SFT-Qwen2.5-Coder-1.5B_v1.1

This model is a fine-tuned version of Qwen/Qwen2.5-Coder-1.5B-Instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.9464

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: 2
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.PAGED_ADAMW_8BIT with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss
1.1739	0.2759	20	1.1330
0.8905	0.5517	40	1.0622
0.9773	0.8276	60	1.0305
0.8092	1.0966	80	1.0049
0.7883	1.3724	100	0.9816
0.7641	1.6483	120	0.9676
0.79	1.9241	140	0.9591
0.8975	2.1931	160	0.9538
0.7537	2.4690	180	0.9526
0.6484	2.7448	200	0.9464

Framework versions

• PEFT 0.18.0
• Transformers 4.57.2
• Pytorch 2.9.0+cu126
• Datasets 4.4.1
• Tokenizers 0.22.1