phi4-speech-en_sl-1epoch

This model is a fine-tuned version of microsoft/Phi-4-multimodal-instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 4.5536
• Wer: 67.5613

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: 4e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.95) and epsilon=1e-07 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 50
• num_epochs: 1
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Wer
No log	0.3077	1	4.5536	67.5613
No log	0.6154	2	4.5536	67.5613
No log	0.9231	3	4.5536	67.5613

Framework versions

• PEFT 0.13.2
• Transformers 4.46.1
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.20.3