README.md

metadata

library_name: transformers
license: apache-2.0
base_model: google/byt5-small
tags:
  - generated_from_trainer
model-index:
  - name: browndw/morphoseg-en-byt5
    results: []

browndw/morphoseg-en-byt5

This model is a fine-tuned version of google/byt5-small on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.0404
• Exact Match: 0.9474

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: 16
• eval_batch_size: 16
• seed: 99
• gradient_accumulation_steps: 2
• total_train_batch_size: 32
• 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: 0.1
• num_epochs: 24

Training results

Training Loss	Epoch	Step	Validation Loss	Exact Match
0.1311	1.0	9104	0.0541	0.8477
0.0946	2.0	18208	0.0372	0.8839
0.0790	3.0	27312	0.0332	0.8960
0.0631	4.0	36416	0.0276	0.9125
0.0478	5.0	45520	0.0255	0.9176
0.0327	6.0	54624	0.0262	0.9246
0.0298	7.0	63728	0.0250	0.9300
0.0284	8.0	72832	0.0253	0.9300
0.0258	9.0	81936	0.0262	0.9335
0.0178	10.0	91040	0.0256	0.9369
0.0144	11.0	100144	0.0272	0.9372
0.0127	12.0	109248	0.0275	0.9404
0.0133	13.0	118352	0.0277	0.9408
0.0117	14.0	127456	0.0280	0.9418
0.0093	15.0	136560	0.0309	0.9424
0.0084	16.0	145664	0.0306	0.9424
0.0062	17.0	154768	0.0322	0.9430
0.0041	18.0	163872	0.0330	0.9446
0.0036	19.0	172976	0.0354	0.9453
0.0030	20.0	182080	0.0351	0.9459
0.0023	21.0	191184	0.0364	0.9469
0.0010	22.0	200288	0.0380	0.9478
0.0007	23.0	209392	0.0396	0.9476
0.0004	24.0	218496	0.0404	0.9474

Framework versions

• Transformers 5.0.0
• Pytorch 2.10.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2