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

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: 0.2663
• Bleu: {'bleu': 0.8973677595827475, 'precisions': [0.9445357190260831, 0.9043357485277211, 0.884048801354545, 0.8699324916460348], 'brevity_penalty': 0.9967654892802698, 'length_ratio': 0.9967707090483877, 'translation_length': 1235588, 'reference_length': 1239591}
• Rouge: {'rouge1': 0.94002209733584, 'rouge2': 0.8959242425644911, 'rougeL': 0.931506639182089, 'rougeLsum': 0.9379602925725274}
• 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: 8
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Bleu	Rouge	Exact Match
0.0486	1.0	953	0.2746	{'bleu': 0.8908185969537198, 'precisions': [0.9404575669873605, 0.8973236909340203, 0.8757372017868406, 0.8612917392143906], 'brevity_penalty': 0.9973237695370935, 'length_ratio': 0.9973273442611313, 'translation_length': 1236278, 'reference_length': 1239591}	{'rouge1': 0.9362204684890618, 'rouge2': 0.8885056700228431, 'rougeL': 0.9265631818483623, 'rougeLsum': 0.9336844419843153}	{'exact_match': 0.0}
0.0491	2.0	1906	0.2663	{'bleu': 0.8973677595827475, 'precisions': [0.9445357190260831, 0.9043357485277211, 0.884048801354545, 0.8699324916460348], 'brevity_penalty': 0.9967654892802698, 'length_ratio': 0.9967707090483877, 'translation_length': 1235588, 'reference_length': 1239591}	{'rouge1': 0.94002209733584, 'rouge2': 0.8959242425644911, 'rougeL': 0.931506639182089, 'rougeLsum': 0.9379602925725274}	{'exact_match': 0.0}
0.0457	3.0	2859	0.2713	{'bleu': 0.9011515587348646, 'precisions': [0.9455932972222963, 0.9073373719165547, 0.8876381124008869, 0.8739477250662966], 'brevity_penalty': 0.9976982094967736, 'length_ratio': 0.9977008545560592, 'translation_length': 1236741, 'reference_length': 1239591}	{'rouge1': 0.9413126529676463, 'rouge2': 0.899299253704354, 'rougeL': 0.9334048668464673, 'rougeLsum': 0.9393911208574579}	{'exact_match': 0.0}
0.0427	4.0	3812	0.2973	{'bleu': 0.90301748257064, 'precisions': [0.9459337521093097, 0.908889094215937, 0.8900854627247164, 0.8768615812629189], 'brevity_penalty': 0.9977289348603028, 'length_ratio': 0.9977315098286451, 'translation_length': 1236779, 'reference_length': 1239591}	{'rouge1': 0.9416505951538487, 'rouge2': 0.9013278888630185, 'rougeL': 0.9338287357833286, 'rougeLsum': 0.9396676143080369}	{'exact_match': 0.0}
0.0387	5.0	4765	0.3174	{'bleu': 0.9027081754085503, 'precisions': [0.9453687163030563, 0.9082821849767655, 0.8894881333354993, 0.8763610309353894], 'brevity_penalty': 0.9980126956655654, 'length_ratio': 0.9980146677412146, 'translation_length': 1237130, 'reference_length': 1239591}	{'rouge1': 0.9410039582969136, 'rouge2': 0.9003660562981124, 'rougeL': 0.9333230818742395, 'rougeLsum': 0.9390562109303144}	{'exact_match': 0.0}
0.0375	6.0	5718	0.3550	{'bleu': 0.9020801569628037, 'precisions': [0.9452423604006518, 0.9080153383434265, 0.889329438008671, 0.8762846622568347], 'brevity_penalty': 0.9974911930257172, 'length_ratio': 0.9974943348249543, 'translation_length': 1236485, 'reference_length': 1239591}	{'rouge1': 0.9405264680129318, 'rouge2': 0.8997693944544493, 'rougeL': 0.9326541656452509, 'rougeLsum': 0.9385618029898112}	{'exact_match': 0.0}
0.0361	7.0	6671	0.3875	{'bleu': 0.9012292336144602, 'precisions': [0.9445250811645649, 0.907065679940991, 0.8882131409279821, 0.8750920375317651], 'brevity_penalty': 0.9976529282968476, 'length_ratio': 0.99765567836488, 'translation_length': 1236685, 'reference_length': 1239591}	{'rouge1': 0.9398793690117034, 'rouge2': 0.8988761356423343, 'rougeL': 0.9319409400581564, 'rougeLsum': 0.9378780059486569}	{'exact_match': 0.0}
0.0336	8.0	7624	0.4056	{'bleu': 0.9006875075046153, 'precisions': [0.9440281128914361, 0.906425354183536, 0.8876174257258417, 0.874531174789441], 'brevity_penalty': 0.9976876979719658, 'length_ratio': 0.997690367225964, 'translation_length': 1236728, 'reference_length': 1239591}	{'rouge1': 0.9394107383916215, 'rouge2': 0.8982745501459677, 'rougeL': 0.9313967228794204, 'rougeLsum': 0.9373525606500615}	{'exact_match': 0.0}

Framework versions

• PEFT 0.10.0
• Transformers 4.40.1
• Pytorch 2.3.0+cu121
• Datasets 2.19.0
• Tokenizers 0.19.1