You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this model content.

dense_eng_hom_100m_mult_reseg_ep20_spm

This model is a fine-tuned version of on the arrow dataset. It achieves the following results on the evaluation set:

  • Loss: 5.0896

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: 0.0001
  • train_batch_size: 8
  • eval_batch_size: 32
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 32
  • optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_steps: 1352
  • training_steps: 13525
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss
7.4746 0.7394 500 6.5613
5.8724 1.4776 1000 5.7328
5.4524 2.2159 1500 5.2701
5.0152 2.9553 2000 4.9518
4.7695 3.6935 2500 4.7523
4.5097 4.4318 3000 4.6241
4.4453 5.1701 3500 4.5447
4.282 5.9094 4000 4.4772
4.1115 6.6477 4500 4.4538
3.9488 7.3860 5000 4.4516
3.9684 8.1242 5500 4.4631
3.8292 8.8636 6000 4.4464
3.6565 9.6018 6500 4.4896
3.5011 10.3401 7000 4.5425
3.5573 11.0784 7500 4.5800
3.4257 11.8177 8000 4.5976
3.2619 12.5560 8500 4.6698
3.1353 13.2943 9000 4.7375
3.1905 14.0325 9500 4.7786
3.0776 14.7719 10000 4.8166
2.9398 15.5102 10500 4.8848
2.874 16.2484 11000 4.9440
2.8896 16.9878 11500 4.9569
2.8135 17.7261 12000 5.0176
2.7288 18.4643 12500 5.0580
2.7072 19.2026 13000 5.0837
2.6928 19.9420 13500 5.0898

Framework versions

  • Transformers 4.57.1
  • Pytorch 2.9.1+cu128
  • Datasets 3.6.0
  • Tokenizers 0.22.1
Downloads last month
-
Safetensors
Model size
0.2B params
Tensor type
F32
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support