Hugging Face's logo

Marcusxx
/
gwanju_large_model

TensorBoard
Safetensors
Korean
whisper
hf-asr-leaderboard
Generated from Trainer
Model card Files Metrics Community
Marcusxx commited on
Commit
91782ae
·
verified ·
1 Parent(s): e2027b3

Model save

Browse files
Files changed (3) hide show
  1. README.md +10 -21
  2. model-00001-of-00002.safetensors +1 -1
  3. model-00002-of-00002.safetensors +1 -1
README.md CHANGED
@@ -1,13 +1,8 @@
1
  ---
2
- language:
3
- - ko
4
  license: apache-2.0
 
5
  tags:
6
- - hf-asr-leaderboard
7
  - generated_from_trainer
8
- base_model: openai/whisper-large
9
- datasets:
10
- - Marcusxx/gwanju
11
  model-index:
12
  - name: gwanju_large_model
13
  results: []
@@ -18,10 +13,10 @@ should probably proofread and complete it, then remove this comment. -->
18
 
19
  # gwanju_large_model
20
 
21
- This model is a fine-tuned version of [openai/whisper-large](https://huggingface.co/openai/whisper-large) on the Marcusxx/gwanju dataset.
22
  It achieves the following results on the evaluation set:
23
- - Loss: 0.3547
24
- - Cer: 395.4097
25
 
26
  ## Model description
27
 
@@ -46,24 +41,18 @@ The following hyperparameters were used during training:
46
  - seed: 42
47
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
48
  - lr_scheduler_type: linear
49
- - lr_scheduler_warmup_steps: 250
50
- - training_steps: 10000
51
  - mixed_precision_training: Native AMP
52
 
53
  ### Training results
54
 
55
  | Training Loss | Epoch | Step | Validation Loss | Cer |
56
  |:-------------:|:------:|:-----:|:---------------:|:--------:|
57
- | 0.4409 | 0.2964 | 1000 | 0.4335 | 106.5804 |
58
- | 0.3824 | 0.5928 | 2000 | 0.3930 | 442.5527 |
59
- | 0.3757 | 0.8892 | 3000 | 0.3727 | 455.4447 |
60
- | 0.2271 | 1.1855 | 4000 | 0.3712 | 463.1818 |
61
- | 0.2528 | 1.4819 | 5000 | 0.3600 | 468.4532 |
62
- | 0.2068 | 1.7783 | 6000 | 0.3523 | 468.7220 |
63
- | 0.1221 | 2.0747 | 7000 | 0.3592 | 480.3038 |
64
- | 0.1157 | 2.3711 | 8000 | 0.3614 | 377.9116 |
65
- | 0.121 | 2.6675 | 9000 | 0.3579 | 401.4220 |
66
- | 0.1046 | 2.9638 | 10000 | 0.3547 | 395.4097 |
67
 
68
 
69
  ### Framework versions
 
1
  ---
 
 
2
  license: apache-2.0
3
+ base_model: openai/whisper-large
4
  tags:
 
5
  - generated_from_trainer
 
 
 
6
  model-index:
7
  - name: gwanju_large_model
8
  results: []
 
13
 
14
  # gwanju_large_model
15
 
16
+ This model is a fine-tuned version of [openai/whisper-large](https://huggingface.co/openai/whisper-large) on an unknown dataset.
17
  It achieves the following results on the evaluation set:
18
+ - Loss: 0.5238
19
+ - Cer: 238.8739
20
 
21
  ## Model description
22
 
 
41
  - seed: 42
42
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
43
  - lr_scheduler_type: linear
44
+ - lr_scheduler_warmup_steps: 500
45
+ - training_steps: 20000
46
  - mixed_precision_training: Native AMP
47
 
48
  ### Training results
49
 
50
  | Training Loss | Epoch | Step | Validation Loss | Cer |
51
  |:-------------:|:------:|:-----:|:---------------:|:--------:|
52
+ | 0.2792 | 1.4819 | 5000 | 0.3803 | 453.3834 |
53
+ | 0.1361 | 2.9638 | 10000 | 0.3823 | 336.9865 |
54
+ | 0.0335 | 4.4458 | 15000 | 0.4644 | 344.2898 |
55
+ | 0.0097 | 5.9277 | 20000 | 0.5238 | 238.8739 |
 
 
 
 
 
 
56
 
57
 
58
  ### Framework versions
model-00001-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3a40ba7ebb8dd77e6a85c262dc3fe92781f2d2c4bcdee83cb810abaa32fe5349
3
  size 4992706480
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:394d667b061226183e33e560b29a0bed3c6875a74923036c2b03dd31c846a0ff
3
  size 4992706480
model-00002-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e1e7af97a740b1db8f6c5e4c402b23dcb57fc75dfcc6b0f2f21b20924b42c474
3
  size 1180663192
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:317e0a88ef950419b677f8088433f55d13df84bdb58371d6442d745c09b98c58
3
  size 1180663192