hyperparam-rust-lora-32-4k

This model is a fine-tuned version of deepseek-ai/deepseek-coder-1.3b-base on the generator dataset. It achieves the following results on the evaluation set:

• Loss: 0.4542

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: 0.0002
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 32
• total_train_batch_size: 32
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.05
• lr_scheduler_warmup_steps: 1
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
0.6673	0.59	25	0.5242
0.4766	1.18	50	0.4720
0.4482	1.77	75	0.4582
0.4442	2.37	100	0.4545
0.4384	2.96	125	0.4542

Framework versions

• PEFT 0.10.0
• Transformers 4.39.3
• Pytorch 2.2.1
• Datasets 2.18.0
• Tokenizers 0.15.2