[Won't Merge] Test auto-round-lib

.azure-pipelines/scripts/ut/3x/run_3x_pt_fp8.sh

@@ -13,7 +13,8 @@ echo "##[section]import check pass"
 echo "##[group]set up UT env..."
 export LD_LIBRARY_PATH=/usr/local/lib/:$LD_LIBRARY_PATH
 sed -i '/^intel_extension_for_pytorch/d' /neural-compressor/test/3x/torch/requirements.txt
-sed -i '/^auto_round/d' /neural-compressor/test/3x/torch/requirements.txt
+# Install auto_round_lib for auto-round-hpu UTs
+# sed -i '/^auto_round/d' /neural-compressor/test/3x/torch/requirements.txt
 cat /neural-compressor/test/3x/torch/requirements.txt
 pip install -r /neural-compressor/test/3x/torch/requirements.txt
 pip install pytest-cov
@@ -31,7 +32,7 @@ mkdir -p ${LOG_DIR}
 ut_log_name=${LOG_DIR}/ut_3x_pt_fp8.log
 pytest --cov="${inc_path}" -vs --disable-warnings --html=report_1.html --self-contained-html torch/quantization/weight_only/test_load.py 2>&1 | tee -a ${ut_log_name}
 pytest --cov="${inc_path}" -vs --disable-warnings --html=report_2.html --self-contained-html torch/quantization/weight_only/test_rtn.py 2>&1 | tee -a ${ut_log_name}
-# pytest --cov="${inc_path}" -vs --disable-warnings --html=report_3.html --self-contained-html torch/quantization/weight_only/test_autoround.py 2>&1 | tee -a ${ut_log_name}
+pytest --cov="${inc_path}" -vs --disable-warnings --html=report_3.html --self-contained-html torch/quantization/weight_only/test_autoround.py 2>&1 | tee -a ${ut_log_name}
 
 # Below folder contains some special configuration for pytest so we need to enter the path and run it separately
 cd /neural-compressor/test/3x/torch/algorithms/fp8_quant

test/3x/torch/quantization/weight_only/test_rtn.py

@@ -320,6 +320,9 @@ def test_rtn_with_quantize_API(self):
         ],
     )
     def test_conv1d(self, bits, use_sym, group_size, group_dim):
+        # skip 8-True--1-1
+        if bits == 8 and use_sym and group_size == -1 and group_dim == 1:
+            pytest.skip(f"Not stable on HPU")
         model = ModelConv1d().to(device)
         input = torch.randn(1, 32).to(device)
         quant_config = RTNConfig(

test/3x/torch/requirements.txt

@@ -1,4 +1,4 @@
-auto_round
+auto-round-lib==0.4.4
 deepspeed @ git+https://github.com/HabanaAI/[email protected]
 expecttest
 intel_extension_for_pytorch