branch: master
Commits on master
- b5fd160 hotfix: increase rtol on simple_matmul 2 years ago
- 4feaaa2 ensure shrink is valid (#2717) 2 years ago
- a43bc78 fix dtypes helpers for integers (#2716) 2 years ago
- bc3c4ce cuda set context before sync (#2715) 2 years ago
- 8d206f6 fix help message (#2705) 2 years ago
- 59ab367 faster mixtral + green for new kernels (#2701) 2 years ago
- 2ee6f68 simpler einsum (#2700) 2 years ago
- b01e390 mixtral touch up: two lines 2 years ago
- b398218 Mixtral Example (#2691) 2 years ago
- 0fd4425 bf16 fix + cleanups from mixtral (#2698) 2 years ago
- 7fbebb3 Implement einsum (#2686) 2 years ago
- 181b097 slightly better extra/to_movement_ops dedups (#2695) 2 years ago
- ef18d79 remove noop from to_movement_ops (#2693) 2 years ago
- 2d0e38e fix jit input_rawbuffers check wrt consts (#2689) 2 years ago
- 67ff2b2 Formatted test_indexing (#2688) 2 years ago
- 1e7823e combine GROUP and GROUPTOP to a single block (#2687) 2 years ago
- 0fb1d47 two linearizer fuzzer failed test case for webgpu (#2685) 2 years ago
- fae5394 validate llama output (#2681) 2 years ago
- 182d067 Update yolov3.py (#2680) 2 years ago
- 73b067f Bitcast p2 bfloat16 tests + clang fix (#2635) 2 years ago
- a29538a green more dtypes tests (#2656) 2 years ago
- 4164d0e multitensor start (#2676) 2 years ago
- 4b01839 support vals on WebGPU, run more tests (#2668) 2 years ago
- d02ff21 enable test_index and test_advancedindex (#2648) 2 years ago
- 00d9eda FROM -> COPY, move vars_from_ast (#2675) 2 years ago
- 51af993 fix fuzz_linearizer using new device Buffer (#2674) 2 years ago
- 650117a split large jit into several graphs (#2650) 2 years ago
- 29f2653 add graph (#2670) 2 years ago
- 539b00a move llama getenv("JIT") from models to examples (#2671) 2 years ago
- fd21ece reduce gpt2 kernel count in test_real_world (#2663) 2 years ago