Files
Qwen2.5-0.5B-Instruct-quant…/recipe.yaml
ModelHub XC 7b3fb4d66e 初始化项目,由ModelHub XC社区提供模型
Model: RedHatAI/Qwen2.5-0.5B-Instruct-quantized.w8a8
Source: Original Platform
2026-04-11 11:34:55 +08:00

10 lines
192 B
YAML

quant_stage:
quant_modifiers:
GPTQModifier:
sequential_update: true
dampening_frac: 0.01
ignore: [lm_head]
scheme: W8A8
targets: Linear
observer: mse