End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"eval_cer": 0.07765640286883035,
|
| 4 |
+
"eval_loss": 0.30630072951316833,
|
| 5 |
+
"eval_runtime": 217.4345,
|
| 6 |
+
"eval_samples": 5425,
|
| 7 |
+
"eval_samples_per_second": 24.95,
|
| 8 |
+
"eval_steps_per_second": 3.123,
|
| 9 |
+
"eval_wer": 0.38515273502249586,
|
| 10 |
+
"train_loss": 0.9527478711016767,
|
| 11 |
+
"train_runtime": 101833.8069,
|
| 12 |
+
"train_samples": 19726,
|
| 13 |
+
"train_samples_per_second": 19.371,
|
| 14 |
+
"train_steps_per_second": 0.151
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"eval_cer": 0.07765640286883035,
|
| 4 |
+
"eval_loss": 0.30630072951316833,
|
| 5 |
+
"eval_runtime": 217.4345,
|
| 6 |
+
"eval_samples": 5425,
|
| 7 |
+
"eval_samples_per_second": 24.95,
|
| 8 |
+
"eval_steps_per_second": 3.123,
|
| 9 |
+
"eval_wer": 0.38515273502249586
|
| 10 |
+
}
|
runs/Feb06_18-52-28_job-0a778896-a7e2-46e9-bcf5-016f91f242cf/events.out.tfevents.1644275949.job-0a778896-a7e2-46e9-bcf5-016f91f242cf.841782.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ea10dd975cbcebc10c4a94057cb6d1b559046748402b9ed86df5415543268ab8
|
| 3 |
+
size 405
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 100.0,
|
| 3 |
+
"train_loss": 0.9527478711016767,
|
| 4 |
+
"train_runtime": 101833.8069,
|
| 5 |
+
"train_samples": 19726,
|
| 6 |
+
"train_samples_per_second": 19.371,
|
| 7 |
+
"train_steps_per_second": 0.151
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1249 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 99.99837925445705,
|
| 5 |
+
"global_step": 15400,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.65,
|
| 12 |
+
"learning_rate": 5.940000000000001e-06,
|
| 13 |
+
"loss": 11.9235,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.3,
|
| 18 |
+
"learning_rate": 1.1940000000000001e-05,
|
| 19 |
+
"loss": 5.1849,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 1.95,
|
| 24 |
+
"learning_rate": 1.794e-05,
|
| 25 |
+
"loss": 3.7405,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 2.6,
|
| 30 |
+
"learning_rate": 2.394e-05,
|
| 31 |
+
"loss": 3.3052,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 3.25,
|
| 36 |
+
"learning_rate": 2.994e-05,
|
| 37 |
+
"loss": 3.1401,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 3.25,
|
| 42 |
+
"eval_cer": 1.0,
|
| 43 |
+
"eval_loss": 3.114637851715088,
|
| 44 |
+
"eval_runtime": 219.3333,
|
| 45 |
+
"eval_samples_per_second": 24.734,
|
| 46 |
+
"eval_steps_per_second": 3.096,
|
| 47 |
+
"eval_wer": 1.0,
|
| 48 |
+
"step": 500
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 3.89,
|
| 52 |
+
"learning_rate": 2.98006711409396e-05,
|
| 53 |
+
"loss": 3.0799,
|
| 54 |
+
"step": 600
|
| 55 |
+
},
|
| 56 |
+
{
|
| 57 |
+
"epoch": 4.54,
|
| 58 |
+
"learning_rate": 2.9599328859060405e-05,
|
| 59 |
+
"loss": 3.0534,
|
| 60 |
+
"step": 700
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 5.19,
|
| 64 |
+
"learning_rate": 2.9397986577181207e-05,
|
| 65 |
+
"loss": 2.994,
|
| 66 |
+
"step": 800
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 5.84,
|
| 70 |
+
"learning_rate": 2.9196644295302013e-05,
|
| 71 |
+
"loss": 2.9428,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 6.49,
|
| 76 |
+
"learning_rate": 2.8995302013422818e-05,
|
| 77 |
+
"loss": 2.7484,
|
| 78 |
+
"step": 1000
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 6.49,
|
| 82 |
+
"eval_cer": 0.706927100586246,
|
| 83 |
+
"eval_loss": 2.284235954284668,
|
| 84 |
+
"eval_runtime": 222.6226,
|
| 85 |
+
"eval_samples_per_second": 24.369,
|
| 86 |
+
"eval_steps_per_second": 3.05,
|
| 87 |
+
"eval_wer": 1.0065119583234667,
|
| 88 |
+
"step": 1000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 7.14,
|
| 92 |
+
"learning_rate": 2.8793959731543624e-05,
|
| 93 |
+
"loss": 2.0772,
|
| 94 |
+
"step": 1100
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 7.79,
|
| 98 |
+
"learning_rate": 2.859261744966443e-05,
|
| 99 |
+
"loss": 1.4967,
|
| 100 |
+
"step": 1200
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 8.44,
|
| 104 |
+
"learning_rate": 2.8391275167785235e-05,
|
| 105 |
+
"loss": 1.2731,
|
| 106 |
+
"step": 1300
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 9.09,
|
| 110 |
+
"learning_rate": 2.818993288590604e-05,
|
| 111 |
+
"loss": 1.1742,
|
| 112 |
+
"step": 1400
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 9.74,
|
| 116 |
+
"learning_rate": 2.7988590604026846e-05,
|
| 117 |
+
"loss": 1.0899,
|
| 118 |
+
"step": 1500
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 9.74,
|
| 122 |
+
"eval_cer": 0.13506382454221202,
|
| 123 |
+
"eval_loss": 0.5414122343063354,
|
| 124 |
+
"eval_runtime": 217.7702,
|
| 125 |
+
"eval_samples_per_second": 24.912,
|
| 126 |
+
"eval_steps_per_second": 3.118,
|
| 127 |
+
"eval_wer": 0.6124792801326071,
|
| 128 |
+
"step": 1500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 10.39,
|
| 132 |
+
"learning_rate": 2.778724832214765e-05,
|
| 133 |
+
"loss": 1.0544,
|
| 134 |
+
"step": 1600
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 11.04,
|
| 138 |
+
"learning_rate": 2.7585906040268457e-05,
|
| 139 |
+
"loss": 1.0284,
|
| 140 |
+
"step": 1700
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 11.69,
|
| 144 |
+
"learning_rate": 2.7384563758389263e-05,
|
| 145 |
+
"loss": 0.9865,
|
| 146 |
+
"step": 1800
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 12.34,
|
| 150 |
+
"learning_rate": 2.7183221476510065e-05,
|
| 151 |
+
"loss": 0.9705,
|
| 152 |
+
"step": 1900
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 12.99,
|
| 156 |
+
"learning_rate": 2.698187919463087e-05,
|
| 157 |
+
"loss": 0.9465,
|
| 158 |
+
"step": 2000
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 12.99,
|
| 162 |
+
"eval_cer": 0.12229891609980818,
|
| 163 |
+
"eval_loss": 0.4565887749195099,
|
| 164 |
+
"eval_runtime": 219.4838,
|
| 165 |
+
"eval_samples_per_second": 24.717,
|
| 166 |
+
"eval_steps_per_second": 3.094,
|
| 167 |
+
"eval_wer": 0.5634619938432394,
|
| 168 |
+
"step": 2000
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 13.64,
|
| 172 |
+
"learning_rate": 2.6780536912751676e-05,
|
| 173 |
+
"loss": 0.9395,
|
| 174 |
+
"step": 2100
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 14.29,
|
| 178 |
+
"learning_rate": 2.6579194630872482e-05,
|
| 179 |
+
"loss": 0.9216,
|
| 180 |
+
"step": 2200
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 14.93,
|
| 184 |
+
"learning_rate": 2.6377852348993287e-05,
|
| 185 |
+
"loss": 0.9011,
|
| 186 |
+
"step": 2300
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 15.58,
|
| 190 |
+
"learning_rate": 2.6176510067114093e-05,
|
| 191 |
+
"loss": 0.9011,
|
| 192 |
+
"step": 2400
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 16.23,
|
| 196 |
+
"learning_rate": 2.59751677852349e-05,
|
| 197 |
+
"loss": 0.8771,
|
| 198 |
+
"step": 2500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 16.23,
|
| 202 |
+
"eval_cer": 0.11608208983968754,
|
| 203 |
+
"eval_loss": 0.42124298214912415,
|
| 204 |
+
"eval_runtime": 218.6532,
|
| 205 |
+
"eval_samples_per_second": 24.811,
|
| 206 |
+
"eval_steps_per_second": 3.105,
|
| 207 |
+
"eval_wer": 0.5365557660430973,
|
| 208 |
+
"step": 2500
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 16.88,
|
| 212 |
+
"learning_rate": 2.5773825503355704e-05,
|
| 213 |
+
"loss": 0.8719,
|
| 214 |
+
"step": 2600
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 17.53,
|
| 218 |
+
"learning_rate": 2.557248322147651e-05,
|
| 219 |
+
"loss": 0.8616,
|
| 220 |
+
"step": 2700
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 18.18,
|
| 224 |
+
"learning_rate": 2.5371140939597315e-05,
|
| 225 |
+
"loss": 0.8573,
|
| 226 |
+
"step": 2800
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 18.83,
|
| 230 |
+
"learning_rate": 2.5169798657718124e-05,
|
| 231 |
+
"loss": 0.8467,
|
| 232 |
+
"step": 2900
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 19.48,
|
| 236 |
+
"learning_rate": 2.496845637583893e-05,
|
| 237 |
+
"loss": 0.8346,
|
| 238 |
+
"step": 3000
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 19.48,
|
| 242 |
+
"eval_cer": 0.11022733400611667,
|
| 243 |
+
"eval_loss": 0.3994467258453369,
|
| 244 |
+
"eval_runtime": 213.3023,
|
| 245 |
+
"eval_samples_per_second": 25.433,
|
| 246 |
+
"eval_steps_per_second": 3.183,
|
| 247 |
+
"eval_wer": 0.5143855079327492,
|
| 248 |
+
"step": 3000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 20.13,
|
| 252 |
+
"learning_rate": 2.4767114093959732e-05,
|
| 253 |
+
"loss": 0.8386,
|
| 254 |
+
"step": 3100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 20.78,
|
| 258 |
+
"learning_rate": 2.4565771812080538e-05,
|
| 259 |
+
"loss": 0.8283,
|
| 260 |
+
"step": 3200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 21.43,
|
| 264 |
+
"learning_rate": 2.4364429530201343e-05,
|
| 265 |
+
"loss": 0.8198,
|
| 266 |
+
"step": 3300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 22.08,
|
| 270 |
+
"learning_rate": 2.416308724832215e-05,
|
| 271 |
+
"loss": 0.8164,
|
| 272 |
+
"step": 3400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 22.73,
|
| 276 |
+
"learning_rate": 2.3961744966442955e-05,
|
| 277 |
+
"loss": 0.8127,
|
| 278 |
+
"step": 3500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 22.73,
|
| 282 |
+
"eval_cer": 0.10510827446479058,
|
| 283 |
+
"eval_loss": 0.3818908929824829,
|
| 284 |
+
"eval_runtime": 218.4137,
|
| 285 |
+
"eval_samples_per_second": 24.838,
|
| 286 |
+
"eval_steps_per_second": 3.109,
|
| 287 |
+
"eval_wer": 0.49437603599336966,
|
| 288 |
+
"step": 3500
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 23.38,
|
| 292 |
+
"learning_rate": 2.376040268456376e-05,
|
| 293 |
+
"loss": 0.8081,
|
| 294 |
+
"step": 3600
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 24.03,
|
| 298 |
+
"learning_rate": 2.3559060402684566e-05,
|
| 299 |
+
"loss": 0.7988,
|
| 300 |
+
"step": 3700
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 24.67,
|
| 304 |
+
"learning_rate": 2.335771812080537e-05,
|
| 305 |
+
"loss": 0.7903,
|
| 306 |
+
"step": 3800
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 25.32,
|
| 310 |
+
"learning_rate": 2.3156375838926177e-05,
|
| 311 |
+
"loss": 0.788,
|
| 312 |
+
"step": 3900
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 25.97,
|
| 316 |
+
"learning_rate": 2.2955033557046982e-05,
|
| 317 |
+
"loss": 0.7833,
|
| 318 |
+
"step": 4000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 25.97,
|
| 322 |
+
"eval_cer": 0.10109853708140422,
|
| 323 |
+
"eval_loss": 0.3704679310321808,
|
| 324 |
+
"eval_runtime": 216.2288,
|
| 325 |
+
"eval_samples_per_second": 25.089,
|
| 326 |
+
"eval_steps_per_second": 3.14,
|
| 327 |
+
"eval_wer": 0.4797537295761307,
|
| 328 |
+
"step": 4000
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 26.62,
|
| 332 |
+
"learning_rate": 2.2753691275167788e-05,
|
| 333 |
+
"loss": 0.7724,
|
| 334 |
+
"step": 4100
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 27.27,
|
| 338 |
+
"learning_rate": 2.255234899328859e-05,
|
| 339 |
+
"loss": 0.7695,
|
| 340 |
+
"step": 4200
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 27.92,
|
| 344 |
+
"learning_rate": 2.2351006711409396e-05,
|
| 345 |
+
"loss": 0.7653,
|
| 346 |
+
"step": 4300
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 28.57,
|
| 350 |
+
"learning_rate": 2.2151677852348994e-05,
|
| 351 |
+
"loss": 0.7728,
|
| 352 |
+
"step": 4400
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 29.22,
|
| 356 |
+
"learning_rate": 2.19503355704698e-05,
|
| 357 |
+
"loss": 0.7603,
|
| 358 |
+
"step": 4500
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 29.22,
|
| 362 |
+
"eval_cer": 0.0992381113790261,
|
| 363 |
+
"eval_loss": 0.36611661314964294,
|
| 364 |
+
"eval_runtime": 217.7866,
|
| 365 |
+
"eval_samples_per_second": 24.91,
|
| 366 |
+
"eval_steps_per_second": 3.118,
|
| 367 |
+
"eval_wer": 0.4704001894387876,
|
| 368 |
+
"step": 4500
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 29.87,
|
| 372 |
+
"learning_rate": 2.1748993288590605e-05,
|
| 373 |
+
"loss": 0.7563,
|
| 374 |
+
"step": 4600
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 30.52,
|
| 378 |
+
"learning_rate": 2.154765100671141e-05,
|
| 379 |
+
"loss": 0.7618,
|
| 380 |
+
"step": 4700
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 31.17,
|
| 384 |
+
"learning_rate": 2.1346308724832217e-05,
|
| 385 |
+
"loss": 0.7467,
|
| 386 |
+
"step": 4800
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 31.82,
|
| 390 |
+
"learning_rate": 2.114697986577181e-05,
|
| 391 |
+
"loss": 0.7514,
|
| 392 |
+
"step": 4900
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 32.47,
|
| 396 |
+
"learning_rate": 2.0945637583892617e-05,
|
| 397 |
+
"loss": 0.7424,
|
| 398 |
+
"step": 5000
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 32.47,
|
| 402 |
+
"eval_cer": 0.09569444337449638,
|
| 403 |
+
"eval_loss": 0.3528956174850464,
|
| 404 |
+
"eval_runtime": 218.9072,
|
| 405 |
+
"eval_samples_per_second": 24.782,
|
| 406 |
+
"eval_steps_per_second": 3.102,
|
| 407 |
+
"eval_wer": 0.4577314705185887,
|
| 408 |
+
"step": 5000
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 33.12,
|
| 412 |
+
"learning_rate": 2.0744295302013423e-05,
|
| 413 |
+
"loss": 0.748,
|
| 414 |
+
"step": 5100
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 33.76,
|
| 418 |
+
"learning_rate": 2.054295302013423e-05,
|
| 419 |
+
"loss": 0.7357,
|
| 420 |
+
"step": 5200
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"epoch": 34.41,
|
| 424 |
+
"learning_rate": 2.0341610738255034e-05,
|
| 425 |
+
"loss": 0.7357,
|
| 426 |
+
"step": 5300
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 35.06,
|
| 430 |
+
"learning_rate": 2.014026845637584e-05,
|
| 431 |
+
"loss": 0.735,
|
| 432 |
+
"step": 5400
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 35.71,
|
| 436 |
+
"learning_rate": 1.9938926174496645e-05,
|
| 437 |
+
"loss": 0.7251,
|
| 438 |
+
"step": 5500
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 35.71,
|
| 442 |
+
"eval_cer": 0.09283254627953377,
|
| 443 |
+
"eval_loss": 0.34103500843048096,
|
| 444 |
+
"eval_runtime": 218.8353,
|
| 445 |
+
"eval_samples_per_second": 24.79,
|
| 446 |
+
"eval_steps_per_second": 3.103,
|
| 447 |
+
"eval_wer": 0.4472827373904807,
|
| 448 |
+
"step": 5500
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 36.36,
|
| 452 |
+
"learning_rate": 1.973758389261745e-05,
|
| 453 |
+
"loss": 0.7306,
|
| 454 |
+
"step": 5600
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 37.01,
|
| 458 |
+
"learning_rate": 1.9536241610738256e-05,
|
| 459 |
+
"loss": 0.7185,
|
| 460 |
+
"step": 5700
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 37.66,
|
| 464 |
+
"learning_rate": 1.9334899328859062e-05,
|
| 465 |
+
"loss": 0.7135,
|
| 466 |
+
"step": 5800
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 38.31,
|
| 470 |
+
"learning_rate": 1.9133557046979864e-05,
|
| 471 |
+
"loss": 0.726,
|
| 472 |
+
"step": 5900
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 38.96,
|
| 476 |
+
"learning_rate": 1.893221476510067e-05,
|
| 477 |
+
"loss": 0.7106,
|
| 478 |
+
"step": 6000
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 38.96,
|
| 482 |
+
"eval_cer": 0.09190425933486893,
|
| 483 |
+
"eval_loss": 0.3401394486427307,
|
| 484 |
+
"eval_runtime": 218.6772,
|
| 485 |
+
"eval_samples_per_second": 24.808,
|
| 486 |
+
"eval_steps_per_second": 3.105,
|
| 487 |
+
"eval_wer": 0.4427539663746152,
|
| 488 |
+
"step": 6000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 39.61,
|
| 492 |
+
"learning_rate": 1.8730872483221475e-05,
|
| 493 |
+
"loss": 0.7157,
|
| 494 |
+
"step": 6100
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 40.26,
|
| 498 |
+
"learning_rate": 1.852953020134228e-05,
|
| 499 |
+
"loss": 0.7055,
|
| 500 |
+
"step": 6200
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 40.91,
|
| 504 |
+
"learning_rate": 1.8328187919463086e-05,
|
| 505 |
+
"loss": 0.7108,
|
| 506 |
+
"step": 6300
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 41.56,
|
| 510 |
+
"learning_rate": 1.8126845637583892e-05,
|
| 511 |
+
"loss": 0.705,
|
| 512 |
+
"step": 6400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 42.21,
|
| 516 |
+
"learning_rate": 1.7925503355704698e-05,
|
| 517 |
+
"loss": 0.7027,
|
| 518 |
+
"step": 6500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 42.21,
|
| 522 |
+
"eval_cer": 0.09045597762866982,
|
| 523 |
+
"eval_loss": 0.3354834318161011,
|
| 524 |
+
"eval_runtime": 219.2927,
|
| 525 |
+
"eval_samples_per_second": 24.739,
|
| 526 |
+
"eval_steps_per_second": 3.096,
|
| 527 |
+
"eval_wer": 0.4352652143026285,
|
| 528 |
+
"step": 6500
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 42.86,
|
| 532 |
+
"learning_rate": 1.7724161073825503e-05,
|
| 533 |
+
"loss": 0.7103,
|
| 534 |
+
"step": 6600
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 43.51,
|
| 538 |
+
"learning_rate": 1.752281879194631e-05,
|
| 539 |
+
"loss": 0.7011,
|
| 540 |
+
"step": 6700
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 44.16,
|
| 544 |
+
"learning_rate": 1.7321476510067114e-05,
|
| 545 |
+
"loss": 0.7004,
|
| 546 |
+
"step": 6800
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 44.8,
|
| 550 |
+
"learning_rate": 1.712013422818792e-05,
|
| 551 |
+
"loss": 0.702,
|
| 552 |
+
"step": 6900
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 45.45,
|
| 556 |
+
"learning_rate": 1.6918791946308722e-05,
|
| 557 |
+
"loss": 0.6927,
|
| 558 |
+
"step": 7000
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 45.45,
|
| 562 |
+
"eval_cer": 0.08853392291751727,
|
| 563 |
+
"eval_loss": 0.33077025413513184,
|
| 564 |
+
"eval_runtime": 215.9851,
|
| 565 |
+
"eval_samples_per_second": 25.117,
|
| 566 |
+
"eval_steps_per_second": 3.144,
|
| 567 |
+
"eval_wer": 0.4296412502959981,
|
| 568 |
+
"step": 7000
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 46.1,
|
| 572 |
+
"learning_rate": 1.6717449664429528e-05,
|
| 573 |
+
"loss": 0.691,
|
| 574 |
+
"step": 7100
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 46.75,
|
| 578 |
+
"learning_rate": 1.6516107382550333e-05,
|
| 579 |
+
"loss": 0.6833,
|
| 580 |
+
"step": 7200
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 47.4,
|
| 584 |
+
"learning_rate": 1.631476510067114e-05,
|
| 585 |
+
"loss": 0.692,
|
| 586 |
+
"step": 7300
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 48.05,
|
| 590 |
+
"learning_rate": 1.6113422818791948e-05,
|
| 591 |
+
"loss": 0.6871,
|
| 592 |
+
"step": 7400
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 48.7,
|
| 596 |
+
"learning_rate": 1.5914093959731546e-05,
|
| 597 |
+
"loss": 0.6828,
|
| 598 |
+
"step": 7500
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 48.7,
|
| 602 |
+
"eval_cer": 0.08625364959286336,
|
| 603 |
+
"eval_loss": 0.324627548456192,
|
| 604 |
+
"eval_runtime": 221.033,
|
| 605 |
+
"eval_samples_per_second": 24.544,
|
| 606 |
+
"eval_steps_per_second": 3.072,
|
| 607 |
+
"eval_wer": 0.42043570921146106,
|
| 608 |
+
"step": 7500
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 49.35,
|
| 612 |
+
"learning_rate": 1.5712751677852352e-05,
|
| 613 |
+
"loss": 0.6789,
|
| 614 |
+
"step": 7600
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 50.0,
|
| 618 |
+
"learning_rate": 1.5511409395973158e-05,
|
| 619 |
+
"loss": 0.6811,
|
| 620 |
+
"step": 7700
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 50.65,
|
| 624 |
+
"learning_rate": 1.531006711409396e-05,
|
| 625 |
+
"loss": 0.683,
|
| 626 |
+
"step": 7800
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 51.3,
|
| 630 |
+
"learning_rate": 1.5108724832214764e-05,
|
| 631 |
+
"loss": 0.6765,
|
| 632 |
+
"step": 7900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 51.95,
|
| 636 |
+
"learning_rate": 1.4907382550335571e-05,
|
| 637 |
+
"loss": 0.6706,
|
| 638 |
+
"step": 8000
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 51.95,
|
| 642 |
+
"eval_cer": 0.08681216248488163,
|
| 643 |
+
"eval_loss": 0.32503771781921387,
|
| 644 |
+
"eval_runtime": 215.1259,
|
| 645 |
+
"eval_samples_per_second": 25.218,
|
| 646 |
+
"eval_steps_per_second": 3.156,
|
| 647 |
+
"eval_wer": 0.42327729102533745,
|
| 648 |
+
"step": 8000
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 52.6,
|
| 652 |
+
"learning_rate": 1.4706040268456375e-05,
|
| 653 |
+
"loss": 0.6777,
|
| 654 |
+
"step": 8100
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 53.25,
|
| 658 |
+
"learning_rate": 1.450469798657718e-05,
|
| 659 |
+
"loss": 0.6754,
|
| 660 |
+
"step": 8200
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 53.89,
|
| 664 |
+
"learning_rate": 1.4303355704697986e-05,
|
| 665 |
+
"loss": 0.6675,
|
| 666 |
+
"step": 8300
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 54.54,
|
| 670 |
+
"learning_rate": 1.4102013422818792e-05,
|
| 671 |
+
"loss": 0.6627,
|
| 672 |
+
"step": 8400
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 55.19,
|
| 676 |
+
"learning_rate": 1.3900671140939599e-05,
|
| 677 |
+
"loss": 0.6629,
|
| 678 |
+
"step": 8500
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 55.19,
|
| 682 |
+
"eval_cer": 0.0849055150259227,
|
| 683 |
+
"eval_loss": 0.3263927102088928,
|
| 684 |
+
"eval_runtime": 213.0987,
|
| 685 |
+
"eval_samples_per_second": 25.458,
|
| 686 |
+
"eval_steps_per_second": 3.186,
|
| 687 |
+
"eval_wer": 0.4158773383850343,
|
| 688 |
+
"step": 8500
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 55.84,
|
| 692 |
+
"learning_rate": 1.3699328859060405e-05,
|
| 693 |
+
"loss": 0.6632,
|
| 694 |
+
"step": 8600
|
| 695 |
+
},
|
| 696 |
+
{
|
| 697 |
+
"epoch": 56.49,
|
| 698 |
+
"learning_rate": 1.3497986577181208e-05,
|
| 699 |
+
"loss": 0.6558,
|
| 700 |
+
"step": 8700
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 57.14,
|
| 704 |
+
"learning_rate": 1.3296644295302014e-05,
|
| 705 |
+
"loss": 0.6691,
|
| 706 |
+
"step": 8800
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 57.79,
|
| 710 |
+
"learning_rate": 1.309530201342282e-05,
|
| 711 |
+
"loss": 0.6633,
|
| 712 |
+
"step": 8900
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 58.44,
|
| 716 |
+
"learning_rate": 1.2893959731543625e-05,
|
| 717 |
+
"loss": 0.6556,
|
| 718 |
+
"step": 9000
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 58.44,
|
| 722 |
+
"eval_cer": 0.08353812139374003,
|
| 723 |
+
"eval_loss": 0.3212815821170807,
|
| 724 |
+
"eval_runtime": 220.228,
|
| 725 |
+
"eval_samples_per_second": 24.634,
|
| 726 |
+
"eval_steps_per_second": 3.083,
|
| 727 |
+
"eval_wer": 0.40995737627279183,
|
| 728 |
+
"step": 9000
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 59.09,
|
| 732 |
+
"learning_rate": 1.269261744966443e-05,
|
| 733 |
+
"loss": 0.6584,
|
| 734 |
+
"step": 9100
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 59.74,
|
| 738 |
+
"learning_rate": 1.2491275167785236e-05,
|
| 739 |
+
"loss": 0.6537,
|
| 740 |
+
"step": 9200
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 60.39,
|
| 744 |
+
"learning_rate": 1.228993288590604e-05,
|
| 745 |
+
"loss": 0.6633,
|
| 746 |
+
"step": 9300
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 61.04,
|
| 750 |
+
"learning_rate": 1.2088590604026846e-05,
|
| 751 |
+
"loss": 0.6474,
|
| 752 |
+
"step": 9400
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 61.69,
|
| 756 |
+
"learning_rate": 1.1887248322147651e-05,
|
| 757 |
+
"loss": 0.6484,
|
| 758 |
+
"step": 9500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 61.69,
|
| 762 |
+
"eval_cer": 0.0836613794112889,
|
| 763 |
+
"eval_loss": 0.31816166639328003,
|
| 764 |
+
"eval_runtime": 216.5133,
|
| 765 |
+
"eval_samples_per_second": 25.056,
|
| 766 |
+
"eval_steps_per_second": 3.136,
|
| 767 |
+
"eval_wer": 0.41241416054937247,
|
| 768 |
+
"step": 9500
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 62.34,
|
| 772 |
+
"learning_rate": 1.1687919463087248e-05,
|
| 773 |
+
"loss": 0.6547,
|
| 774 |
+
"step": 9600
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 62.99,
|
| 778 |
+
"learning_rate": 1.1486577181208054e-05,
|
| 779 |
+
"loss": 0.6481,
|
| 780 |
+
"step": 9700
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 63.64,
|
| 784 |
+
"learning_rate": 1.128523489932886e-05,
|
| 785 |
+
"loss": 0.648,
|
| 786 |
+
"step": 9800
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 64.29,
|
| 790 |
+
"learning_rate": 1.1083892617449665e-05,
|
| 791 |
+
"loss": 0.6471,
|
| 792 |
+
"step": 9900
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 64.93,
|
| 796 |
+
"learning_rate": 1.088255033557047e-05,
|
| 797 |
+
"loss": 0.6407,
|
| 798 |
+
"step": 10000
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 64.93,
|
| 802 |
+
"eval_cer": 0.08248272461847792,
|
| 803 |
+
"eval_loss": 0.3171332776546478,
|
| 804 |
+
"eval_runtime": 217.1179,
|
| 805 |
+
"eval_samples_per_second": 24.986,
|
| 806 |
+
"eval_steps_per_second": 3.127,
|
| 807 |
+
"eval_wer": 0.4050142079090694,
|
| 808 |
+
"step": 10000
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 65.58,
|
| 812 |
+
"learning_rate": 1.0681208053691274e-05,
|
| 813 |
+
"loss": 0.6446,
|
| 814 |
+
"step": 10100
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 66.23,
|
| 818 |
+
"learning_rate": 1.047986577181208e-05,
|
| 819 |
+
"loss": 0.6383,
|
| 820 |
+
"step": 10200
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"epoch": 66.88,
|
| 824 |
+
"learning_rate": 1.0278523489932886e-05,
|
| 825 |
+
"loss": 0.6413,
|
| 826 |
+
"step": 10300
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 67.53,
|
| 830 |
+
"learning_rate": 1.0077181208053691e-05,
|
| 831 |
+
"loss": 0.6494,
|
| 832 |
+
"step": 10400
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 68.18,
|
| 836 |
+
"learning_rate": 9.875838926174497e-06,
|
| 837 |
+
"loss": 0.6375,
|
| 838 |
+
"step": 10500
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 68.18,
|
| 842 |
+
"eval_cer": 0.0822362085833802,
|
| 843 |
+
"eval_loss": 0.31498104333877563,
|
| 844 |
+
"eval_runtime": 218.6829,
|
| 845 |
+
"eval_samples_per_second": 24.808,
|
| 846 |
+
"eval_steps_per_second": 3.105,
|
| 847 |
+
"eval_wer": 0.4038598152971821,
|
| 848 |
+
"step": 10500
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 68.83,
|
| 852 |
+
"learning_rate": 9.6744966442953e-06,
|
| 853 |
+
"loss": 0.6359,
|
| 854 |
+
"step": 10600
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 69.48,
|
| 858 |
+
"learning_rate": 9.473154362416108e-06,
|
| 859 |
+
"loss": 0.638,
|
| 860 |
+
"step": 10700
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 70.13,
|
| 864 |
+
"learning_rate": 9.271812080536914e-06,
|
| 865 |
+
"loss": 0.6405,
|
| 866 |
+
"step": 10800
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 70.78,
|
| 870 |
+
"learning_rate": 9.070469798657719e-06,
|
| 871 |
+
"loss": 0.6388,
|
| 872 |
+
"step": 10900
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 71.43,
|
| 876 |
+
"learning_rate": 8.869127516778525e-06,
|
| 877 |
+
"loss": 0.6363,
|
| 878 |
+
"step": 11000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 71.43,
|
| 882 |
+
"eval_cer": 0.08095355483826237,
|
| 883 |
+
"eval_loss": 0.3129253089427948,
|
| 884 |
+
"eval_runtime": 218.9143,
|
| 885 |
+
"eval_samples_per_second": 24.781,
|
| 886 |
+
"eval_steps_per_second": 3.102,
|
| 887 |
+
"eval_wer": 0.3991238456073881,
|
| 888 |
+
"step": 11000
|
| 889 |
+
},
|
| 890 |
+
{
|
| 891 |
+
"epoch": 72.08,
|
| 892 |
+
"learning_rate": 8.66778523489933e-06,
|
| 893 |
+
"loss": 0.6369,
|
| 894 |
+
"step": 11100
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 72.73,
|
| 898 |
+
"learning_rate": 8.466442953020134e-06,
|
| 899 |
+
"loss": 0.635,
|
| 900 |
+
"step": 11200
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 73.38,
|
| 904 |
+
"learning_rate": 8.26510067114094e-06,
|
| 905 |
+
"loss": 0.6337,
|
| 906 |
+
"step": 11300
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 74.03,
|
| 910 |
+
"learning_rate": 8.063758389261745e-06,
|
| 911 |
+
"loss": 0.6308,
|
| 912 |
+
"step": 11400
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 74.67,
|
| 916 |
+
"learning_rate": 7.862416107382551e-06,
|
| 917 |
+
"loss": 0.6307,
|
| 918 |
+
"step": 11500
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 74.67,
|
| 922 |
+
"eval_cer": 0.08074170512060026,
|
| 923 |
+
"eval_loss": 0.3114279508590698,
|
| 924 |
+
"eval_runtime": 219.6542,
|
| 925 |
+
"eval_samples_per_second": 24.698,
|
| 926 |
+
"eval_steps_per_second": 3.091,
|
| 927 |
+
"eval_wer": 0.3986206488278475,
|
| 928 |
+
"step": 11500
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 75.32,
|
| 932 |
+
"learning_rate": 7.661073825503357e-06,
|
| 933 |
+
"loss": 0.6335,
|
| 934 |
+
"step": 11600
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 75.97,
|
| 938 |
+
"learning_rate": 7.459731543624161e-06,
|
| 939 |
+
"loss": 0.628,
|
| 940 |
+
"step": 11700
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 76.62,
|
| 944 |
+
"learning_rate": 7.260402684563759e-06,
|
| 945 |
+
"loss": 0.6324,
|
| 946 |
+
"step": 11800
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 77.27,
|
| 950 |
+
"learning_rate": 7.059060402684564e-06,
|
| 951 |
+
"loss": 0.6317,
|
| 952 |
+
"step": 11900
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 77.92,
|
| 956 |
+
"learning_rate": 6.857718120805369e-06,
|
| 957 |
+
"loss": 0.6232,
|
| 958 |
+
"step": 12000
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 77.92,
|
| 962 |
+
"eval_cer": 0.07899683380967422,
|
| 963 |
+
"eval_loss": 0.31030353903770447,
|
| 964 |
+
"eval_runtime": 220.4971,
|
| 965 |
+
"eval_samples_per_second": 24.603,
|
| 966 |
+
"eval_steps_per_second": 3.079,
|
| 967 |
+
"eval_wer": 0.3895335069855553,
|
| 968 |
+
"step": 12000
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 78.57,
|
| 972 |
+
"learning_rate": 6.656375838926175e-06,
|
| 973 |
+
"loss": 0.6295,
|
| 974 |
+
"step": 12100
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 79.22,
|
| 978 |
+
"learning_rate": 6.4550335570469795e-06,
|
| 979 |
+
"loss": 0.6234,
|
| 980 |
+
"step": 12200
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 79.87,
|
| 984 |
+
"learning_rate": 6.255704697986578e-06,
|
| 985 |
+
"loss": 0.6172,
|
| 986 |
+
"step": 12300
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 80.52,
|
| 990 |
+
"learning_rate": 6.054362416107383e-06,
|
| 991 |
+
"loss": 0.6203,
|
| 992 |
+
"step": 12400
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 81.17,
|
| 996 |
+
"learning_rate": 5.853020134228188e-06,
|
| 997 |
+
"loss": 0.6216,
|
| 998 |
+
"step": 12500
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 81.17,
|
| 1002 |
+
"eval_cer": 0.0789506120530934,
|
| 1003 |
+
"eval_loss": 0.30863967537879944,
|
| 1004 |
+
"eval_runtime": 218.4142,
|
| 1005 |
+
"eval_samples_per_second": 24.838,
|
| 1006 |
+
"eval_steps_per_second": 3.109,
|
| 1007 |
+
"eval_wer": 0.3891191096376983,
|
| 1008 |
+
"step": 12500
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 81.82,
|
| 1012 |
+
"learning_rate": 5.651677852348994e-06,
|
| 1013 |
+
"loss": 0.6203,
|
| 1014 |
+
"step": 12600
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 82.47,
|
| 1018 |
+
"learning_rate": 5.4503355704697986e-06,
|
| 1019 |
+
"loss": 0.6209,
|
| 1020 |
+
"step": 12700
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 83.12,
|
| 1024 |
+
"learning_rate": 5.248993288590604e-06,
|
| 1025 |
+
"loss": 0.6257,
|
| 1026 |
+
"step": 12800
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 83.76,
|
| 1030 |
+
"learning_rate": 5.04765100671141e-06,
|
| 1031 |
+
"loss": 0.6245,
|
| 1032 |
+
"step": 12900
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 84.41,
|
| 1036 |
+
"learning_rate": 4.8463087248322145e-06,
|
| 1037 |
+
"loss": 0.6174,
|
| 1038 |
+
"step": 13000
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 84.41,
|
| 1042 |
+
"eval_cer": 0.07851150536557558,
|
| 1043 |
+
"eval_loss": 0.3082079291343689,
|
| 1044 |
+
"eval_runtime": 215.0269,
|
| 1045 |
+
"eval_samples_per_second": 25.229,
|
| 1046 |
+
"eval_steps_per_second": 3.158,
|
| 1047 |
+
"eval_wer": 0.3880535164574947,
|
| 1048 |
+
"step": 13000
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 85.06,
|
| 1052 |
+
"learning_rate": 4.64496644295302e-06,
|
| 1053 |
+
"loss": 0.6222,
|
| 1054 |
+
"step": 13100
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 85.71,
|
| 1058 |
+
"learning_rate": 4.443624161073826e-06,
|
| 1059 |
+
"loss": 0.6113,
|
| 1060 |
+
"step": 13200
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 86.36,
|
| 1064 |
+
"learning_rate": 4.2422818791946304e-06,
|
| 1065 |
+
"loss": 0.6238,
|
| 1066 |
+
"step": 13300
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 87.01,
|
| 1070 |
+
"learning_rate": 4.040939597315437e-06,
|
| 1071 |
+
"loss": 0.618,
|
| 1072 |
+
"step": 13400
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 87.66,
|
| 1076 |
+
"learning_rate": 3.8395973154362425e-06,
|
| 1077 |
+
"loss": 0.6196,
|
| 1078 |
+
"step": 13500
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 87.66,
|
| 1082 |
+
"eval_cer": 0.07821876757389704,
|
| 1083 |
+
"eval_loss": 0.30590009689331055,
|
| 1084 |
+
"eval_runtime": 213.2382,
|
| 1085 |
+
"eval_samples_per_second": 25.441,
|
| 1086 |
+
"eval_steps_per_second": 3.184,
|
| 1087 |
+
"eval_wer": 0.3874911200568316,
|
| 1088 |
+
"step": 13500
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 88.31,
|
| 1092 |
+
"learning_rate": 3.6382550335570468e-06,
|
| 1093 |
+
"loss": 0.6174,
|
| 1094 |
+
"step": 13600
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 88.96,
|
| 1098 |
+
"learning_rate": 3.4369127516778524e-06,
|
| 1099 |
+
"loss": 0.6128,
|
| 1100 |
+
"step": 13700
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 89.61,
|
| 1104 |
+
"learning_rate": 3.235570469798658e-06,
|
| 1105 |
+
"loss": 0.6246,
|
| 1106 |
+
"step": 13800
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 90.26,
|
| 1110 |
+
"learning_rate": 3.034228187919463e-06,
|
| 1111 |
+
"loss": 0.6097,
|
| 1112 |
+
"step": 13900
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 90.91,
|
| 1116 |
+
"learning_rate": 2.8328859060402687e-06,
|
| 1117 |
+
"loss": 0.6174,
|
| 1118 |
+
"step": 14000
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 90.91,
|
| 1122 |
+
"eval_cer": 0.07799151060404132,
|
| 1123 |
+
"eval_loss": 0.30842480063438416,
|
| 1124 |
+
"eval_runtime": 212.6251,
|
| 1125 |
+
"eval_samples_per_second": 25.514,
|
| 1126 |
+
"eval_steps_per_second": 3.193,
|
| 1127 |
+
"eval_wer": 0.3862479280132607,
|
| 1128 |
+
"step": 14000
|
| 1129 |
+
},
|
| 1130 |
+
{
|
| 1131 |
+
"epoch": 91.56,
|
| 1132 |
+
"learning_rate": 2.631543624161074e-06,
|
| 1133 |
+
"loss": 0.6194,
|
| 1134 |
+
"step": 14100
|
| 1135 |
+
},
|
| 1136 |
+
{
|
| 1137 |
+
"epoch": 92.21,
|
| 1138 |
+
"learning_rate": 2.430201342281879e-06,
|
| 1139 |
+
"loss": 0.6167,
|
| 1140 |
+
"step": 14200
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 92.86,
|
| 1144 |
+
"learning_rate": 2.2288590604026842e-06,
|
| 1145 |
+
"loss": 0.614,
|
| 1146 |
+
"step": 14300
|
| 1147 |
+
},
|
| 1148 |
+
{
|
| 1149 |
+
"epoch": 93.51,
|
| 1150 |
+
"learning_rate": 2.0275167785234902e-06,
|
| 1151 |
+
"loss": 0.615,
|
| 1152 |
+
"step": 14400
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 94.16,
|
| 1156 |
+
"learning_rate": 1.8261744966442954e-06,
|
| 1157 |
+
"loss": 0.6169,
|
| 1158 |
+
"step": 14500
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 94.16,
|
| 1162 |
+
"eval_cer": 0.07787595621258926,
|
| 1163 |
+
"eval_loss": 0.30701127648353577,
|
| 1164 |
+
"eval_runtime": 215.1709,
|
| 1165 |
+
"eval_samples_per_second": 25.213,
|
| 1166 |
+
"eval_steps_per_second": 3.156,
|
| 1167 |
+
"eval_wer": 0.3859519299076486,
|
| 1168 |
+
"step": 14500
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 94.8,
|
| 1172 |
+
"learning_rate": 1.6248322147651008e-06,
|
| 1173 |
+
"loss": 0.6123,
|
| 1174 |
+
"step": 14600
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 95.45,
|
| 1178 |
+
"learning_rate": 1.423489932885906e-06,
|
| 1179 |
+
"loss": 0.6133,
|
| 1180 |
+
"step": 14700
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 96.1,
|
| 1184 |
+
"learning_rate": 1.2221476510067115e-06,
|
| 1185 |
+
"loss": 0.6068,
|
| 1186 |
+
"step": 14800
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 96.75,
|
| 1190 |
+
"learning_rate": 1.020805369127517e-06,
|
| 1191 |
+
"loss": 0.6135,
|
| 1192 |
+
"step": 14900
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 97.4,
|
| 1196 |
+
"learning_rate": 8.194630872483221e-07,
|
| 1197 |
+
"loss": 0.6166,
|
| 1198 |
+
"step": 15000
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 97.4,
|
| 1202 |
+
"eval_cer": 0.07776810544723402,
|
| 1203 |
+
"eval_loss": 0.30662447214126587,
|
| 1204 |
+
"eval_runtime": 214.9469,
|
| 1205 |
+
"eval_samples_per_second": 25.239,
|
| 1206 |
+
"eval_steps_per_second": 3.159,
|
| 1207 |
+
"eval_wer": 0.3855079327492304,
|
| 1208 |
+
"step": 15000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 98.05,
|
| 1212 |
+
"learning_rate": 6.181208053691276e-07,
|
| 1213 |
+
"loss": 0.6189,
|
| 1214 |
+
"step": 15100
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 98.7,
|
| 1218 |
+
"learning_rate": 4.167785234899329e-07,
|
| 1219 |
+
"loss": 0.6085,
|
| 1220 |
+
"step": 15200
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 99.35,
|
| 1224 |
+
"learning_rate": 2.1543624161073826e-07,
|
| 1225 |
+
"loss": 0.6135,
|
| 1226 |
+
"step": 15300
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 100.0,
|
| 1230 |
+
"learning_rate": 1.4093959731543625e-08,
|
| 1231 |
+
"loss": 0.6047,
|
| 1232 |
+
"step": 15400
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 100.0,
|
| 1236 |
+
"step": 15400,
|
| 1237 |
+
"total_flos": 2.652008062738907e+20,
|
| 1238 |
+
"train_loss": 0.9527478711016767,
|
| 1239 |
+
"train_runtime": 101833.8069,
|
| 1240 |
+
"train_samples_per_second": 19.371,
|
| 1241 |
+
"train_steps_per_second": 0.151
|
| 1242 |
+
}
|
| 1243 |
+
],
|
| 1244 |
+
"max_steps": 15400,
|
| 1245 |
+
"num_train_epochs": 100,
|
| 1246 |
+
"total_flos": 2.652008062738907e+20,
|
| 1247 |
+
"trial_name": null,
|
| 1248 |
+
"trial_params": null
|
| 1249 |
+
}
|