CodePhi-3-mini-4k-instruct-pythonAPPSLORA3k

This model is a fine-tuned version of microsoft/Phi-3-mini-4k-instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.6463

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: 5e-06
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 200
• training_steps: 3000

Training results

Training Loss	Epoch	Step	Validation Loss
0.664	0.0667	200	0.7185
0.642	0.1333	400	0.6786
0.6163	0.2	600	0.6618
0.5852	0.2667	800	0.6543
0.5608	0.3333	1000	0.6519
0.6224	0.4	1200	0.6499
0.5984	0.4667	1400	0.6484
0.5635	0.5333	1600	0.6474
0.5687	0.6	1800	0.6468
0.6199	0.6667	2000	0.6465
0.5785	0.7333	2200	0.6463
0.594	0.8	2400	0.6463
0.5622	0.8667	2600	0.6463
0.5791	0.9333	2800	0.6462
0.5964	1.0	3000	0.6463

Framework versions

• PEFT 0.11.0
• Transformers 4.40.2
• Pytorch 2.3.0+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1