nllb-srt-translate

This model is a fine-tuned version of facebook/nllb-200-distilled-600M on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 1.7495

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: 5e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• 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: 50
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss
3.4046	0.3419	50	3.1991
2.8715	0.6838	100	2.6569
2.5797	1.0205	150	2.3924
2.4477	1.3624	200	2.2326
2.2755	1.7043	250	2.1179
2.192	2.0410	300	2.0329
2.1271	2.3829	350	1.9603
2.0209	2.7248	400	1.8972
1.937	3.0615	450	1.8475
1.9208	3.4034	500	1.8107
1.9622	3.7453	550	1.7854
1.8933	4.0821	600	1.7669
1.9055	4.4239	650	1.7553
1.8653	4.7658	700	1.7495

Framework versions

• PEFT 0.18.0
• Transformers 4.57.3
• Pytorch 2.9.1+cu128
• Datasets 4.4.1
• Tokenizers 0.22.1