Hugging Face's logo

syp1229
/
whisper-small-Youngfreq_E

Automatic Speech Recognition
Transformers
PyTorch
Hindi
whisper
hf-asr-leaderboard
Generated from Trainer
Inference Endpoints
Model card Files Community
1
syp1229 commited on
Commit
d137748
·
1 Parent(s): d133356

End of training

Browse files
Files changed (2) hide show
  1. README.md +21 -36
  2. pytorch_model.bin +1 -1
README.md CHANGED
@@ -18,8 +18,8 @@ should probably proofread and complete it, then remove this comment. -->
18
 
19
  This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the aihub Y E dialogue dataset.
20
  It achieves the following results on the evaluation set:
21
- - Loss: 0.2258
22
- - Cer: 0.3581
23
 
24
  ## Model description
25
 
@@ -42,48 +42,33 @@ The following hyperparameters were used during training:
42
  - train_batch_size: 8
43
  - eval_batch_size: 8
44
  - seed: 42
 
 
45
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
46
  - lr_scheduler_type: linear
47
- - lr_scheduler_warmup_steps: 100
48
  - num_epochs: 5
49
 
50
  ### Training results
51
 
52
  | Training Loss | Epoch | Step | Validation Loss | Cer |
53
  |:-------------:|:-----:|:----:|:---------------:|:------:|
54
- | 0.2621 | 0.15 | 100 | 0.2922 | 0.0749 |
55
- | 0.2142 | 0.3 | 200 | 0.2691 | 0.0689 |
56
- | 0.2793 | 0.44 | 300 | 0.2678 | 0.0703 |
57
- | 0.2804 | 0.59 | 400 | 0.2688 | 0.0884 |
58
- | 0.2686 | 0.74 | 500 | 0.2498 | 0.0910 |
59
- | 0.2047 | 0.89 | 600 | 0.2415 | 0.2238 |
60
- | 0.0935 | 1.04 | 700 | 0.2438 | 0.0816 |
61
- | 0.0714 | 1.19 | 800 | 0.2456 | 0.1135 |
62
- | 0.0692 | 1.33 | 900 | 0.2473 | 0.0906 |
63
- | 0.0851 | 1.48 | 1000 | 0.2421 | 0.3472 |
64
- | 0.075 | 1.63 | 1100 | 0.2362 | 0.1613 |
65
- | 0.1515 | 1.78 | 1200 | 0.2349 | 0.0970 |
66
- | 0.0651 | 1.93 | 1300 | 0.2380 | 0.2633 |
67
- | 0.027 | 2.07 | 1400 | 0.2350 | 0.1955 |
68
- | 0.026 | 2.22 | 1500 | 0.2391 | 0.1991 |
69
- | 0.0341 | 2.37 | 1600 | 0.2375 | 0.1891 |
70
- | 0.0355 | 2.52 | 1700 | 0.2345 | 0.2059 |
71
- | 0.0413 | 2.67 | 1800 | 0.2239 | 0.3248 |
72
- | 0.0282 | 2.81 | 1900 | 0.2334 | 0.6045 |
73
- | 0.0228 | 2.96 | 2000 | 0.2272 | 0.3440 |
74
- | 0.006 | 3.11 | 2100 | 0.2239 | 0.3789 |
75
- | 0.009 | 3.26 | 2200 | 0.2283 | 0.4540 |
76
- | 0.0082 | 3.41 | 2300 | 0.2271 | 0.3503 |
77
- | 0.0222 | 3.56 | 2400 | 0.2276 | 0.4661 |
78
- | 0.009 | 3.7 | 2500 | 0.2281 | 0.4465 |
79
- | 0.0034 | 3.85 | 2600 | 0.2247 | 0.3773 |
80
- | 0.0099 | 4.0 | 2700 | 0.2191 | 0.4596 |
81
- | 0.0051 | 4.15 | 2800 | 0.2220 | 0.3566 |
82
- | 0.0032 | 4.3 | 2900 | 0.2258 | 0.3479 |
83
- | 0.0019 | 4.44 | 3000 | 0.2264 | 0.4057 |
84
- | 0.0024 | 4.59 | 3100 | 0.2262 | 0.4251 |
85
- | 0.0021 | 4.74 | 3200 | 0.2268 | 0.3512 |
86
- | 0.0019 | 4.89 | 3300 | 0.2258 | 0.3581 |
87
 
88
 
89
  ### Framework versions
 
18
 
19
  This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the aihub Y E dialogue dataset.
20
  It achieves the following results on the evaluation set:
21
+ - Loss: 0.2157
22
+ - Cer: 0.0491
23
 
24
  ## Model description
25
 
 
42
  - train_batch_size: 8
43
  - eval_batch_size: 8
44
  - seed: 42
45
+ - gradient_accumulation_steps: 2
46
+ - total_train_batch_size: 16
47
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
  - lr_scheduler_type: linear
49
+ - lr_scheduler_warmup_steps: 50
50
  - num_epochs: 5
51
 
52
  ### Training results
53
 
54
  | Training Loss | Epoch | Step | Validation Loss | Cer |
55
  |:-------------:|:-----:|:----:|:---------------:|:------:|
56
+ | 0.2455 | 0.3 | 100 | 0.2528 | 0.0663 |
57
+ | 0.2591 | 0.59 | 200 | 0.2452 | 0.0646 |
58
+ | 0.1702 | 0.89 | 300 | 0.2298 | 0.0628 |
59
+ | 0.0738 | 1.19 | 400 | 0.2136 | 0.0923 |
60
+ | 0.0957 | 1.48 | 500 | 0.2263 | 0.0618 |
61
+ | 0.0729 | 1.78 | 600 | 0.2139 | 0.0565 |
62
+ | 0.0242 | 2.07 | 700 | 0.2073 | 0.0520 |
63
+ | 0.028 | 2.37 | 800 | 0.2063 | 0.0482 |
64
+ | 0.0351 | 2.67 | 900 | 0.2162 | 0.0506 |
65
+ | 0.0239 | 2.96 | 1000 | 0.2075 | 0.0513 |
66
+ | 0.0088 | 3.26 | 1100 | 0.2194 | 0.0495 |
67
+ | 0.0079 | 3.56 | 1200 | 0.2187 | 0.0508 |
68
+ | 0.0072 | 3.85 | 1300 | 0.2217 | 0.0510 |
69
+ | 0.0046 | 4.15 | 1400 | 0.2164 | 0.0488 |
70
+ | 0.0038 | 4.44 | 1500 | 0.2149 | 0.0490 |
71
+ | 0.003 | 4.74 | 1600 | 0.2157 | 0.0491 |
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
72
 
73
 
74
  ### Framework versions
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b742e46e0505837d1f7988c1dd65bc1df9b1becd3cf5dd12f32baf6ca7b6d0b9
3
  size 967102729
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:97e007a9dd87b8450e9585181af792b70e1b5d08d7bed25b8b00a9017ea39d96
3
  size 967102729