mixtral-8x7b-instruct-v0.1-english-to-hinglish-translation

This model is a fine-tuned version of mistralai/Mixtral-8x7B-Instruct-v0.1 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8719
• Rouge Scores: {'rouge1': 93.54182106152926, 'rouge2': 86.05446823709033, 'rougeL': 88.96922992123118, 'rougeLsum': 93.53351107529416}
• Bleu Scores: [0.0005895581385084092, 0.0005810015690343468, 0.0005667645554887882, 0.0005509301891024064]
• Gen Len: 2047.998

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
0.9944	1.0	500	0.8572	{'rouge1': 93.66433639177112, 'rouge2': 86.16312209617104, 'rougeL': 88.96569028246026, 'rougeLsum': 93.65082127462976}	[0.000581715028380836, 0.0005735131190908485, 0.000559640714092523, 0.0005441766968373427]	2047.998
0.6294	2.0	1000	0.8719	{'rouge1': 93.54182106152926, 'rouge2': 86.05446823709033, 'rougeL': 88.96922992123118, 'rougeLsum': 93.53351107529416}	[0.0005895581385084092, 0.0005810015690343468, 0.0005667645554887882, 0.0005509301891024064]	2047.998

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