End of training
Browse files- README.md +6 -3
- adapter.bem.safetensors +3 -0
- all_results.json +15 -0
- eval_results.json +9 -0
- train_results.json +9 -0
- trainer_state.json +803 -0
README.md
CHANGED
|
@@ -3,6 +3,9 @@ library_name: transformers
|
|
| 3 |
license: cc-by-nc-4.0
|
| 4 |
base_model: facebook/mms-1b-all
|
| 5 |
tags:
|
|
|
|
|
|
|
|
|
|
| 6 |
- generated_from_trainer
|
| 7 |
metrics:
|
| 8 |
- wer
|
|
@@ -16,10 +19,10 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 16 |
|
| 17 |
# mms-1b-all-bem-genbed-f-model
|
| 18 |
|
| 19 |
-
This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on
|
| 20 |
It achieves the following results on the evaluation set:
|
| 21 |
-
- Loss: 0.
|
| 22 |
-
- Wer: 0.
|
| 23 |
|
| 24 |
## Model description
|
| 25 |
|
|
|
|
| 3 |
license: cc-by-nc-4.0
|
| 4 |
base_model: facebook/mms-1b-all
|
| 5 |
tags:
|
| 6 |
+
- automatic-speech-recognition
|
| 7 |
+
- genbed
|
| 8 |
+
- mms
|
| 9 |
- generated_from_trainer
|
| 10 |
metrics:
|
| 11 |
- wer
|
|
|
|
| 19 |
|
| 20 |
# mms-1b-all-bem-genbed-f-model
|
| 21 |
|
| 22 |
+
This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the GENBED - BEM dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 0.1823
|
| 25 |
+
- Wer: 0.3431
|
| 26 |
|
| 27 |
## Model description
|
| 28 |
|
adapter.bem.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b8d7d648e76b575d0c523d28e5000ed361a827601cfba7144ae62f815c1b2564
|
| 3 |
+
size 8798532
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.438356164383562,
|
| 3 |
+
"eval_loss": 0.18234074115753174,
|
| 4 |
+
"eval_runtime": 65.9372,
|
| 5 |
+
"eval_samples": 970,
|
| 6 |
+
"eval_samples_per_second": 14.711,
|
| 7 |
+
"eval_steps_per_second": 1.85,
|
| 8 |
+
"eval_wer": 0.34310881942206467,
|
| 9 |
+
"total_flos": 1.108496718380256e+19,
|
| 10 |
+
"train_loss": 0.39921932220458983,
|
| 11 |
+
"train_runtime": 6096.3108,
|
| 12 |
+
"train_samples": 2917,
|
| 13 |
+
"train_samples_per_second": 14.355,
|
| 14 |
+
"train_steps_per_second": 3.592
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.438356164383562,
|
| 3 |
+
"eval_loss": 0.18234074115753174,
|
| 4 |
+
"eval_runtime": 65.9372,
|
| 5 |
+
"eval_samples": 970,
|
| 6 |
+
"eval_samples_per_second": 14.711,
|
| 7 |
+
"eval_steps_per_second": 1.85,
|
| 8 |
+
"eval_wer": 0.34310881942206467
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 6.438356164383562,
|
| 3 |
+
"total_flos": 1.108496718380256e+19,
|
| 4 |
+
"train_loss": 0.39921932220458983,
|
| 5 |
+
"train_runtime": 6096.3108,
|
| 6 |
+
"train_samples": 2917,
|
| 7 |
+
"train_samples_per_second": 14.355,
|
| 8 |
+
"train_steps_per_second": 3.592
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,803 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.18234074115753174,
|
| 3 |
+
"best_model_checkpoint": "/scratch/skscla001/results/mms-1b-all-bem-genbed-f-model/checkpoint-4400",
|
| 4 |
+
"epoch": 6.438356164383562,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 4700,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.136986301369863,
|
| 13 |
+
"grad_norm": 2.5438549518585205,
|
| 14 |
+
"learning_rate": 0.00028799999999999995,
|
| 15 |
+
"loss": 6.6556,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.136986301369863,
|
| 20 |
+
"eval_loss": 0.5950818061828613,
|
| 21 |
+
"eval_runtime": 66.923,
|
| 22 |
+
"eval_samples_per_second": 14.494,
|
| 23 |
+
"eval_steps_per_second": 1.823,
|
| 24 |
+
"eval_wer": 0.6643033623375229,
|
| 25 |
+
"step": 100
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.273972602739726,
|
| 29 |
+
"grad_norm": 1.7334426641464233,
|
| 30 |
+
"learning_rate": 0.0002986788990825688,
|
| 31 |
+
"loss": 0.4415,
|
| 32 |
+
"step": 200
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.273972602739726,
|
| 36 |
+
"eval_loss": 0.27343931794166565,
|
| 37 |
+
"eval_runtime": 66.3712,
|
| 38 |
+
"eval_samples_per_second": 14.615,
|
| 39 |
+
"eval_steps_per_second": 1.838,
|
| 40 |
+
"eval_wer": 0.45654742722096897,
|
| 41 |
+
"step": 200
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.410958904109589,
|
| 45 |
+
"grad_norm": 2.5106163024902344,
|
| 46 |
+
"learning_rate": 0.00029730275229357796,
|
| 47 |
+
"loss": 0.3448,
|
| 48 |
+
"step": 300
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.410958904109589,
|
| 52 |
+
"eval_loss": 0.248216450214386,
|
| 53 |
+
"eval_runtime": 65.6824,
|
| 54 |
+
"eval_samples_per_second": 14.768,
|
| 55 |
+
"eval_steps_per_second": 1.857,
|
| 56 |
+
"eval_wer": 0.4289397357396068,
|
| 57 |
+
"step": 300
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.547945205479452,
|
| 61 |
+
"grad_norm": 1.4925711154937744,
|
| 62 |
+
"learning_rate": 0.00029592660550458713,
|
| 63 |
+
"loss": 0.3459,
|
| 64 |
+
"step": 400
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.547945205479452,
|
| 68 |
+
"eval_loss": 0.23923924565315247,
|
| 69 |
+
"eval_runtime": 65.8079,
|
| 70 |
+
"eval_samples_per_second": 14.74,
|
| 71 |
+
"eval_steps_per_second": 1.854,
|
| 72 |
+
"eval_wer": 0.4148673326887958,
|
| 73 |
+
"step": 400
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.684931506849315,
|
| 77 |
+
"grad_norm": 3.7458858489990234,
|
| 78 |
+
"learning_rate": 0.0002945504587155963,
|
| 79 |
+
"loss": 0.3184,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.684931506849315,
|
| 84 |
+
"eval_loss": 0.23041340708732605,
|
| 85 |
+
"eval_runtime": 66.6329,
|
| 86 |
+
"eval_samples_per_second": 14.557,
|
| 87 |
+
"eval_steps_per_second": 1.831,
|
| 88 |
+
"eval_wer": 0.40852938016972823,
|
| 89 |
+
"step": 500
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.821917808219178,
|
| 93 |
+
"grad_norm": 2.308128595352173,
|
| 94 |
+
"learning_rate": 0.00029317431192660547,
|
| 95 |
+
"loss": 0.3058,
|
| 96 |
+
"step": 600
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.821917808219178,
|
| 100 |
+
"eval_loss": 0.2372046560049057,
|
| 101 |
+
"eval_runtime": 66.0123,
|
| 102 |
+
"eval_samples_per_second": 14.694,
|
| 103 |
+
"eval_steps_per_second": 1.848,
|
| 104 |
+
"eval_wer": 0.4107852615748201,
|
| 105 |
+
"step": 600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.958904109589041,
|
| 109 |
+
"grad_norm": 2.1398093700408936,
|
| 110 |
+
"learning_rate": 0.00029181192660550457,
|
| 111 |
+
"loss": 0.3077,
|
| 112 |
+
"step": 700
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.958904109589041,
|
| 116 |
+
"eval_loss": 0.22706551849842072,
|
| 117 |
+
"eval_runtime": 65.7716,
|
| 118 |
+
"eval_samples_per_second": 14.748,
|
| 119 |
+
"eval_steps_per_second": 1.855,
|
| 120 |
+
"eval_wer": 0.4172306370179396,
|
| 121 |
+
"step": 700
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.095890410958904,
|
| 125 |
+
"grad_norm": 1.1179248094558716,
|
| 126 |
+
"learning_rate": 0.00029043577981651374,
|
| 127 |
+
"loss": 0.2812,
|
| 128 |
+
"step": 800
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.095890410958904,
|
| 132 |
+
"eval_loss": 0.22172598540782928,
|
| 133 |
+
"eval_runtime": 65.6625,
|
| 134 |
+
"eval_samples_per_second": 14.773,
|
| 135 |
+
"eval_steps_per_second": 1.858,
|
| 136 |
+
"eval_wer": 0.39832420238478894,
|
| 137 |
+
"step": 800
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.2328767123287672,
|
| 141 |
+
"grad_norm": 0.7546290159225464,
|
| 142 |
+
"learning_rate": 0.0002890596330275229,
|
| 143 |
+
"loss": 0.3297,
|
| 144 |
+
"step": 900
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.2328767123287672,
|
| 148 |
+
"eval_loss": 0.22088122367858887,
|
| 149 |
+
"eval_runtime": 66.7304,
|
| 150 |
+
"eval_samples_per_second": 14.536,
|
| 151 |
+
"eval_steps_per_second": 1.828,
|
| 152 |
+
"eval_wer": 0.3984316253088409,
|
| 153 |
+
"step": 900
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 1.36986301369863,
|
| 157 |
+
"grad_norm": 0.5848907232284546,
|
| 158 |
+
"learning_rate": 0.0002876834862385321,
|
| 159 |
+
"loss": 0.2817,
|
| 160 |
+
"step": 1000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 1.36986301369863,
|
| 164 |
+
"eval_loss": 0.21634483337402344,
|
| 165 |
+
"eval_runtime": 66.3551,
|
| 166 |
+
"eval_samples_per_second": 14.618,
|
| 167 |
+
"eval_steps_per_second": 1.839,
|
| 168 |
+
"eval_wer": 0.41239660543559997,
|
| 169 |
+
"step": 1000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 1.5068493150684932,
|
| 173 |
+
"grad_norm": 1.0077754259109497,
|
| 174 |
+
"learning_rate": 0.00028630733944954125,
|
| 175 |
+
"loss": 0.2927,
|
| 176 |
+
"step": 1100
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.5068493150684932,
|
| 180 |
+
"eval_loss": 0.21464507281780243,
|
| 181 |
+
"eval_runtime": 66.426,
|
| 182 |
+
"eval_samples_per_second": 14.603,
|
| 183 |
+
"eval_steps_per_second": 1.837,
|
| 184 |
+
"eval_wer": 0.3862928348909657,
|
| 185 |
+
"step": 1100
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.643835616438356,
|
| 189 |
+
"grad_norm": 0.9891207218170166,
|
| 190 |
+
"learning_rate": 0.0002849311926605504,
|
| 191 |
+
"loss": 0.2806,
|
| 192 |
+
"step": 1200
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.643835616438356,
|
| 196 |
+
"eval_loss": 0.21057289838790894,
|
| 197 |
+
"eval_runtime": 65.8444,
|
| 198 |
+
"eval_samples_per_second": 14.732,
|
| 199 |
+
"eval_steps_per_second": 1.853,
|
| 200 |
+
"eval_wer": 0.38511118272639383,
|
| 201 |
+
"step": 1200
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 1.7808219178082192,
|
| 205 |
+
"grad_norm": 0.7662916779518127,
|
| 206 |
+
"learning_rate": 0.0002835550458715596,
|
| 207 |
+
"loss": 0.2574,
|
| 208 |
+
"step": 1300
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 1.7808219178082192,
|
| 212 |
+
"eval_loss": 0.20976857841014862,
|
| 213 |
+
"eval_runtime": 66.4117,
|
| 214 |
+
"eval_samples_per_second": 14.606,
|
| 215 |
+
"eval_steps_per_second": 1.837,
|
| 216 |
+
"eval_wer": 0.3866151036631217,
|
| 217 |
+
"step": 1300
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.9178082191780823,
|
| 221 |
+
"grad_norm": 0.8207290768623352,
|
| 222 |
+
"learning_rate": 0.00028217889908256877,
|
| 223 |
+
"loss": 0.2829,
|
| 224 |
+
"step": 1400
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.9178082191780823,
|
| 228 |
+
"eval_loss": 0.20674893260002136,
|
| 229 |
+
"eval_runtime": 66.1157,
|
| 230 |
+
"eval_samples_per_second": 14.671,
|
| 231 |
+
"eval_steps_per_second": 1.845,
|
| 232 |
+
"eval_wer": 0.37716188634654635,
|
| 233 |
+
"step": 1400
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 2.0547945205479454,
|
| 237 |
+
"grad_norm": 1.1322788000106812,
|
| 238 |
+
"learning_rate": 0.000280802752293578,
|
| 239 |
+
"loss": 0.2764,
|
| 240 |
+
"step": 1500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 2.0547945205479454,
|
| 244 |
+
"eval_loss": 0.20762823522090912,
|
| 245 |
+
"eval_runtime": 66.5673,
|
| 246 |
+
"eval_samples_per_second": 14.572,
|
| 247 |
+
"eval_steps_per_second": 1.833,
|
| 248 |
+
"eval_wer": 0.37888065313137825,
|
| 249 |
+
"step": 1500
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.191780821917808,
|
| 253 |
+
"grad_norm": 1.9598041772842407,
|
| 254 |
+
"learning_rate": 0.0002794266055045871,
|
| 255 |
+
"loss": 0.2635,
|
| 256 |
+
"step": 1600
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.191780821917808,
|
| 260 |
+
"eval_loss": 0.20759092271327972,
|
| 261 |
+
"eval_runtime": 66.2145,
|
| 262 |
+
"eval_samples_per_second": 14.649,
|
| 263 |
+
"eval_steps_per_second": 1.842,
|
| 264 |
+
"eval_wer": 0.37694704049844235,
|
| 265 |
+
"step": 1600
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 2.328767123287671,
|
| 269 |
+
"grad_norm": 1.0646597146987915,
|
| 270 |
+
"learning_rate": 0.0002780504587155963,
|
| 271 |
+
"loss": 0.2761,
|
| 272 |
+
"step": 1700
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.328767123287671,
|
| 276 |
+
"eval_loss": 0.20681439340114594,
|
| 277 |
+
"eval_runtime": 66.1689,
|
| 278 |
+
"eval_samples_per_second": 14.659,
|
| 279 |
+
"eval_steps_per_second": 1.844,
|
| 280 |
+
"eval_wer": 0.38006230529595014,
|
| 281 |
+
"step": 1700
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 2.4657534246575343,
|
| 285 |
+
"grad_norm": 2.4462172985076904,
|
| 286 |
+
"learning_rate": 0.0002766743119266055,
|
| 287 |
+
"loss": 0.2854,
|
| 288 |
+
"step": 1800
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.4657534246575343,
|
| 292 |
+
"eval_loss": 0.19937340915203094,
|
| 293 |
+
"eval_runtime": 66.2725,
|
| 294 |
+
"eval_samples_per_second": 14.637,
|
| 295 |
+
"eval_steps_per_second": 1.841,
|
| 296 |
+
"eval_wer": 0.3644859813084112,
|
| 297 |
+
"step": 1800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.602739726027397,
|
| 301 |
+
"grad_norm": 1.0019258260726929,
|
| 302 |
+
"learning_rate": 0.0002752981651376147,
|
| 303 |
+
"loss": 0.2557,
|
| 304 |
+
"step": 1900
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 2.602739726027397,
|
| 308 |
+
"eval_loss": 0.20158326625823975,
|
| 309 |
+
"eval_runtime": 66.179,
|
| 310 |
+
"eval_samples_per_second": 14.657,
|
| 311 |
+
"eval_steps_per_second": 1.843,
|
| 312 |
+
"eval_wer": 0.3860779890428617,
|
| 313 |
+
"step": 1900
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 2.73972602739726,
|
| 317 |
+
"grad_norm": 1.5527012348175049,
|
| 318 |
+
"learning_rate": 0.0002739220183486238,
|
| 319 |
+
"loss": 0.2717,
|
| 320 |
+
"step": 2000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 2.73972602739726,
|
| 324 |
+
"eval_loss": 0.2010820060968399,
|
| 325 |
+
"eval_runtime": 66.4692,
|
| 326 |
+
"eval_samples_per_second": 14.593,
|
| 327 |
+
"eval_steps_per_second": 1.835,
|
| 328 |
+
"eval_wer": 0.37340208400472663,
|
| 329 |
+
"step": 2000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 2.8767123287671232,
|
| 333 |
+
"grad_norm": 1.1809673309326172,
|
| 334 |
+
"learning_rate": 0.000272545871559633,
|
| 335 |
+
"loss": 0.2504,
|
| 336 |
+
"step": 2100
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 2.8767123287671232,
|
| 340 |
+
"eval_loss": 0.19885794818401337,
|
| 341 |
+
"eval_runtime": 66.3483,
|
| 342 |
+
"eval_samples_per_second": 14.62,
|
| 343 |
+
"eval_steps_per_second": 1.839,
|
| 344 |
+
"eval_wer": 0.367386400257815,
|
| 345 |
+
"step": 2100
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 3.0136986301369864,
|
| 349 |
+
"grad_norm": 1.5838390588760376,
|
| 350 |
+
"learning_rate": 0.0002711697247706422,
|
| 351 |
+
"loss": 0.2606,
|
| 352 |
+
"step": 2200
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 3.0136986301369864,
|
| 356 |
+
"eval_loss": 0.19898578524589539,
|
| 357 |
+
"eval_runtime": 66.1937,
|
| 358 |
+
"eval_samples_per_second": 14.654,
|
| 359 |
+
"eval_steps_per_second": 1.843,
|
| 360 |
+
"eval_wer": 0.3834998388656139,
|
| 361 |
+
"step": 2200
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 3.1506849315068495,
|
| 365 |
+
"grad_norm": 0.551179051399231,
|
| 366 |
+
"learning_rate": 0.00026979357798165136,
|
| 367 |
+
"loss": 0.2583,
|
| 368 |
+
"step": 2300
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 3.1506849315068495,
|
| 372 |
+
"eval_loss": 0.20280733704566956,
|
| 373 |
+
"eval_runtime": 66.4916,
|
| 374 |
+
"eval_samples_per_second": 14.588,
|
| 375 |
+
"eval_steps_per_second": 1.835,
|
| 376 |
+
"eval_wer": 0.3666344397894511,
|
| 377 |
+
"step": 2300
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 3.287671232876712,
|
| 381 |
+
"grad_norm": 0.7909404635429382,
|
| 382 |
+
"learning_rate": 0.00026841743119266053,
|
| 383 |
+
"loss": 0.2591,
|
| 384 |
+
"step": 2400
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 3.287671232876712,
|
| 388 |
+
"eval_loss": 0.19522929191589355,
|
| 389 |
+
"eval_runtime": 66.4208,
|
| 390 |
+
"eval_samples_per_second": 14.604,
|
| 391 |
+
"eval_steps_per_second": 1.837,
|
| 392 |
+
"eval_wer": 0.35073584702975613,
|
| 393 |
+
"step": 2400
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 3.4246575342465753,
|
| 397 |
+
"grad_norm": 0.699164867401123,
|
| 398 |
+
"learning_rate": 0.0002670412844036697,
|
| 399 |
+
"loss": 0.2408,
|
| 400 |
+
"step": 2500
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 3.4246575342465753,
|
| 404 |
+
"eval_loss": 0.19884993135929108,
|
| 405 |
+
"eval_runtime": 66.6309,
|
| 406 |
+
"eval_samples_per_second": 14.558,
|
| 407 |
+
"eval_steps_per_second": 1.831,
|
| 408 |
+
"eval_wer": 0.36373402084004725,
|
| 409 |
+
"step": 2500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 3.5616438356164384,
|
| 413 |
+
"grad_norm": 0.6602646112442017,
|
| 414 |
+
"learning_rate": 0.0002656651376146789,
|
| 415 |
+
"loss": 0.2485,
|
| 416 |
+
"step": 2600
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.5616438356164384,
|
| 420 |
+
"eval_loss": 0.19720342755317688,
|
| 421 |
+
"eval_runtime": 66.8315,
|
| 422 |
+
"eval_samples_per_second": 14.514,
|
| 423 |
+
"eval_steps_per_second": 1.825,
|
| 424 |
+
"eval_wer": 0.35932968095391554,
|
| 425 |
+
"step": 2600
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 3.6986301369863015,
|
| 429 |
+
"grad_norm": 0.8821811676025391,
|
| 430 |
+
"learning_rate": 0.00026428899082568804,
|
| 431 |
+
"loss": 0.2474,
|
| 432 |
+
"step": 2700
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 3.6986301369863015,
|
| 436 |
+
"eval_loss": 0.1949489265680313,
|
| 437 |
+
"eval_runtime": 66.4835,
|
| 438 |
+
"eval_samples_per_second": 14.59,
|
| 439 |
+
"eval_steps_per_second": 1.835,
|
| 440 |
+
"eval_wer": 0.353421420131056,
|
| 441 |
+
"step": 2700
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 3.8356164383561646,
|
| 445 |
+
"grad_norm": 0.9004201292991638,
|
| 446 |
+
"learning_rate": 0.0002629128440366972,
|
| 447 |
+
"loss": 0.2398,
|
| 448 |
+
"step": 2800
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 3.8356164383561646,
|
| 452 |
+
"eval_loss": 0.19590793550014496,
|
| 453 |
+
"eval_runtime": 66.4984,
|
| 454 |
+
"eval_samples_per_second": 14.587,
|
| 455 |
+
"eval_steps_per_second": 1.835,
|
| 456 |
+
"eval_wer": 0.3697497045869589,
|
| 457 |
+
"step": 2800
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 3.9726027397260273,
|
| 461 |
+
"grad_norm": 0.9938905835151672,
|
| 462 |
+
"learning_rate": 0.0002615366972477064,
|
| 463 |
+
"loss": 0.2512,
|
| 464 |
+
"step": 2900
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 3.9726027397260273,
|
| 468 |
+
"eval_loss": 0.19063439965248108,
|
| 469 |
+
"eval_runtime": 66.702,
|
| 470 |
+
"eval_samples_per_second": 14.542,
|
| 471 |
+
"eval_steps_per_second": 1.829,
|
| 472 |
+
"eval_wer": 0.3558921473842518,
|
| 473 |
+
"step": 2900
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 4.109589041095891,
|
| 477 |
+
"grad_norm": 53.745086669921875,
|
| 478 |
+
"learning_rate": 0.0002601743119266055,
|
| 479 |
+
"loss": 0.2266,
|
| 480 |
+
"step": 3000
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 4.109589041095891,
|
| 484 |
+
"eval_loss": 0.19052593410015106,
|
| 485 |
+
"eval_runtime": 66.4298,
|
| 486 |
+
"eval_samples_per_second": 14.602,
|
| 487 |
+
"eval_steps_per_second": 1.837,
|
| 488 |
+
"eval_wer": 0.3481576968525083,
|
| 489 |
+
"step": 3000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 4.2465753424657535,
|
| 493 |
+
"grad_norm": 0.7372691035270691,
|
| 494 |
+
"learning_rate": 0.00025879816513761465,
|
| 495 |
+
"loss": 0.2538,
|
| 496 |
+
"step": 3100
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 4.2465753424657535,
|
| 500 |
+
"eval_loss": 0.19160659611225128,
|
| 501 |
+
"eval_runtime": 66.6561,
|
| 502 |
+
"eval_samples_per_second": 14.552,
|
| 503 |
+
"eval_steps_per_second": 1.83,
|
| 504 |
+
"eval_wer": 0.35213234504243207,
|
| 505 |
+
"step": 3100
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 4.383561643835616,
|
| 509 |
+
"grad_norm": 0.9205912947654724,
|
| 510 |
+
"learning_rate": 0.0002574220183486238,
|
| 511 |
+
"loss": 0.2268,
|
| 512 |
+
"step": 3200
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 4.383561643835616,
|
| 516 |
+
"eval_loss": 0.1914052814245224,
|
| 517 |
+
"eval_runtime": 66.9123,
|
| 518 |
+
"eval_samples_per_second": 14.497,
|
| 519 |
+
"eval_steps_per_second": 1.823,
|
| 520 |
+
"eval_wer": 0.38951552261252553,
|
| 521 |
+
"step": 3200
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 4.52054794520548,
|
| 525 |
+
"grad_norm": 0.7206313610076904,
|
| 526 |
+
"learning_rate": 0.000256045871559633,
|
| 527 |
+
"loss": 0.2249,
|
| 528 |
+
"step": 3300
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 4.52054794520548,
|
| 532 |
+
"eval_loss": 0.18965989351272583,
|
| 533 |
+
"eval_runtime": 66.7204,
|
| 534 |
+
"eval_samples_per_second": 14.538,
|
| 535 |
+
"eval_steps_per_second": 1.829,
|
| 536 |
+
"eval_wer": 0.3417123214093888,
|
| 537 |
+
"step": 3300
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.657534246575342,
|
| 541 |
+
"grad_norm": 0.8181961178779602,
|
| 542 |
+
"learning_rate": 0.00025466972477064217,
|
| 543 |
+
"loss": 0.2416,
|
| 544 |
+
"step": 3400
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.657534246575342,
|
| 548 |
+
"eval_loss": 0.18771447241306305,
|
| 549 |
+
"eval_runtime": 66.4891,
|
| 550 |
+
"eval_samples_per_second": 14.589,
|
| 551 |
+
"eval_steps_per_second": 1.835,
|
| 552 |
+
"eval_wer": 0.34579439252336447,
|
| 553 |
+
"step": 3400
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 4.794520547945205,
|
| 557 |
+
"grad_norm": 0.9659631848335266,
|
| 558 |
+
"learning_rate": 0.00025329357798165134,
|
| 559 |
+
"loss": 0.2421,
|
| 560 |
+
"step": 3500
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 4.794520547945205,
|
| 564 |
+
"eval_loss": 0.18715131282806396,
|
| 565 |
+
"eval_runtime": 66.81,
|
| 566 |
+
"eval_samples_per_second": 14.519,
|
| 567 |
+
"eval_steps_per_second": 1.826,
|
| 568 |
+
"eval_wer": 0.3411752067891288,
|
| 569 |
+
"step": 3500
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 4.931506849315069,
|
| 573 |
+
"grad_norm": 0.8366265296936035,
|
| 574 |
+
"learning_rate": 0.0002519174311926605,
|
| 575 |
+
"loss": 0.244,
|
| 576 |
+
"step": 3600
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 4.931506849315069,
|
| 580 |
+
"eval_loss": 0.1854562610387802,
|
| 581 |
+
"eval_runtime": 66.011,
|
| 582 |
+
"eval_samples_per_second": 14.695,
|
| 583 |
+
"eval_steps_per_second": 1.848,
|
| 584 |
+
"eval_wer": 0.352776882586744,
|
| 585 |
+
"step": 3600
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 5.068493150684931,
|
| 589 |
+
"grad_norm": 1.7714653015136719,
|
| 590 |
+
"learning_rate": 0.00025054128440366974,
|
| 591 |
+
"loss": 0.2371,
|
| 592 |
+
"step": 3700
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 5.068493150684931,
|
| 596 |
+
"eval_loss": 0.18706142902374268,
|
| 597 |
+
"eval_runtime": 66.1616,
|
| 598 |
+
"eval_samples_per_second": 14.661,
|
| 599 |
+
"eval_steps_per_second": 1.844,
|
| 600 |
+
"eval_wer": 0.34472016328284455,
|
| 601 |
+
"step": 3700
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 5.205479452054795,
|
| 605 |
+
"grad_norm": 2.374345302581787,
|
| 606 |
+
"learning_rate": 0.00024916513761467885,
|
| 607 |
+
"loss": 0.2383,
|
| 608 |
+
"step": 3800
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 5.205479452054795,
|
| 612 |
+
"eval_loss": 0.18328477442264557,
|
| 613 |
+
"eval_runtime": 66.7399,
|
| 614 |
+
"eval_samples_per_second": 14.534,
|
| 615 |
+
"eval_steps_per_second": 1.828,
|
| 616 |
+
"eval_wer": 0.35234719089053607,
|
| 617 |
+
"step": 3800
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 5.342465753424658,
|
| 621 |
+
"grad_norm": 1.9919145107269287,
|
| 622 |
+
"learning_rate": 0.000247788990825688,
|
| 623 |
+
"loss": 0.2409,
|
| 624 |
+
"step": 3900
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 5.342465753424658,
|
| 628 |
+
"eval_loss": 0.18861110508441925,
|
| 629 |
+
"eval_runtime": 66.6306,
|
| 630 |
+
"eval_samples_per_second": 14.558,
|
| 631 |
+
"eval_steps_per_second": 1.831,
|
| 632 |
+
"eval_wer": 0.34869481147276826,
|
| 633 |
+
"step": 3900
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 5.47945205479452,
|
| 637 |
+
"grad_norm": 2.0718414783477783,
|
| 638 |
+
"learning_rate": 0.00024641284403669725,
|
| 639 |
+
"loss": 0.2312,
|
| 640 |
+
"step": 4000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 5.47945205479452,
|
| 644 |
+
"eval_loss": 0.18479366600513458,
|
| 645 |
+
"eval_runtime": 66.8869,
|
| 646 |
+
"eval_samples_per_second": 14.502,
|
| 647 |
+
"eval_steps_per_second": 1.824,
|
| 648 |
+
"eval_wer": 0.3437533569663766,
|
| 649 |
+
"step": 4000
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 5.616438356164384,
|
| 653 |
+
"grad_norm": 2.795865774154663,
|
| 654 |
+
"learning_rate": 0.0002450366972477064,
|
| 655 |
+
"loss": 0.2261,
|
| 656 |
+
"step": 4100
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 5.616438356164384,
|
| 660 |
+
"eval_loss": 0.18661826848983765,
|
| 661 |
+
"eval_runtime": 67.285,
|
| 662 |
+
"eval_samples_per_second": 14.416,
|
| 663 |
+
"eval_steps_per_second": 1.813,
|
| 664 |
+
"eval_wer": 0.3468686217638844,
|
| 665 |
+
"step": 4100
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 5.7534246575342465,
|
| 669 |
+
"grad_norm": 2.0424089431762695,
|
| 670 |
+
"learning_rate": 0.00024366055045871556,
|
| 671 |
+
"loss": 0.2169,
|
| 672 |
+
"step": 4200
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 5.7534246575342465,
|
| 676 |
+
"eval_loss": 0.18412768840789795,
|
| 677 |
+
"eval_runtime": 66.3619,
|
| 678 |
+
"eval_samples_per_second": 14.617,
|
| 679 |
+
"eval_steps_per_second": 1.838,
|
| 680 |
+
"eval_wer": 0.33763025029541305,
|
| 681 |
+
"step": 4200
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 5.890410958904109,
|
| 685 |
+
"grad_norm": 1.5393688678741455,
|
| 686 |
+
"learning_rate": 0.00024228440366972474,
|
| 687 |
+
"loss": 0.2283,
|
| 688 |
+
"step": 4300
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 5.890410958904109,
|
| 692 |
+
"eval_loss": 0.1864866465330124,
|
| 693 |
+
"eval_runtime": 66.1121,
|
| 694 |
+
"eval_samples_per_second": 14.672,
|
| 695 |
+
"eval_steps_per_second": 1.845,
|
| 696 |
+
"eval_wer": 0.3411752067891288,
|
| 697 |
+
"step": 4300
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 6.027397260273973,
|
| 701 |
+
"grad_norm": 0.7328805327415466,
|
| 702 |
+
"learning_rate": 0.00024090825688073393,
|
| 703 |
+
"loss": 0.2182,
|
| 704 |
+
"step": 4400
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 6.027397260273973,
|
| 708 |
+
"eval_loss": 0.18234074115753174,
|
| 709 |
+
"eval_runtime": 66.2722,
|
| 710 |
+
"eval_samples_per_second": 14.637,
|
| 711 |
+
"eval_steps_per_second": 1.841,
|
| 712 |
+
"eval_wer": 0.34310881942206467,
|
| 713 |
+
"step": 4400
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 6.164383561643835,
|
| 717 |
+
"grad_norm": 0.7929720878601074,
|
| 718 |
+
"learning_rate": 0.0002395321100917431,
|
| 719 |
+
"loss": 0.2141,
|
| 720 |
+
"step": 4500
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 6.164383561643835,
|
| 724 |
+
"eval_loss": 0.18578113615512848,
|
| 725 |
+
"eval_runtime": 66.1558,
|
| 726 |
+
"eval_samples_per_second": 14.662,
|
| 727 |
+
"eval_steps_per_second": 1.844,
|
| 728 |
+
"eval_wer": 0.34031582339671285,
|
| 729 |
+
"step": 4500
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 6.301369863013699,
|
| 733 |
+
"grad_norm": 0.8493778705596924,
|
| 734 |
+
"learning_rate": 0.00023815596330275225,
|
| 735 |
+
"loss": 0.2127,
|
| 736 |
+
"step": 4600
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 6.301369863013699,
|
| 740 |
+
"eval_loss": 0.18759718537330627,
|
| 741 |
+
"eval_runtime": 66.2656,
|
| 742 |
+
"eval_samples_per_second": 14.638,
|
| 743 |
+
"eval_steps_per_second": 1.841,
|
| 744 |
+
"eval_wer": 0.3355892147384252,
|
| 745 |
+
"step": 4600
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 6.438356164383562,
|
| 749 |
+
"grad_norm": 0.8041571378707886,
|
| 750 |
+
"learning_rate": 0.00023677981651376145,
|
| 751 |
+
"loss": 0.229,
|
| 752 |
+
"step": 4700
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 6.438356164383562,
|
| 756 |
+
"eval_loss": 0.18627603352069855,
|
| 757 |
+
"eval_runtime": 66.3029,
|
| 758 |
+
"eval_samples_per_second": 14.63,
|
| 759 |
+
"eval_steps_per_second": 1.84,
|
| 760 |
+
"eval_wer": 0.33612632935868514,
|
| 761 |
+
"step": 4700
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 6.438356164383562,
|
| 765 |
+
"step": 4700,
|
| 766 |
+
"total_flos": 1.108496718380256e+19,
|
| 767 |
+
"train_loss": 0.39921932220458983,
|
| 768 |
+
"train_runtime": 6096.3108,
|
| 769 |
+
"train_samples_per_second": 14.355,
|
| 770 |
+
"train_steps_per_second": 3.592
|
| 771 |
+
}
|
| 772 |
+
],
|
| 773 |
+
"logging_steps": 100,
|
| 774 |
+
"max_steps": 21900,
|
| 775 |
+
"num_input_tokens_seen": 0,
|
| 776 |
+
"num_train_epochs": 30,
|
| 777 |
+
"save_steps": 400,
|
| 778 |
+
"stateful_callbacks": {
|
| 779 |
+
"EarlyStoppingCallback": {
|
| 780 |
+
"args": {
|
| 781 |
+
"early_stopping_patience": 3,
|
| 782 |
+
"early_stopping_threshold": 0.0
|
| 783 |
+
},
|
| 784 |
+
"attributes": {
|
| 785 |
+
"early_stopping_patience_counter": 0
|
| 786 |
+
}
|
| 787 |
+
},
|
| 788 |
+
"TrainerControl": {
|
| 789 |
+
"args": {
|
| 790 |
+
"should_epoch_stop": false,
|
| 791 |
+
"should_evaluate": false,
|
| 792 |
+
"should_log": false,
|
| 793 |
+
"should_save": true,
|
| 794 |
+
"should_training_stop": false
|
| 795 |
+
},
|
| 796 |
+
"attributes": {}
|
| 797 |
+
}
|
| 798 |
+
},
|
| 799 |
+
"total_flos": 1.108496718380256e+19,
|
| 800 |
+
"train_batch_size": 4,
|
| 801 |
+
"trial_name": null,
|
| 802 |
+
"trial_params": null
|
| 803 |
+
}
|