wav2vec2-xls-r-1b-hi-cv7

This model is a fine-tuned version of facebook/wav2vec2-xls-r-1b on the MOZILLA-FOUNDATION/COMMON_VOICE_7_0 - HI dataset. It achieves the following results on the evaluation set:

• Loss: 0.5878
• Wer: 0.3419

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 7.5e-05
• train_batch_size: 8
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 2000
• num_epochs: 100.0
• mixed_precision_training: Native AMP

Training results

Framework versions

• Transformers 4.16.0.dev0
• Pytorch 1.10.1+cu102
• Datasets 1.17.1.dev0
• Tokenizers 0.11.0

Evaluation Commands

1. To evaluate on mozilla-foundation/common_voice_7_0 with split test

python eval.py --model_id anuragshas/wav2vec2-xls-r-1b-hi --dataset mozilla-foundation/common_voice_7_0 --config hi --split test

Inference With LM

import torch
from datasets import load_dataset
from transformers import AutoModelForCTC, AutoProcessor
import torchaudio.functional as F
model_id = "anuragshas/wav2vec2-xls-r-1b-hi"
sample_iter = iter(load_dataset("mozilla-foundation/common_voice_7_0", "hi", split="test", streaming=True, use_auth_token=True))
sample = next(sample_iter)
resampled_audio = F.resample(torch.tensor(sample["audio"]["array"]), 48_000, 16_000).numpy()
model = AutoModelForCTC.from_pretrained(model_id)
processor = AutoProcessor.from_pretrained(model_id)
input_values = processor(resampled_audio, return_tensors="pt").input_values
with torch.no_grad():
    logits = model(input_values).logits
transcription = processor.batch_decode(logits.numpy()).text
# => "तुम्हारे पास तीन महीने बचे हैं"

Eval results on Common Voice 7 "test" (WER):