phi3.5-hugcoder

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: 0.6832

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: 0.0003
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 16
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.1
• training_steps: 100

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.1	10	0.5937
No log	0.2	20	0.4756
No log	0.3	30	0.4394
No log	0.4	40	0.4419
0.5311	0.5	50	0.4827
0.5311	0.6	60	0.5897
0.5311	0.7	70	0.5495
0.5311	0.8	80	0.6268
0.5311	0.9	90	0.6744
0.1769	1.0	100	0.6832

Framework versions

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