yusei926
/

qwen3-4b-sft-merged-v2-20260207-1148

Text Generation

structured-output

text-generation-inference

Model card Files Files and versions

qwen3-4b-sft-merged-v2-20260207-1148

Merged 16-bit model from unsloth/Qwen3-4B-Instruct-2507 (SFT only).

Training

SFT: LR=5e-05, Epochs=2, LoRA r=64
DPO: disabled

Downloads last month: 7

Safetensors

Model size

4B params

Tensor type

BF16

·

Model tree for yusei926/qwen3-4b-sft-merged-v2-20260207-1148

Base model

Qwen/Qwen3-4B-Instruct-2507

Finetuned

unsloth/Qwen3-4B-Instruct-2507

Finetuned

(388)

this model

Datasets used to train yusei926/qwen3-4b-sft-merged-v2-20260207-1148