nllb-200-distilled-1.3B-finetuned-finetuned-finetuned

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

• Loss: 0.3940
• Rouge: 0.1765
• Gen Len: 13.5

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• training_steps: 2000
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge	Gen Len
0.0469	200.0	200	0.3920	0.0476	20.0
0.0069	400.0	400	0.3806	0.1364	17.5
0.0034	600.0	600	0.3799	0.1364	17.5
0.0022	800.0	800	0.3908	0.1364	109.0
0.0016	1000.0	1000	0.3874	0.1765	13.5
0.0013	1200.0	1200	0.3904	0.1765	13.5
0.0011	1400.0	1400	0.3920	0.1765	13.5
0.001	1600.0	1600	0.3930	0.1765	13.5
0.0009	1800.0	1800	0.3939	0.1765	13.5
0.0008	2000.0	2000	0.3940	0.1765	13.5

Framework versions

• Transformers 4.39.2
• Pytorch 2.2.2+cu121
• Datasets 2.21.0
• Tokenizers 0.15.2