meiiny00 commited on
Commit
0b988a4
1 Parent(s): 5a83ac3

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +83 -0
README.md ADDED
@@ -0,0 +1,83 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: mit
4
+ base_model: openai/whisper-large-v3-turbo
5
+ tags:
6
+ - generated_from_trainer
7
+ datasets:
8
+ - fsicoli/common_voice_18_0
9
+ metrics:
10
+ - wer
11
+ model-index:
12
+ - name: Whisper Turbo Train
13
+ results:
14
+ - task:
15
+ name: Automatic Speech Recognition
16
+ type: automatic-speech-recognition
17
+ dataset:
18
+ name: Common Voice 18.0
19
+ type: fsicoli/common_voice_18_0
20
+ split: None
21
+ metrics:
22
+ - name: Wer
23
+ type: wer
24
+ value: 15.246076710047603
25
+ ---
26
+
27
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
28
+ should probably proofread and complete it, then remove this comment. -->
29
+
30
+ # Whisper Turbo Train
31
+
32
+ This model is a fine-tuned version of [openai/whisper-large-v3-turbo](https://huggingface.co/openai/whisper-large-v3-turbo) on the Common Voice 18.0 dataset.
33
+ It achieves the following results on the evaluation set:
34
+ - Loss: 0.1156
35
+ - Wer: 15.2461
36
+
37
+ ## Model description
38
+
39
+ More information needed
40
+
41
+ ## Intended uses & limitations
42
+
43
+ More information needed
44
+
45
+ ## Training and evaluation data
46
+
47
+ More information needed
48
+
49
+ ## Training procedure
50
+
51
+ ### Training hyperparameters
52
+
53
+ The following hyperparameters were used during training:
54
+ - learning_rate: 1e-05
55
+ - train_batch_size: 32
56
+ - eval_batch_size: 32
57
+ - seed: 42
58
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
59
+ - lr_scheduler_type: linear
60
+ - lr_scheduler_warmup_steps: 1000
61
+ - training_steps: 8000
62
+ - mixed_precision_training: Native AMP
63
+
64
+ ### Training results
65
+
66
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
67
+ |:-------------:|:------:|:----:|:---------------:|:-------:|
68
+ | 0.3715 | 0.4257 | 1000 | 0.3457 | 40.4692 |
69
+ | 0.251 | 0.8514 | 2000 | 0.2181 | 27.7065 |
70
+ | 0.1569 | 1.2771 | 3000 | 0.1814 | 24.1533 |
71
+ | 0.1436 | 1.7029 | 4000 | 0.1531 | 20.3812 |
72
+ | 0.0931 | 2.1286 | 5000 | 0.1374 | 18.4662 |
73
+ | 0.0891 | 2.5543 | 6000 | 0.1252 | 16.9349 |
74
+ | 0.0738 | 2.9800 | 7000 | 0.1199 | 15.5610 |
75
+ | 0.0544 | 3.4057 | 8000 | 0.1156 | 15.2461 |
76
+
77
+
78
+ ### Framework versions
79
+
80
+ - Transformers 4.45.1
81
+ - Pytorch 2.1.0
82
+ - Datasets 3.0.1
83
+ - Tokenizers 0.20.0