Albus-Chen/Qwen3-4B_General-Domain-RAG-SFT

SFT_General_Domain

This model is a fine-tuned version of Qwen/Qwen3-4B on the HTX_RAG_General_train dataset. It achieves the following results on the evaluation set:

• Loss: 0.9163

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: 1e-05
• train_batch_size: 16
• eval_batch_size: 8
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 2
• total_train_batch_size: 128
• total_eval_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 0.1
• num_epochs: 5.0

Training results

Training Loss	Epoch	Step	Validation Loss
0.6974	0.4570	250	0.7169
0.6683	0.9141	500	0.6801
0.5631	1.3711	750	0.6760
0.5473	1.8282	1000	0.6649
0.372	2.2852	1250	0.7165
0.3771	2.7422	1500	0.7131
0.2403	3.1993	1750	0.8032
0.2243	3.6563	2000	0.8147
0.1584	4.1133	2250	0.9107
0.1448	4.5704	2500	0.9161

Framework versions

• Transformers 4.57.1
• Pytorch 2.9.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.1