metadata

library_name: transformers
license: cc-by-nc-4.0
base_model: facebook/nllb-200-1.3B
tags:
  - generated_from_trainer
metrics:
  - bleu
model-index:
  - name: nllb-200-1.3B-ft-eng-to-cym
    results: []

nllb-200-1.3B-ft-eng-to-cym

This model is a fine-tuned version of facebook/nllb-200-1.3B on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.6294
Bleu: 31.9664
Gen Len: 56.0969

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: 32
eval_batch_size: 32
seed: 42
optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
training_steps: 15000

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Gen Len
0.9749	0.0731	1000	0.8471	26.4405	64.7467
0.8876	0.1463	2000	0.7820	25.7663	72.4875
0.8296	0.2194	3000	0.7455	29.7194	64.5048
0.7953	0.2926	4000	0.7169	26.1134	62.5134
0.7682	0.3657	5000	0.6996	32.6703	55.0
0.7499	0.4389	6000	0.6835	30.9855	57.704
0.7238	0.5120	7000	0.6696	30.3129	54.3538
0.7184	0.5851	8000	0.6597	33.7707	53.5875
0.7171	0.6583	9000	0.6511	32.3995	53.0923
0.7062	0.7314	10000	0.6440	31.099	56.6207
0.691	0.8046	11000	0.6386	32.5796	55.578
0.6851	0.8777	12000	0.6343	32.4382	55.1046
0.6892	0.9508	13000	0.6317	31.7749	55.8827
0.6586	1.0240	14000	0.6304	31.799	56.5098
0.6659	1.0971	15000	0.6294	31.9664	56.0969

Framework versions

Transformers 4.49.0
Pytorch 2.6.0+cu124
Datasets 3.3.2
Tokenizers 0.21.0