gchhablani commited on
Commit
fd49820
1 Parent(s): 970317f

Create README.md

Browse files
Files changed (1) hide show
  1. README.md +123 -0
README.md ADDED
@@ -0,0 +1,123 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language: rm-sursilv
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: Wav2Vec2 Large 53 Romansh Sursilvan by Gunjan Chhablani
15
+ results:
16
+ - task:
17
+ name: Speech Recognition
18
+ type: automatic-speech-recognition
19
+ dataset:
20
+ name: Common Voice rm-sursilv
21
+ type: common_voice
22
+ args: rm-sursilv
23
+ metrics:
24
+ - name: Test WER
25
+ type: wer
26
+ value: 25.16
27
+ ---
28
+
29
+ # Wav2Vec2-Large-XLSR-53-Romansh-Sursilvan
30
+
31
+ Fine-tuned [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on Romansh Sursilvan using the [Common Voice](https://huggingface.co/datasets/common_voice) dataset.
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", "rm-sursilv", split="test[:2%]")
45
+
46
+ processor = Wav2Vec2Processor.from_pretrained("gchhablani/wav2vec2-large-xlsr-rm-sursilv")
47
+ model = Wav2Vec2ForCTC.from_pretrained("gchhablani/wav2vec2-large-xlsr-rm-sursilv")
48
+
49
+ resampler = torchaudio.transforms.Resample(48_000, 16_000)
50
+
51
+ # Preprocessing the datasets.
52
+ # We need to read the aduio 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["speech"][:2], 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 Portuguese 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", "rm-sursilv", split="test")
84
+
85
+ processor = Wav2Vec2Processor.from_pretrained("gchhablani/wav2vec2-large-xlsr-rm-sursilv")
86
+ model = Wav2Vec2ForCTC.from_pretrained("gchhablani/wav2vec2-large-xlsr-rm-sursilv")
87
+ model.to("cuda")
88
+
89
+ chars_to_ignore_regex = '[\,\?\.\!\-\;\:\"\“\%\‘\”\�\…\«\»\–]'
90
+ resampler = torchaudio.transforms.Resample(48_000, 16_000)
91
+
92
+ # Preprocessing the datasets.
93
+ # We need to read the aduio files as arrays
94
+ def speech_file_to_array_fn(batch):
95
+ batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
96
+ speech_array, sampling_rate = torchaudio.load(batch["path"])
97
+ batch["speech"] = resampler(speech_array).squeeze().numpy()
98
+ return batch
99
+
100
+ test_dataset = test_dataset.map(speech_file_to_array_fn)
101
+
102
+ # Preprocessing the datasets.
103
+ # We need to read the aduio files as arrays
104
+ def evaluate(batch):
105
+ inputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
106
+
107
+ with torch.no_grad():
108
+ logits = model(inputs.input_values.to("cuda"), attention_mask=inputs.attention_mask.to("cuda")).logits
109
+
110
+ pred_ids = torch.argmax(logits, dim=-1)
111
+ batch["pred_strings"] = processor.batch_decode(pred_ids)
112
+ return batch
113
+
114
+ result = test_dataset.map(evaluate, batched=True, batch_size=8)
115
+
116
+ print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"], references=result["sentence"])))
117
+ ```
118
+
119
+ **Test Result**: 25.16 %
120
+
121
+ ## Training
122
+
123
+ The Common Voice `train` and `validation` datasets were used for training. The code can be found [here](https://colab.research.google.com/drive/1dpZr_GzRowCciUbzM3GnW04TNKnB7vrP?usp=sharing).