Skip to content

PTQ-Level Efficiency, QAT-Level Accuracy | 训练后量化的效率,量化感知训练的精度

fix: update cleanup logic for quantized model path in tests (#1419)

Xin Heebb534a29558
diff --git a/test/test_cuda/quantization/test_mxfp_nvfp.py b/test/test_cuda/quantization/test_mxfp_nvfp.py
index 11843249d..e5ff3a636 100644
--- a/test/test_cuda/quantization/test_mxfp_nvfp.py
+++ b/test/test_cuda/quantization/test_mxfp_nvfp.py
@@ -84,7 +84,7 @@ class TestAutoRound:
quantization_config["format"] == "nvfp4-pack-quantized"
and quantization_config["config_groups"]["group_0"]["input_activations"]["num_bits"] == 4
), f"Invalid NVFP4 quantization configuration: {quantization_config}"
- shutil.rmtree("./saved", ignore_errors=True)
+ shutil.rmtree(quantized_model_path, ignore_errors=True)
# from vllm import LLM, SamplingParams
# prompts = [
# "The capital of France is",
@@ -122,6 +122,7 @@ class TestAutoRound:
autoround.quantize()
quantized_model_path = self.save_dir
autoround.save_quantized(output_dir=quantized_model_path, inplace=False, format="auto_round")
+ shutil.rmtree(quantized_model_path, ignore_errors=True)
def test_nvfp4_moe_actmax_ar(self, tiny_deepseek_v2_model_path, dataloader):
scheme = "nvfp4"
@@ -136,6 +137,7 @@ class TestAutoRound:
autoround.quantize()
quantized_model_path = self.save_dir
autoround.save_quantized(output_dir=quantized_model_path, inplace=False, format="auto_round")
+ shutil.rmtree(quantized_model_path, ignore_errors=True)
def test_qwen_moe_quant_infer(self, dataloader):
model_name = get_model_path("qwen/Qwen1.5-MoE-A2.7B")