File size: 4,567 Bytes
a69148f |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 |
---
language: hi
#datasets:
#- Interspeech 2021
metrics:
- wer
tags:
- audio
- automatic-speech-recognition
- speech
license: mit
model-index:
- name: Wav2Vec2 Vakyansh Hindi Model by Harveen Chadha
results:
- task:
name: Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice hi
type: common_voice
args: hi
metrics:
- name: Test WER
type: wer
value: 33.17
---
## Spaces Demo
Check the spaces demo [here](https://huggingface.co/spaces/Harveenchadha/wav2vec2-vakyansh-hindi/tree/main)
## Pretrained Model
Fine-tuned on Multilingual Pretrained Model [CLSRIL-23](https://arxiv.org/abs/2107.07402). The original fairseq checkpoint is present [here](https://github.com/Open-Speech-EkStep/vakyansh-models). When using this model, make sure that your speech input is sampled at 16kHz.
**Note: The result from this model is without a language model so you may witness a higher WER in some cases.**
## Dataset
This model was trained on 4200 hours of Hindi Labelled Data. The labelled data is not present in public domain as of now.
## Training Script
Models were trained using experimental platform setup by Vakyansh team at Ekstep. Here is the [training repository](https://github.com/Open-Speech-EkStep/vakyansh-wav2vec2-experimentation).
In case you want to explore training logs on wandb they are [here](https://wandb.ai/harveenchadha/hindi_finetuning_multilingual?workspace=user-harveenchadha).
## [Colab Demo](https://colab.research.google.com/github/harveenchadha/bol/blob/main/demos/hf/hindi/hf_hindi_him_4200_demo.ipynb)
## Usage
The model can be used directly (without a language model) as follows:
```python
import soundfile as sf
import torch
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
import argparse
def parse_transcription(wav_file):
# load pretrained model
processor = Wav2Vec2Processor.from_pretrained("Harveenchadha/vakyansh-wav2vec2-hindi-him-4200")
model = Wav2Vec2ForCTC.from_pretrained("Harveenchadha/vakyansh-wav2vec2-hindi-him-4200")
# load audio
audio_input, sample_rate = sf.read(wav_file)
# pad input values and return pt tensor
input_values = processor(audio_input, sampling_rate=sample_rate, return_tensors="pt").input_values
# INFERENCE
# retrieve logits & take argmax
logits = model(input_values).logits
predicted_ids = torch.argmax(logits, dim=-1)
# transcribe
transcription = processor.decode(predicted_ids[0], skip_special_tokens=True)
print(transcription)
```
## Evaluation
The model can be evaluated as follows on the hindi test data of Common Voice.
```python
import torch
import torchaudio
from datasets import load_dataset, load_metric
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
import re
test_dataset = load_dataset("common_voice", "hi", split="test")
wer = load_metric("wer")
processor = Wav2Vec2Processor.from_pretrained("Harveenchadha/vakyansh-wav2vec2-hindi-him-4200")
model = Wav2Vec2ForCTC.from_pretrained("Harveenchadha/vakyansh-wav2vec2-hindi-him-4200")
model.to("cuda")
resampler = torchaudio.transforms.Resample(48_000, 16_000)
chars_to_ignore_regex = '[\,\?\.\!\-\;\:\"\“]'
# Preprocessing the datasets.
# We need to read the aduio files as arrays
def speech_file_to_array_fn(batch):
batch["sentence"] = re.sub(chars_to_ignore_regex, '', batch["sentence"]).lower()
speech_array, sampling_rate = torchaudio.load(batch["path"])
batch["speech"] = resampler(speech_array).squeeze().numpy()
return batch
test_dataset = test_dataset.map(speech_file_to_array_fn)
# Preprocessing the datasets.
# We need to read the aduio files as arrays
def evaluate(batch):
inputs = processor(batch["speech"], sampling_rate=16_000, return_tensors="pt", padding=True)
with torch.no_grad():
logits = model(inputs.input_values.to("cuda")).logits
pred_ids = torch.argmax(logits, dim=-1)
batch["pred_strings"] = processor.batch_decode(pred_ids, skip_special_tokens=True)
return batch
result = test_dataset.map(evaluate, batched=True, batch_size=8)
print("WER: {:2f}".format(100 * wer.compute(predictions=result["pred_strings"], references=result["sentence"])))
```
**Test Result**: 33.17 %
[**Colab Evaluation**](https://colab.research.google.com/github/harveenchadha/bol/blob/main/demos/hf/hindi/hf_vakyansh_hindi_him_4200_evaluation_common_voice.ipynb)
## Credits
Thanks to Ekstep Foundation for making this possible. The vakyansh team will be open sourcing speech models in all the Indic Languages. |