Configuration Parsing Warning:In adapter_config.json: "peft.task_type" must be a string

kurdish_mms_1b_final_safe

This model is a fine-tuned version of facebook/mms-1b-all on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.1168
  • Wer: 0.2834

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: 0.0003
  • train_batch_size: 2
  • eval_batch_size: 8
  • seed: 42
  • gradient_accumulation_steps: 8
  • total_train_batch_size: 16
  • optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 500
  • num_epochs: 3
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Wer
0.1983 0.7323 500 0.1608 0.3344
0.1744 1.4643 1000 0.1375 0.3003
0.1522 2.1963 1500 0.1233 0.2868
0.1535 2.9286 2000 0.1168 0.2834

Framework versions

  • PEFT 0.18.0
  • Transformers 4.57.3
  • Pytorch 2.9.1+cu128
  • Datasets 2.19.0
  • Tokenizers 0.22.1
Downloads last month
1
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for kamalaziz/kurdish_mms_1b_final_safe

Adapter
(5)
this model