File size: 3,002 Bytes
b20b68a c879a3e b20b68a c879a3e b20b68a c879a3e b20b68a c879a3e b20b68a |
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 |
---
language:
- ur
license: apache-2.0
tags:
- automatic-speech-recognition
- generated_from_trainer
- hf-asr-leaderboard
- model_for_talk
- mozilla-foundation/common_voice_7_0
- robust-speech-event
- ur
datasets:
- mozilla-foundation/common_voice_7_0
model-index:
- name: XLS-R-300M - Urdu
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 7
type: mozilla-foundation/common_voice_7_0
args: ur
metrics:
- name: Test WER
type: wer
value: 105.66
- name: Test CER
type: cer
value: 434.011
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
infinitejoy/wav2vec2-large-xls-r-300m-urdu
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the MOZILLA-FOUNDATION/COMMON_VOICE_7_0 - -UR dataset.
It achieves the following results on the evaluation set:
- Loss: NA
- Wer: NA
## 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: 8
- 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: 50.0
- mixed_precision_training: Native AMP
### Training results
### Framework versions
- Transformers 4.16.0.dev0
- Pytorch 1.10.0+cu102
- Datasets 1.17.1.dev0
- Tokenizers 0.10.3
#### Evaluation Commands
1. To evaluate on `mozilla-foundation/common_voice_7_0` with split `test`
```bash
python eval.py \
--model_id infinitejoy/wav2vec2-large-xls-r-300m-urdu --dataset speech-recognition-community-v2/dev_data \
--config ur --split validation --chunk_length_s 10 --stride_length_s 1
```
### Inference
```python
import torch
from datasets import load_dataset
from transformers import AutoModelForCTC, AutoProcessor
import torchaudio.functional as F
model_id = "infinitejoy/wav2vec2-large-xls-r-300m-urdu"
sample_iter = iter(load_dataset("mozilla-foundation/common_voice_7_0", "ur", 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
```
### Eval results on Common Voice 7 "test" (WER):
|