PTQ-Level Efficiency, QAT-Level Accuracy | 训练后量化的效率,量化感知训练的精度
fix merge error (#325)
wenhuach21dadb95ad14af
diff --git a/auto_round/backend.py b/auto_round/backend.pyindex 6afbf1ef0..66c5e6671 100644--- a/auto_round/backend.py+++ b/auto_round/backend.py@@ -404,7 +404,7 @@ def get_autogptq_infer_linear(backend, bits=4, group_size=128, sym=False):from packaging.version import Version# Import the appropriate QuantLinear based on the version of auto_gptq- if Version(version) <= Version("0.7.1"):+ if Version(version) < Version("0.7.2"):QuantLinear = dynamically_import_QuantLinear(use_triton=use_triton,desc_act=False,