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 +755 -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-natbed-combined
|
| 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-natbed-combined
|
| 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: 0.5565
|
| 25 |
+
- Wer: 0.4959
|
| 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:d9b4b54421e7838e372b14b9496570d562606f0cc264742d23128d641564b9c0
|
| 3 |
+
size 8824152
|
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.506883604505632,
|
| 3 |
+
"eval_loss": 0.5565288662910461,
|
| 4 |
+
"eval_runtime": 95.2893,
|
| 5 |
+
"eval_samples": 1358,
|
| 6 |
+
"eval_samples_per_second": 14.251,
|
| 7 |
+
"eval_steps_per_second": 1.784,
|
| 8 |
+
"eval_wer": 0.4959462779461142,
|
| 9 |
+
"total_flos": 2.278783878958534e+19,
|
| 10 |
+
"train_loss": 0.858475610559637,
|
| 11 |
+
"train_runtime": 9525.4326,
|
| 12 |
+
"train_samples": 6387,
|
| 13 |
+
"train_samples_per_second": 20.116,
|
| 14 |
+
"train_steps_per_second": 2.516
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.506883604505632,
|
| 3 |
+
"eval_loss": 0.5565288662910461,
|
| 4 |
+
"eval_runtime": 95.2893,
|
| 5 |
+
"eval_samples": 1358,
|
| 6 |
+
"eval_samples_per_second": 14.251,
|
| 7 |
+
"eval_steps_per_second": 1.784,
|
| 8 |
+
"eval_wer": 0.4959462779461142
|
| 9 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.506883604505632,
|
| 3 |
+
"total_flos": 2.278783878958534e+19,
|
| 4 |
+
"train_loss": 0.858475610559637,
|
| 5 |
+
"train_runtime": 9525.4326,
|
| 6 |
+
"train_samples": 6387,
|
| 7 |
+
"train_samples_per_second": 20.116,
|
| 8 |
+
"train_steps_per_second": 2.516
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,755 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5565288662910461,
|
| 3 |
+
"best_model_checkpoint": "/scratch/skscla001/results/mms-1b-all-bem-natbed-combined/checkpoint-3800",
|
| 4 |
+
"epoch": 5.506883604505632,
|
| 5 |
+
"eval_steps": 100,
|
| 6 |
+
"global_step": 4400,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1251564455569462,
|
| 13 |
+
"grad_norm": 4.582801342010498,
|
| 14 |
+
"learning_rate": 0.00029099999999999997,
|
| 15 |
+
"loss": 8.2239,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.1251564455569462,
|
| 20 |
+
"eval_loss": 1.3301187753677368,
|
| 21 |
+
"eval_runtime": 95.4493,
|
| 22 |
+
"eval_samples_per_second": 14.227,
|
| 23 |
+
"eval_steps_per_second": 1.781,
|
| 24 |
+
"eval_wer": 0.8959954139710098,
|
| 25 |
+
"step": 100
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.2503128911138924,
|
| 29 |
+
"grad_norm": 2.8414461612701416,
|
| 30 |
+
"learning_rate": 0.000298780896522832,
|
| 31 |
+
"loss": 0.8992,
|
| 32 |
+
"step": 200
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.2503128911138924,
|
| 36 |
+
"eval_loss": 0.7774138450622559,
|
| 37 |
+
"eval_runtime": 95.7815,
|
| 38 |
+
"eval_samples_per_second": 14.178,
|
| 39 |
+
"eval_steps_per_second": 1.775,
|
| 40 |
+
"eval_wer": 0.6657112439603636,
|
| 41 |
+
"step": 200
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.37546933667083854,
|
| 45 |
+
"grad_norm": 1.6420665979385376,
|
| 46 |
+
"learning_rate": 0.00029752408881441136,
|
| 47 |
+
"loss": 0.8263,
|
| 48 |
+
"step": 300
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.37546933667083854,
|
| 52 |
+
"eval_loss": 0.7549673914909363,
|
| 53 |
+
"eval_runtime": 95.0731,
|
| 54 |
+
"eval_samples_per_second": 14.284,
|
| 55 |
+
"eval_steps_per_second": 1.788,
|
| 56 |
+
"eval_wer": 0.5891409384980755,
|
| 57 |
+
"step": 300
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.5006257822277848,
|
| 61 |
+
"grad_norm": 1.838030219078064,
|
| 62 |
+
"learning_rate": 0.000296279849183075,
|
| 63 |
+
"loss": 0.7846,
|
| 64 |
+
"step": 400
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.5006257822277848,
|
| 68 |
+
"eval_loss": 0.6995427012443542,
|
| 69 |
+
"eval_runtime": 94.8499,
|
| 70 |
+
"eval_samples_per_second": 14.317,
|
| 71 |
+
"eval_steps_per_second": 1.792,
|
| 72 |
+
"eval_wer": 0.5665383670461059,
|
| 73 |
+
"step": 400
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.6257822277847309,
|
| 77 |
+
"grad_norm": 3.311655282974243,
|
| 78 |
+
"learning_rate": 0.00029502304147465433,
|
| 79 |
+
"loss": 0.9046,
|
| 80 |
+
"step": 500
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.6257822277847309,
|
| 84 |
+
"eval_loss": 0.6796234846115112,
|
| 85 |
+
"eval_runtime": 96.1056,
|
| 86 |
+
"eval_samples_per_second": 14.13,
|
| 87 |
+
"eval_steps_per_second": 1.769,
|
| 88 |
+
"eval_wer": 0.5545819343215134,
|
| 89 |
+
"step": 500
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.7509386733416771,
|
| 93 |
+
"grad_norm": 1.0524609088897705,
|
| 94 |
+
"learning_rate": 0.00029376623376623374,
|
| 95 |
+
"loss": 0.7688,
|
| 96 |
+
"step": 600
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.7509386733416771,
|
| 100 |
+
"eval_loss": 0.6881176829338074,
|
| 101 |
+
"eval_runtime": 95.4601,
|
| 102 |
+
"eval_samples_per_second": 14.226,
|
| 103 |
+
"eval_steps_per_second": 1.781,
|
| 104 |
+
"eval_wer": 0.5436082220948325,
|
| 105 |
+
"step": 600
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.8760951188986232,
|
| 109 |
+
"grad_norm": 1.534467339515686,
|
| 110 |
+
"learning_rate": 0.00029250942605781315,
|
| 111 |
+
"loss": 0.7139,
|
| 112 |
+
"step": 700
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.8760951188986232,
|
| 116 |
+
"eval_loss": 0.6998778581619263,
|
| 117 |
+
"eval_runtime": 96.7901,
|
| 118 |
+
"eval_samples_per_second": 14.03,
|
| 119 |
+
"eval_steps_per_second": 1.756,
|
| 120 |
+
"eval_wer": 0.5557284415690771,
|
| 121 |
+
"step": 700
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.0012515644555695,
|
| 125 |
+
"grad_norm": 4.051555633544922,
|
| 126 |
+
"learning_rate": 0.0002912526183493925,
|
| 127 |
+
"loss": 0.7922,
|
| 128 |
+
"step": 800
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.0012515644555695,
|
| 132 |
+
"eval_loss": 0.6810752153396606,
|
| 133 |
+
"eval_runtime": 96.2586,
|
| 134 |
+
"eval_samples_per_second": 14.108,
|
| 135 |
+
"eval_steps_per_second": 1.766,
|
| 136 |
+
"eval_wer": 0.542134141347965,
|
| 137 |
+
"step": 800
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 1.1264080100125156,
|
| 141 |
+
"grad_norm": 3.281071186065674,
|
| 142 |
+
"learning_rate": 0.0002899958106409719,
|
| 143 |
+
"loss": 0.7929,
|
| 144 |
+
"step": 900
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.1264080100125156,
|
| 148 |
+
"eval_loss": 0.6760120391845703,
|
| 149 |
+
"eval_runtime": 97.2734,
|
| 150 |
+
"eval_samples_per_second": 13.961,
|
| 151 |
+
"eval_steps_per_second": 1.748,
|
| 152 |
+
"eval_wer": 0.5433625419703546,
|
| 153 |
+
"step": 900
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 1.2515644555694618,
|
| 157 |
+
"grad_norm": 6.815141201019287,
|
| 158 |
+
"learning_rate": 0.0002887515710096355,
|
| 159 |
+
"loss": 0.7508,
|
| 160 |
+
"step": 1000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 1.2515644555694618,
|
| 164 |
+
"eval_loss": 0.6555261015892029,
|
| 165 |
+
"eval_runtime": 96.3987,
|
| 166 |
+
"eval_samples_per_second": 14.087,
|
| 167 |
+
"eval_steps_per_second": 1.764,
|
| 168 |
+
"eval_wer": 0.5660470067971501,
|
| 169 |
+
"step": 1000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 1.3767209011264079,
|
| 173 |
+
"grad_norm": 2.6557281017303467,
|
| 174 |
+
"learning_rate": 0.0002874947633012149,
|
| 175 |
+
"loss": 0.7534,
|
| 176 |
+
"step": 1100
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.3767209011264079,
|
| 180 |
+
"eval_loss": 0.6411116719245911,
|
| 181 |
+
"eval_runtime": 96.0443,
|
| 182 |
+
"eval_samples_per_second": 14.139,
|
| 183 |
+
"eval_steps_per_second": 1.77,
|
| 184 |
+
"eval_wer": 0.5372205388584064,
|
| 185 |
+
"step": 1100
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.5018773466833542,
|
| 189 |
+
"grad_norm": 19.327106475830078,
|
| 190 |
+
"learning_rate": 0.00028623795559279424,
|
| 191 |
+
"loss": 0.7316,
|
| 192 |
+
"step": 1200
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.5018773466833542,
|
| 196 |
+
"eval_loss": 0.6420451998710632,
|
| 197 |
+
"eval_runtime": 97.2234,
|
| 198 |
+
"eval_samples_per_second": 13.968,
|
| 199 |
+
"eval_steps_per_second": 1.749,
|
| 200 |
+
"eval_wer": 0.5319793628695438,
|
| 201 |
+
"step": 1200
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 1.6270337922403004,
|
| 205 |
+
"grad_norm": 4.125315189361572,
|
| 206 |
+
"learning_rate": 0.00028498114788437365,
|
| 207 |
+
"loss": 0.7147,
|
| 208 |
+
"step": 1300
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 1.6270337922403004,
|
| 212 |
+
"eval_loss": 0.6725718975067139,
|
| 213 |
+
"eval_runtime": 97.3227,
|
| 214 |
+
"eval_samples_per_second": 13.954,
|
| 215 |
+
"eval_steps_per_second": 1.747,
|
| 216 |
+
"eval_wer": 0.5293587748751126,
|
| 217 |
+
"step": 1300
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.7521902377972465,
|
| 221 |
+
"grad_norm": 10.938685417175293,
|
| 222 |
+
"learning_rate": 0.00028372434017595306,
|
| 223 |
+
"loss": 0.6734,
|
| 224 |
+
"step": 1400
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.7521902377972465,
|
| 228 |
+
"eval_loss": 0.6308336853981018,
|
| 229 |
+
"eval_runtime": 96.6381,
|
| 230 |
+
"eval_samples_per_second": 14.052,
|
| 231 |
+
"eval_steps_per_second": 1.759,
|
| 232 |
+
"eval_wer": 0.5252641061338138,
|
| 233 |
+
"step": 1400
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.8773466833541927,
|
| 237 |
+
"grad_norm": 7.951139450073242,
|
| 238 |
+
"learning_rate": 0.00028246753246753247,
|
| 239 |
+
"loss": 0.7084,
|
| 240 |
+
"step": 1500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 1.8773466833541927,
|
| 244 |
+
"eval_loss": 0.6205306649208069,
|
| 245 |
+
"eval_runtime": 97.9744,
|
| 246 |
+
"eval_samples_per_second": 13.861,
|
| 247 |
+
"eval_steps_per_second": 1.735,
|
| 248 |
+
"eval_wer": 0.5438539022193104,
|
| 249 |
+
"step": 1500
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 2.002503128911139,
|
| 253 |
+
"grad_norm": 2.1400046348571777,
|
| 254 |
+
"learning_rate": 0.00028121072475911183,
|
| 255 |
+
"loss": 0.6714,
|
| 256 |
+
"step": 1600
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.002503128911139,
|
| 260 |
+
"eval_loss": 0.6118720769882202,
|
| 261 |
+
"eval_runtime": 97.507,
|
| 262 |
+
"eval_samples_per_second": 13.927,
|
| 263 |
+
"eval_steps_per_second": 1.743,
|
| 264 |
+
"eval_wer": 0.5231348783883384,
|
| 265 |
+
"step": 1600
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 2.127659574468085,
|
| 269 |
+
"grad_norm": 8.502281188964844,
|
| 270 |
+
"learning_rate": 0.00027995391705069124,
|
| 271 |
+
"loss": 0.6888,
|
| 272 |
+
"step": 1700
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.127659574468085,
|
| 276 |
+
"eval_loss": 0.6349774599075317,
|
| 277 |
+
"eval_runtime": 96.997,
|
| 278 |
+
"eval_samples_per_second": 14.0,
|
| 279 |
+
"eval_steps_per_second": 1.753,
|
| 280 |
+
"eval_wer": 0.5166653017770862,
|
| 281 |
+
"step": 1700
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 2.252816020025031,
|
| 285 |
+
"grad_norm": 1.059962272644043,
|
| 286 |
+
"learning_rate": 0.0002786971093422706,
|
| 287 |
+
"loss": 0.6871,
|
| 288 |
+
"step": 1800
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.252816020025031,
|
| 292 |
+
"eval_loss": 0.6183043718338013,
|
| 293 |
+
"eval_runtime": 97.927,
|
| 294 |
+
"eval_samples_per_second": 13.867,
|
| 295 |
+
"eval_steps_per_second": 1.736,
|
| 296 |
+
"eval_wer": 0.5118335926623536,
|
| 297 |
+
"step": 1800
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 2.3779724655819776,
|
| 301 |
+
"grad_norm": 1.3032373189926147,
|
| 302 |
+
"learning_rate": 0.00027744030163385,
|
| 303 |
+
"loss": 0.6882,
|
| 304 |
+
"step": 1900
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 2.3779724655819776,
|
| 308 |
+
"eval_loss": 0.5973983407020569,
|
| 309 |
+
"eval_runtime": 97.1298,
|
| 310 |
+
"eval_samples_per_second": 13.981,
|
| 311 |
+
"eval_steps_per_second": 1.75,
|
| 312 |
+
"eval_wer": 0.5332896568667594,
|
| 313 |
+
"step": 1900
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 2.5031289111389237,
|
| 317 |
+
"grad_norm": 1.9106544256210327,
|
| 318 |
+
"learning_rate": 0.00027618349392542936,
|
| 319 |
+
"loss": 0.6769,
|
| 320 |
+
"step": 2000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 2.5031289111389237,
|
| 324 |
+
"eval_loss": 0.5994674563407898,
|
| 325 |
+
"eval_runtime": 96.7136,
|
| 326 |
+
"eval_samples_per_second": 14.041,
|
| 327 |
+
"eval_steps_per_second": 1.758,
|
| 328 |
+
"eval_wer": 0.5300958152485464,
|
| 329 |
+
"step": 2000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 2.6282853566958697,
|
| 333 |
+
"grad_norm": 1.2871947288513184,
|
| 334 |
+
"learning_rate": 0.00027492668621700877,
|
| 335 |
+
"loss": 0.6801,
|
| 336 |
+
"step": 2100
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 2.6282853566958697,
|
| 340 |
+
"eval_loss": 0.5880154371261597,
|
| 341 |
+
"eval_runtime": 97.9989,
|
| 342 |
+
"eval_samples_per_second": 13.857,
|
| 343 |
+
"eval_steps_per_second": 1.735,
|
| 344 |
+
"eval_wer": 0.5377937924821882,
|
| 345 |
+
"step": 2100
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 2.7534418022528158,
|
| 349 |
+
"grad_norm": 0.619396448135376,
|
| 350 |
+
"learning_rate": 0.0002736698785085881,
|
| 351 |
+
"loss": 0.6695,
|
| 352 |
+
"step": 2200
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 2.7534418022528158,
|
| 356 |
+
"eval_loss": 0.5972831845283508,
|
| 357 |
+
"eval_runtime": 97.3411,
|
| 358 |
+
"eval_samples_per_second": 13.951,
|
| 359 |
+
"eval_steps_per_second": 1.746,
|
| 360 |
+
"eval_wer": 0.5051183359266236,
|
| 361 |
+
"step": 2200
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 2.8785982478097623,
|
| 365 |
+
"grad_norm": 1.0727863311767578,
|
| 366 |
+
"learning_rate": 0.0002724130708001676,
|
| 367 |
+
"loss": 0.6557,
|
| 368 |
+
"step": 2300
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.8785982478097623,
|
| 372 |
+
"eval_loss": 0.6026735901832581,
|
| 373 |
+
"eval_runtime": 97.4796,
|
| 374 |
+
"eval_samples_per_second": 13.931,
|
| 375 |
+
"eval_steps_per_second": 1.744,
|
| 376 |
+
"eval_wer": 0.5056096961755794,
|
| 377 |
+
"step": 2300
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 3.0037546933667083,
|
| 381 |
+
"grad_norm": 0.626075029373169,
|
| 382 |
+
"learning_rate": 0.00027115626309174695,
|
| 383 |
+
"loss": 0.6525,
|
| 384 |
+
"step": 2400
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 3.0037546933667083,
|
| 388 |
+
"eval_loss": 0.59461510181427,
|
| 389 |
+
"eval_runtime": 97.9487,
|
| 390 |
+
"eval_samples_per_second": 13.864,
|
| 391 |
+
"eval_steps_per_second": 1.736,
|
| 392 |
+
"eval_wer": 0.4996314798132831,
|
| 393 |
+
"step": 2400
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 3.1289111389236544,
|
| 397 |
+
"grad_norm": 4.777908802032471,
|
| 398 |
+
"learning_rate": 0.00026989945538332636,
|
| 399 |
+
"loss": 0.6829,
|
| 400 |
+
"step": 2500
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 3.1289111389236544,
|
| 404 |
+
"eval_loss": 0.5881961584091187,
|
| 405 |
+
"eval_runtime": 97.8698,
|
| 406 |
+
"eval_samples_per_second": 13.876,
|
| 407 |
+
"eval_steps_per_second": 1.737,
|
| 408 |
+
"eval_wer": 0.49979526656293505,
|
| 409 |
+
"step": 2500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 3.254067584480601,
|
| 413 |
+
"grad_norm": 5.189509391784668,
|
| 414 |
+
"learning_rate": 0.0002686426476749057,
|
| 415 |
+
"loss": 0.6627,
|
| 416 |
+
"step": 2600
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.254067584480601,
|
| 420 |
+
"eval_loss": 0.6009677052497864,
|
| 421 |
+
"eval_runtime": 95.9556,
|
| 422 |
+
"eval_samples_per_second": 14.152,
|
| 423 |
+
"eval_steps_per_second": 1.772,
|
| 424 |
+
"eval_wer": 0.49848497256571944,
|
| 425 |
+
"step": 2600
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 3.379224030037547,
|
| 429 |
+
"grad_norm": 6.410580158233643,
|
| 430 |
+
"learning_rate": 0.0002673858399664851,
|
| 431 |
+
"loss": 0.6146,
|
| 432 |
+
"step": 2700
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 3.379224030037547,
|
| 436 |
+
"eval_loss": 0.5770368576049805,
|
| 437 |
+
"eval_runtime": 95.4343,
|
| 438 |
+
"eval_samples_per_second": 14.23,
|
| 439 |
+
"eval_steps_per_second": 1.781,
|
| 440 |
+
"eval_wer": 0.5009417738104988,
|
| 441 |
+
"step": 2700
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 3.504380475594493,
|
| 445 |
+
"grad_norm": 1.4566117525100708,
|
| 446 |
+
"learning_rate": 0.0002661290322580645,
|
| 447 |
+
"loss": 0.6205,
|
| 448 |
+
"step": 2800
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 3.504380475594493,
|
| 452 |
+
"eval_loss": 0.5738538503646851,
|
| 453 |
+
"eval_runtime": 95.9905,
|
| 454 |
+
"eval_samples_per_second": 14.147,
|
| 455 |
+
"eval_steps_per_second": 1.771,
|
| 456 |
+
"eval_wer": 0.5020882810580624,
|
| 457 |
+
"step": 2800
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 3.6295369211514394,
|
| 461 |
+
"grad_norm": 5.905794143676758,
|
| 462 |
+
"learning_rate": 0.0002648722245496439,
|
| 463 |
+
"loss": 0.7025,
|
| 464 |
+
"step": 2900
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 3.6295369211514394,
|
| 468 |
+
"eval_loss": 0.5806447863578796,
|
| 469 |
+
"eval_runtime": 96.9069,
|
| 470 |
+
"eval_samples_per_second": 14.013,
|
| 471 |
+
"eval_steps_per_second": 1.754,
|
| 472 |
+
"eval_wer": 0.5223978380149046,
|
| 473 |
+
"step": 2900
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 3.7546933667083855,
|
| 477 |
+
"grad_norm": 7.341220855712891,
|
| 478 |
+
"learning_rate": 0.00026361541684122324,
|
| 479 |
+
"loss": 0.6379,
|
| 480 |
+
"step": 3000
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 3.7546933667083855,
|
| 484 |
+
"eval_loss": 0.6210225820541382,
|
| 485 |
+
"eval_runtime": 96.7041,
|
| 486 |
+
"eval_samples_per_second": 14.043,
|
| 487 |
+
"eval_steps_per_second": 1.758,
|
| 488 |
+
"eval_wer": 0.5064286299238392,
|
| 489 |
+
"step": 3000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 3.8798498122653315,
|
| 493 |
+
"grad_norm": 2.2515931129455566,
|
| 494 |
+
"learning_rate": 0.00026235860913280265,
|
| 495 |
+
"loss": 0.6104,
|
| 496 |
+
"step": 3100
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 3.8798498122653315,
|
| 500 |
+
"eval_loss": 0.5702349543571472,
|
| 501 |
+
"eval_runtime": 95.9116,
|
| 502 |
+
"eval_samples_per_second": 14.159,
|
| 503 |
+
"eval_steps_per_second": 1.772,
|
| 504 |
+
"eval_wer": 0.5033985750552781,
|
| 505 |
+
"step": 3100
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 4.005006257822278,
|
| 509 |
+
"grad_norm": 0.5629591941833496,
|
| 510 |
+
"learning_rate": 0.00026110180142438206,
|
| 511 |
+
"loss": 0.6607,
|
| 512 |
+
"step": 3200
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 4.005006257822278,
|
| 516 |
+
"eval_loss": 0.5755676627159119,
|
| 517 |
+
"eval_runtime": 96.4548,
|
| 518 |
+
"eval_samples_per_second": 14.079,
|
| 519 |
+
"eval_steps_per_second": 1.762,
|
| 520 |
+
"eval_wer": 0.4891491278355581,
|
| 521 |
+
"step": 3200
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 4.130162703379224,
|
| 525 |
+
"grad_norm": 6.9959259033203125,
|
| 526 |
+
"learning_rate": 0.0002598449937159615,
|
| 527 |
+
"loss": 0.6776,
|
| 528 |
+
"step": 3300
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 4.130162703379224,
|
| 532 |
+
"eval_loss": 0.5678644180297852,
|
| 533 |
+
"eval_runtime": 95.8409,
|
| 534 |
+
"eval_samples_per_second": 14.169,
|
| 535 |
+
"eval_steps_per_second": 1.774,
|
| 536 |
+
"eval_wer": 0.4885758742117763,
|
| 537 |
+
"step": 3300
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 4.25531914893617,
|
| 541 |
+
"grad_norm": 4.9695658683776855,
|
| 542 |
+
"learning_rate": 0.00025860075408462504,
|
| 543 |
+
"loss": 0.6343,
|
| 544 |
+
"step": 3400
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.25531914893617,
|
| 548 |
+
"eval_loss": 0.5597887635231018,
|
| 549 |
+
"eval_runtime": 95.5822,
|
| 550 |
+
"eval_samples_per_second": 14.208,
|
| 551 |
+
"eval_steps_per_second": 1.779,
|
| 552 |
+
"eval_wer": 0.4898861682089919,
|
| 553 |
+
"step": 3400
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 4.380475594493117,
|
| 557 |
+
"grad_norm": 1.5676358938217163,
|
| 558 |
+
"learning_rate": 0.0002573439463762044,
|
| 559 |
+
"loss": 0.5818,
|
| 560 |
+
"step": 3500
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 4.380475594493117,
|
| 564 |
+
"eval_loss": 0.5806618928909302,
|
| 565 |
+
"eval_runtime": 95.8037,
|
| 566 |
+
"eval_samples_per_second": 14.175,
|
| 567 |
+
"eval_steps_per_second": 1.774,
|
| 568 |
+
"eval_wer": 0.49643763819507003,
|
| 569 |
+
"step": 3500
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 4.505632040050062,
|
| 573 |
+
"grad_norm": 1.323522925376892,
|
| 574 |
+
"learning_rate": 0.0002560871386677838,
|
| 575 |
+
"loss": 0.6085,
|
| 576 |
+
"step": 3600
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 4.505632040050062,
|
| 580 |
+
"eval_loss": 0.5931637287139893,
|
| 581 |
+
"eval_runtime": 95.2763,
|
| 582 |
+
"eval_samples_per_second": 14.253,
|
| 583 |
+
"eval_steps_per_second": 1.784,
|
| 584 |
+
"eval_wer": 0.4915240357055114,
|
| 585 |
+
"step": 3600
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 4.630788485607009,
|
| 589 |
+
"grad_norm": 1.9220880270004272,
|
| 590 |
+
"learning_rate": 0.0002548303309593632,
|
| 591 |
+
"loss": 0.6648,
|
| 592 |
+
"step": 3700
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 4.630788485607009,
|
| 596 |
+
"eval_loss": 0.5579622983932495,
|
| 597 |
+
"eval_runtime": 96.2129,
|
| 598 |
+
"eval_samples_per_second": 14.115,
|
| 599 |
+
"eval_steps_per_second": 1.767,
|
| 600 |
+
"eval_wer": 0.486037179592171,
|
| 601 |
+
"step": 3700
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 4.755944931163955,
|
| 605 |
+
"grad_norm": 0.6678237915039062,
|
| 606 |
+
"learning_rate": 0.00025357352325094257,
|
| 607 |
+
"loss": 0.6359,
|
| 608 |
+
"step": 3800
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 4.755944931163955,
|
| 612 |
+
"eval_loss": 0.5565288662910461,
|
| 613 |
+
"eval_runtime": 95.7617,
|
| 614 |
+
"eval_samples_per_second": 14.181,
|
| 615 |
+
"eval_steps_per_second": 1.775,
|
| 616 |
+
"eval_wer": 0.4959462779461142,
|
| 617 |
+
"step": 3800
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 4.881101376720901,
|
| 621 |
+
"grad_norm": 0.8339139819145203,
|
| 622 |
+
"learning_rate": 0.000252316715542522,
|
| 623 |
+
"loss": 0.6139,
|
| 624 |
+
"step": 3900
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 4.881101376720901,
|
| 628 |
+
"eval_loss": 0.5604941248893738,
|
| 629 |
+
"eval_runtime": 95.8364,
|
| 630 |
+
"eval_samples_per_second": 14.17,
|
| 631 |
+
"eval_steps_per_second": 1.774,
|
| 632 |
+
"eval_wer": 0.4885758742117763,
|
| 633 |
+
"step": 3900
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 5.006257822277847,
|
| 637 |
+
"grad_norm": 0.6905695199966431,
|
| 638 |
+
"learning_rate": 0.00025105990783410133,
|
| 639 |
+
"loss": 0.5995,
|
| 640 |
+
"step": 4000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 5.006257822277847,
|
| 644 |
+
"eval_loss": 0.5719751715660095,
|
| 645 |
+
"eval_runtime": 95.3867,
|
| 646 |
+
"eval_samples_per_second": 14.237,
|
| 647 |
+
"eval_steps_per_second": 1.782,
|
| 648 |
+
"eval_wer": 0.48030464335435263,
|
| 649 |
+
"step": 4000
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 5.131414267834794,
|
| 653 |
+
"grad_norm": 1.2827470302581787,
|
| 654 |
+
"learning_rate": 0.00024980310012568074,
|
| 655 |
+
"loss": 0.6349,
|
| 656 |
+
"step": 4100
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 5.131414267834794,
|
| 660 |
+
"eval_loss": 0.5505784749984741,
|
| 661 |
+
"eval_runtime": 96.3073,
|
| 662 |
+
"eval_samples_per_second": 14.101,
|
| 663 |
+
"eval_steps_per_second": 1.765,
|
| 664 |
+
"eval_wer": 0.5011874539349767,
|
| 665 |
+
"step": 4100
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 5.256570713391739,
|
| 669 |
+
"grad_norm": 1.4041779041290283,
|
| 670 |
+
"learning_rate": 0.00024854629241726015,
|
| 671 |
+
"loss": 0.6134,
|
| 672 |
+
"step": 4200
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 5.256570713391739,
|
| 676 |
+
"eval_loss": 0.5602907538414001,
|
| 677 |
+
"eval_runtime": 96.7539,
|
| 678 |
+
"eval_samples_per_second": 14.036,
|
| 679 |
+
"eval_steps_per_second": 1.757,
|
| 680 |
+
"eval_wer": 0.4784210957333552,
|
| 681 |
+
"step": 4200
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 5.381727158948686,
|
| 685 |
+
"grad_norm": 16.846094131469727,
|
| 686 |
+
"learning_rate": 0.0002472894847088395,
|
| 687 |
+
"loss": 0.5989,
|
| 688 |
+
"step": 4300
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 5.381727158948686,
|
| 692 |
+
"eval_loss": 0.5714141726493835,
|
| 693 |
+
"eval_runtime": 96.2578,
|
| 694 |
+
"eval_samples_per_second": 14.108,
|
| 695 |
+
"eval_steps_per_second": 1.766,
|
| 696 |
+
"eval_wer": 0.48439931209565146,
|
| 697 |
+
"step": 4300
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 5.506883604505632,
|
| 701 |
+
"grad_norm": 1.1656825542449951,
|
| 702 |
+
"learning_rate": 0.0002460326770004189,
|
| 703 |
+
"loss": 0.6083,
|
| 704 |
+
"step": 4400
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 5.506883604505632,
|
| 708 |
+
"eval_loss": 0.5697967410087585,
|
| 709 |
+
"eval_runtime": 96.7222,
|
| 710 |
+
"eval_samples_per_second": 14.04,
|
| 711 |
+
"eval_steps_per_second": 1.758,
|
| 712 |
+
"eval_wer": 0.4758005077389239,
|
| 713 |
+
"step": 4400
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 5.506883604505632,
|
| 717 |
+
"step": 4400,
|
| 718 |
+
"total_flos": 2.278783878958534e+19,
|
| 719 |
+
"train_loss": 0.858475610559637,
|
| 720 |
+
"train_runtime": 9525.4326,
|
| 721 |
+
"train_samples_per_second": 20.116,
|
| 722 |
+
"train_steps_per_second": 2.516
|
| 723 |
+
}
|
| 724 |
+
],
|
| 725 |
+
"logging_steps": 100,
|
| 726 |
+
"max_steps": 23970,
|
| 727 |
+
"num_input_tokens_seen": 0,
|
| 728 |
+
"num_train_epochs": 30,
|
| 729 |
+
"save_steps": 200,
|
| 730 |
+
"stateful_callbacks": {
|
| 731 |
+
"EarlyStoppingCallback": {
|
| 732 |
+
"args": {
|
| 733 |
+
"early_stopping_patience": 3,
|
| 734 |
+
"early_stopping_threshold": 0.0
|
| 735 |
+
},
|
| 736 |
+
"attributes": {
|
| 737 |
+
"early_stopping_patience_counter": 3
|
| 738 |
+
}
|
| 739 |
+
},
|
| 740 |
+
"TrainerControl": {
|
| 741 |
+
"args": {
|
| 742 |
+
"should_epoch_stop": false,
|
| 743 |
+
"should_evaluate": false,
|
| 744 |
+
"should_log": false,
|
| 745 |
+
"should_save": true,
|
| 746 |
+
"should_training_stop": true
|
| 747 |
+
},
|
| 748 |
+
"attributes": {}
|
| 749 |
+
}
|
| 750 |
+
},
|
| 751 |
+
"total_flos": 2.278783878958534e+19,
|
| 752 |
+
"train_batch_size": 8,
|
| 753 |
+
"trial_name": null,
|
| 754 |
+
"trial_params": null
|
| 755 |
+
}
|