metadata

license: apache-2.0
base_model: openai/whisper-large-v2
tags:
  - generated_from_trainer
model-index:
  - name: whisper-large-v2-atcosim_corpus
    results: []

whisper-large-v2-atcosim_corpus

This model is a fine-tuned version of openai/whisper-large-v2 on an unknown dataset.

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: 1e-05
train_batch_size: 256
eval_batch_size: 8
seed: 42
distributed_type: multi-GPU
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
lr_scheduler_warmup_steps: 250
num_epochs: 5

Training results

Framework versions

Transformers 4.42.3
Pytorch 2.3.1+cu118
Datasets 2.20.0
Tokenizers 0.19.1