update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,106 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
license: apache-2.0
|
3 |
+
tags:
|
4 |
+
- generated_from_trainer
|
5 |
+
datasets:
|
6 |
+
- librispeech_asr
|
7 |
+
model-index:
|
8 |
+
- name: wav2vec2-large-xls-r-300m-en-colab
|
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 |
+
# wav2vec2-large-xls-r-300m-en-colab
|
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 librispeech_asr dataset.
|
18 |
+
It achieves the following results on the evaluation set:
|
19 |
+
- Loss: 0.1169
|
20 |
+
- Wer: 0.0597
|
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: 0.0003
|
40 |
+
- train_batch_size: 32
|
41 |
+
- eval_batch_size: 32
|
42 |
+
- seed: 42
|
43 |
+
- gradient_accumulation_steps: 2
|
44 |
+
- total_train_batch_size: 64
|
45 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
46 |
+
- lr_scheduler_type: linear
|
47 |
+
- lr_scheduler_warmup_steps: 500
|
48 |
+
- num_epochs: 10
|
49 |
+
- mixed_precision_training: Native AMP
|
50 |
+
|
51 |
+
### Training results
|
52 |
+
|
53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
54 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
55 |
+
| 5.6951 | 0.22 | 100 | 3.1606 | 1.0 |
|
56 |
+
| 2.924 | 0.45 | 200 | 2.9297 | 1.0 |
|
57 |
+
| 2.5328 | 0.67 | 300 | 1.4339 | 0.8953 |
|
58 |
+
| 0.8611 | 0.9 | 400 | 0.6104 | 0.5306 |
|
59 |
+
| 0.3714 | 1.12 | 500 | 0.2497 | 0.2150 |
|
60 |
+
| 0.2015 | 1.35 | 600 | 0.1853 | 0.1615 |
|
61 |
+
| 0.1593 | 1.57 | 700 | 0.1613 | 0.1366 |
|
62 |
+
| 0.1436 | 1.79 | 800 | 0.1503 | 0.1311 |
|
63 |
+
| 0.1249 | 2.02 | 900 | 0.1374 | 0.1038 |
|
64 |
+
| 0.0936 | 2.24 | 1000 | 0.1328 | 0.1016 |
|
65 |
+
| 0.0896 | 2.47 | 1100 | 0.1234 | 0.0942 |
|
66 |
+
| 0.0872 | 2.69 | 1200 | 0.1148 | 0.0922 |
|
67 |
+
| 0.0859 | 2.91 | 1300 | 0.1140 | 0.0892 |
|
68 |
+
| 0.0733 | 3.14 | 1400 | 0.1134 | 0.0839 |
|
69 |
+
| 0.0633 | 3.36 | 1500 | 0.1085 | 0.0802 |
|
70 |
+
| 0.0567 | 3.59 | 1600 | 0.1103 | 0.0807 |
|
71 |
+
| 0.0604 | 3.81 | 1700 | 0.1088 | 0.0809 |
|
72 |
+
| 0.0586 | 4.04 | 1800 | 0.1113 | 0.0804 |
|
73 |
+
| 0.0516 | 4.26 | 1900 | 0.1123 | 0.0808 |
|
74 |
+
| 0.055 | 4.48 | 2000 | 0.1130 | 0.0764 |
|
75 |
+
| 0.0568 | 4.71 | 2100 | 0.1128 | 0.0807 |
|
76 |
+
| 0.0529 | 4.93 | 2200 | 0.1009 | 0.0727 |
|
77 |
+
| 0.0455 | 5.16 | 2300 | 0.1050 | 0.0726 |
|
78 |
+
| 0.0443 | 5.38 | 2400 | 0.1078 | 0.0720 |
|
79 |
+
| 0.0434 | 5.61 | 2500 | 0.1027 | 0.0702 |
|
80 |
+
| 0.0418 | 5.83 | 2600 | 0.1009 | 0.0693 |
|
81 |
+
| 0.0381 | 6.05 | 2700 | 0.1079 | 0.0689 |
|
82 |
+
| 0.0344 | 6.28 | 2800 | 0.1062 | 0.0678 |
|
83 |
+
| 0.0353 | 6.5 | 2900 | 0.1054 | 0.0682 |
|
84 |
+
| 0.0342 | 6.73 | 3000 | 0.1030 | 0.0661 |
|
85 |
+
| 0.0329 | 6.95 | 3100 | 0.1021 | 0.0659 |
|
86 |
+
| 0.0316 | 7.17 | 3200 | 0.1085 | 0.0667 |
|
87 |
+
| 0.0275 | 7.4 | 3300 | 0.1089 | 0.0645 |
|
88 |
+
| 0.0275 | 7.62 | 3400 | 0.1064 | 0.0645 |
|
89 |
+
| 0.0268 | 7.85 | 3500 | 0.1109 | 0.0639 |
|
90 |
+
| 0.0259 | 8.07 | 3600 | 0.1123 | 0.0636 |
|
91 |
+
| 0.024 | 8.3 | 3700 | 0.1169 | 0.0631 |
|
92 |
+
| 0.0225 | 8.52 | 3800 | 0.1170 | 0.0617 |
|
93 |
+
| 0.0229 | 8.74 | 3900 | 0.1153 | 0.0614 |
|
94 |
+
| 0.0214 | 8.97 | 4000 | 0.1143 | 0.0610 |
|
95 |
+
| 0.02 | 9.19 | 4100 | 0.1162 | 0.0606 |
|
96 |
+
| 0.0194 | 9.42 | 4200 | 0.1173 | 0.0603 |
|
97 |
+
| 0.0193 | 9.64 | 4300 | 0.1184 | 0.0601 |
|
98 |
+
| 0.0177 | 9.87 | 4400 | 0.1169 | 0.0597 |
|
99 |
+
|
100 |
+
|
101 |
+
### Framework versions
|
102 |
+
|
103 |
+
- Transformers 4.11.3
|
104 |
+
- Pytorch 1.10.0+cu113
|
105 |
+
- Datasets 1.18.3
|
106 |
+
- Tokenizers 0.10.3
|