Views
No views yet
| 项目 | 值 |
|---|---|
| 基座模型 | Qwen/Qwen3.6-27B |
| 量化格式 | EXL3 (ExLlamaV3 v1.0.0) |
| 平均位宽 | 2.00 bpw |
| GPU | NVIDIA A100 |
| 项目 | 值 |
|---|---|
| OS | Ubuntu 20.04.3 LTS |
| Kernel | 5.15.0-56-generic |
| GCC | 12.4.0 (conda-forge) |
| NVIDIA Driver | 570.211.01 |
| NVCC | 12.4.131 |
| Python | 3.10.20 |
| PyTorch | 2.5.1+cu121 |
| Triton | 3.1.0 |
| Flash-Attn | 未安装(triton 正常) |
triton_paged.py 中 if X and not Y and not Z 需拆分中间变量绕过 Triton 3.1 限制bc_attn.py 中 constexprs= 改为 constants=CUDA_HOME 需正确指向 $CONDA_PREFIX,CPLUS_INCLUDE_PATH 设置 targets/x86_64-linux/include1EXL3_PREFER_FA2=1 python convert.py \
2 -i /root/qwen3.6-27b \
3 -o /root/qwen3.6-27b-exl3-2.00bpw \
4 -w /root/temp/qwen-exl3 \
5 -b 2.00