README.md

metadata

language:
  - hi
license: apache-2.0
base_model: CKSINGH/whisper-small-hi-iiib
tags:
  - hf-asr-leaderboard
  - generated_from_trainer
metrics:
  - wer
model-index:
  - name: CKSINGH whisper-small-hi-fiefox
    results: []

CKSINGH whisper-small-hi-fiefox

This model is a fine-tuned version of CKSINGH/whisper-small-hi-iiib on the Common Voice 13.0 dataset. It achieves the following results on the evaluation set:

• Loss: 0.3975
• Wer: 33.5625
• Cer: 10.9848

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: 1.75e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 1
• training_steps: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer	Cer
0.4544	0.0	1	0.5773	34.5894	11.3627
0.5049	0.01	2	0.3975	33.5625	10.9848

Framework versions

• Transformers 4.36.0.dev0
• Pytorch 2.1.0+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0