master

MMuneeb Khanlibentry: fast-path cache and O(1) set lookups to reduce dispatch overhead

22e9219d创建于 5月20日历史提交

文件	最后提交记录	最后更新时间
models	Add FlagGems v4.2.0	4 个月前
__init__.py	Add FlagGems v4.2.0	4 个月前
code_cache.py	Add FlagGems v4.2.0	4 个月前
code_utils.py	Add FlagGems v4.2.0	4 个月前
codegen_config_utils.py	Add FlagGems v4.2.0	4 个月前
libentry.py	libentry: fast-path cache and O(1) set lookups to reduce dispatch overhead	2 个月前
limits.py	Add FlagGems v4.2.0	4 个月前
pointwise_dynamic.py	[FlagGems] fix the error like: 1. AttributeError: 'dict' object has no attribute 'total_memory' (test_pointwise_dynamic.py) 2. AttributeError: module 'torch.cpu' has no attribute 'get_device_capability' (test_pointwise_dynamic.py) 3. AssertionError: Torch not compiled with CUDA enabled (test_quant.py) 4. TypeError: unsupported operand type(s) for //: 'int' and 'NoneType' (test_general_reduction_ops.py)	4 个月前
random_utils.py	[FlagGems] fix errs like module 'torch.cpu' has no attribute 'default_generators'	4 个月前
shape_utils.py	Add FlagGems v4.2.0	4 个月前
tensor_wrapper.py	Add FlagGems v4.2.0	4 个月前
triton_driver_helper.py	Add FlagGems v4.2.0	4 个月前
triton_lang_extension.py	Add FlagGems v4.2.0	4 个月前
triton_lang_helper.py	Add FlagGems v4.2.0	4 个月前
type_utils.py	Add FlagGems v4.2.0	4 个月前