patrickvonplaten
commited on
Commit
•
43b9e58
1
Parent(s):
eaa092f
Update README.md
Browse files
README.md
CHANGED
@@ -11,170 +11,42 @@ tags:
|
|
11 |
- speech
|
12 |
- xlsr-fine-tuning-week
|
13 |
license: apache-2.0
|
14 |
-
model-index:
|
15 |
-
- name: XLSR Wav2Vec2 Spanish by Jonatas Grosman
|
16 |
-
results:
|
17 |
-
- task:
|
18 |
-
name: Speech Recognition
|
19 |
-
type: automatic-speech-recognition
|
20 |
-
dataset:
|
21 |
-
name: Common Voice es
|
22 |
-
type: common_voice
|
23 |
-
args: es
|
24 |
-
metrics:
|
25 |
-
- name: Test WER
|
26 |
-
type: wer
|
27 |
-
value: 8.81
|
28 |
-
- name: Test CER
|
29 |
-
type: cer
|
30 |
-
value: 2.70
|
31 |
---
|
32 |
|
33 |
-
# Wav2Vec2-Large-XLSR-53-Spanish
|
34 |
|
35 |
-
|
36 |
-
|
37 |
|
38 |
-
This model
|
|
|
|
|
39 |
|
40 |
-
|
41 |
|
42 |
-
|
43 |
|
44 |
-
|
45 |
-
|
46 |
-
Using the [ASRecognition](https://github.com/jonatasgrosman/asrecognition) library:
|
47 |
-
|
48 |
-
```python
|
49 |
-
from asrecognition import ASREngine
|
50 |
-
|
51 |
-
asr = ASREngine("es", model_path="jonatasgrosman/wav2vec2-large-xlsr-53-spanish")
|
52 |
-
|
53 |
-
audio_paths = ["/path/to/file.mp3", "/path/to/another_file.wav"]
|
54 |
-
transcriptions = asr.transcribe(audio_paths)
|
55 |
-
```
|
56 |
-
|
57 |
-
Writing your own inference script:
|
58 |
-
|
59 |
-
```python
|
60 |
-
import torch
|
61 |
-
import librosa
|
62 |
-
from datasets import load_dataset
|
63 |
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
|
|
|
64 |
|
65 |
-
|
66 |
-
MODEL_ID = "jonatasgrosman/wav2vec2-large-xlsr-53-spanish"
|
67 |
-
SAMPLES = 10
|
68 |
-
|
69 |
-
test_dataset = load_dataset("common_voice", LANG_ID, split=f"test[:{SAMPLES}]")
|
70 |
-
|
71 |
-
processor = Wav2Vec2Processor.from_pretrained(MODEL_ID)
|
72 |
-
model = Wav2Vec2ForCTC.from_pretrained(MODEL_ID)
|
73 |
-
|
74 |
-
# Preprocessing the datasets.
|
75 |
-
# We need to read the audio files as arrays
|
76 |
-
def speech_file_to_array_fn(batch):
|
77 |
-
speech_array, sampling_rate = librosa.load(batch["path"], sr=16_000)
|
78 |
-
batch["speech"] = speech_array
|
79 |
-
batch["sentence"] = batch["sentence"].upper()
|
80 |
-
return batch
|
81 |
-
|
82 |
-
test_dataset = test_dataset.map(speech_file_to_array_fn)
|
83 |
-
inputs = processor(test_dataset["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
|
84 |
-
|
85 |
-
with torch.no_grad():
|
86 |
-
logits = model(inputs.input_values, attention_mask=inputs.attention_mask).logits
|
87 |
-
|
88 |
-
predicted_ids = torch.argmax(logits, dim=-1)
|
89 |
-
predicted_sentences = processor.batch_decode(predicted_ids)
|
90 |
-
|
91 |
-
for i, predicted_sentence in enumerate(predicted_sentences):
|
92 |
-
print("-" * 100)
|
93 |
-
print("Reference:", test_dataset[i]["sentence"])
|
94 |
-
print("Prediction:", predicted_sentence)
|
95 |
-
```
|
96 |
-
|
97 |
-
| Reference | Prediction |
|
98 |
-
| ------------- | ------------- |
|
99 |
-
| HABITA EN AGUAS POCO PROFUNDAS Y ROCOSAS. | HABITAN AGUAS POCO PROFUNDAS Y ROCOSAS |
|
100 |
-
| OPERA PRINCIPALMENTE VUELOS DE CABOTAJE Y REGIONALES DE CARGA. | OPERA PRINCIPALMENTE VUELO DE CARBOTAJES Y REGIONALES DE CARGAN |
|
101 |
-
| PARA VISITAR CONTACTAR PRIMERO CON LA DIRECCIÓN. | PARA VISITAR CONTACTAR PRIMERO CON LA DIRECCIÓN |
|
102 |
-
| TRES | TRES |
|
103 |
-
| REALIZÓ LOS ESTUDIOS PRIMARIOS EN FRANCIA, PARA CONTINUAR LUEGO EN ESPAÑA. | REALIZÓ LOS ESTUDIOS PRIMARIOS EN FRANCIA PARA CONTINUAR LUEGO EN ESPAÑA |
|
104 |
-
| EN LOS AÑOS QUE SIGUIERON, ESTE TRABAJO ESPARTA PRODUJO DOCENAS DE BUENOS JUGADORES. | EN LOS AÑOS QUE SIGUIERON ESTE TRABAJO ESPARTA PRODUJO DOCENA DE BUENOS JUGADORES |
|
105 |
-
| SE ESTÁ TRATANDO DE RECUPERAR SU CULTIVO EN LAS ISLAS CANARIAS. | SE ESTÓ TRATANDO DE RECUPERAR SU CULTIVO EN LAS ISLAS CANARIAS |
|
106 |
-
| SÍ | SÍ |
|
107 |
-
| "FUE ""SACADA"" DE LA SERIE EN EL EPISODIO ""LEAD"", EN QUE ALEXANDRA CABOT REGRESÓ." | FUE SACADA DE LA SERIE EN EL EPISODIO LEED EN QUE ALEXANDRA KAOT REGRESÓ |
|
108 |
-
| SE UBICAN ESPECÍFICAMENTE EN EL VALLE DE MOKA, EN LA PROVINCIA DE BIOKO SUR. | SE UBICAN ESPECÍFICAMENTE EN EL VALLE DE MOCA EN LA PROVINCIA DE PÍOCOSUR |
|
109 |
-
|
110 |
-
## Evaluation
|
111 |
-
|
112 |
-
The model can be evaluated as follows on the Spanish test data of Common Voice.
|
113 |
-
|
114 |
-
```python
|
115 |
-
import torch
|
116 |
-
import re
|
117 |
-
import librosa
|
118 |
-
from datasets import load_dataset, load_metric
|
119 |
-
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
|
120 |
-
|
121 |
-
LANG_ID = "es"
|
122 |
-
MODEL_ID = "jonatasgrosman/wav2vec2-large-xlsr-53-spanish"
|
123 |
-
DEVICE = "cuda"
|
124 |
-
|
125 |
-
CHARS_TO_IGNORE = [",", "?", "¿", ".", "!", "¡", ";", ";", ":", '""', "%", '"', "�", "ʿ", "·", "჻", "~", "՞",
|
126 |
-
"؟", "،", "।", "॥", "«", "»", "„", "“", "”", "「", "」", "‘", "’", "《", "》", "(", ")", "[", "]",
|
127 |
-
"{", "}", "=", "`", "_", "+", "<", ">", "…", "–", "°", "´", "ʾ", "‹", "›", "©", "®", "—", "→", "。",
|
128 |
-
"、", "﹂", "﹁", "‧", "~", "﹏", ",", "{", "}", "(", ")", "[", "]", "【", "】", "‥", "〽",
|
129 |
-
"『", "』", "〝", "〟", "⟨", "⟩", "〜", ":", "!", "?", "♪", "؛", "/", "\\", "º", "−", "^", "ʻ", "ˆ"]
|
130 |
-
|
131 |
-
test_dataset = load_dataset("common_voice", LANG_ID, split="test")
|
132 |
-
|
133 |
-
wer = load_metric("wer.py") # https://github.com/jonatasgrosman/wav2vec2-sprint/blob/main/wer.py
|
134 |
-
cer = load_metric("cer.py") # https://github.com/jonatasgrosman/wav2vec2-sprint/blob/main/cer.py
|
135 |
-
|
136 |
-
chars_to_ignore_regex = f"[{re.escape(''.join(CHARS_TO_IGNORE))}]"
|
137 |
-
|
138 |
-
processor = Wav2Vec2Processor.from_pretrained(MODEL_ID)
|
139 |
-
model = Wav2Vec2ForCTC.from_pretrained(MODEL_ID)
|
140 |
-
model.to(DEVICE)
|
141 |
-
|
142 |
-
# Preprocessing the datasets.
|
143 |
-
# We need to read the audio files as arrays
|
144 |
-
def speech_file_to_array_fn(batch):
|
145 |
-
with warnings.catch_warnings():
|
146 |
-
warnings.simplefilter("ignore")
|
147 |
-
speech_array, sampling_rate = librosa.load(batch["path"], sr=16_000)
|
148 |
-
batch["speech"] = speech_array
|
149 |
-
batch["sentence"] = re.sub(chars_to_ignore_regex, "", batch["sentence"]).upper()
|
150 |
-
return batch
|
151 |
-
|
152 |
-
test_dataset = test_dataset.map(speech_file_to_array_fn)
|
153 |
|
154 |
-
|
155 |
-
# We need to read the audio files as arrays
|
156 |
-
def evaluate(batch):
|
157 |
-
inputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
|
158 |
|
159 |
-
|
160 |
-
|
161 |
|
162 |
-
|
163 |
-
batch["pred_strings"] = processor.batch_decode(pred_ids)
|
164 |
-
return batch
|
165 |
|
166 |
-
|
|
|
167 |
|
168 |
-
|
169 |
-
references = [x.upper() for x in result["sentence"]]
|
170 |
|
171 |
-
print(
|
172 |
-
print(f"CER: {cer.compute(predictions=predictions, references=references, chunk_size=1000) * 100}")
|
173 |
```
|
174 |
|
175 |
-
**Test Result**:
|
176 |
-
|
177 |
-
In the table below I report the Word Error Rate (WER) and the Character Error Rate (CER) of the model. I ran the evaluation script described above on other models as well (on 2021-04-22). Note that the table below may show different results from those already reported, this may have been caused due to some specificity of the other evaluation scripts used.
|
178 |
|
179 |
| Model | WER | CER |
|
180 |
| ------------- | ------------- | ------------- |
|
|
|
11 |
- speech
|
12 |
- xlsr-fine-tuning-week
|
13 |
license: apache-2.0
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
14 |
---
|
15 |
|
16 |
+
# Wav2Vec2-Large-XLSR-53-Spanish-With-LM
|
17 |
|
18 |
+
This is a model copy of [Wav2Vec2-Large-XLSR-53-Spanish](https://huggingface.co/jonatasgrosman/wav2vec2-large-xlsr-53-spanish)
|
19 |
+
that has language model support.
|
20 |
|
21 |
+
This model card can be seen as a demo for the [pyctcdecode](https://github.com/kensho-technologies/pyctcdecode) integration
|
22 |
+
with Transformers led by [this PR](https://github.com/huggingface/transformers/pull/14339). The PR explains in-detail how the
|
23 |
+
integration works.
|
24 |
|
25 |
+
In a nutshell: This PR adds a new Wav2Vec2WithLMProcessor class as drop-in replacement for Wav2Vec2Processor.
|
26 |
|
27 |
+
The only change from the existing ASR pipeline will be:
|
28 |
|
29 |
+
```diff
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
30 |
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
|
31 |
+
from datasets import load_dataset
|
32 |
|
33 |
+
ds = load_dataset("common_voice", "es", split="test", streaming=True)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
34 |
|
35 |
+
sample = next(iter(ds))
|
|
|
|
|
|
|
36 |
|
37 |
+
model = Wav2Vec2ForCTC.from_pretrained("patrickvonplaten/wav2vec2-large-xlsr-53-spanish-with-lm")
|
38 |
+
processor = Wav2Vec2Processor.from_pretrained("patrickvonplaten/wav2vec2-large-xlsr-53-spanish-with-lm")
|
39 |
|
40 |
+
input_values = processor(sample["audio"]["array"], return_tensors="pt").input_values
|
|
|
|
|
41 |
|
42 |
+
logits = model(input_values).logits
|
43 |
+
prediction_ids = torch.argmax(logits, dim=-1)
|
44 |
|
45 |
+
transcription = processor.batch_decode(prediction_ids)
|
|
|
46 |
|
47 |
+
print(transcription)
|
|
|
48 |
```
|
49 |
|
|
|
|
|
|
|
50 |
|
51 |
| Model | WER | CER |
|
52 |
| ------------- | ------------- | ------------- |
|