nllb-200-1.3B-ICFOSS_Tamil_Malayalam_Translator

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: 0.7754
• Bleu: 29.8537
• Rouge: {'rouge1': 0.20075757575757575, 'rouge2': 0.09794372294372294, 'rougeL': 0.1974025974025974, 'rougeLsum': 0.19978354978354979}
• Chrf: {'score': 64.95358370898863, '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
0.9004	1.0	3806	0.8154	27.1742	{'rouge1': 0.19772727272727275, 'rouge2': 0.09577922077922078, 'rougeL': 0.1945887445887446, 'rougeLsum': 0.19675324675324676}	{'score': 63.3462918203542, 'char_order': 6, 'word_order': 0, 'beta': 2}
0.7946	2.0	7612	0.7844	28.0891	{'rouge1': 0.2006395120031484, 'rouge2': 0.09577922077922078, 'rougeL': 0.19744195198740652, 'rougeLsum': 0.200137741046832}	{'score': 64.30399837657693, 'char_order': 6, 'word_order': 0, 'beta': 2}
0.7549	3.0	11418	0.7754	29.3679	{'rouge1': 0.20058048012593466, 'rouge2': 0.09794372294372293, 'rougeL': 0.1974025974025974, 'rougeLsum': 0.1996064541519087}	{'score': 64.84398894116521, 'char_order': 6, 'word_order': 0, 'beta': 2}
0.7403	4.0	15224	0.7754	29.1387	{'rouge1': 0.20058048012593466, 'rouge2': 0.09794372294372293, 'rougeL': 0.1974025974025974, 'rougeLsum': 0.1996064541519087}	{'score': 64.64187026525788, 'char_order': 6, 'word_order': 0, 'beta': 2}
0.7374	5.0	19030	0.7754	29.8537	{'rouge1': 0.20075757575757575, 'rouge2': 0.09794372294372294, 'rougeL': 0.1974025974025974, 'rougeLsum': 0.19978354978354979}	{'score': 64.95358370898863, 'char_order': 6, 'word_order': 0, 'beta': 2}

Framework versions

• PEFT 0.10.0
• Transformers 4.40.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1