PTQ-Level Efficiency, QAT-Level Accuracy | 训练后量化的效率,量化感知训练的精度
test: skip HPU FP8 test with transformers 5.16.x (#2247)
Wang, Changc3a74f3077cf
diff --git a/test/unit/test_hpu/test_quant_fp8.py b/test/unit/test_hpu/test_quant_fp8.pyindex 274e47905..fc9737dd7 100644--- a/test/unit/test_hpu/test_quant_fp8.py+++ b/test/unit/test_hpu/test_quant_fp8.py@@ -1,13 +1,21 @@import osimport shutil+from test.helpers import transformers_versionimport pytestimport torch+from packaging import versionfrom auto_round import AutoRoundMODEL_LIST = (- "Qwen/Qwen3-0.6B-FP8",+ pytest.param(+ "Qwen/Qwen3-0.6B-FP8",+ marks=pytest.mark.skipif(+ version.parse("5.16.0") <= transformers_version < version.parse("5.17.0"),+ reason="fails with transformers 5.16.x",+ ),+ ),"Qwen/Qwen3-0.6B",)