naleraphael
commited on
Commit
•
a95817a
1
Parent(s):
ad6efd3
update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,96 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
datasets:
|
6 |
+
- common_voice
|
7 |
+
model-index:
|
8 |
+
- name: rasr_sample
|
9 |
+
results: []
|
10 |
+
---
|
11 |
+
|
12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
13 |
+
should probably proofread and complete it, then remove this comment. -->
|
14 |
+
|
15 |
+
# rasr_sample
|
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: 0.3149
|
20 |
+
- Wer: 0.2679
|
21 |
+
|
22 |
+
## Model description
|
23 |
+
|
24 |
+
More information needed
|
25 |
+
|
26 |
+
## Intended uses & limitations
|
27 |
+
|
28 |
+
More information needed
|
29 |
+
|
30 |
+
## Training and evaluation data
|
31 |
+
|
32 |
+
More information needed
|
33 |
+
|
34 |
+
## Training procedure
|
35 |
+
|
36 |
+
### Training hyperparameters
|
37 |
+
|
38 |
+
The following hyperparameters were used during training:
|
39 |
+
- learning_rate: 7.5e-05
|
40 |
+
- train_batch_size: 8
|
41 |
+
- eval_batch_size: 8
|
42 |
+
- seed: 42
|
43 |
+
- gradient_accumulation_steps: 4
|
44 |
+
- total_train_batch_size: 32
|
45 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
46 |
+
- lr_scheduler_type: linear
|
47 |
+
- lr_scheduler_warmup_steps: 2000
|
48 |
+
- num_epochs: 50.0
|
49 |
+
- mixed_precision_training: Native AMP
|
50 |
+
|
51 |
+
### Training results
|
52 |
+
|
53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
54 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|
|
55 |
+
| 3.3332 | 1.45 | 500 | 3.3031 | 1.0 |
|
56 |
+
| 2.9272 | 2.91 | 1000 | 2.9353 | 0.9970 |
|
57 |
+
| 2.0736 | 4.36 | 1500 | 1.1565 | 0.8714 |
|
58 |
+
| 1.7339 | 5.81 | 2000 | 0.7156 | 0.6688 |
|
59 |
+
| 1.5989 | 7.27 | 2500 | 0.5791 | 0.5519 |
|
60 |
+
| 1.4916 | 8.72 | 3000 | 0.5038 | 0.5169 |
|
61 |
+
| 1.4562 | 10.17 | 3500 | 0.4861 | 0.4805 |
|
62 |
+
| 1.3893 | 11.63 | 4000 | 0.4584 | 0.4761 |
|
63 |
+
| 1.3797 | 13.08 | 4500 | 0.4298 | 0.4686 |
|
64 |
+
| 1.3508 | 14.53 | 5000 | 0.4138 | 0.3744 |
|
65 |
+
| 1.3165 | 15.99 | 5500 | 0.4015 | 0.3578 |
|
66 |
+
| 1.281 | 17.44 | 6000 | 0.3883 | 0.3472 |
|
67 |
+
| 1.2682 | 18.89 | 6500 | 0.3904 | 0.3434 |
|
68 |
+
| 1.2477 | 20.35 | 7000 | 0.3726 | 0.3321 |
|
69 |
+
| 1.2364 | 21.8 | 7500 | 0.3685 | 0.3281 |
|
70 |
+
| 1.2041 | 23.26 | 8000 | 0.3597 | 0.3194 |
|
71 |
+
| 1.1901 | 24.71 | 8500 | 0.3542 | 0.3203 |
|
72 |
+
| 1.1903 | 26.16 | 9000 | 0.3500 | 0.3138 |
|
73 |
+
| 1.1677 | 27.61 | 9500 | 0.3458 | 0.3067 |
|
74 |
+
| 1.1718 | 29.07 | 10000 | 0.3595 | 0.3112 |
|
75 |
+
| 1.1562 | 30.52 | 10500 | 0.3433 | 0.3022 |
|
76 |
+
| 1.1392 | 31.97 | 11000 | 0.3440 | 0.2936 |
|
77 |
+
| 1.1258 | 33.43 | 11500 | 0.3396 | 0.2950 |
|
78 |
+
| 1.1067 | 34.88 | 12000 | 0.3379 | 0.2939 |
|
79 |
+
| 1.0953 | 36.34 | 12500 | 0.3370 | 0.2868 |
|
80 |
+
| 1.0835 | 37.79 | 13000 | 0.3317 | 0.2860 |
|
81 |
+
| 1.0772 | 39.24 | 13500 | 0.3302 | 0.2854 |
|
82 |
+
| 1.0853 | 40.7 | 14000 | 0.3265 | 0.2783 |
|
83 |
+
| 1.0689 | 42.15 | 14500 | 0.3306 | 0.2770 |
|
84 |
+
| 1.0394 | 43.6 | 15000 | 0.3233 | 0.2757 |
|
85 |
+
| 1.0581 | 45.06 | 15500 | 0.3199 | 0.2713 |
|
86 |
+
| 1.0362 | 46.51 | 16000 | 0.3154 | 0.2683 |
|
87 |
+
| 1.0406 | 47.96 | 16500 | 0.3176 | 0.2688 |
|
88 |
+
| 1.0082 | 49.42 | 17000 | 0.3149 | 0.2679 |
|
89 |
+
|
90 |
+
|
91 |
+
### Framework versions
|
92 |
+
|
93 |
+
- Transformers 4.17.0.dev0
|
94 |
+
- Pytorch 1.10.2+cu102
|
95 |
+
- Datasets 1.18.2.dev0
|
96 |
+
- Tokenizers 0.11.0
|