cheers again
Browse files- README.md +1 -1
- trainer_state.json +1180 -0
README.md
CHANGED
|
@@ -15,7 +15,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
# whisper2
|
| 17 |
|
| 18 |
-
This model is a fine-tuned version of [openai/whisper-tiny.en](https://huggingface.co/openai/whisper-tiny.en) on
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 0.5233
|
| 21 |
- Wer: 31.1083
|
|
|
|
| 15 |
|
| 16 |
# whisper2
|
| 17 |
|
| 18 |
+
This model is a fine-tuned version of [openai/whisper-tiny.en](https://huggingface.co/openai/whisper-tiny.en) on the tiny dataset.
|
| 19 |
It achieves the following results on the evaluation set:
|
| 20 |
- Loss: 0.5233
|
| 21 |
- Wer: 31.1083
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1180 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 28.05415617128463,
|
| 3 |
+
"best_model_checkpoint": "whisper2/checkpoint-430",
|
| 4 |
+
"epoch": 7.042253521126761,
|
| 5 |
+
"eval_steps": 10,
|
| 6 |
+
"global_step": 500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.07042253521126761,
|
| 13 |
+
"grad_norm": 43.82194137573242,
|
| 14 |
+
"learning_rate": 1.0000000000000001e-07,
|
| 15 |
+
"loss": 3.9547,
|
| 16 |
+
"step": 5
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.14084507042253522,
|
| 20 |
+
"grad_norm": 45.53117370605469,
|
| 21 |
+
"learning_rate": 2.0000000000000002e-07,
|
| 22 |
+
"loss": 3.9553,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.14084507042253522,
|
| 27 |
+
"eval_loss": 3.964555501937866,
|
| 28 |
+
"eval_runtime": 264.1292,
|
| 29 |
+
"eval_samples_per_second": 1.893,
|
| 30 |
+
"eval_steps_per_second": 0.239,
|
| 31 |
+
"eval_wer": 74.87405541561712,
|
| 32 |
+
"step": 10
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.2112676056338028,
|
| 36 |
+
"grad_norm": 46.162776947021484,
|
| 37 |
+
"learning_rate": 3.0000000000000004e-07,
|
| 38 |
+
"loss": 3.882,
|
| 39 |
+
"step": 15
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.28169014084507044,
|
| 43 |
+
"grad_norm": 46.07596206665039,
|
| 44 |
+
"learning_rate": 4.0000000000000003e-07,
|
| 45 |
+
"loss": 3.9548,
|
| 46 |
+
"step": 20
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.28169014084507044,
|
| 50 |
+
"eval_loss": 3.8793957233428955,
|
| 51 |
+
"eval_runtime": 256.6948,
|
| 52 |
+
"eval_samples_per_second": 1.948,
|
| 53 |
+
"eval_steps_per_second": 0.245,
|
| 54 |
+
"eval_wer": 77.67632241813602,
|
| 55 |
+
"step": 20
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.352112676056338,
|
| 59 |
+
"grad_norm": 45.13657760620117,
|
| 60 |
+
"learning_rate": 5.000000000000001e-07,
|
| 61 |
+
"loss": 3.9469,
|
| 62 |
+
"step": 25
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.4225352112676056,
|
| 66 |
+
"grad_norm": 44.565940856933594,
|
| 67 |
+
"learning_rate": 6.000000000000001e-07,
|
| 68 |
+
"loss": 3.8127,
|
| 69 |
+
"step": 30
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.4225352112676056,
|
| 73 |
+
"eval_loss": 3.740476608276367,
|
| 74 |
+
"eval_runtime": 257.3378,
|
| 75 |
+
"eval_samples_per_second": 1.943,
|
| 76 |
+
"eval_steps_per_second": 0.245,
|
| 77 |
+
"eval_wer": 76.4168765743073,
|
| 78 |
+
"step": 30
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 0.49295774647887325,
|
| 82 |
+
"grad_norm": 44.24871826171875,
|
| 83 |
+
"learning_rate": 7.000000000000001e-07,
|
| 84 |
+
"loss": 3.7507,
|
| 85 |
+
"step": 35
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.5633802816901409,
|
| 89 |
+
"grad_norm": 42.1717529296875,
|
| 90 |
+
"learning_rate": 8.000000000000001e-07,
|
| 91 |
+
"loss": 3.6178,
|
| 92 |
+
"step": 40
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.5633802816901409,
|
| 96 |
+
"eval_loss": 3.5547332763671875,
|
| 97 |
+
"eval_runtime": 256.8157,
|
| 98 |
+
"eval_samples_per_second": 1.947,
|
| 99 |
+
"eval_steps_per_second": 0.245,
|
| 100 |
+
"eval_wer": 75.31486146095719,
|
| 101 |
+
"step": 40
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.6338028169014085,
|
| 105 |
+
"grad_norm": 44.667205810546875,
|
| 106 |
+
"learning_rate": 9.000000000000001e-07,
|
| 107 |
+
"loss": 3.4825,
|
| 108 |
+
"step": 45
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.704225352112676,
|
| 112 |
+
"grad_norm": 43.76979064941406,
|
| 113 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 114 |
+
"loss": 3.3992,
|
| 115 |
+
"step": 50
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.704225352112676,
|
| 119 |
+
"eval_loss": 3.323503255844116,
|
| 120 |
+
"eval_runtime": 255.1809,
|
| 121 |
+
"eval_samples_per_second": 1.959,
|
| 122 |
+
"eval_steps_per_second": 0.247,
|
| 123 |
+
"eval_wer": 70.27707808564232,
|
| 124 |
+
"step": 50
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.7746478873239436,
|
| 128 |
+
"grad_norm": 41.28179168701172,
|
| 129 |
+
"learning_rate": 1.1e-06,
|
| 130 |
+
"loss": 3.3124,
|
| 131 |
+
"step": 55
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.8450704225352113,
|
| 135 |
+
"grad_norm": 40.813392639160156,
|
| 136 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 137 |
+
"loss": 3.1416,
|
| 138 |
+
"step": 60
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 0.8450704225352113,
|
| 142 |
+
"eval_loss": 3.040179491043091,
|
| 143 |
+
"eval_runtime": 255.4069,
|
| 144 |
+
"eval_samples_per_second": 1.958,
|
| 145 |
+
"eval_steps_per_second": 0.247,
|
| 146 |
+
"eval_wer": 67.85264483627203,
|
| 147 |
+
"step": 60
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 0.9154929577464789,
|
| 151 |
+
"grad_norm": 40.00282287597656,
|
| 152 |
+
"learning_rate": 1.3e-06,
|
| 153 |
+
"loss": 2.88,
|
| 154 |
+
"step": 65
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.9859154929577465,
|
| 158 |
+
"grad_norm": 40.60588455200195,
|
| 159 |
+
"learning_rate": 1.4000000000000001e-06,
|
| 160 |
+
"loss": 2.8052,
|
| 161 |
+
"step": 70
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.9859154929577465,
|
| 165 |
+
"eval_loss": 2.6852359771728516,
|
| 166 |
+
"eval_runtime": 254.3541,
|
| 167 |
+
"eval_samples_per_second": 1.966,
|
| 168 |
+
"eval_steps_per_second": 0.248,
|
| 169 |
+
"eval_wer": 65.96347607052897,
|
| 170 |
+
"step": 70
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 1.056338028169014,
|
| 174 |
+
"grad_norm": 44.205726623535156,
|
| 175 |
+
"learning_rate": 1.5e-06,
|
| 176 |
+
"loss": 2.4894,
|
| 177 |
+
"step": 75
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 1.1267605633802817,
|
| 181 |
+
"grad_norm": 40.45851516723633,
|
| 182 |
+
"learning_rate": 1.6000000000000001e-06,
|
| 183 |
+
"loss": 2.3513,
|
| 184 |
+
"step": 80
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.1267605633802817,
|
| 188 |
+
"eval_loss": 2.223541021347046,
|
| 189 |
+
"eval_runtime": 256.3144,
|
| 190 |
+
"eval_samples_per_second": 1.951,
|
| 191 |
+
"eval_steps_per_second": 0.246,
|
| 192 |
+
"eval_wer": 68.3249370277078,
|
| 193 |
+
"step": 80
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.1971830985915493,
|
| 197 |
+
"grad_norm": 37.049591064453125,
|
| 198 |
+
"learning_rate": 1.7000000000000002e-06,
|
| 199 |
+
"loss": 2.2021,
|
| 200 |
+
"step": 85
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.267605633802817,
|
| 204 |
+
"grad_norm": 32.15092468261719,
|
| 205 |
+
"learning_rate": 1.8000000000000001e-06,
|
| 206 |
+
"loss": 1.893,
|
| 207 |
+
"step": 90
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 1.267605633802817,
|
| 211 |
+
"eval_loss": 1.6707711219787598,
|
| 212 |
+
"eval_runtime": 254.2495,
|
| 213 |
+
"eval_samples_per_second": 1.967,
|
| 214 |
+
"eval_steps_per_second": 0.248,
|
| 215 |
+
"eval_wer": 63.822418136020154,
|
| 216 |
+
"step": 90
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 1.3380281690140845,
|
| 220 |
+
"grad_norm": 29.11300277709961,
|
| 221 |
+
"learning_rate": 1.9000000000000002e-06,
|
| 222 |
+
"loss": 1.6227,
|
| 223 |
+
"step": 95
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 1.408450704225352,
|
| 227 |
+
"grad_norm": 19.466663360595703,
|
| 228 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 229 |
+
"loss": 1.2871,
|
| 230 |
+
"step": 100
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.408450704225352,
|
| 234 |
+
"eval_loss": 1.164486050605774,
|
| 235 |
+
"eval_runtime": 254.5126,
|
| 236 |
+
"eval_samples_per_second": 1.965,
|
| 237 |
+
"eval_steps_per_second": 0.248,
|
| 238 |
+
"eval_wer": 63.25566750629723,
|
| 239 |
+
"step": 100
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.4788732394366197,
|
| 243 |
+
"grad_norm": 15.238794326782227,
|
| 244 |
+
"learning_rate": 2.1000000000000002e-06,
|
| 245 |
+
"loss": 1.09,
|
| 246 |
+
"step": 105
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 1.5492957746478875,
|
| 250 |
+
"grad_norm": 10.725071907043457,
|
| 251 |
+
"learning_rate": 2.2e-06,
|
| 252 |
+
"loss": 0.9146,
|
| 253 |
+
"step": 110
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 1.5492957746478875,
|
| 257 |
+
"eval_loss": 0.8784648776054382,
|
| 258 |
+
"eval_runtime": 256.185,
|
| 259 |
+
"eval_samples_per_second": 1.952,
|
| 260 |
+
"eval_steps_per_second": 0.246,
|
| 261 |
+
"eval_wer": 56.83249370277078,
|
| 262 |
+
"step": 110
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 1.619718309859155,
|
| 266 |
+
"grad_norm": 7.8202009201049805,
|
| 267 |
+
"learning_rate": 2.3000000000000004e-06,
|
| 268 |
+
"loss": 0.8882,
|
| 269 |
+
"step": 115
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.6901408450704225,
|
| 273 |
+
"grad_norm": 8.60835075378418,
|
| 274 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 275 |
+
"loss": 0.8044,
|
| 276 |
+
"step": 120
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 1.6901408450704225,
|
| 280 |
+
"eval_loss": 0.7906607985496521,
|
| 281 |
+
"eval_runtime": 255.9613,
|
| 282 |
+
"eval_samples_per_second": 1.953,
|
| 283 |
+
"eval_steps_per_second": 0.246,
|
| 284 |
+
"eval_wer": 46.977329974811084,
|
| 285 |
+
"step": 120
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.76056338028169,
|
| 289 |
+
"grad_norm": 9.780821800231934,
|
| 290 |
+
"learning_rate": 2.5e-06,
|
| 291 |
+
"loss": 0.6849,
|
| 292 |
+
"step": 125
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 1.8309859154929577,
|
| 296 |
+
"grad_norm": 9.33056926727295,
|
| 297 |
+
"learning_rate": 2.6e-06,
|
| 298 |
+
"loss": 0.6634,
|
| 299 |
+
"step": 130
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.8309859154929577,
|
| 303 |
+
"eval_loss": 0.7425487637519836,
|
| 304 |
+
"eval_runtime": 255.5846,
|
| 305 |
+
"eval_samples_per_second": 1.956,
|
| 306 |
+
"eval_steps_per_second": 0.246,
|
| 307 |
+
"eval_wer": 47.48110831234257,
|
| 308 |
+
"step": 130
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.9014084507042255,
|
| 312 |
+
"grad_norm": 8.966361999511719,
|
| 313 |
+
"learning_rate": 2.7000000000000004e-06,
|
| 314 |
+
"loss": 0.7421,
|
| 315 |
+
"step": 135
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.971830985915493,
|
| 319 |
+
"grad_norm": 7.636435031890869,
|
| 320 |
+
"learning_rate": 2.8000000000000003e-06,
|
| 321 |
+
"loss": 0.6722,
|
| 322 |
+
"step": 140
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 1.971830985915493,
|
| 326 |
+
"eval_loss": 0.7099979519844055,
|
| 327 |
+
"eval_runtime": 253.8483,
|
| 328 |
+
"eval_samples_per_second": 1.97,
|
| 329 |
+
"eval_steps_per_second": 0.248,
|
| 330 |
+
"eval_wer": 45.90680100755667,
|
| 331 |
+
"step": 140
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 2.0422535211267605,
|
| 335 |
+
"grad_norm": 8.085705757141113,
|
| 336 |
+
"learning_rate": 2.9e-06,
|
| 337 |
+
"loss": 0.6865,
|
| 338 |
+
"step": 145
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.112676056338028,
|
| 342 |
+
"grad_norm": 8.131012916564941,
|
| 343 |
+
"learning_rate": 3e-06,
|
| 344 |
+
"loss": 0.6823,
|
| 345 |
+
"step": 150
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.112676056338028,
|
| 349 |
+
"eval_loss": 0.6854478120803833,
|
| 350 |
+
"eval_runtime": 255.8245,
|
| 351 |
+
"eval_samples_per_second": 1.954,
|
| 352 |
+
"eval_steps_per_second": 0.246,
|
| 353 |
+
"eval_wer": 42.41183879093199,
|
| 354 |
+
"step": 150
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.183098591549296,
|
| 358 |
+
"grad_norm": 8.054609298706055,
|
| 359 |
+
"learning_rate": 3.1000000000000004e-06,
|
| 360 |
+
"loss": 0.6001,
|
| 361 |
+
"step": 155
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.2535211267605635,
|
| 365 |
+
"grad_norm": 6.9759063720703125,
|
| 366 |
+
"learning_rate": 3.2000000000000003e-06,
|
| 367 |
+
"loss": 0.5802,
|
| 368 |
+
"step": 160
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.2535211267605635,
|
| 372 |
+
"eval_loss": 0.6659273505210876,
|
| 373 |
+
"eval_runtime": 254.855,
|
| 374 |
+
"eval_samples_per_second": 1.962,
|
| 375 |
+
"eval_steps_per_second": 0.247,
|
| 376 |
+
"eval_wer": 40.42821158690176,
|
| 377 |
+
"step": 160
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.323943661971831,
|
| 381 |
+
"grad_norm": 8.077522277832031,
|
| 382 |
+
"learning_rate": 3.3000000000000006e-06,
|
| 383 |
+
"loss": 0.6065,
|
| 384 |
+
"step": 165
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.3943661971830985,
|
| 388 |
+
"grad_norm": 6.6878228187561035,
|
| 389 |
+
"learning_rate": 3.4000000000000005e-06,
|
| 390 |
+
"loss": 0.6084,
|
| 391 |
+
"step": 170
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.3943661971830985,
|
| 395 |
+
"eval_loss": 0.6503352522850037,
|
| 396 |
+
"eval_runtime": 253.7567,
|
| 397 |
+
"eval_samples_per_second": 1.97,
|
| 398 |
+
"eval_steps_per_second": 0.248,
|
| 399 |
+
"eval_wer": 40.8375314861461,
|
| 400 |
+
"step": 170
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 2.464788732394366,
|
| 404 |
+
"grad_norm": 7.941697597503662,
|
| 405 |
+
"learning_rate": 3.5e-06,
|
| 406 |
+
"loss": 0.5972,
|
| 407 |
+
"step": 175
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 2.535211267605634,
|
| 411 |
+
"grad_norm": 7.986533164978027,
|
| 412 |
+
"learning_rate": 3.6000000000000003e-06,
|
| 413 |
+
"loss": 0.6038,
|
| 414 |
+
"step": 180
|
| 415 |
+
},
|
| 416 |
+
{
|
| 417 |
+
"epoch": 2.535211267605634,
|
| 418 |
+
"eval_loss": 0.6345599889755249,
|
| 419 |
+
"eval_runtime": 254.9306,
|
| 420 |
+
"eval_samples_per_second": 1.961,
|
| 421 |
+
"eval_steps_per_second": 0.247,
|
| 422 |
+
"eval_wer": 41.49874055415617,
|
| 423 |
+
"step": 180
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 2.6056338028169015,
|
| 427 |
+
"grad_norm": 6.744418144226074,
|
| 428 |
+
"learning_rate": 3.7e-06,
|
| 429 |
+
"loss": 0.5007,
|
| 430 |
+
"step": 185
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 2.676056338028169,
|
| 434 |
+
"grad_norm": 6.323821544647217,
|
| 435 |
+
"learning_rate": 3.8000000000000005e-06,
|
| 436 |
+
"loss": 0.5095,
|
| 437 |
+
"step": 190
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 2.676056338028169,
|
| 441 |
+
"eval_loss": 0.6247134804725647,
|
| 442 |
+
"eval_runtime": 257.1561,
|
| 443 |
+
"eval_samples_per_second": 1.944,
|
| 444 |
+
"eval_steps_per_second": 0.245,
|
| 445 |
+
"eval_wer": 42.03400503778337,
|
| 446 |
+
"step": 190
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 2.7464788732394365,
|
| 450 |
+
"grad_norm": 6.979465961456299,
|
| 451 |
+
"learning_rate": 3.900000000000001e-06,
|
| 452 |
+
"loss": 0.5943,
|
| 453 |
+
"step": 195
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 2.816901408450704,
|
| 457 |
+
"grad_norm": 6.675357818603516,
|
| 458 |
+
"learning_rate": 4.000000000000001e-06,
|
| 459 |
+
"loss": 0.5251,
|
| 460 |
+
"step": 200
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 2.816901408450704,
|
| 464 |
+
"eval_loss": 0.6154741644859314,
|
| 465 |
+
"eval_runtime": 255.2235,
|
| 466 |
+
"eval_samples_per_second": 1.959,
|
| 467 |
+
"eval_steps_per_second": 0.247,
|
| 468 |
+
"eval_wer": 39.357682619647356,
|
| 469 |
+
"step": 200
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 2.887323943661972,
|
| 473 |
+
"grad_norm": 6.802981853485107,
|
| 474 |
+
"learning_rate": 4.1e-06,
|
| 475 |
+
"loss": 0.5528,
|
| 476 |
+
"step": 205
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.9577464788732395,
|
| 480 |
+
"grad_norm": 6.836462497711182,
|
| 481 |
+
"learning_rate": 4.2000000000000004e-06,
|
| 482 |
+
"loss": 0.5699,
|
| 483 |
+
"step": 210
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 2.9577464788732395,
|
| 487 |
+
"eval_loss": 0.6045908331871033,
|
| 488 |
+
"eval_runtime": 254.5675,
|
| 489 |
+
"eval_samples_per_second": 1.964,
|
| 490 |
+
"eval_steps_per_second": 0.247,
|
| 491 |
+
"eval_wer": 38.350125944584384,
|
| 492 |
+
"step": 210
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 3.028169014084507,
|
| 496 |
+
"grad_norm": 6.114952087402344,
|
| 497 |
+
"learning_rate": 4.3e-06,
|
| 498 |
+
"loss": 0.478,
|
| 499 |
+
"step": 215
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 3.0985915492957745,
|
| 503 |
+
"grad_norm": 5.803236961364746,
|
| 504 |
+
"learning_rate": 4.4e-06,
|
| 505 |
+
"loss": 0.4839,
|
| 506 |
+
"step": 220
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 3.0985915492957745,
|
| 510 |
+
"eval_loss": 0.5944731831550598,
|
| 511 |
+
"eval_runtime": 254.5629,
|
| 512 |
+
"eval_samples_per_second": 1.964,
|
| 513 |
+
"eval_steps_per_second": 0.247,
|
| 514 |
+
"eval_wer": 37.27959697732997,
|
| 515 |
+
"step": 220
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 3.169014084507042,
|
| 519 |
+
"grad_norm": 5.95841646194458,
|
| 520 |
+
"learning_rate": 4.5e-06,
|
| 521 |
+
"loss": 0.4982,
|
| 522 |
+
"step": 225
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 3.23943661971831,
|
| 526 |
+
"grad_norm": 6.992792129516602,
|
| 527 |
+
"learning_rate": 4.600000000000001e-06,
|
| 528 |
+
"loss": 0.4843,
|
| 529 |
+
"step": 230
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 3.23943661971831,
|
| 533 |
+
"eval_loss": 0.5861312747001648,
|
| 534 |
+
"eval_runtime": 257.6573,
|
| 535 |
+
"eval_samples_per_second": 1.941,
|
| 536 |
+
"eval_steps_per_second": 0.245,
|
| 537 |
+
"eval_wer": 48.394206549118394,
|
| 538 |
+
"step": 230
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 3.3098591549295775,
|
| 542 |
+
"grad_norm": 5.872804164886475,
|
| 543 |
+
"learning_rate": 4.7e-06,
|
| 544 |
+
"loss": 0.4471,
|
| 545 |
+
"step": 235
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 3.380281690140845,
|
| 549 |
+
"grad_norm": 6.013182640075684,
|
| 550 |
+
"learning_rate": 4.800000000000001e-06,
|
| 551 |
+
"loss": 0.4538,
|
| 552 |
+
"step": 240
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 3.380281690140845,
|
| 556 |
+
"eval_loss": 0.5793710350990295,
|
| 557 |
+
"eval_runtime": 254.563,
|
| 558 |
+
"eval_samples_per_second": 1.964,
|
| 559 |
+
"eval_steps_per_second": 0.247,
|
| 560 |
+
"eval_wer": 34.66624685138539,
|
| 561 |
+
"step": 240
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 3.4507042253521125,
|
| 565 |
+
"grad_norm": 6.745495319366455,
|
| 566 |
+
"learning_rate": 4.9000000000000005e-06,
|
| 567 |
+
"loss": 0.4932,
|
| 568 |
+
"step": 245
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 3.52112676056338,
|
| 572 |
+
"grad_norm": 5.320774078369141,
|
| 573 |
+
"learning_rate": 5e-06,
|
| 574 |
+
"loss": 0.4741,
|
| 575 |
+
"step": 250
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 3.52112676056338,
|
| 579 |
+
"eval_loss": 0.5736850500106812,
|
| 580 |
+
"eval_runtime": 255.3883,
|
| 581 |
+
"eval_samples_per_second": 1.958,
|
| 582 |
+
"eval_steps_per_second": 0.247,
|
| 583 |
+
"eval_wer": 33.816120906801004,
|
| 584 |
+
"step": 250
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 3.591549295774648,
|
| 588 |
+
"grad_norm": 6.753683090209961,
|
| 589 |
+
"learning_rate": 5.1e-06,
|
| 590 |
+
"loss": 0.5025,
|
| 591 |
+
"step": 255
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 3.6619718309859155,
|
| 595 |
+
"grad_norm": 7.474066257476807,
|
| 596 |
+
"learning_rate": 5.2e-06,
|
| 597 |
+
"loss": 0.4542,
|
| 598 |
+
"step": 260
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 3.6619718309859155,
|
| 602 |
+
"eval_loss": 0.5662725567817688,
|
| 603 |
+
"eval_runtime": 255.3299,
|
| 604 |
+
"eval_samples_per_second": 1.958,
|
| 605 |
+
"eval_steps_per_second": 0.247,
|
| 606 |
+
"eval_wer": 41.97103274559194,
|
| 607 |
+
"step": 260
|
| 608 |
+
},
|
| 609 |
+
{
|
| 610 |
+
"epoch": 3.732394366197183,
|
| 611 |
+
"grad_norm": 5.626581192016602,
|
| 612 |
+
"learning_rate": 5.300000000000001e-06,
|
| 613 |
+
"loss": 0.4639,
|
| 614 |
+
"step": 265
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 3.802816901408451,
|
| 618 |
+
"grad_norm": 5.518383026123047,
|
| 619 |
+
"learning_rate": 5.400000000000001e-06,
|
| 620 |
+
"loss": 0.4163,
|
| 621 |
+
"step": 270
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 3.802816901408451,
|
| 625 |
+
"eval_loss": 0.5622957944869995,
|
| 626 |
+
"eval_runtime": 256.1828,
|
| 627 |
+
"eval_samples_per_second": 1.952,
|
| 628 |
+
"eval_steps_per_second": 0.246,
|
| 629 |
+
"eval_wer": 46.095717884130984,
|
| 630 |
+
"step": 270
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 3.873239436619718,
|
| 634 |
+
"grad_norm": 6.132260799407959,
|
| 635 |
+
"learning_rate": 5.500000000000001e-06,
|
| 636 |
+
"loss": 0.3922,
|
| 637 |
+
"step": 275
|
| 638 |
+
},
|
| 639 |
+
{
|
| 640 |
+
"epoch": 3.943661971830986,
|
| 641 |
+
"grad_norm": 5.8338942527771,
|
| 642 |
+
"learning_rate": 5.600000000000001e-06,
|
| 643 |
+
"loss": 0.3496,
|
| 644 |
+
"step": 280
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 3.943661971830986,
|
| 648 |
+
"eval_loss": 0.560535192489624,
|
| 649 |
+
"eval_runtime": 255.0016,
|
| 650 |
+
"eval_samples_per_second": 1.961,
|
| 651 |
+
"eval_steps_per_second": 0.247,
|
| 652 |
+
"eval_wer": 42.2544080604534,
|
| 653 |
+
"step": 280
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 4.014084507042254,
|
| 657 |
+
"grad_norm": 4.769192695617676,
|
| 658 |
+
"learning_rate": 5.7e-06,
|
| 659 |
+
"loss": 0.4389,
|
| 660 |
+
"step": 285
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 4.084507042253521,
|
| 664 |
+
"grad_norm": 5.79905366897583,
|
| 665 |
+
"learning_rate": 5.8e-06,
|
| 666 |
+
"loss": 0.3835,
|
| 667 |
+
"step": 290
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 4.084507042253521,
|
| 671 |
+
"eval_loss": 0.5556859374046326,
|
| 672 |
+
"eval_runtime": 255.3987,
|
| 673 |
+
"eval_samples_per_second": 1.958,
|
| 674 |
+
"eval_steps_per_second": 0.247,
|
| 675 |
+
"eval_wer": 41.656171284634766,
|
| 676 |
+
"step": 290
|
| 677 |
+
},
|
| 678 |
+
{
|
| 679 |
+
"epoch": 4.154929577464789,
|
| 680 |
+
"grad_norm": 5.353799819946289,
|
| 681 |
+
"learning_rate": 5.9e-06,
|
| 682 |
+
"loss": 0.385,
|
| 683 |
+
"step": 295
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 4.225352112676056,
|
| 687 |
+
"grad_norm": 5.164504528045654,
|
| 688 |
+
"learning_rate": 6e-06,
|
| 689 |
+
"loss": 0.3462,
|
| 690 |
+
"step": 300
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 4.225352112676056,
|
| 694 |
+
"eval_loss": 0.550672173500061,
|
| 695 |
+
"eval_runtime": 255.5806,
|
| 696 |
+
"eval_samples_per_second": 1.956,
|
| 697 |
+
"eval_steps_per_second": 0.246,
|
| 698 |
+
"eval_wer": 36.39798488664987,
|
| 699 |
+
"step": 300
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 4.295774647887324,
|
| 703 |
+
"grad_norm": 5.903466701507568,
|
| 704 |
+
"learning_rate": 6.1e-06,
|
| 705 |
+
"loss": 0.3733,
|
| 706 |
+
"step": 305
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 4.366197183098592,
|
| 710 |
+
"grad_norm": 6.308957099914551,
|
| 711 |
+
"learning_rate": 6.200000000000001e-06,
|
| 712 |
+
"loss": 0.3133,
|
| 713 |
+
"step": 310
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 4.366197183098592,
|
| 717 |
+
"eval_loss": 0.5452054738998413,
|
| 718 |
+
"eval_runtime": 255.9204,
|
| 719 |
+
"eval_samples_per_second": 1.954,
|
| 720 |
+
"eval_steps_per_second": 0.246,
|
| 721 |
+
"eval_wer": 42.56926952141058,
|
| 722 |
+
"step": 310
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 4.436619718309859,
|
| 726 |
+
"grad_norm": 4.767759323120117,
|
| 727 |
+
"learning_rate": 6.300000000000001e-06,
|
| 728 |
+
"loss": 0.3544,
|
| 729 |
+
"step": 315
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 4.507042253521127,
|
| 733 |
+
"grad_norm": 5.711643695831299,
|
| 734 |
+
"learning_rate": 6.4000000000000006e-06,
|
| 735 |
+
"loss": 0.3638,
|
| 736 |
+
"step": 320
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 4.507042253521127,
|
| 740 |
+
"eval_loss": 0.5434854030609131,
|
| 741 |
+
"eval_runtime": 253.7024,
|
| 742 |
+
"eval_samples_per_second": 1.971,
|
| 743 |
+
"eval_steps_per_second": 0.248,
|
| 744 |
+
"eval_wer": 35.957178841309826,
|
| 745 |
+
"step": 320
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 4.577464788732394,
|
| 749 |
+
"grad_norm": 5.667789936065674,
|
| 750 |
+
"learning_rate": 6.5000000000000004e-06,
|
| 751 |
+
"loss": 0.3974,
|
| 752 |
+
"step": 325
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 4.647887323943662,
|
| 756 |
+
"grad_norm": 6.108503341674805,
|
| 757 |
+
"learning_rate": 6.600000000000001e-06,
|
| 758 |
+
"loss": 0.3826,
|
| 759 |
+
"step": 330
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 4.647887323943662,
|
| 763 |
+
"eval_loss": 0.5396420955657959,
|
| 764 |
+
"eval_runtime": 252.7138,
|
| 765 |
+
"eval_samples_per_second": 1.979,
|
| 766 |
+
"eval_steps_per_second": 0.249,
|
| 767 |
+
"eval_wer": 31.95843828715365,
|
| 768 |
+
"step": 330
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 4.71830985915493,
|
| 772 |
+
"grad_norm": 5.889377117156982,
|
| 773 |
+
"learning_rate": 6.700000000000001e-06,
|
| 774 |
+
"loss": 0.3813,
|
| 775 |
+
"step": 335
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 4.788732394366197,
|
| 779 |
+
"grad_norm": 5.469658851623535,
|
| 780 |
+
"learning_rate": 6.800000000000001e-06,
|
| 781 |
+
"loss": 0.3581,
|
| 782 |
+
"step": 340
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 4.788732394366197,
|
| 786 |
+
"eval_loss": 0.5361477136611938,
|
| 787 |
+
"eval_runtime": 251.8728,
|
| 788 |
+
"eval_samples_per_second": 1.985,
|
| 789 |
+
"eval_steps_per_second": 0.25,
|
| 790 |
+
"eval_wer": 33.78463476070529,
|
| 791 |
+
"step": 340
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 4.859154929577465,
|
| 795 |
+
"grad_norm": 5.188804626464844,
|
| 796 |
+
"learning_rate": 6.9e-06,
|
| 797 |
+
"loss": 0.3351,
|
| 798 |
+
"step": 345
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 4.929577464788732,
|
| 802 |
+
"grad_norm": 5.103167533874512,
|
| 803 |
+
"learning_rate": 7e-06,
|
| 804 |
+
"loss": 0.3127,
|
| 805 |
+
"step": 350
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 4.929577464788732,
|
| 809 |
+
"eval_loss": 0.5339432954788208,
|
| 810 |
+
"eval_runtime": 252.7571,
|
| 811 |
+
"eval_samples_per_second": 1.978,
|
| 812 |
+
"eval_steps_per_second": 0.249,
|
| 813 |
+
"eval_wer": 37.342569269521405,
|
| 814 |
+
"step": 350
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 5.0,
|
| 818 |
+
"grad_norm": 9.485374450683594,
|
| 819 |
+
"learning_rate": 7.100000000000001e-06,
|
| 820 |
+
"loss": 0.3265,
|
| 821 |
+
"step": 355
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 5.070422535211268,
|
| 825 |
+
"grad_norm": 5.010895252227783,
|
| 826 |
+
"learning_rate": 7.2000000000000005e-06,
|
| 827 |
+
"loss": 0.2988,
|
| 828 |
+
"step": 360
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 5.070422535211268,
|
| 832 |
+
"eval_loss": 0.5347580909729004,
|
| 833 |
+
"eval_runtime": 253.3761,
|
| 834 |
+
"eval_samples_per_second": 1.973,
|
| 835 |
+
"eval_steps_per_second": 0.249,
|
| 836 |
+
"eval_wer": 38.727959697733,
|
| 837 |
+
"step": 360
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 5.140845070422535,
|
| 841 |
+
"grad_norm": 5.113419055938721,
|
| 842 |
+
"learning_rate": 7.3e-06,
|
| 843 |
+
"loss": 0.2953,
|
| 844 |
+
"step": 365
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 5.211267605633803,
|
| 848 |
+
"grad_norm": 5.5772247314453125,
|
| 849 |
+
"learning_rate": 7.4e-06,
|
| 850 |
+
"loss": 0.2807,
|
| 851 |
+
"step": 370
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 5.211267605633803,
|
| 855 |
+
"eval_loss": 0.5343714952468872,
|
| 856 |
+
"eval_runtime": 252.932,
|
| 857 |
+
"eval_samples_per_second": 1.977,
|
| 858 |
+
"eval_steps_per_second": 0.249,
|
| 859 |
+
"eval_wer": 35.51637279596977,
|
| 860 |
+
"step": 370
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 5.28169014084507,
|
| 864 |
+
"grad_norm": 5.650921821594238,
|
| 865 |
+
"learning_rate": 7.500000000000001e-06,
|
| 866 |
+
"loss": 0.3147,
|
| 867 |
+
"step": 375
|
| 868 |
+
},
|
| 869 |
+
{
|
| 870 |
+
"epoch": 5.352112676056338,
|
| 871 |
+
"grad_norm": 5.143499374389648,
|
| 872 |
+
"learning_rate": 7.600000000000001e-06,
|
| 873 |
+
"loss": 0.2612,
|
| 874 |
+
"step": 380
|
| 875 |
+
},
|
| 876 |
+
{
|
| 877 |
+
"epoch": 5.352112676056338,
|
| 878 |
+
"eval_loss": 0.5304917097091675,
|
| 879 |
+
"eval_runtime": 252.0942,
|
| 880 |
+
"eval_samples_per_second": 1.983,
|
| 881 |
+
"eval_steps_per_second": 0.25,
|
| 882 |
+
"eval_wer": 34.66624685138539,
|
| 883 |
+
"step": 380
|
| 884 |
+
},
|
| 885 |
+
{
|
| 886 |
+
"epoch": 5.422535211267606,
|
| 887 |
+
"grad_norm": 5.593881607055664,
|
| 888 |
+
"learning_rate": 7.7e-06,
|
| 889 |
+
"loss": 0.2606,
|
| 890 |
+
"step": 385
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 5.492957746478873,
|
| 894 |
+
"grad_norm": 5.4485392570495605,
|
| 895 |
+
"learning_rate": 7.800000000000002e-06,
|
| 896 |
+
"loss": 0.2762,
|
| 897 |
+
"step": 390
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 5.492957746478873,
|
| 901 |
+
"eval_loss": 0.5305802226066589,
|
| 902 |
+
"eval_runtime": 252.0179,
|
| 903 |
+
"eval_samples_per_second": 1.984,
|
| 904 |
+
"eval_steps_per_second": 0.25,
|
| 905 |
+
"eval_wer": 32.27329974811083,
|
| 906 |
+
"step": 390
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 5.563380281690141,
|
| 910 |
+
"grad_norm": 4.250403881072998,
|
| 911 |
+
"learning_rate": 7.9e-06,
|
| 912 |
+
"loss": 0.2609,
|
| 913 |
+
"step": 395
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 5.633802816901408,
|
| 917 |
+
"grad_norm": 5.564484596252441,
|
| 918 |
+
"learning_rate": 8.000000000000001e-06,
|
| 919 |
+
"loss": 0.299,
|
| 920 |
+
"step": 400
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 5.633802816901408,
|
| 924 |
+
"eval_loss": 0.5266876220703125,
|
| 925 |
+
"eval_runtime": 251.1581,
|
| 926 |
+
"eval_samples_per_second": 1.991,
|
| 927 |
+
"eval_steps_per_second": 0.251,
|
| 928 |
+
"eval_wer": 36.87027707808564,
|
| 929 |
+
"step": 400
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 5.704225352112676,
|
| 933 |
+
"grad_norm": 4.646668910980225,
|
| 934 |
+
"learning_rate": 8.1e-06,
|
| 935 |
+
"loss": 0.2368,
|
| 936 |
+
"step": 405
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"epoch": 5.774647887323944,
|
| 940 |
+
"grad_norm": 5.00687313079834,
|
| 941 |
+
"learning_rate": 8.2e-06,
|
| 942 |
+
"loss": 0.2718,
|
| 943 |
+
"step": 410
|
| 944 |
+
},
|
| 945 |
+
{
|
| 946 |
+
"epoch": 5.774647887323944,
|
| 947 |
+
"eval_loss": 0.5231830477714539,
|
| 948 |
+
"eval_runtime": 252.6711,
|
| 949 |
+
"eval_samples_per_second": 1.979,
|
| 950 |
+
"eval_steps_per_second": 0.249,
|
| 951 |
+
"eval_wer": 41.68765743073048,
|
| 952 |
+
"step": 410
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 5.845070422535211,
|
| 956 |
+
"grad_norm": 4.078917503356934,
|
| 957 |
+
"learning_rate": 8.3e-06,
|
| 958 |
+
"loss": 0.252,
|
| 959 |
+
"step": 415
|
| 960 |
+
},
|
| 961 |
+
{
|
| 962 |
+
"epoch": 5.915492957746479,
|
| 963 |
+
"grad_norm": 4.877511501312256,
|
| 964 |
+
"learning_rate": 8.400000000000001e-06,
|
| 965 |
+
"loss": 0.2618,
|
| 966 |
+
"step": 420
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 5.915492957746479,
|
| 970 |
+
"eval_loss": 0.5207710266113281,
|
| 971 |
+
"eval_runtime": 252.1118,
|
| 972 |
+
"eval_samples_per_second": 1.983,
|
| 973 |
+
"eval_steps_per_second": 0.25,
|
| 974 |
+
"eval_wer": 34.09949622166247,
|
| 975 |
+
"step": 420
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 5.985915492957746,
|
| 979 |
+
"grad_norm": 5.141012191772461,
|
| 980 |
+
"learning_rate": 8.5e-06,
|
| 981 |
+
"loss": 0.3232,
|
| 982 |
+
"step": 425
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 6.056338028169014,
|
| 986 |
+
"grad_norm": 4.299196243286133,
|
| 987 |
+
"learning_rate": 8.6e-06,
|
| 988 |
+
"loss": 0.2121,
|
| 989 |
+
"step": 430
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 6.056338028169014,
|
| 993 |
+
"eval_loss": 0.5220197439193726,
|
| 994 |
+
"eval_runtime": 252.0399,
|
| 995 |
+
"eval_samples_per_second": 1.984,
|
| 996 |
+
"eval_steps_per_second": 0.25,
|
| 997 |
+
"eval_wer": 28.05415617128463,
|
| 998 |
+
"step": 430
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 6.126760563380282,
|
| 1002 |
+
"grad_norm": 3.769075393676758,
|
| 1003 |
+
"learning_rate": 8.700000000000001e-06,
|
| 1004 |
+
"loss": 0.2119,
|
| 1005 |
+
"step": 435
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 6.197183098591549,
|
| 1009 |
+
"grad_norm": 4.311405181884766,
|
| 1010 |
+
"learning_rate": 8.8e-06,
|
| 1011 |
+
"loss": 0.1929,
|
| 1012 |
+
"step": 440
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 6.197183098591549,
|
| 1016 |
+
"eval_loss": 0.5256190299987793,
|
| 1017 |
+
"eval_runtime": 253.2092,
|
| 1018 |
+
"eval_samples_per_second": 1.975,
|
| 1019 |
+
"eval_steps_per_second": 0.249,
|
| 1020 |
+
"eval_wer": 35.79974811083124,
|
| 1021 |
+
"step": 440
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 6.267605633802817,
|
| 1025 |
+
"grad_norm": 3.735041618347168,
|
| 1026 |
+
"learning_rate": 8.900000000000001e-06,
|
| 1027 |
+
"loss": 0.2104,
|
| 1028 |
+
"step": 445
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 6.338028169014084,
|
| 1032 |
+
"grad_norm": 6.507180690765381,
|
| 1033 |
+
"learning_rate": 9e-06,
|
| 1034 |
+
"loss": 0.2504,
|
| 1035 |
+
"step": 450
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 6.338028169014084,
|
| 1039 |
+
"eval_loss": 0.529583215713501,
|
| 1040 |
+
"eval_runtime": 252.3402,
|
| 1041 |
+
"eval_samples_per_second": 1.981,
|
| 1042 |
+
"eval_steps_per_second": 0.25,
|
| 1043 |
+
"eval_wer": 32.87153652392947,
|
| 1044 |
+
"step": 450
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 6.408450704225352,
|
| 1048 |
+
"grad_norm": 4.1670355796813965,
|
| 1049 |
+
"learning_rate": 9.100000000000001e-06,
|
| 1050 |
+
"loss": 0.1931,
|
| 1051 |
+
"step": 455
|
| 1052 |
+
},
|
| 1053 |
+
{
|
| 1054 |
+
"epoch": 6.47887323943662,
|
| 1055 |
+
"grad_norm": 4.260618209838867,
|
| 1056 |
+
"learning_rate": 9.200000000000002e-06,
|
| 1057 |
+
"loss": 0.2064,
|
| 1058 |
+
"step": 460
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 6.47887323943662,
|
| 1062 |
+
"eval_loss": 0.5265011191368103,
|
| 1063 |
+
"eval_runtime": 253.5935,
|
| 1064 |
+
"eval_samples_per_second": 1.972,
|
| 1065 |
+
"eval_steps_per_second": 0.248,
|
| 1066 |
+
"eval_wer": 35.3904282115869,
|
| 1067 |
+
"step": 460
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 6.549295774647887,
|
| 1071 |
+
"grad_norm": 4.580427169799805,
|
| 1072 |
+
"learning_rate": 9.3e-06,
|
| 1073 |
+
"loss": 0.2099,
|
| 1074 |
+
"step": 465
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 6.619718309859155,
|
| 1078 |
+
"grad_norm": 5.135242938995361,
|
| 1079 |
+
"learning_rate": 9.4e-06,
|
| 1080 |
+
"loss": 0.2044,
|
| 1081 |
+
"step": 470
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 6.619718309859155,
|
| 1085 |
+
"eval_loss": 0.5266779065132141,
|
| 1086 |
+
"eval_runtime": 253.6172,
|
| 1087 |
+
"eval_samples_per_second": 1.971,
|
| 1088 |
+
"eval_steps_per_second": 0.248,
|
| 1089 |
+
"eval_wer": 38.31863979848866,
|
| 1090 |
+
"step": 470
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 6.690140845070422,
|
| 1094 |
+
"grad_norm": 4.770451545715332,
|
| 1095 |
+
"learning_rate": 9.5e-06,
|
| 1096 |
+
"loss": 0.2118,
|
| 1097 |
+
"step": 475
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 6.76056338028169,
|
| 1101 |
+
"grad_norm": 4.276612758636475,
|
| 1102 |
+
"learning_rate": 9.600000000000001e-06,
|
| 1103 |
+
"loss": 0.1844,
|
| 1104 |
+
"step": 480
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 6.76056338028169,
|
| 1108 |
+
"eval_loss": 0.5231460332870483,
|
| 1109 |
+
"eval_runtime": 253.7339,
|
| 1110 |
+
"eval_samples_per_second": 1.971,
|
| 1111 |
+
"eval_steps_per_second": 0.248,
|
| 1112 |
+
"eval_wer": 35.107052896725435,
|
| 1113 |
+
"step": 480
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 6.830985915492958,
|
| 1117 |
+
"grad_norm": 6.741299152374268,
|
| 1118 |
+
"learning_rate": 9.7e-06,
|
| 1119 |
+
"loss": 0.2276,
|
| 1120 |
+
"step": 485
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 6.901408450704225,
|
| 1124 |
+
"grad_norm": 5.4448370933532715,
|
| 1125 |
+
"learning_rate": 9.800000000000001e-06,
|
| 1126 |
+
"loss": 0.1867,
|
| 1127 |
+
"step": 490
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 6.901408450704225,
|
| 1131 |
+
"eval_loss": 0.5235409140586853,
|
| 1132 |
+
"eval_runtime": 252.1039,
|
| 1133 |
+
"eval_samples_per_second": 1.983,
|
| 1134 |
+
"eval_steps_per_second": 0.25,
|
| 1135 |
+
"eval_wer": 31.580604534005037,
|
| 1136 |
+
"step": 490
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 6.971830985915493,
|
| 1140 |
+
"grad_norm": 5.26415491104126,
|
| 1141 |
+
"learning_rate": 9.9e-06,
|
| 1142 |
+
"loss": 0.2232,
|
| 1143 |
+
"step": 495
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 7.042253521126761,
|
| 1147 |
+
"grad_norm": 3.9737112522125244,
|
| 1148 |
+
"learning_rate": 0.0,
|
| 1149 |
+
"loss": 0.1562,
|
| 1150 |
+
"step": 500
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 7.042253521126761,
|
| 1154 |
+
"eval_loss": 0.5233400464057922,
|
| 1155 |
+
"eval_runtime": 252.8742,
|
| 1156 |
+
"eval_samples_per_second": 1.977,
|
| 1157 |
+
"eval_steps_per_second": 0.249,
|
| 1158 |
+
"eval_wer": 31.10831234256927,
|
| 1159 |
+
"step": 500
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 7.042253521126761,
|
| 1163 |
+
"step": 500,
|
| 1164 |
+
"total_flos": 7.8022170722304e+17,
|
| 1165 |
+
"train_loss": 0.9523631989955902,
|
| 1166 |
+
"train_runtime": 13251.4495,
|
| 1167 |
+
"train_samples_per_second": 2.415,
|
| 1168 |
+
"train_steps_per_second": 0.038
|
| 1169 |
+
}
|
| 1170 |
+
],
|
| 1171 |
+
"logging_steps": 5,
|
| 1172 |
+
"max_steps": 500,
|
| 1173 |
+
"num_input_tokens_seen": 0,
|
| 1174 |
+
"num_train_epochs": 8,
|
| 1175 |
+
"save_steps": 10,
|
| 1176 |
+
"total_flos": 7.8022170722304e+17,
|
| 1177 |
+
"train_batch_size": 64,
|
| 1178 |
+
"trial_name": null,
|
| 1179 |
+
"trial_params": null
|
| 1180 |
+
}
|