llama-pro-8b-english-to-hinglish-translation

This model is a fine-tuned version of TencentARC/LLaMA-Pro-8B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7581
• Rouge Scores: {'rouge1': 0.920395098520514, 'rouge2': 0.8260982220795346, 'rougeL': 0.8629699886603178, 'rougeLsum': 0.9203938314651259}
• Bleu Scores: [0.0799874668395633, 0.07835910254936027, 0.07659241243282754, 0.0747653244473694]
• Gen Len: 2048.0

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: 0.0001
• train_batch_size: 2
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 4
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 2
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Rouge Scores	Bleu Scores	Gen Len
0.8367	1.0	500	0.7705	{'rouge1': 0.9217534575496169, 'rouge2': 0.826045962480547, 'rougeL': 0.861202510859852, 'rougeLsum': 0.9217216485625217}	[0.07995984189186581, 0.07829715249738517, 0.07650957414005091, 0.07466840050021681]	2048.0
0.581	2.0	1000	0.7581	{'rouge1': 0.920395098520514, 'rouge2': 0.8260982220795346, 'rougeL': 0.8629699886603178, 'rougeLsum': 0.9203938314651259}	[0.0799874668395633, 0.07835910254936027, 0.07659241243282754, 0.0747653244473694]	2048.0

Framework versions

• PEFT 0.8.2
• Transformers 4.38.0.dev0
• Pytorch 2.1.0+cu118
• Datasets 2.16.2.dev0
• Tokenizers 0.15.1