LoRA-Q-30B-R8

This model is a fine-tuned version of Qwen/Qwen3-Coder-30B-A3B-Instruct on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.9320

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: 2e-05
• train_batch_size: 4
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_ratio: 0.03
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss
1.0568	0.9231	30	1.0044
0.9918	1.8308	60	0.9818
1.1982	2.7385	90	0.9649
1.4336	3.6462	120	0.9525
1.394	4.5538	150	0.9441
1.2829	5.4615	180	0.9385
1.0959	6.3692	210	0.9352
1.0356	7.2769	240	0.9334
0.961	8.1846	270	0.9323
0.9198	9.0923	300	0.9320
0.901	10.0	330	0.9321

Framework versions

• PEFT 0.17.1
• Transformers 4.57.1
• Pytorch 2.8.0+cu126
• Datasets 4.0.0
• Tokenizers 0.22.1