Commit
•
d2840ad
1
Parent(s):
c1bc781
update model card README.md
Browse files
README.md
CHANGED
|
@@ -15,8 +15,8 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 15 |
|
| 16 |
This model was trained from scratch on the librispeech_asr dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
-
- Loss:
|
| 19 |
-
- Wer:
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
@@ -39,8 +39,8 @@ The following hyperparameters were used during training:
|
|
| 39 |
- train_batch_size: 8
|
| 40 |
- eval_batch_size: 8
|
| 41 |
- seed: 42
|
| 42 |
-
- gradient_accumulation_steps:
|
| 43 |
-
- total_train_batch_size:
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: linear
|
| 46 |
- lr_scheduler_warmup_steps: 500
|
|
@@ -51,9 +51,7 @@ The following hyperparameters were used during training:
|
|
| 51 |
|
| 52 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 53 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 54 |
-
| 6.
|
| 55 |
-
| 0.0 | 0.56 | 1000 | nan | 2.0097 |
|
| 56 |
-
| 0.0 | 0.84 | 1500 | nan | 2.0097 |
|
| 57 |
|
| 58 |
|
| 59 |
### Framework versions
|
|
|
|
| 15 |
|
| 16 |
This model was trained from scratch on the librispeech_asr dataset.
|
| 17 |
It achieves the following results on the evaluation set:
|
| 18 |
+
- Loss: 6.8012
|
| 19 |
+
- Wer: 1.0130
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
|
|
| 39 |
- train_batch_size: 8
|
| 40 |
- eval_batch_size: 8
|
| 41 |
- seed: 42
|
| 42 |
+
- gradient_accumulation_steps: 4
|
| 43 |
+
- total_train_batch_size: 32
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: linear
|
| 46 |
- lr_scheduler_warmup_steps: 500
|
|
|
|
| 51 |
|
| 52 |
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 53 |
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 54 |
+
| 6.302 | 0.56 | 500 | 6.8012 | 1.0130 |
|
|
|
|
|
|
|
| 55 |
|
| 56 |
|
| 57 |
### Framework versions
|