Mistral-7B-Instruct-v0.2-advisegpt-v0.5

This model is a fine-tuned version of mistralai/Mistral-7B-Instruct-v0.2 on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 0.0840
• Bleu: {'bleu': 0.9537910015397628, 'precisions': [0.9763005593772222, 0.9591762297332277, 0.9471223357463351, 0.9370695448087227], 'brevity_penalty': 0.9989373668126428, 'length_ratio': 0.9989379310075293, 'translation_length': 1022387, 'reference_length': 1023474}
• Rouge: {'rouge1': 0.9741038510844018, 'rouge2': 0.9550445541823809, 'rougeL': 0.9723656951648176, 'rougeLsum': 0.9736935611588988}
• Exact Match: {'exact_match': 0.0}

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: 2e-05
• train_batch_size: 3
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 10
• total_train_batch_size: 30
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 3
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Bleu	Exact Match	Validation Loss	Rouge
0.069	0.9999	829	{'bleu': 0.9459206747141892, 'brevity_penalty': 0.998656611989374, 'length_ratio': 0.9986575135274565, 'precisions': [0.9726768417963018, 0.9521542081327253, 0.9380288226144853, 0.9265355643009697], 'reference_length': 1023474, 'translation_length': 1022100}	{'exact_match': 0.0}	0.0990	{'rouge1': 0.9702189356306301, 'rouge2': 0.9472171244648081, 'rougeL': 0.9677029434775739, 'rougeLsum': 0.9695684693436178}
0.0501	1.9999	1658	{'bleu': 0.9537910015397628, 'brevity_penalty': 0.9989373668126428, 'length_ratio': 0.9989379310075293, 'precisions': [0.9763005593772222, 0.9591762297332277, 0.9471223357463351, 0.9370695448087227], 'reference_length': 1023474, 'translation_length': 1022387}	{'exact_match': 0.0}	0.0840	{'rouge1': 0.9741105562035488, 'rouge2': 0.9550205654651982, 'rougeL': 0.9723363685950056, 'rougeLsum': 0.9737013621980013}
0.0479	2.9999	2487	0.0850	{'bleu': 0.9548514568958526, 'precisions': [0.9767648848122783, 0.9601353822381405, 0.9483682511725553, 0.9385079979703334], 'brevity_penalty': 0.9989676875356347, 'length_ratio': 0.9989682200036347, 'translation_length': 1022418, 'reference_length': 1023474}	{'rouge1': 0.9746456572659052, 'rouge2': 0.9560608145101823, 'rougeL': 0.9729518327172596, 'rougeLsum': 0.9742472834405176}	{'exact_match': 0.0}

Framework versions

• PEFT 0.10.0
• Transformers 4.40.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1