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 +723 -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-bemgen-combined-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 |
+
- bemgen
|
| 8 |
+
- mms
|
| 9 |
- generated_from_trainer
|
| 10 |
metrics:
|
| 11 |
- wer
|
|
|
|
| 19 |
|
| 20 |
# mms-1b-bemgen-combined-model
|
| 21 |
|
| 22 |
+
This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the BEMGEN - BEM dataset.
|
| 23 |
It achieves the following results on the evaluation set:
|
| 24 |
+
- Loss: 0.2477
|
| 25 |
+
- Wer: 0.3897
|
| 26 |
|
| 27 |
## Model description
|
| 28 |
|
adapter.bem.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4d8f12e7839c9ce6542abdd511dc1961f50144460b7aa9494dce1341ef703f69
|
| 3 |
+
size 8798532
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.1660649819494586,
|
| 3 |
+
"eval_loss": 0.24774542450904846,
|
| 4 |
+
"eval_runtime": 61.1885,
|
| 5 |
+
"eval_samples": 976,
|
| 6 |
+
"eval_samples_per_second": 15.951,
|
| 7 |
+
"eval_steps_per_second": 3.988,
|
| 8 |
+
"eval_wer": 0.3897474967348716,
|
| 9 |
+
"total_flos": 1.0016492328632693e+19,
|
| 10 |
+
"train_loss": 0.5294508952186221,
|
| 11 |
+
"train_runtime": 6287.6058,
|
| 12 |
+
"train_samples": 7756,
|
| 13 |
+
"train_samples_per_second": 37.006,
|
| 14 |
+
"train_steps_per_second": 9.252
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.1660649819494586,
|
| 3 |
+
"eval_loss": 0.24774542450904846,
|
| 4 |
+
"eval_runtime": 61.1885,
|
| 5 |
+
"eval_samples": 976,
|
| 6 |
+
"eval_samples_per_second": 15.951,
|
| 7 |
+
"eval_steps_per_second": 3.988,
|
| 8 |
+
"eval_wer": 0.3897474967348716
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 2.1660649819494586,
|
| 3 |
+
"total_flos": 1.0016492328632693e+19,
|
| 4 |
+
"train_loss": 0.5294508952186221,
|
| 5 |
+
"train_runtime": 6287.6058,
|
| 6 |
+
"train_samples": 7756,
|
| 7 |
+
"train_samples_per_second": 37.006,
|
| 8 |
+
"train_steps_per_second": 9.252
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,723 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.24565543234348297,
|
| 3 |
+
"best_model_checkpoint": "/scratch/skscla001/speech/results/mms-1b-bemgen-combined-model/checkpoint-3900",
|
| 4 |
+
"epoch": 2.1660649819494586,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 4200,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.05157297576070139,
|
| 13 |
+
"grad_norm": 2.867751359939575,
|
| 14 |
+
"learning_rate": 0.000285,
|
| 15 |
+
"loss": 6.8762,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.05157297576070139,
|
| 20 |
+
"eval_loss": 0.9800576567649841,
|
| 21 |
+
"eval_runtime": 61.3726,
|
| 22 |
+
"eval_samples_per_second": 15.903,
|
| 23 |
+
"eval_steps_per_second": 3.976,
|
| 24 |
+
"eval_wer": 0.9386155855463648,
|
| 25 |
+
"step": 100
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.10314595152140278,
|
| 29 |
+
"grad_norm": 4.205641269683838,
|
| 30 |
+
"learning_rate": 0.0002995092130187704,
|
| 31 |
+
"loss": 0.5788,
|
| 32 |
+
"step": 200
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.10314595152140278,
|
| 36 |
+
"eval_loss": 0.34664157032966614,
|
| 37 |
+
"eval_runtime": 60.8325,
|
| 38 |
+
"eval_samples_per_second": 16.044,
|
| 39 |
+
"eval_steps_per_second": 4.011,
|
| 40 |
+
"eval_wer": 0.5014148889856335,
|
| 41 |
+
"step": 200
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.15471892728210418,
|
| 45 |
+
"grad_norm": 16.612009048461914,
|
| 46 |
+
"learning_rate": 0.00029899259514379193,
|
| 47 |
+
"loss": 0.4891,
|
| 48 |
+
"step": 300
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.15471892728210418,
|
| 52 |
+
"eval_loss": 0.3219561278820038,
|
| 53 |
+
"eval_runtime": 60.7824,
|
| 54 |
+
"eval_samples_per_second": 16.057,
|
| 55 |
+
"eval_steps_per_second": 4.014,
|
| 56 |
+
"eval_wer": 0.48204179364388333,
|
| 57 |
+
"step": 300
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.20629190304280556,
|
| 61 |
+
"grad_norm": 15.71129035949707,
|
| 62 |
+
"learning_rate": 0.0002984759772688135,
|
| 63 |
+
"loss": 0.4386,
|
| 64 |
+
"step": 400
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.20629190304280556,
|
| 68 |
+
"eval_loss": 0.30709779262542725,
|
| 69 |
+
"eval_runtime": 60.8186,
|
| 70 |
+
"eval_samples_per_second": 16.048,
|
| 71 |
+
"eval_steps_per_second": 4.012,
|
| 72 |
+
"eval_wer": 0.4801915542011319,
|
| 73 |
+
"step": 400
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.25786487880350695,
|
| 77 |
+
"grad_norm": 2.6226158142089844,
|
| 78 |
+
"learning_rate": 0.000297959359393835,
|
| 79 |
+
"loss": 0.4272,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.25786487880350695,
|
| 84 |
+
"eval_loss": 0.30557531118392944,
|
| 85 |
+
"eval_runtime": 61.7053,
|
| 86 |
+
"eval_samples_per_second": 15.817,
|
| 87 |
+
"eval_steps_per_second": 3.954,
|
| 88 |
+
"eval_wer": 0.49880278624292557,
|
| 89 |
+
"step": 500
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.30943785456420836,
|
| 93 |
+
"grad_norm": 2.265719413757324,
|
| 94 |
+
"learning_rate": 0.00029744274151885655,
|
| 95 |
+
"loss": 0.3982,
|
| 96 |
+
"step": 600
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.30943785456420836,
|
| 100 |
+
"eval_loss": 0.2980726957321167,
|
| 101 |
+
"eval_runtime": 61.2582,
|
| 102 |
+
"eval_samples_per_second": 15.933,
|
| 103 |
+
"eval_steps_per_second": 3.983,
|
| 104 |
+
"eval_wer": 0.4625598606878537,
|
| 105 |
+
"step": 600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.36101083032490977,
|
| 109 |
+
"grad_norm": 1.4846241474151611,
|
| 110 |
+
"learning_rate": 0.00029692612364387805,
|
| 111 |
+
"loss": 0.425,
|
| 112 |
+
"step": 700
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.36101083032490977,
|
| 116 |
+
"eval_loss": 0.2976619601249695,
|
| 117 |
+
"eval_runtime": 61.0566,
|
| 118 |
+
"eval_samples_per_second": 15.985,
|
| 119 |
+
"eval_steps_per_second": 3.996,
|
| 120 |
+
"eval_wer": 0.4631040487592512,
|
| 121 |
+
"step": 700
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.4125838060856111,
|
| 125 |
+
"grad_norm": 1.5233323574066162,
|
| 126 |
+
"learning_rate": 0.0002964095057688996,
|
| 127 |
+
"loss": 0.4036,
|
| 128 |
+
"step": 800
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.4125838060856111,
|
| 132 |
+
"eval_loss": 0.2897385358810425,
|
| 133 |
+
"eval_runtime": 60.9529,
|
| 134 |
+
"eval_samples_per_second": 16.012,
|
| 135 |
+
"eval_steps_per_second": 4.003,
|
| 136 |
+
"eval_wer": 0.44383979103178056,
|
| 137 |
+
"step": 800
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.46415678184631254,
|
| 141 |
+
"grad_norm": 1.4918992519378662,
|
| 142 |
+
"learning_rate": 0.0002958928878939211,
|
| 143 |
+
"loss": 0.3903,
|
| 144 |
+
"step": 900
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.46415678184631254,
|
| 148 |
+
"eval_loss": 0.28775253891944885,
|
| 149 |
+
"eval_runtime": 61.5398,
|
| 150 |
+
"eval_samples_per_second": 15.86,
|
| 151 |
+
"eval_steps_per_second": 3.965,
|
| 152 |
+
"eval_wer": 0.4626686983021332,
|
| 153 |
+
"step": 900
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.5157297576070139,
|
| 157 |
+
"grad_norm": 35.91661071777344,
|
| 158 |
+
"learning_rate": 0.0002953762700189426,
|
| 159 |
+
"loss": 0.3758,
|
| 160 |
+
"step": 1000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.5157297576070139,
|
| 164 |
+
"eval_loss": 0.29262155294418335,
|
| 165 |
+
"eval_runtime": 61.6256,
|
| 166 |
+
"eval_samples_per_second": 15.838,
|
| 167 |
+
"eval_steps_per_second": 3.959,
|
| 168 |
+
"eval_wer": 0.4523291249455812,
|
| 169 |
+
"step": 1000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.5673027333677153,
|
| 173 |
+
"grad_norm": 2.335728645324707,
|
| 174 |
+
"learning_rate": 0.0002948596521439642,
|
| 175 |
+
"loss": 0.3861,
|
| 176 |
+
"step": 1100
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.5673027333677153,
|
| 180 |
+
"eval_loss": 0.28073564171791077,
|
| 181 |
+
"eval_runtime": 60.9884,
|
| 182 |
+
"eval_samples_per_second": 16.003,
|
| 183 |
+
"eval_steps_per_second": 4.001,
|
| 184 |
+
"eval_wer": 0.44101001306051374,
|
| 185 |
+
"step": 1100
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.6188757091284167,
|
| 189 |
+
"grad_norm": 1.2785513401031494,
|
| 190 |
+
"learning_rate": 0.0002943430342689857,
|
| 191 |
+
"loss": 0.3763,
|
| 192 |
+
"step": 1200
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.6188757091284167,
|
| 196 |
+
"eval_loss": 0.2789745032787323,
|
| 197 |
+
"eval_runtime": 61.0183,
|
| 198 |
+
"eval_samples_per_second": 15.995,
|
| 199 |
+
"eval_steps_per_second": 3.999,
|
| 200 |
+
"eval_wer": 0.4330648672181106,
|
| 201 |
+
"step": 1200
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.6704486848891181,
|
| 205 |
+
"grad_norm": 4.705647945404053,
|
| 206 |
+
"learning_rate": 0.0002938264163940072,
|
| 207 |
+
"loss": 0.3984,
|
| 208 |
+
"step": 1300
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.6704486848891181,
|
| 212 |
+
"eval_loss": 0.2803143262863159,
|
| 213 |
+
"eval_runtime": 61.7547,
|
| 214 |
+
"eval_samples_per_second": 15.804,
|
| 215 |
+
"eval_steps_per_second": 3.951,
|
| 216 |
+
"eval_wer": 0.4312146277753592,
|
| 217 |
+
"step": 1300
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.7220216606498195,
|
| 221 |
+
"grad_norm": 1.780588984489441,
|
| 222 |
+
"learning_rate": 0.00029330979851902874,
|
| 223 |
+
"loss": 0.373,
|
| 224 |
+
"step": 1400
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.7220216606498195,
|
| 228 |
+
"eval_loss": 0.2802477478981018,
|
| 229 |
+
"eval_runtime": 61.6178,
|
| 230 |
+
"eval_samples_per_second": 15.84,
|
| 231 |
+
"eval_steps_per_second": 3.96,
|
| 232 |
+
"eval_wer": 0.42457553330430997,
|
| 233 |
+
"step": 1400
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.7735946364105208,
|
| 237 |
+
"grad_norm": 2.3556313514709473,
|
| 238 |
+
"learning_rate": 0.0002927931806440503,
|
| 239 |
+
"loss": 0.3848,
|
| 240 |
+
"step": 1500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.7735946364105208,
|
| 244 |
+
"eval_loss": 0.2759012281894684,
|
| 245 |
+
"eval_runtime": 61.2513,
|
| 246 |
+
"eval_samples_per_second": 15.934,
|
| 247 |
+
"eval_steps_per_second": 3.984,
|
| 248 |
+
"eval_wer": 0.47518502394427514,
|
| 249 |
+
"step": 1500
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.8251676121712223,
|
| 253 |
+
"grad_norm": 4.117414951324463,
|
| 254 |
+
"learning_rate": 0.0002922765627690718,
|
| 255 |
+
"loss": 0.4235,
|
| 256 |
+
"step": 1600
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.8251676121712223,
|
| 260 |
+
"eval_loss": 0.2738034725189209,
|
| 261 |
+
"eval_runtime": 60.9855,
|
| 262 |
+
"eval_samples_per_second": 16.004,
|
| 263 |
+
"eval_steps_per_second": 4.001,
|
| 264 |
+
"eval_wer": 0.42675228558989986,
|
| 265 |
+
"step": 1600
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.8767405879319237,
|
| 269 |
+
"grad_norm": 7.5644683837890625,
|
| 270 |
+
"learning_rate": 0.0002917651110728431,
|
| 271 |
+
"loss": 0.3704,
|
| 272 |
+
"step": 1700
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.8767405879319237,
|
| 276 |
+
"eval_loss": 0.26875266432762146,
|
| 277 |
+
"eval_runtime": 61.6159,
|
| 278 |
+
"eval_samples_per_second": 15.84,
|
| 279 |
+
"eval_steps_per_second": 3.96,
|
| 280 |
+
"eval_wer": 0.4218545929473226,
|
| 281 |
+
"step": 1700
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.9283135636926251,
|
| 285 |
+
"grad_norm": 2.0241034030914307,
|
| 286 |
+
"learning_rate": 0.00029124849319786463,
|
| 287 |
+
"loss": 0.3911,
|
| 288 |
+
"step": 1800
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.9283135636926251,
|
| 292 |
+
"eval_loss": 0.2653037905693054,
|
| 293 |
+
"eval_runtime": 61.6914,
|
| 294 |
+
"eval_samples_per_second": 15.821,
|
| 295 |
+
"eval_steps_per_second": 3.955,
|
| 296 |
+
"eval_wer": 0.42011319111885065,
|
| 297 |
+
"step": 1800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.9798865394533265,
|
| 301 |
+
"grad_norm": 1.9250996112823486,
|
| 302 |
+
"learning_rate": 0.00029073187532288613,
|
| 303 |
+
"loss": 0.3954,
|
| 304 |
+
"step": 1900
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.9798865394533265,
|
| 308 |
+
"eval_loss": 0.26971080899238586,
|
| 309 |
+
"eval_runtime": 61.0364,
|
| 310 |
+
"eval_samples_per_second": 15.99,
|
| 311 |
+
"eval_steps_per_second": 3.998,
|
| 312 |
+
"eval_wer": 0.4481932956029604,
|
| 313 |
+
"step": 1900
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 1.0314595152140278,
|
| 317 |
+
"grad_norm": 2.152578353881836,
|
| 318 |
+
"learning_rate": 0.0002902152574479077,
|
| 319 |
+
"loss": 0.352,
|
| 320 |
+
"step": 2000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.0314595152140278,
|
| 324 |
+
"eval_loss": 0.26541659235954285,
|
| 325 |
+
"eval_runtime": 61.1446,
|
| 326 |
+
"eval_samples_per_second": 15.962,
|
| 327 |
+
"eval_steps_per_second": 3.991,
|
| 328 |
+
"eval_wer": 0.4154331737048324,
|
| 329 |
+
"step": 2000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 1.0830324909747293,
|
| 333 |
+
"grad_norm": 1.2193535566329956,
|
| 334 |
+
"learning_rate": 0.0002896986395729292,
|
| 335 |
+
"loss": 0.3808,
|
| 336 |
+
"step": 2100
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 1.0830324909747293,
|
| 340 |
+
"eval_loss": 0.2631310522556305,
|
| 341 |
+
"eval_runtime": 61.7949,
|
| 342 |
+
"eval_samples_per_second": 15.794,
|
| 343 |
+
"eval_steps_per_second": 3.949,
|
| 344 |
+
"eval_wer": 0.40509360034828035,
|
| 345 |
+
"step": 2100
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.1346054667354306,
|
| 349 |
+
"grad_norm": 0.9607815742492676,
|
| 350 |
+
"learning_rate": 0.0002891820216979507,
|
| 351 |
+
"loss": 0.3681,
|
| 352 |
+
"step": 2200
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 1.1346054667354306,
|
| 356 |
+
"eval_loss": 0.26097217202186584,
|
| 357 |
+
"eval_runtime": 61.6359,
|
| 358 |
+
"eval_samples_per_second": 15.835,
|
| 359 |
+
"eval_steps_per_second": 3.959,
|
| 360 |
+
"eval_wer": 0.4218545929473226,
|
| 361 |
+
"step": 2200
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.1861784424961321,
|
| 365 |
+
"grad_norm": 1.6105040311813354,
|
| 366 |
+
"learning_rate": 0.00028866540382297226,
|
| 367 |
+
"loss": 0.3355,
|
| 368 |
+
"step": 2300
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.1861784424961321,
|
| 372 |
+
"eval_loss": 0.26081275939941406,
|
| 373 |
+
"eval_runtime": 61.1026,
|
| 374 |
+
"eval_samples_per_second": 15.973,
|
| 375 |
+
"eval_steps_per_second": 3.993,
|
| 376 |
+
"eval_wer": 0.40977361776229865,
|
| 377 |
+
"step": 2300
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 1.2377514182568334,
|
| 381 |
+
"grad_norm": 2.3081679344177246,
|
| 382 |
+
"learning_rate": 0.0002881487859479938,
|
| 383 |
+
"loss": 0.342,
|
| 384 |
+
"step": 2400
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 1.2377514182568334,
|
| 388 |
+
"eval_loss": 0.2601791322231293,
|
| 389 |
+
"eval_runtime": 61.0993,
|
| 390 |
+
"eval_samples_per_second": 15.974,
|
| 391 |
+
"eval_steps_per_second": 3.993,
|
| 392 |
+
"eval_wer": 0.40824989116238575,
|
| 393 |
+
"step": 2400
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 1.2893243940175347,
|
| 397 |
+
"grad_norm": 0.6918842792510986,
|
| 398 |
+
"learning_rate": 0.0002876321680730153,
|
| 399 |
+
"loss": 0.347,
|
| 400 |
+
"step": 2500
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 1.2893243940175347,
|
| 404 |
+
"eval_loss": 0.26280567049980164,
|
| 405 |
+
"eval_runtime": 61.7023,
|
| 406 |
+
"eval_samples_per_second": 15.818,
|
| 407 |
+
"eval_steps_per_second": 3.954,
|
| 408 |
+
"eval_wer": 0.40552895080539836,
|
| 409 |
+
"step": 2500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.3408973697782363,
|
| 413 |
+
"grad_norm": 0.920050323009491,
|
| 414 |
+
"learning_rate": 0.0002871155501980368,
|
| 415 |
+
"loss": 0.3409,
|
| 416 |
+
"step": 2600
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.3408973697782363,
|
| 420 |
+
"eval_loss": 0.25879552960395813,
|
| 421 |
+
"eval_runtime": 61.69,
|
| 422 |
+
"eval_samples_per_second": 15.821,
|
| 423 |
+
"eval_steps_per_second": 3.955,
|
| 424 |
+
"eval_wer": 0.412929908576404,
|
| 425 |
+
"step": 2600
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 1.3924703455389376,
|
| 429 |
+
"grad_norm": 0.7028564810752869,
|
| 430 |
+
"learning_rate": 0.0002865989323230584,
|
| 431 |
+
"loss": 0.3423,
|
| 432 |
+
"step": 2700
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 1.3924703455389376,
|
| 436 |
+
"eval_loss": 0.2616526484489441,
|
| 437 |
+
"eval_runtime": 61.13,
|
| 438 |
+
"eval_samples_per_second": 15.966,
|
| 439 |
+
"eval_steps_per_second": 3.991,
|
| 440 |
+
"eval_wer": 0.41924249020461474,
|
| 441 |
+
"step": 2700
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.444043321299639,
|
| 445 |
+
"grad_norm": 3.5141775608062744,
|
| 446 |
+
"learning_rate": 0.0002860823144480799,
|
| 447 |
+
"loss": 0.3341,
|
| 448 |
+
"step": 2800
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 1.444043321299639,
|
| 452 |
+
"eval_loss": 0.25779473781585693,
|
| 453 |
+
"eval_runtime": 61.3161,
|
| 454 |
+
"eval_samples_per_second": 15.918,
|
| 455 |
+
"eval_steps_per_second": 3.979,
|
| 456 |
+
"eval_wer": 0.40552895080539836,
|
| 457 |
+
"step": 2800
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 1.4956162970603404,
|
| 461 |
+
"grad_norm": 1.0055650472640991,
|
| 462 |
+
"learning_rate": 0.0002855656965731014,
|
| 463 |
+
"loss": 0.3425,
|
| 464 |
+
"step": 2900
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.4956162970603404,
|
| 468 |
+
"eval_loss": 0.2579568922519684,
|
| 469 |
+
"eval_runtime": 61.9639,
|
| 470 |
+
"eval_samples_per_second": 15.751,
|
| 471 |
+
"eval_steps_per_second": 3.938,
|
| 472 |
+
"eval_wer": 0.39878101872006966,
|
| 473 |
+
"step": 2900
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.5471892728210417,
|
| 477 |
+
"grad_norm": 1.4293900728225708,
|
| 478 |
+
"learning_rate": 0.00028504907869812294,
|
| 479 |
+
"loss": 0.337,
|
| 480 |
+
"step": 3000
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.5471892728210417,
|
| 484 |
+
"eval_loss": 0.25681352615356445,
|
| 485 |
+
"eval_runtime": 61.6019,
|
| 486 |
+
"eval_samples_per_second": 15.844,
|
| 487 |
+
"eval_steps_per_second": 3.961,
|
| 488 |
+
"eval_wer": 0.40705267740531126,
|
| 489 |
+
"step": 3000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 1.5987622485817432,
|
| 493 |
+
"grad_norm": 0.6135945916175842,
|
| 494 |
+
"learning_rate": 0.00028453246082314445,
|
| 495 |
+
"loss": 0.3412,
|
| 496 |
+
"step": 3100
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 1.5987622485817432,
|
| 500 |
+
"eval_loss": 0.25524020195007324,
|
| 501 |
+
"eval_runtime": 61.308,
|
| 502 |
+
"eval_samples_per_second": 15.92,
|
| 503 |
+
"eval_steps_per_second": 3.98,
|
| 504 |
+
"eval_wer": 0.39932520679146716,
|
| 505 |
+
"step": 3100
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 1.6503352243424445,
|
| 509 |
+
"grad_norm": 4.757889270782471,
|
| 510 |
+
"learning_rate": 0.00028401584294816595,
|
| 511 |
+
"loss": 0.3837,
|
| 512 |
+
"step": 3200
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 1.6503352243424445,
|
| 516 |
+
"eval_loss": 0.26221156120300293,
|
| 517 |
+
"eval_runtime": 61.5959,
|
| 518 |
+
"eval_samples_per_second": 15.845,
|
| 519 |
+
"eval_steps_per_second": 3.961,
|
| 520 |
+
"eval_wer": 0.40835872877666524,
|
| 521 |
+
"step": 3200
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.701908200103146,
|
| 525 |
+
"grad_norm": 1.1227970123291016,
|
| 526 |
+
"learning_rate": 0.0002834992250731875,
|
| 527 |
+
"loss": 0.3372,
|
| 528 |
+
"step": 3300
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.701908200103146,
|
| 532 |
+
"eval_loss": 0.2548165023326874,
|
| 533 |
+
"eval_runtime": 61.9858,
|
| 534 |
+
"eval_samples_per_second": 15.746,
|
| 535 |
+
"eval_steps_per_second": 3.936,
|
| 536 |
+
"eval_wer": 0.3991075315629081,
|
| 537 |
+
"step": 3300
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 1.7534811758638473,
|
| 541 |
+
"grad_norm": 0.8613722324371338,
|
| 542 |
+
"learning_rate": 0.00028298260719820907,
|
| 543 |
+
"loss": 0.3394,
|
| 544 |
+
"step": 3400
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 1.7534811758638473,
|
| 548 |
+
"eval_loss": 0.2535094916820526,
|
| 549 |
+
"eval_runtime": 61.5961,
|
| 550 |
+
"eval_samples_per_second": 15.845,
|
| 551 |
+
"eval_steps_per_second": 3.961,
|
| 552 |
+
"eval_wer": 0.4060731388767958,
|
| 553 |
+
"step": 3400
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 1.8050541516245486,
|
| 557 |
+
"grad_norm": 0.5101200938224792,
|
| 558 |
+
"learning_rate": 0.00028246598932323057,
|
| 559 |
+
"loss": 0.3542,
|
| 560 |
+
"step": 3500
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.8050541516245486,
|
| 564 |
+
"eval_loss": 0.25123441219329834,
|
| 565 |
+
"eval_runtime": 61.2846,
|
| 566 |
+
"eval_samples_per_second": 15.926,
|
| 567 |
+
"eval_steps_per_second": 3.981,
|
| 568 |
+
"eval_wer": 0.39268611232041795,
|
| 569 |
+
"step": 3500
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 1.8566271273852502,
|
| 573 |
+
"grad_norm": 1.271552324295044,
|
| 574 |
+
"learning_rate": 0.0002819493714482521,
|
| 575 |
+
"loss": 0.3368,
|
| 576 |
+
"step": 3600
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 1.8566271273852502,
|
| 580 |
+
"eval_loss": 0.258027583360672,
|
| 581 |
+
"eval_runtime": 61.4346,
|
| 582 |
+
"eval_samples_per_second": 15.887,
|
| 583 |
+
"eval_steps_per_second": 3.972,
|
| 584 |
+
"eval_wer": 0.4004135829342621,
|
| 585 |
+
"step": 3600
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.9082001031459517,
|
| 589 |
+
"grad_norm": 1.8105818033218384,
|
| 590 |
+
"learning_rate": 0.00028143275357327363,
|
| 591 |
+
"loss": 0.3807,
|
| 592 |
+
"step": 3700
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.9082001031459517,
|
| 596 |
+
"eval_loss": 0.24900555610656738,
|
| 597 |
+
"eval_runtime": 61.7426,
|
| 598 |
+
"eval_samples_per_second": 15.808,
|
| 599 |
+
"eval_steps_per_second": 3.952,
|
| 600 |
+
"eval_wer": 0.39747496734871574,
|
| 601 |
+
"step": 3700
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 1.959773078906653,
|
| 605 |
+
"grad_norm": 1.8133718967437744,
|
| 606 |
+
"learning_rate": 0.00028091613569829514,
|
| 607 |
+
"loss": 0.3454,
|
| 608 |
+
"step": 3800
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 1.959773078906653,
|
| 612 |
+
"eval_loss": 0.2513742446899414,
|
| 613 |
+
"eval_runtime": 61.6449,
|
| 614 |
+
"eval_samples_per_second": 15.833,
|
| 615 |
+
"eval_steps_per_second": 3.958,
|
| 616 |
+
"eval_wer": 0.40019590770570307,
|
| 617 |
+
"step": 3800
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 2.0113460546673543,
|
| 621 |
+
"grad_norm": 0.7546507120132446,
|
| 622 |
+
"learning_rate": 0.00028039951782331664,
|
| 623 |
+
"loss": 0.3456,
|
| 624 |
+
"step": 3900
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 2.0113460546673543,
|
| 628 |
+
"eval_loss": 0.24565543234348297,
|
| 629 |
+
"eval_runtime": 61.2207,
|
| 630 |
+
"eval_samples_per_second": 15.942,
|
| 631 |
+
"eval_steps_per_second": 3.986,
|
| 632 |
+
"eval_wer": 0.3931214627775359,
|
| 633 |
+
"step": 3900
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 2.0629190304280556,
|
| 637 |
+
"grad_norm": 0.9227738976478577,
|
| 638 |
+
"learning_rate": 0.0002798828999483382,
|
| 639 |
+
"loss": 0.3202,
|
| 640 |
+
"step": 4000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 2.0629190304280556,
|
| 644 |
+
"eval_loss": 0.24660241603851318,
|
| 645 |
+
"eval_runtime": 61.4456,
|
| 646 |
+
"eval_samples_per_second": 15.884,
|
| 647 |
+
"eval_steps_per_second": 3.971,
|
| 648 |
+
"eval_wer": 0.391597736177623,
|
| 649 |
+
"step": 4000
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 2.114492006188757,
|
| 653 |
+
"grad_norm": 1.0991692543029785,
|
| 654 |
+
"learning_rate": 0.00027937144825210947,
|
| 655 |
+
"loss": 0.3233,
|
| 656 |
+
"step": 4100
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 2.114492006188757,
|
| 660 |
+
"eval_loss": 0.2494671791791916,
|
| 661 |
+
"eval_runtime": 61.9678,
|
| 662 |
+
"eval_samples_per_second": 15.75,
|
| 663 |
+
"eval_steps_per_second": 3.938,
|
| 664 |
+
"eval_wer": 0.39747496734871574,
|
| 665 |
+
"step": 4100
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 2.1660649819494586,
|
| 669 |
+
"grad_norm": 0.6972938776016235,
|
| 670 |
+
"learning_rate": 0.000278854830377131,
|
| 671 |
+
"loss": 0.3052,
|
| 672 |
+
"step": 4200
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 2.1660649819494586,
|
| 676 |
+
"eval_loss": 0.247751384973526,
|
| 677 |
+
"eval_runtime": 61.6785,
|
| 678 |
+
"eval_samples_per_second": 15.824,
|
| 679 |
+
"eval_steps_per_second": 3.956,
|
| 680 |
+
"eval_wer": 0.38985633434915107,
|
| 681 |
+
"step": 4200
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 2.1660649819494586,
|
| 685 |
+
"step": 4200,
|
| 686 |
+
"total_flos": 1.0016492328632693e+19,
|
| 687 |
+
"train_loss": 0.5294508952186221,
|
| 688 |
+
"train_runtime": 6287.6058,
|
| 689 |
+
"train_samples_per_second": 37.006,
|
| 690 |
+
"train_steps_per_second": 9.252
|
| 691 |
+
}
|
| 692 |
+
],
|
| 693 |
+
"logging_steps": 100,
|
| 694 |
+
"max_steps": 58170,
|
| 695 |
+
"num_input_tokens_seen": 0,
|
| 696 |
+
"num_train_epochs": 30,
|
| 697 |
+
"save_steps": 400,
|
| 698 |
+
"stateful_callbacks": {
|
| 699 |
+
"EarlyStoppingCallback": {
|
| 700 |
+
"args": {
|
| 701 |
+
"early_stopping_patience": 3,
|
| 702 |
+
"early_stopping_threshold": 0.0
|
| 703 |
+
},
|
| 704 |
+
"attributes": {
|
| 705 |
+
"early_stopping_patience_counter": 1
|
| 706 |
+
}
|
| 707 |
+
},
|
| 708 |
+
"TrainerControl": {
|
| 709 |
+
"args": {
|
| 710 |
+
"should_epoch_stop": false,
|
| 711 |
+
"should_evaluate": false,
|
| 712 |
+
"should_log": false,
|
| 713 |
+
"should_save": true,
|
| 714 |
+
"should_training_stop": false
|
| 715 |
+
},
|
| 716 |
+
"attributes": {}
|
| 717 |
+
}
|
| 718 |
+
},
|
| 719 |
+
"total_flos": 1.0016492328632693e+19,
|
| 720 |
+
"train_batch_size": 4,
|
| 721 |
+
"trial_name": null,
|
| 722 |
+
"trial_params": null
|
| 723 |
+
}
|