phi2-english-to-hinglish-translation

This model is a fine-tuned version of microsoft/phi-2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.3394
• Rouge Scores: {'rouge1': 0.02194963696306387, 'rouge2': 0.017844397420545253, 'rougeL': 0.017985463648805815, 'rougeLsum': 0.02198801722885821}
• Bleu Scores: [0.0141983812922229, 0.013783602019353523, 0.013237039007079092, 0.012647324457245113]
• 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
1.6688	1.0	500	1.4150	{'rouge1': 0.021944939879946292, 'rouge2': 0.017781155558600512, 'rougeL': 0.017866554441667286, 'rougeLsum': 0.02197862373873669}	[0.014214089766333284, 0.013807603949625002, 0.013250971870467268, 0.012646602626664907]	2048.0
1.2148	2.0	1000	1.3394	{'rouge1': 0.02194963696306387, 'rouge2': 0.017844397420545253, 'rougeL': 0.017985463648805815, 'rougeLsum': 0.02198801722885821}	[0.0141983812922229, 0.013783602019353523, 0.013237039007079092, 0.012647324457245113]	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