lucio commited on
Commit
ab6470e
1 Parent(s): ea798b8

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +126 -0
README.md CHANGED
@@ -0,0 +1,126 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language: rw
3
+ datasets:
4
+ - common_voice
5
+ metrics:
6
+ - wer
7
+ tags:
8
+ - audio
9
+ - automatic-speech-recognition
10
+ - speech
11
+ - xlsr-fine-tuning-week
12
+ license: apache-2.0
13
+ model-index:
14
+ - name: XLSR Wav2Vec2 Large Kinyarwanda by Lucio
15
+ results:
16
+ - task:
17
+ name: Speech Recognition
18
+ type: automatic-speech-recognition
19
+ dataset:
20
+ name: Common Voice rw
21
+ type: common_voice
22
+ args: rw
23
+ metrics:
24
+ - name: Test WER
25
+ type: wer
26
+ value: ??
27
+ ---
28
+
29
+ # Wav2Vec2-Large-XLSR-53-rw
30
+
31
+ Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on Kinyarwanda using the [Common Voice](https://huggingface.co/datasets/common_voice) dataset, using the validation set for training, and taking 12% of the test data for validation.
32
+ When using this model, make sure that your speech input is sampled at 16kHz.
33
+
34
+ ## Usage
35
+
36
+ The model can be used directly (without a language model) as follows:
37
+
38
+ ```python
39
+ import torch
40
+ import torchaudio
41
+ from datasets import load_dataset
42
+ from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
43
+
44
+ test_dataset = load_dataset("common_voice", "rw", split="test[:2%]")
45
+
46
+ processor = Wav2Vec2Processor.from_pretrained("lucio/wav2vec2-large-xlsr-kinyarwanda")
47
+ model = Wav2Vec2ForCTC.from_pretrained("lucio/wav2vec2-large-xlsr-kinyarwanda")
48
+
49
+ resampler = torchaudio.transforms.Resample(48_000, 16_000)
50
+
51
+ # Preprocessing the datasets.
52
+ # We need to read the audio files as arrays
53
+ def speech_file_to_array_fn(batch):
54
+ speech_array, sampling_rate = torchaudio.load(batch["path"])
55
+ batch["speech"] = resampler(speech_array).squeeze().numpy()
56
+ return batch
57
+
58
+ test_dataset = test_dataset.map(speech_file_to_array_fn)
59
+ inputs = processor(test_dataset[:2]["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
60
+
61
+ with torch.no_grad():
62
+ logits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits
63
+
64
+ predicted_ids = torch.argmax(logits, dim=-1)
65
+
66
+ print("Prediction:", processor.batch_decode(predicted_ids))
67
+ print("Reference:", test_dataset["sentence"][:2])
68
+ ```
69
+
70
+
71
+ ## Evaluation
72
+
73
+ The model can be evaluated as follows on the Kinyarwanda test data of Common Voice.
74
+
75
+
76
+ ```python
77
+ import torch
78
+ import torchaudio
79
+ from datasets import load_dataset, load_metric
80
+ from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
81
+ import re
82
+
83
+ test_dataset = load_dataset("common_voice", "rw", split="test")
84
+ wer = load_metric("wer")
85
+
86
+ processor = Wav2Vec2Processor.from_pretrained("lucio/wav2vec2-large-xlsr-kinyarwanda")
87
+ model = Wav2Vec2ForCTC.from_pretrained("lucio/wav2vec2-large-xlsr-kinyarwanda")
88
+ model.to("cuda")
89
+
90
+ chars_to_ignore_regex = '[\[\],?.!;:%\'"‘’“”(){}‟ˮ´ʺ″«»/…‽�–-]'
91
+ resampler = torchaudio.transforms.Resample(48_000, 16_000)
92
+
93
+ # Preprocessing the datasets.
94
+ # We need to read the audio files as arrays
95
+ def speech_file_to_array_fn(batch):
96
+ batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
97
+ speech_array, sampling_rate = torchaudio.load(batch["path"])
98
+ batch["speech"] = resampler(speech_array).squeeze().numpy()
99
+ return batch
100
+
101
+ test_dataset = test_dataset.map(speech_file_to_array_fn)
102
+
103
+ # Preprocessing the datasets.
104
+ # We need to read the audio files as arrays
105
+ def evaluate(batch):
106
+ inputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
107
+
108
+ with torch.no_grad():
109
+ logits = model(inputs.input_values.to("cuda"), attention_mask=inputs.attention_mask.to("cuda")).logits
110
+
111
+ pred_ids = torch.argmax(logits, dim=-1)
112
+ batch["pred_strings"] = processor.batch_decode(pred_ids)
113
+ return batch
114
+
115
+ result = test_dataset.map(evaluate, batched=True, batch_size=8)
116
+
117
+ print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"], references=result["sentence"])))
118
+ ```
119
+
120
+ **Test Result**: ?? %
121
+
122
+ ## Training
123
+
124
+ The Common Voice `validation` dataset was used for training, with 12% of the test dataset used for validation, trained on 1 V100 GPU for 48 hours (20 epochs).
125
+
126
+ The script used for training was just the `run_finetuning.py` script provided in OVHcloud's databuzzword/hf-wav2vec image.