Update for ci (#312)

* Update awq.py

* Update module_utils.py

* Update module_utils.py

* Update base_blockwise_quantization.py

* Update awq.py

* Update base_blockwise_quantization.py

Author: gushiqiao

llmc/compression/quantization/awq.py

@@ -9,7 +9,14 @@
 from llmc.utils.registry_factory import ALGO_REGISTRY
 
 from .base_blockwise_quantization import BaseBlockwiseQuantization
-from .fp8_kernel import weight_dequant, weight_quant
+
+try:
+    from .fp8_kernel import weight_dequant, weight_quant
+except Exception:
+    logger.warning(
+        'import triton error. '
+    )
+
 from .module_utils import (_LLMC_LINEAR_TYPES_, _LLMC_LN_TYPES_,
                            _TRANSFORMERS_LINEAR_TYPES_,
                            _TRANSFORMERS_LN_TYPES_, FakeQuantLinear)

llmc/compression/quantization/base_blockwise_quantization.py

@@ -17,7 +17,14 @@
 from ..blockwise_optimization import BlockwiseOpt
 from .attn_utils import _LLMC_ATTN_MAP_
 from .auto_clip import AutoClipper
-from .fp8_kernel import weight_dequant, weight_quant
+
+try:
+    from .fp8_kernel import weight_dequant, weight_quant
+except Exception:
+    logger.warning(
+        'import triton error. '
+    )
+
 from .hadamard_utils import apply_exact_had_to_linear, get_hadK
 from .module_utils import (_LLMC_LINEAR_TYPES_, _LLMC_LN_TYPES_,
                            _REALQUANT_LINEAR_MAP_, _TRANSFORMERS_LINEAR_TYPES_,

llmc/compression/quantization/module_utils.py

@@ -8,7 +8,12 @@
 from loguru import logger
 from transformers.pytorch_utils import ALL_LAYERNORM_LAYERS
 
-from .fp8_kernel import act_quant, fp8_gemm, weight_dequant, weight_quant
+try:
+    from .fp8_kernel import act_quant, fp8_gemm, weight_dequant, weight_quant
+except Exception:
+    logger.warning(
+        'import triton error. '
+    )
 
 try:
     import fast_hadamard_transform