branch: master
Commits on master
- b7b494e no numpy (#3134) 2 years ago
- 6bb65cd fix off-by-one error in st_equal (#3131) 2 years ago
- 44c0591 dtype fmt (#3132) 2 years ago
- 5ec6693 remove np from metal graph (#3129) 2 years ago
- 2ef09ca update test_ptr_ne (#3130) 2 years ago
- e39cd3e update env_vars.md (#3127) 2 years ago
- 537fb8b separate try except blocks in onnx2torch in model benchmark (#3126) 2 years ago
- 0dba34b Fix backward fn for `<` and `==` (#3037) 2 years ago
- db965a0 remove numpy from ops_torch (#3124) 2 years ago
- 1f9aee8 remove numpy from device (#3123) 2 years ago
- ea58246 move fromcpu out of lazy.py (#3122) 2 years ago
- 9634506 hotfix: ptrdtype compare was broken 2 years ago
- 26e0faf make DType a dataclass (#3111) 2 years ago
- c13d51d add device options for tests in multigpu (#3121) 2 years ago
- 79f4627 fix conversation: llama generates token not prob now (#3120) 2 years ago
- 152ef7f minor cleanups of onnx_ops (#3116) 2 years ago
- fb3f8f7 move sample inside jit for beautiful_mnist (#3115) 2 years ago
- a313e63 add Tensor.var (#3114) 2 years ago
- c658aa4 minor cleanup of test_disk_tensor (#3112) 2 years ago
- 9c73d27 cleanup ops_disk type annotation and redundant str cast (#3110) 2 years ago
- a300fea failed test case due to cast resets shapetracker (#3109) 2 years ago
- cf1d0a6 no exceptions in __del__ when module creation is failed in hip/cuda (#3107) 2 years ago
- 12f28ac catch runtime error in search._time_program (#3106) 2 years ago
- f018a55 update NumNode.__hash__ to be hash(self.b) (#3105) 2 years ago
- c3c35f9 flag to profile mixtral - 1.7 tok/s now (#3104) 2 years ago
- e078e2d add half @ half to mac benchmark (#3103) 2 years ago
- ddbdb52 wmma: enable METAL half tensor cores and clean up cstyle (#3095) 2 years ago
- f96fc6e fix gpt2 with empty prompt take 2 (#3102) 2 years ago
- ca46d35 Revert "fix gpt2 with empty prompt" (#3101) 2 years ago
- 1d7f01b fix gpt2 with empty prompt (#3100) 2 years ago