nllb-200-1.3B-ICFOSS-English_Malayalam_Translation

This model is a fine-tuned version of facebook/nllb-200-1.3B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7986
• Bleu: 25.5240
• Rouge: {'rouge1': 0.41343600278075376, 'rouge2': 0.26557575177744935, 'rougeL': 0.4045593832940667, 'rougeLsum': 0.40500259745407907}
• Chrf: {'score': 63.47912581213381, 'char_order': 6, 'word_order': 0, 'beta': 2}

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: 0.0002
• 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: cosine
• num_epochs: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Rouge	Chrf
0.922	1.0	3000	0.8224	24.7167	{'rouge1': 0.4114291884720509, 'rouge2': 0.2644602223766743, 'rougeL': 0.4023641002206244, 'rougeLsum': 0.4027992964074694}	{'score': 62.82692539709392, 'char_order': 6, 'word_order': 0, 'beta': 2}
0.8622	2.0	6000	0.8031	25.2224	{'rouge1': 0.41228269394912576, 'rouge2': 0.2638244877955537, 'rougeL': 0.4027881671223306, 'rougeLsum': 0.4032496266035372}	{'score': 63.148595621624104, 'char_order': 6, 'word_order': 0, 'beta': 2}
0.847	3.0	9000	0.7991	25.5124	{'rouge1': 0.4131136011924885, 'rouge2': 0.2660486503477698, 'rougeL': 0.40400418780615166, 'rougeLsum': 0.404579016867125}	{'score': 63.46834487204518, 'char_order': 6, 'word_order': 0, 'beta': 2}
0.8407	4.0	12000	0.7986	25.5672	{'rouge1': 0.41280607037014805, 'rouge2': 0.26520458752310194, 'rougeL': 0.40385353186878514, 'rougeLsum': 0.404398367682378}	{'score': 63.500532443749044, 'char_order': 6, 'word_order': 0, 'beta': 2}
0.8414	5.0	15000	0.7986	25.5240	{'rouge1': 0.41343600278075376, 'rouge2': 0.26557575177744935, 'rougeL': 0.4045593832940667, 'rougeLsum': 0.40500259745407907}	{'score': 63.47912581213381, 'char_order': 6, 'word_order': 0, 'beta': 2}

Framework versions

• PEFT 0.7.2.dev0
• Transformers 4.36.1
• Pytorch 2.1.0+cu121
• Datasets 2.16.0
• Tokenizers 0.15.0