Hugging Face's logo

Gemmar
/
LugandaASRwav20Vec1B

Automatic Speech Recognition
Transformers
PyTorch
wav2vec2
Generated from Trainer
Eval Results
Inference Endpoints
Model card Files Community
1
Gemmar commited on
Commit
0e8822d
•
1 Parent(s): 75f1fa3

End of training

Browse files
Files changed (2) hide show
  1. README.md +99 -0
  2. pytorch_model.bin +1 -1
README.md ADDED
@@ -0,0 +1,99 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ base_model: facebook/wav2vec2-xls-r-1b
4
+ tags:
5
+ - generated_from_trainer
6
+ datasets:
7
+ - common_voice_13_0
8
+ metrics:
9
+ - wer
10
+ model-index:
11
+ - name: LugandaASRwav20Vec1B
12
+ results:
13
+ - task:
14
+ name: Automatic Speech Recognition
15
+ type: automatic-speech-recognition
16
+ dataset:
17
+ name: common_voice_13_0
18
+ type: common_voice_13_0
19
+ config: lg
20
+ split: validation
21
+ args: lg
22
+ metrics:
23
+ - name: Wer
24
+ type: wer
25
+ value: 0.23043478260869565
26
+ ---
27
+
28
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
29
+ should probably proofread and complete it, then remove this comment. -->
30
+
31
+ # LugandaASRwav20Vec1B
32
+
33
+ This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the common_voice_13_0 dataset.
34
+ It achieves the following results on the evaluation set:
35
+ - Loss: 0.1854
36
+ - Wer: 0.2304
37
+
38
+ ## Model description
39
+
40
+ More information needed
41
+
42
+ ## Intended uses & limitations
43
+
44
+ More information needed
45
+
46
+ ## Training and evaluation data
47
+
48
+ More information needed
49
+
50
+ ## Training procedure
51
+
52
+ ### Training hyperparameters
53
+
54
+ The following hyperparameters were used during training:
55
+ - learning_rate: 0.0003
56
+ - train_batch_size: 4
57
+ - eval_batch_size: 8
58
+ - seed: 42
59
+ - gradient_accumulation_steps: 24
60
+ - total_train_batch_size: 96
61
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
62
+ - lr_scheduler_type: linear
63
+ - lr_scheduler_warmup_steps: 200
64
+ - num_epochs: 3
65
+
66
+ ### Training results
67
+
68
+ | Training Loss | Epoch | Step | Validation Loss | Wer |
69
+ |:-------------:|:-----:|:----:|:---------------:|:------:|
70
+ | 4.303 | 0.14 | 100 | 2.1141 | 1.0 |
71
+ | 0.7155 | 0.27 | 200 | 0.5656 | 0.6752 |
72
+ | 0.4493 | 0.41 | 300 | 0.4402 | 0.5607 |
73
+ | 0.3964 | 0.54 | 400 | 0.3918 | 0.5114 |
74
+ | 0.3646 | 0.68 | 500 | 0.3601 | 0.4592 |
75
+ | 0.3294 | 0.81 | 600 | 0.3381 | 0.4467 |
76
+ | 0.3339 | 0.95 | 700 | 0.3340 | 0.4266 |
77
+ | 0.2893 | 1.08 | 800 | 0.2913 | 0.3670 |
78
+ | 0.2743 | 1.22 | 900 | 0.2854 | 0.3600 |
79
+ | 0.262 | 1.36 | 1000 | 0.2666 | 0.3318 |
80
+ | 0.2545 | 1.49 | 1100 | 0.2601 | 0.3341 |
81
+ | 0.2437 | 1.63 | 1200 | 0.2488 | 0.3152 |
82
+ | 0.2235 | 1.76 | 1300 | 0.2416 | 0.3015 |
83
+ | 0.2188 | 1.9 | 1400 | 0.2330 | 0.2902 |
84
+ | 0.2054 | 2.03 | 1500 | 0.2218 | 0.2750 |
85
+ | 0.1743 | 2.17 | 1600 | 0.2153 | 0.2672 |
86
+ | 0.1722 | 2.3 | 1700 | 0.2098 | 0.2575 |
87
+ | 0.1656 | 2.44 | 1800 | 0.2011 | 0.2538 |
88
+ | 0.1608 | 2.58 | 1900 | 0.2000 | 0.2475 |
89
+ | 0.1574 | 2.71 | 2000 | 0.1937 | 0.2428 |
90
+ | 0.1531 | 2.85 | 2100 | 0.1882 | 0.2347 |
91
+ | 0.1451 | 2.98 | 2200 | 0.1854 | 0.2304 |
92
+
93
+
94
+ ### Framework versions
95
+
96
+ - Transformers 4.32.0
97
+ - Pytorch 2.0.1+cu118
98
+ - Datasets 2.13.0
99
+ - Tokenizers 0.13.3
pytorch_model.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:3239146bd2fb211efcb1a74ab516b4b8f0eb508a3974198a479e35d28b663baa
3
  size 3850429293
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ab4a62e54e8ba322ec27281bb0ebab166f5cf9939b6ec04c2a065e2159be7dfc
3
  size 3850429293