learning_rate": 6.539999999999999e-05,
|
| 795 |
+
"loss": 2.8793,
|
| 796 |
+
"step": 111
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 2.0,
|
| 800 |
+
"grad_norm": 0.23390421271324158,
|
| 801 |
+
"learning_rate": 6.599999999999999e-05,
|
| 802 |
+
"loss": 2.9017,
|
| 803 |
+
"step": 112
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 2.02,
|
| 807 |
+
"grad_norm": 0.7915804982185364,
|
| 808 |
+
"learning_rate": 6.659999999999999e-05,
|
| 809 |
+
"loss": 2.8696,
|
| 810 |
+
"step": 113
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 2.04,
|
| 814 |
+
"grad_norm": 0.7361267805099487,
|
| 815 |
+
"learning_rate": 6.72e-05,
|
| 816 |
+
"loss": 2.8697,
|
| 817 |
+
"step": 114
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 2.05,
|
| 821 |
+
"grad_norm": 0.29742589592933655,
|
| 822 |
+
"learning_rate": 6.78e-05,
|
| 823 |
+
"loss": 2.8667,
|
| 824 |
+
"step": 115
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.07,
|
| 828 |
+
"grad_norm": 0.3403497338294983,
|
| 829 |
+
"learning_rate": 6.84e-05,
|
| 830 |
+
"loss": 2.868,
|
| 831 |
+
"step": 116
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 2.09,
|
| 835 |
+
"grad_norm": 0.48037293553352356,
|
| 836 |
+
"learning_rate": 6.9e-05,
|
| 837 |
+
"loss": 2.8735,
|
| 838 |
+
"step": 117
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 2.11,
|
| 842 |
+
"grad_norm": 0.2523638904094696,
|
| 843 |
+
"learning_rate": 6.96e-05,
|
| 844 |
+
"loss": 2.8907,
|
| 845 |
+
"step": 118
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 2.12,
|
| 849 |
+
"grad_norm": 0.5274858474731445,
|
| 850 |
+
"learning_rate": 7.02e-05,
|
| 851 |
+
"loss": 2.8753,
|
| 852 |
+
"step": 119
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 2.14,
|
| 856 |
+
"grad_norm": 0.897480309009552,
|
| 857 |
+
"learning_rate": 7.079999999999999e-05,
|
| 858 |
+
"loss": 2.8626,
|
| 859 |
+
"step": 120
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 2.16,
|
| 863 |
+
"grad_norm": 0.4159209132194519,
|
| 864 |
+
"learning_rate": 7.139999999999999e-05,
|
| 865 |
+
"loss": 2.8615,
|
| 866 |
+
"step": 121
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.18,
|
| 870 |
+
"grad_norm": 0.42704087495803833,
|
| 871 |
+
"learning_rate": 7.199999999999999e-05,
|
| 872 |
+
"loss": 2.8638,
|
| 873 |
+
"step": 122
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 2.2,
|
| 877 |
+
"grad_norm": 0.5404195189476013,
|
| 878 |
+
"learning_rate": 7.259999999999999e-05,
|
| 879 |
+
"loss": 2.8668,
|
| 880 |
+
"step": 123
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 2.21,
|
| 884 |
+
"grad_norm": 0.19115819036960602,
|
| 885 |
+
"learning_rate": 7.319999999999999e-05,
|
| 886 |
+
"loss": 2.8642,
|
| 887 |
+
"step": 124
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 2.23,
|
| 891 |
+
"grad_norm": 0.8274586796760559,
|
| 892 |
+
"learning_rate": 7.379999999999999e-05,
|
| 893 |
+
"loss": 2.8839,
|
| 894 |
+
"step": 125
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 2.25,
|
| 898 |
+
"grad_norm": 0.8492292761802673,
|
| 899 |
+
"learning_rate": 7.439999999999999e-05,
|
| 900 |
+
"loss": 2.8577,
|
| 901 |
+
"step": 126
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 2.27,
|
| 905 |
+
"grad_norm": 0.15917377173900604,
|
| 906 |
+
"learning_rate": 7.5e-05,
|
| 907 |
+
"loss": 2.8539,
|
| 908 |
+
"step": 127
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 2.29,
|
| 912 |
+
"grad_norm": 0.8107254505157471,
|
| 913 |
+
"learning_rate": 7.56e-05,
|
| 914 |
+
"loss": 2.8577,
|
| 915 |
+
"step": 128
|
| 916 |
+
},
|
| 917 |
+
{
|
| 918 |
+
"epoch": 2.3,
|
| 919 |
+
"grad_norm": 0.8370546102523804,
|
| 920 |
+
"learning_rate": 7.62e-05,
|
| 921 |
+
"loss": 2.8616,
|
| 922 |
+
"step": 129
|
| 923 |
+
},
|
| 924 |
+
{
|
| 925 |
+
"epoch": 2.32,
|
| 926 |
+
"grad_norm": 0.3829539716243744,
|
| 927 |
+
"learning_rate": 7.68e-05,
|
| 928 |
+
"loss": 2.8597,
|
| 929 |
+
"step": 130
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 2.34,
|
| 933 |
+
"grad_norm": 0.7102254033088684,
|
| 934 |
+
"learning_rate": 7.74e-05,
|
| 935 |
+
"loss": 2.8837,
|
| 936 |
+
"step": 131
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 2.36,
|
| 940 |
+
"grad_norm": 1.2454396486282349,
|
| 941 |
+
"learning_rate": 7.8e-05,
|
| 942 |
+
"loss": 2.8633,
|
| 943 |
+
"step": 132
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 2.38,
|
| 947 |
+
"grad_norm": 0.8274021744728088,
|
| 948 |
+
"learning_rate": 7.86e-05,
|
| 949 |
+
"loss": 2.8548,
|
| 950 |
+
"step": 133
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 2.39,
|
| 954 |
+
"grad_norm": 0.14240995049476624,
|
| 955 |
+
"learning_rate": 7.92e-05,
|
| 956 |
+
"loss": 2.8516,
|
| 957 |
+
"step": 134
|
| 958 |
+
},
|
| 959 |
+
{
|
| 960 |
+
"epoch": 2.41,
|
| 961 |
+
"grad_norm": 0.5579639673233032,
|
| 962 |
+
"learning_rate": 7.98e-05,
|
| 963 |
+
"loss": 2.8549,
|
| 964 |
+
"step": 135
|
| 965 |
+
},
|
| 966 |
+
{
|
| 967 |
+
"epoch": 2.43,
|
| 968 |
+
"grad_norm": 0.7262759208679199,
|
| 969 |
+
"learning_rate": 8.04e-05,
|
| 970 |
+
"loss": 2.8608,
|
| 971 |
+
"step": 136
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 2.45,
|
| 975 |
+
"grad_norm": 0.2196040004491806,
|
| 976 |
+
"learning_rate": 8.1e-05,
|
| 977 |
+
"loss": 2.8894,
|
| 978 |
+
"step": 137
|
| 979 |
+
},
|
| 980 |
+
{
|
| 981 |
+
"epoch": 2.46,
|
| 982 |
+
"grad_norm": 1.0430012941360474,
|
| 983 |
+
"learning_rate": 8.16e-05,
|
| 984 |
+
"loss": 2.8597,
|
| 985 |
+
"step": 138
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 2.48,
|
| 989 |
+
"grad_norm": 1.0342605113983154,
|
| 990 |
+
"learning_rate": 8.22e-05,
|
| 991 |
+
"loss": 2.8544,
|
| 992 |
+
"step": 139
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 2.5,
|
| 996 |
+
"grad_norm": 0.4288654327392578,
|
| 997 |
+
"learning_rate": 8.28e-05,
|
| 998 |
+
"loss": 2.8482,
|
| 999 |
+
"step": 140
|
| 1000 |
+
},
|
| 1001 |
+
{
|
| 1002 |
+
"epoch": 2.52,
|
| 1003 |
+
"grad_norm": 0.38442760705947876,
|
| 1004 |
+
"learning_rate": 8.34e-05,
|
| 1005 |
+
"loss": 2.8521,
|
| 1006 |
+
"step": 141
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 2.54,
|
| 1010 |
+
"grad_norm": 0.8211755752563477,
|
| 1011 |
+
"learning_rate": 8.4e-05,
|
| 1012 |
+
"loss": 2.8542,
|
| 1013 |
+
"step": 142
|
| 1014 |
+
},
|
| 1015 |
+
{
|
| 1016 |
+
"epoch": 2.55,
|
| 1017 |
+
"grad_norm": 0.523823082447052,
|
| 1018 |
+
"learning_rate": 8.459999999999998e-05,
|
| 1019 |
+
"loss": 2.8627,
|
| 1020 |
+
"step": 143
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 2.57,
|
| 1024 |
+
"grad_norm": 0.5984336733818054,
|
| 1025 |
+
"learning_rate": 8.519999999999998e-05,
|
| 1026 |
+
"loss": 2.8704,
|
| 1027 |
+
"step": 144
|
| 1028 |
+
},
|
| 1029 |
+
{
|
| 1030 |
+
"epoch": 2.59,
|
| 1031 |
+
"grad_norm": 0.904820442199707,
|
| 1032 |
+
"learning_rate": 8.579999999999998e-05,
|
| 1033 |
+
"loss": 2.8508,
|
| 1034 |
+
"step": 145
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 2.61,
|
| 1038 |
+
"grad_norm": 0.3196875751018524,
|
| 1039 |
+
"learning_rate": 8.639999999999999e-05,
|
| 1040 |
+
"loss": 2.8515,
|
| 1041 |
+
"step": 146
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 2.62,
|
| 1045 |
+
"grad_norm": 0.4112975597381592,
|
| 1046 |
+
"learning_rate": 8.699999999999999e-05,
|
| 1047 |
+
"loss": 2.8506,
|
| 1048 |
+
"step": 147
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 2.64,
|
| 1052 |
+
"grad_norm": 0.7180864214897156,
|
| 1053 |
+
"learning_rate": 8.759999999999999e-05,
|
| 1054 |
+
"loss": 2.8535,
|
| 1055 |
+
"step": 148
|
| 1056 |
+
},
|
| 1057 |
+
{
|
| 1058 |
+
"epoch": 2.66,
|
| 1059 |
+
"grad_norm": 0.22370034456253052,
|
| 1060 |
+
"learning_rate": 8.819999999999999e-05,
|
| 1061 |
+
"loss": 2.8556,
|
| 1062 |
+
"step": 149
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 2.68,
|
| 1066 |
+
"grad_norm": 0.5966680645942688,
|
| 1067 |
+
"learning_rate": 8.879999999999999e-05,
|
| 1068 |
+
"loss": 2.8687,
|
| 1069 |
+
"step": 150
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 2.7,
|
| 1073 |
+
"grad_norm": 0.6786354780197144,
|
| 1074 |
+
"learning_rate": 8.939999999999999e-05,
|
| 1075 |
+
"loss": 2.8495,
|
| 1076 |
+
"step": 151
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 2.71,
|
| 1080 |
+
"grad_norm": 0.12561751902103424,
|
| 1081 |
+
"learning_rate": 8.999999999999999e-05,
|
| 1082 |
+
"loss": 2.8437,
|
| 1083 |
+
"step": 152
|
| 1084 |
+
},
|
| 1085 |
+
{
|
| 1086 |
+
"epoch": 2.73,
|
| 1087 |
+
"grad_norm": 0.5226555466651917,
|
| 1088 |
+
"learning_rate": 9.059999999999999e-05,
|
| 1089 |
+
"loss": 2.8497,
|
| 1090 |
+
"step": 153
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 2.75,
|
| 1094 |
+
"grad_norm": 0.46070218086242676,
|
| 1095 |
+
"learning_rate": 9.12e-05,
|
| 1096 |
+
"loss": 2.8518,
|
| 1097 |
+
"step": 154
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 2.77,
|
| 1101 |
+
"grad_norm": 0.09678909927606583,
|
| 1102 |
+
"learning_rate": 9.18e-05,
|
| 1103 |
+
"loss": 2.8467,
|
| 1104 |
+
"step": 155
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 2.79,
|
| 1108 |
+
"grad_norm": 0.7364938259124756,
|
| 1109 |
+
"learning_rate": 9.24e-05,
|
| 1110 |
+
"loss": 2.875,
|
| 1111 |
+
"step": 156
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 2.8,
|
| 1115 |
+
"grad_norm": 0.4200565814971924,
|
| 1116 |
+
"learning_rate": 9.3e-05,
|
| 1117 |
+
"loss": 2.8391,
|
| 1118 |
+
"step": 157
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 2.82,
|
| 1122 |
+
"grad_norm": 0.2178226113319397,
|
| 1123 |
+
"learning_rate": 9.36e-05,
|
| 1124 |
+
"loss": 2.836,
|
| 1125 |
+
"step": 158
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 2.84,
|
| 1129 |
+
"grad_norm": 0.4477235972881317,
|
| 1130 |
+
"learning_rate": 9.419999999999999e-05,
|
| 1131 |
+
"loss": 2.8443,
|
| 1132 |
+
"step": 159
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 2.86,
|
| 1136 |
+
"grad_norm": 0.19233360886573792,
|
| 1137 |
+
"learning_rate": 9.479999999999999e-05,
|
| 1138 |
+
"loss": 2.8395,
|
| 1139 |
+
"step": 160
|
| 1140 |
+
},
|
| 1141 |
+
{
|
| 1142 |
+
"epoch": 2.88,
|
| 1143 |
+
"grad_norm": 0.3035629391670227,
|
| 1144 |
+
"learning_rate": 9.539999999999999e-05,
|
| 1145 |
+
"loss": 2.8422,
|
| 1146 |
+
"step": 161
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 2.89,
|
| 1150 |
+
"grad_norm": 0.25835344195365906,
|
| 1151 |
+
"learning_rate": 9.599999999999999e-05,
|
| 1152 |
+
"loss": 2.8615,
|
| 1153 |
+
"step": 162
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 2.91,
|
| 1157 |
+
"grad_norm": 0.31518957018852234,
|
| 1158 |
+
"learning_rate": 9.659999999999999e-05,
|
| 1159 |
+
"loss": 2.8357,
|
| 1160 |
+
"step": 163
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 2.93,
|
| 1164 |
+
"grad_norm": 0.1827758401632309,
|
| 1165 |
+
"learning_rate": 9.719999999999999e-05,
|
| 1166 |
+
"loss": 2.8341,
|
| 1167 |
+
"step": 164
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 2.95,
|
| 1171 |
+
"grad_norm": 0.3359813690185547,
|
| 1172 |
+
"learning_rate": 9.779999999999999e-05,
|
| 1173 |
+
"loss": 2.8345,
|
| 1174 |
+
"step": 165
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 2.96,
|
| 1178 |
+
"grad_norm": 0.16261117160320282,
|
| 1179 |
+
"learning_rate": 9.839999999999999e-05,
|
| 1180 |
+
"loss": 2.8392,
|
| 1181 |
+
"step": 166
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 2.98,
|
| 1185 |
+
"grad_norm": 0.3922206163406372,
|
| 1186 |
+
"learning_rate": 9.9e-05,
|
| 1187 |
+
"loss": 2.8415,
|
| 1188 |
+
"step": 167
|
| 1189 |
+
},
|
| 1190 |
+
{
|
| 1191 |
+
"epoch": 3.0,
|
| 1192 |
+
"grad_norm": 0.18259093165397644,
|
| 1193 |
+
"learning_rate": 9.96e-05,
|
| 1194 |
+
"loss": 2.8537,
|
| 1195 |
+
"step": 168
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"epoch": 3.0,
|
| 1199 |
+
"step": 168,
|
| 1200 |
+
"total_flos": 3.362627535247927e+19,
|
| 1201 |
+
"train_loss": 4.4063241723037905,
|
| 1202 |
+
"train_runtime": 1105.5076,
|
| 1203 |
+
"train_samples_per_second": 77.443,
|
| 1204 |
+
"train_steps_per_second": 0.152
|
| 1205 |
+
}
|
| 1206 |
+
],
|
| 1207 |
+
"logging_steps": 1.0,
|
| 1208 |
+
"max_steps": 168,
|
| 1209 |
+
"num_input_tokens_seen": 0,
|
| 1210 |
+
"num_train_epochs": 3,
|
| 1211 |
+
"save_steps": 400,
|
| 1212 |
+
"total_flos": 3.362627535247927e+19,
|
| 1213 |
+
"train_batch_size": 64,
|
| 1214 |
+
"trial_name": null,
|
| 1215 |
+
"trial_params": null
|
| 1216 |
+
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4539
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea94823cb00e477ae3b82c03c2f82c4c604dcdf83e5ee9d3dada83137ed68c9e
|
| 3 |
size 4539
|