Hugging Face's logo

dongim04
/
whisper-base-zh

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
whisper
Generated from Trainer
Inference Endpoints
Model card Files Metrics Community
dongim04 commited on
Commit
0e6dbee
•
1 Parent(s): 1f8aa16

End of training

Browse files
Files changed (2) hide show
  1. README.md +46 -28
  2. model.safetensors +1 -1
README.md CHANGED
@@ -1,42 +1,25 @@
1
  ---
2
  library_name: transformers
3
- language:
4
- - zh
5
  license: apache-2.0
6
  base_model: openai/whisper-base
7
  tags:
8
- - hf-asr-leaderboard
9
  - generated_from_trainer
10
- datasets:
11
- - aimpower/mandarin_stutter_speech
12
  metrics:
13
  - wer
14
  model-index:
15
- - name: Whisper Base ZH - Dongim Lee
16
- results:
17
- - task:
18
- name: Automatic Speech Recognition
19
- type: automatic-speech-recognition
20
- dataset:
21
- name: AImpower Mandarin Stutter Speech
22
- type: aimpower/mandarin_stutter_speech
23
- config: zh
24
- split: test
25
- metrics:
26
- - name: Wer
27
- type: wer
28
- value: 87.24489795918367
29
  ---
30
 
31
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
32
  should probably proofread and complete it, then remove this comment. -->
33
 
34
- # Whisper Base ZH - Dongim Lee
35
 
36
- This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the AImpower Mandarin Stutter Speech dataset.
37
  It achieves the following results on the evaluation set:
38
- - Loss: 0.4500
39
- - Wer: 87.2449
40
 
41
  ## Model description
42
 
@@ -61,7 +44,6 @@ The following hyperparameters were used during training:
61
  - seed: 42
62
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
63
  - lr_scheduler_type: linear
64
- - lr_scheduler_warmup_steps: 500
65
  - training_steps: 4000
66
  - mixed_precision_training: Native AMP
67
 
@@ -69,10 +51,46 @@ The following hyperparameters were used during training:
69
 
70
  | Training Loss | Epoch | Step | Validation Loss | Wer |
71
  |:-------------:|:------:|:----:|:---------------:|:-------:|
72
- | 0.2036 | 2.0704 | 1000 | 0.3756 | 86.5816 |
73
- | 0.0885 | 4.1408 | 2000 | 0.3903 | 86.2245 |
74
- | 0.0367 | 6.2112 | 3000 | 0.4295 | 86.8878 |
75
- | 0.0242 | 8.2816 | 4000 | 0.4500 | 87.2449 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
76
 
77
 
78
  ### Framework versions
 
1
  ---
2
  library_name: transformers
 
 
3
  license: apache-2.0
4
  base_model: openai/whisper-base
5
  tags:
 
6
  - generated_from_trainer
 
 
7
  metrics:
8
  - wer
9
  model-index:
10
+ - name: whisper-base-zh
11
+ results: []
 
 
 
 
 
 
 
 
 
 
 
 
12
  ---
13
 
14
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
  should probably proofread and complete it, then remove this comment. -->
16
 
17
+ # whisper-base-zh
18
 
19
+ This model is a fine-tuned version of [openai/whisper-base](https://huggingface.co/openai/whisper-base) on the None dataset.
20
  It achieves the following results on the evaluation set:
21
+ - Loss: 0.4316
22
+ - Wer: 85.1032
23
 
24
  ## Model description
25
 
 
44
  - seed: 42
45
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
  - lr_scheduler_type: linear
 
47
  - training_steps: 4000
48
  - mixed_precision_training: Native AMP
49
 
 
51
 
52
  | Training Loss | Epoch | Step | Validation Loss | Wer |
53
  |:-------------:|:------:|:----:|:---------------:|:-------:|
54
+ | 0.5085 | 0.2070 | 100 | 0.4641 | 89.6326 |
55
+ | 0.4047 | 0.4141 | 200 | 0.4315 | 88.2235 |
56
+ | 0.4156 | 0.6211 | 300 | 0.4055 | 86.8646 |
57
+ | 0.3826 | 0.8282 | 400 | 0.3907 | 87.2169 |
58
+ | 0.3087 | 1.0352 | 500 | 0.3806 | 86.9149 |
59
+ | 0.2709 | 1.2422 | 600 | 0.3775 | 87.0659 |
60
+ | 0.2707 | 1.4493 | 700 | 0.3681 | 85.7071 |
61
+ | 0.273 | 1.6563 | 800 | 0.3633 | 85.5561 |
62
+ | 0.2638 | 1.8634 | 900 | 0.3592 | 85.4051 |
63
+ | 0.1698 | 2.0704 | 1000 | 0.3613 | 85.6568 |
64
+ | 0.1849 | 2.2774 | 1100 | 0.3628 | 85.7574 |
65
+ | 0.1819 | 2.4845 | 1200 | 0.3645 | 85.4555 |
66
+ | 0.1712 | 2.6915 | 1300 | 0.3611 | 84.8515 |
67
+ | 0.1768 | 2.8986 | 1400 | 0.3573 | 84.3986 |
68
+ | 0.1265 | 3.1056 | 1500 | 0.3640 | 85.2038 |
69
+ | 0.1275 | 3.3126 | 1600 | 0.3661 | 85.4051 |
70
+ | 0.1277 | 3.5197 | 1700 | 0.3698 | 84.3986 |
71
+ | 0.1195 | 3.7267 | 1800 | 0.3684 | 85.3045 |
72
+ | 0.1176 | 3.9337 | 1900 | 0.3668 | 84.9522 |
73
+ | 0.0843 | 4.1408 | 2000 | 0.3764 | 84.9522 |
74
+ | 0.077 | 4.3478 | 2100 | 0.3801 | 84.8012 |
75
+ | 0.0853 | 4.5549 | 2200 | 0.3804 | 85.4051 |
76
+ | 0.0826 | 4.7619 | 2300 | 0.3827 | 84.8515 |
77
+ | 0.0759 | 4.9689 | 2400 | 0.3803 | 84.9522 |
78
+ | 0.0519 | 5.1760 | 2500 | 0.3959 | 84.9522 |
79
+ | 0.0578 | 5.3830 | 2600 | 0.3946 | 85.0528 |
80
+ | 0.0594 | 5.5901 | 2700 | 0.3956 | 85.0528 |
81
+ | 0.0598 | 5.7971 | 2800 | 0.3971 | 85.1535 |
82
+ | 0.0519 | 6.0041 | 2900 | 0.3980 | 84.7509 |
83
+ | 0.0421 | 6.2112 | 3000 | 0.4103 | 85.5058 |
84
+ | 0.0379 | 6.4182 | 3100 | 0.4124 | 84.9019 |
85
+ | 0.0363 | 6.6253 | 3200 | 0.4131 | 84.9019 |
86
+ | 0.0377 | 6.8323 | 3300 | 0.4149 | 85.2038 |
87
+ | 0.0286 | 7.0393 | 3400 | 0.4205 | 84.8515 |
88
+ | 0.0263 | 7.2464 | 3500 | 0.4257 | 84.6502 |
89
+ | 0.0309 | 7.4534 | 3600 | 0.4276 | 85.3548 |
90
+ | 0.0295 | 7.6605 | 3700 | 0.4282 | 85.3045 |
91
+ | 0.0266 | 7.8675 | 3800 | 0.4291 | 85.2038 |
92
+ | 0.0241 | 8.0745 | 3900 | 0.4308 | 85.1535 |
93
+ | 0.025 | 8.2816 | 4000 | 0.4316 | 85.1032 |
94
 
95
 
96
  ### Framework versions
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3160e0573d98a5a242e955f214d5dc728131965464fc98944ab78d177ffb585a
3
  size 290403936
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b6063f84cecd0840d9b1f3fe4e62f10fe3b753f828df4c5d7e209a3136563197
3
  size 290403936