ymoslem commited on
Commit
a681414
1 Parent(s): a48ad5f

Model save

Browse files
README.md CHANGED
@@ -1,50 +1,19 @@
1
  ---
2
- language:
3
- - ga
4
- - en
5
  license: apache-2.0
6
  base_model: openai/whisper-small
7
  tags:
8
  - generated_from_trainer
9
- datasets:
10
- - ymoslem/IWSLT2023-GA-EN
11
- - ymoslem/FLEURS-GA-EN
12
- - ymoslem/BitesizeIrish-GA-EN
13
- - ymoslem/SpokenWords-GA-EN-MTed
14
- - ymoslem/Tatoeba-Speech-Irish
15
- - ymoslem/Wikimedia-Speech-Irish
16
- metrics:
17
- - bleu
18
- - wer
19
  model-index:
20
- - name: Whisper Small GA-EN Speech Translation
21
- results:
22
- - task:
23
- name: Automatic Speech Recognition
24
- type: automatic-speech-recognition
25
- dataset:
26
- name: IWSLT-2023, FLEURS, BiteSize, SpokenWords, Tatoeba, and Wikimedia
27
- type: ymoslem/IWSLT2023-GA-EN
28
- metrics:
29
- - name: Bleu
30
- type: bleu
31
- value: 23.1
32
- - name: Wer
33
- type: wer
34
- value: 82.89058982440342
35
  ---
36
 
37
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
38
  should probably proofread and complete it, then remove this comment. -->
39
 
40
- # Whisper Small GA-EN Speech Translation
41
 
42
- This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the IWSLT-2023, FLEURS, BiteSize, SpokenWords, Tatoeba, and Wikimedia dataset.
43
- It achieves the following results on the evaluation set:
44
- - Loss: 1.2172
45
- - Bleu: 23.1
46
- - Chrf: 42.54
47
- - Wer: 82.8906
48
 
49
  ## Model description
50
 
@@ -69,26 +38,10 @@ The following hyperparameters were used during training:
69
  - seed: 42
70
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
71
  - lr_scheduler_type: linear
72
- - lr_scheduler_warmup_steps: 0.03
73
- - training_steps: 1000
74
  - mixed_precision_training: Native AMP
75
 
76
- ### Training results
77
-
78
- | Training Loss | Epoch | Step | Validation Loss | Bleu | Chrf | Wer |
79
- |:-------------:|:-----:|:----:|:---------------:|:-----:|:-----:|:--------:|
80
- | 2.8459 | 0.07 | 100 | 2.0769 | 3.28 | 18.43 | 149.0770 |
81
- | 2.3328 | 0.13 | 200 | 1.8396 | 4.5 | 22.06 | 207.7443 |
82
- | 2.1669 | 0.2 | 300 | 1.6215 | 14.6 | 30.8 | 89.1941 |
83
- | 1.8606 | 0.26 | 400 | 1.5030 | 14.65 | 33.33 | 92.4358 |
84
- | 1.7255 | 0.33 | 500 | 1.4085 | 14.9 | 35.14 | 103.8271 |
85
- | 1.5855 | 0.39 | 600 | 1.3587 | 15.78 | 35.02 | 103.0617 |
86
- | 1.5875 | 0.46 | 700 | 1.2986 | 25.3 | 41.37 | 69.4732 |
87
- | 1.44 | 0.53 | 800 | 1.2575 | 25.78 | 42.23 | 70.0585 |
88
- | 1.3317 | 0.59 | 900 | 1.2338 | 23.24 | 41.64 | 79.1085 |
89
- | 1.3166 | 0.66 | 1000 | 1.2172 | 23.1 | 42.54 | 82.8906 |
90
-
91
-
92
  ### Framework versions
93
 
94
  - Transformers 4.39.3
 
1
  ---
 
 
 
2
  license: apache-2.0
3
  base_model: openai/whisper-small
4
  tags:
5
  - generated_from_trainer
 
 
 
 
 
 
 
 
 
 
6
  model-index:
7
+ - name: whisper-small-ga2en-v4
8
+ results: []
 
 
 
 
 
 
 
 
 
 
 
 
 
9
  ---
10
 
11
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
12
  should probably proofread and complete it, then remove this comment. -->
13
 
14
+ # whisper-small-ga2en-v4
15
 
16
+ This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on an unknown dataset.
 
 
 
 
 
17
 
18
  ## Model description
19
 
 
38
  - seed: 42
39
  - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
40
  - lr_scheduler_type: linear
41
+ - lr_scheduler_warmup_steps: 0.01
42
+ - training_steps: 2000
43
  - mixed_precision_training: Native AMP
44
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
45
  ### Framework versions
46
 
47
  - Transformers 4.39.3
runs/Apr18_11-10-25_3f8237e36747/events.out.tfevents.1713438696.3f8237e36747.395.1 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:676460e130e049b3e7e08b03486d13399f6c7d375b22dce9e18fefd7fc3ce167
3
- size 29348
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0f6c6cf6de7a7a5b4fcee8b296b40c7ba98966022d55a62da3364fa6ec2ec502
3
+ size 30960