Automatic Speech Recognition
Transformers
PyTorch
Safetensors
Urdu
wav2vec2
mozilla-foundation/common_voice_7_0
Generated from Trainer
sv
robust-speech-event
model_for_talk
hf-asr-leaderboard
Instructions to use Maniac/wav2vec2-xls-r-urdu with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use Maniac/wav2vec2-xls-r-urdu with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="Maniac/wav2vec2-xls-r-urdu")# Load model directly from transformers import AutoProcessor, AutoModelForCTC processor = AutoProcessor.from_pretrained("Maniac/wav2vec2-xls-r-urdu") model = AutoModelForCTC.from_pretrained("Maniac/wav2vec2-xls-r-urdu", device_map="auto") - Notebooks
- Google Colab
- Kaggle
File size: 704 Bytes
12c4f3e 6af6c19 12c4f3e e1fb9b0 12c4f3e e1fb9b0 12c4f3e | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | python run_speech_recognition_ctc.py \
--dataset_name="mozilla-foundation/common_voice_7_0" \
--model_name_or_path="facebook/wav2vec2-large-xlsr-53" \
--dataset_config_name="ur" \
--output_dir="./" \
--overwrite_output_dir \
--max_steps="1000" \
--per_device_train_batch_size="16" \
--learning_rate="3e-4" \
--save_total_limit="10" \
--evaluation_strategy="steps" \
--text_column_name="sentence" \
--length_column_name="input_length" \
--chars_to_ignore ,\?\.\!\-\;\:\"\“\%\‘\”\�\ّ\ٔ\َِ,ً\
--save_steps="100" \
--layerdrop="0.0" \
--freeze_feature_encoder \
--gradient_checkpointing \
--fp16 \
--group_by_length \
--push_to_hub \
--use_auth_token \
--do_train --do_eval |