Hugging Face's logo

ymoslem
/
whisper-medium-ga2en-v4

Automatic Speech Recognition
Transformers
TensorBoard
Safetensors
6 datasets
Irish
English
whisper
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Metrics Community
ymoslem commited on
Commit
8c87fbd
•
1 Parent(s): 1e730da

End of training

Browse files
Files changed (2) hide show
  1. README.md +72 -4
  2. model.safetensors +1 -1
README.md CHANGED
@@ -1,19 +1,51 @@
1
  ---
 
 
 
2
  license: apache-2.0
3
  base_model: openai/whisper-medium
4
  tags:
5
  - generated_from_trainer
 
 
 
 
 
 
 
 
 
 
6
  model-index:
7
- - name: whisper-medium-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-medium-ga2en-v4
15
 
16
- This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on an unknown dataset.
 
 
 
 
 
17
 
18
  ## Model description
19
 
@@ -44,6 +76,42 @@ The following hyperparameters were used during training:
44
  - training_steps: 3000
45
  - mixed_precision_training: Native AMP
46
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
47
  ### Framework versions
48
 
49
  - Transformers 4.40.0
 
1
  ---
2
+ language:
3
+ - ga
4
+ - en
5
  license: apache-2.0
6
  base_model: openai/whisper-medium
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 Medium 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, augmented
27
+ with noise
28
+ type: ymoslem/IWSLT2023-GA-EN
29
+ metrics:
30
+ - name: Bleu
31
+ type: bleu
32
+ value: 32.01
33
+ - name: Wer
34
+ type: wer
35
+ value: 62.76452048626745
36
  ---
37
 
38
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
39
  should probably proofread and complete it, then remove this comment. -->
40
 
41
+ # Whisper Medium GA-EN Speech Translation
42
 
43
+ This model is a fine-tuned version of [openai/whisper-medium](https://huggingface.co/openai/whisper-medium) on the IWSLT-2023, FLEURS, BiteSize, SpokenWords, Tatoeba, and Wikimedia, augmented with noise dataset.
44
+ It achieves the following results on the evaluation set:
45
+ - Loss: 1.1903
46
+ - Bleu: 32.01
47
+ - Chrf: 50.38
48
+ - Wer: 62.7645
49
 
50
  ## Model description
51
 
 
76
  - training_steps: 3000
77
  - mixed_precision_training: Native AMP
78
 
79
+ ### Training results
80
+
81
+ | Training Loss | Epoch | Step | Validation Loss | Bleu | Chrf | Wer |
82
+ |:-------------:|:------:|:----:|:---------------:|:-----:|:-----:|:--------:|
83
+ | 2.4487 | 0.0219 | 100 | 1.9518 | 8.34 | 24.49 | 117.2445 |
84
+ | 2.11 | 0.0438 | 200 | 1.6630 | 15.32 | 32.12 | 84.0612 |
85
+ | 1.9757 | 0.0657 | 300 | 1.5366 | 10.86 | 33.42 | 131.7875 |
86
+ | 1.7964 | 0.0876 | 400 | 1.4825 | 19.81 | 36.71 | 81.9451 |
87
+ | 1.6422 | 0.1095 | 500 | 1.4432 | 18.83 | 40.4 | 84.0162 |
88
+ | 1.3839 | 0.1314 | 600 | 1.4169 | 24.91 | 40.87 | 69.0230 |
89
+ | 1.352 | 0.1533 | 700 | 1.4340 | 25.01 | 41.57 | 71.5894 |
90
+ | 1.2434 | 0.1752 | 800 | 1.3813 | 24.05 | 41.29 | 73.7506 |
91
+ | 1.2223 | 0.1970 | 900 | 1.3578 | 25.89 | 41.61 | 70.5988 |
92
+ | 1.0414 | 0.2189 | 1000 | 1.3075 | 27.45 | 44.17 | 68.2575 |
93
+ | 0.9199 | 0.2408 | 1100 | 1.3022 | 23.14 | 44.3 | 84.1513 |
94
+ | 0.8648 | 0.2627 | 1200 | 1.3050 | 23.36 | 43.37 | 72.4448 |
95
+ | 0.8469 | 0.2846 | 1300 | 1.2853 | 28.37 | 45.97 | 67.1319 |
96
+ | 0.7649 | 0.3065 | 1400 | 1.2755 | 28.56 | 46.76 | 66.0964 |
97
+ | 0.7321 | 0.3284 | 1500 | 1.2750 | 27.23 | 46.1 | 69.3381 |
98
+ | 0.6541 | 0.3503 | 1600 | 1.2557 | 30.02 | 48.06 | 65.6011 |
99
+ | 0.6107 | 0.3722 | 1700 | 1.2520 | 30.41 | 49.23 | 64.2954 |
100
+ | 0.5738 | 0.3941 | 1800 | 1.2435 | 32.45 | 50.27 | 63.4399 |
101
+ | 0.4983 | 0.4160 | 1900 | 1.2007 | 31.17 | 48.58 | 64.0702 |
102
+ | 0.4439 | 0.4379 | 2000 | 1.2140 | 32.29 | 50.37 | 60.6033 |
103
+ | 0.367 | 0.4598 | 2100 | 1.2230 | 29.54 | 49.14 | 67.7172 |
104
+ | 0.2807 | 0.4817 | 2200 | 1.2277 | 33.1 | 51.21 | 62.9446 |
105
+ | 0.2621 | 0.5036 | 2300 | 1.2441 | 30.59 | 49.49 | 64.8807 |
106
+ | 0.2965 | 0.5255 | 2400 | 1.1969 | 31.82 | 49.67 | 63.5299 |
107
+ | 0.236 | 0.5473 | 2500 | 1.2275 | 31.17 | 50.29 | 65.1959 |
108
+ | 0.229 | 0.5692 | 2600 | 1.2008 | 30.02 | 50.27 | 70.6439 |
109
+ | 0.164 | 0.5911 | 2700 | 1.2192 | 31.37 | 50.57 | 63.6200 |
110
+ | 0.1786 | 0.6130 | 2800 | 1.1965 | 31.81 | 50.13 | 62.8546 |
111
+ | 0.1987 | 0.6349 | 2900 | 1.1883 | 32.88 | 51.52 | 62.0441 |
112
+ | 0.1633 | 0.6568 | 3000 | 1.1903 | 32.01 | 50.38 | 62.7645 |
113
+
114
+
115
  ### Framework versions
116
 
117
  - Transformers 4.40.0
model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:5a38309f9820caa2e1a2079fa55306df3535d0516c1640d79e2e5d6eb11713f7
3
  size 3055544304
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fefb5ac7957661d50786856c151b9bb8e2a3f6db59741661a07b5f184d9f1865
3
  size 3055544304