update model card README.md
Browse files
README.md
CHANGED
@@ -1,10 +1,6 @@
|
|
1 |
---
|
2 |
-
language:
|
3 |
-
- tr
|
4 |
license: apache-2.0
|
5 |
tags:
|
6 |
-
- automatic-speech-recognition
|
7 |
-
- common_voice
|
8 |
- generated_from_trainer
|
9 |
datasets:
|
10 |
- common_voice
|
@@ -18,11 +14,11 @@ should probably proofread and complete it, then remove this comment. -->
|
|
18 |
|
19 |
#
|
20 |
|
21 |
-
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the
|
22 |
It achieves the following results on the evaluation set:
|
23 |
-
- Loss:
|
24 |
-
- Wer: 0.
|
25 |
-
- Cer: 0.
|
26 |
|
27 |
## Model description
|
28 |
|
@@ -42,17 +38,30 @@ More information needed
|
|
42 |
|
43 |
The following hyperparameters were used during training:
|
44 |
- learning_rate: 0.0005
|
45 |
-
- train_batch_size:
|
46 |
- eval_batch_size: 8
|
47 |
- seed: 42
|
48 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
49 |
- lr_scheduler_type: linear
|
50 |
- lr_scheduler_warmup_steps: 500
|
51 |
-
- num_epochs:
|
52 |
- mixed_precision_training: Native AMP
|
53 |
|
54 |
### Training results
|
55 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
56 |
|
57 |
|
58 |
### Framework versions
|
|
|
1 |
---
|
|
|
|
|
2 |
license: apache-2.0
|
3 |
tags:
|
|
|
|
|
4 |
- generated_from_trainer
|
5 |
datasets:
|
6 |
- common_voice
|
|
|
14 |
|
15 |
#
|
16 |
|
17 |
+
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the common_voice dataset.
|
18 |
It achieves the following results on the evaluation set:
|
19 |
+
- Loss: 0.4164
|
20 |
+
- Wer: 0.3098
|
21 |
+
- Cer: 0.0764
|
22 |
|
23 |
## Model description
|
24 |
|
|
|
38 |
|
39 |
The following hyperparameters were used during training:
|
40 |
- learning_rate: 0.0005
|
41 |
+
- train_batch_size: 64
|
42 |
- eval_batch_size: 8
|
43 |
- seed: 42
|
44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
45 |
- lr_scheduler_type: linear
|
46 |
- lr_scheduler_warmup_steps: 500
|
47 |
+
- num_epochs: 100.0
|
48 |
- mixed_precision_training: Native AMP
|
49 |
|
50 |
### Training results
|
51 |
|
52 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|
53 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|:------:|
|
54 |
+
| 0.6356 | 9.09 | 500 | 0.5055 | 0.5536 | 0.1381 |
|
55 |
+
| 0.3847 | 18.18 | 1000 | 0.4002 | 0.4247 | 0.1065 |
|
56 |
+
| 0.3377 | 27.27 | 1500 | 0.4193 | 0.4167 | 0.1078 |
|
57 |
+
| 0.2175 | 36.36 | 2000 | 0.4351 | 0.3861 | 0.0974 |
|
58 |
+
| 0.2074 | 45.45 | 2500 | 0.3962 | 0.3622 | 0.0916 |
|
59 |
+
| 0.159 | 54.55 | 3000 | 0.4062 | 0.3526 | 0.0888 |
|
60 |
+
| 0.1882 | 63.64 | 3500 | 0.3991 | 0.3445 | 0.0850 |
|
61 |
+
| 0.1766 | 72.73 | 4000 | 0.4214 | 0.3396 | 0.0847 |
|
62 |
+
| 0.116 | 81.82 | 4500 | 0.4182 | 0.3265 | 0.0812 |
|
63 |
+
| 0.0718 | 90.91 | 5000 | 0.4259 | 0.3191 | 0.0781 |
|
64 |
+
| 0.019 | 100.0 | 5500 | 0.4164 | 0.3098 | 0.0764 |
|
65 |
|
66 |
|
67 |
### Framework versions
|