oosawy commited on
Commit
9d2c8d7
·
verified ·
1 Parent(s): 04e2820

Model save

Browse files
Files changed (1) hide show
  1. README.md +106 -0
README.md ADDED
@@ -0,0 +1,106 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: apache-2.0
4
+ base_model: facebook/wav2vec2-base
5
+ tags:
6
+ - generated_from_trainer
7
+ datasets:
8
+ - timit_asr
9
+ metrics:
10
+ - wer
11
+ model-index:
12
+ - name: repo_name
13
+ results:
14
+ - task:
15
+ name: Automatic Speech Recognition
16
+ type: automatic-speech-recognition
17
+ dataset:
18
+ name: timit_asr
19
+ type: timit_asr
20
+ config: clean
21
+ split: None
22
+ args: clean
23
+ metrics:
24
+ - name: Wer
25
+ type: wer
26
+ value: 0.22107366825167116
27
+ ---
28
+
29
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
30
+ should probably proofread and complete it, then remove this comment. -->
31
+
32
+ # repo_name
33
+
34
+ This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the timit_asr dataset.
35
+ It achieves the following results on the evaluation set:
36
+ - Loss: 0.5351
37
+ - Wer: 0.2211
38
+
39
+ ## Model description
40
+
41
+ More information needed
42
+
43
+ ## Intended uses & limitations
44
+
45
+ More information needed
46
+
47
+ ## Training and evaluation data
48
+
49
+ More information needed
50
+
51
+ ## Training procedure
52
+
53
+ ### Training hyperparameters
54
+
55
+ The following hyperparameters were used during training:
56
+ - learning_rate: 0.0001
57
+ - train_batch_size: 8
58
+ - eval_batch_size: 8
59
+ - seed: 42
60
+ - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
61
+ - lr_scheduler_type: linear
62
+ - lr_scheduler_warmup_steps: 1000
63
+ - num_epochs: 30
64
+ - mixed_precision_training: Native AMP
65
+
66
+ ### Training results
67
+
68
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
69
+ |:-------------:|:-------:|:-----:|:---------------:|:------:|
70
+ | 3.5252 | 1.0040 | 500 | 1.6991 | 0.9701 |
71
+ | 0.854 | 2.0080 | 1000 | 0.5187 | 0.4025 |
72
+ | 0.4211 | 3.0120 | 1500 | 0.4289 | 0.3326 |
73
+ | 0.2871 | 4.0161 | 2000 | 0.3947 | 0.2896 |
74
+ | 0.2266 | 5.0201 | 2500 | 0.4034 | 0.2881 |
75
+ | 0.1789 | 6.0241 | 3000 | 0.4833 | 0.2926 |
76
+ | 0.1638 | 7.0281 | 3500 | 0.4342 | 0.2776 |
77
+ | 0.15 | 8.0321 | 4000 | 0.4643 | 0.2750 |
78
+ | 0.1251 | 9.0361 | 4500 | 0.4449 | 0.2642 |
79
+ | 0.1064 | 10.0402 | 5000 | 0.4785 | 0.2578 |
80
+ | 0.0986 | 11.0442 | 5500 | 0.4480 | 0.2627 |
81
+ | 0.0883 | 12.0482 | 6000 | 0.4876 | 0.2603 |
82
+ | 0.0784 | 13.0522 | 6500 | 0.5100 | 0.2519 |
83
+ | 0.0721 | 14.0562 | 7000 | 0.4795 | 0.2536 |
84
+ | 0.0696 | 15.0602 | 7500 | 0.4797 | 0.2456 |
85
+ | 0.0598 | 16.0643 | 8000 | 0.5064 | 0.2410 |
86
+ | 0.0575 | 17.0683 | 8500 | 0.5075 | 0.2362 |
87
+ | 0.0508 | 18.0723 | 9000 | 0.5062 | 0.2420 |
88
+ | 0.048 | 19.0763 | 9500 | 0.5078 | 0.2397 |
89
+ | 0.0402 | 20.0803 | 10000 | 0.5511 | 0.2341 |
90
+ | 0.0429 | 21.0843 | 10500 | 0.4835 | 0.2330 |
91
+ | 0.0362 | 22.0884 | 11000 | 0.5800 | 0.2308 |
92
+ | 0.0333 | 23.0924 | 11500 | 0.5250 | 0.2306 |
93
+ | 0.0285 | 24.0964 | 12000 | 0.5242 | 0.2288 |
94
+ | 0.0296 | 25.1004 | 12500 | 0.4995 | 0.2238 |
95
+ | 0.0264 | 26.1044 | 13000 | 0.5296 | 0.2236 |
96
+ | 0.0245 | 27.1084 | 13500 | 0.5530 | 0.2233 |
97
+ | 0.0214 | 28.1124 | 14000 | 0.5376 | 0.2209 |
98
+ | 0.0214 | 29.1165 | 14500 | 0.5351 | 0.2211 |
99
+
100
+
101
+ ### Framework versions
102
+
103
+ - Transformers 4.56.2
104
+ - Pytorch 2.8.0+cu126
105
+ - Datasets 2.21.0
106
+ - Tokenizers 0.22.1