Phi-3.5-MultiCap

This model is a fine-tuned version of microsoft/Phi-3.5-mini-instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.5367

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: 16
• eval_batch_size: 16
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 128
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 2

Training results

Training Loss	Epoch	Step	Validation Loss
1.0783	0.1354	30	1.0955
0.716	0.2707	60	0.7190
0.6167	0.4061	90	0.6266
0.6226	0.5415	120	0.5929
0.5665	0.6768	150	0.5737
0.5834	0.8122	180	0.5621
0.5931	0.9475	210	0.5549
0.5431	1.0829	240	0.5496
0.5678	1.2183	270	0.5458
0.5336	1.3536	300	0.5425
0.5292	1.4890	330	0.5403
0.5627	1.6244	360	0.5384
0.5493	1.7597	390	0.5374
0.5154	1.8951	420	0.5367

Framework versions

• PEFT 0.12.0
• Transformers 4.44.2
• Pytorch 2.4.0+cu124
• Datasets 2.21.0
• Tokenizers 0.19.1