branch: master
Commits on master
- 6090cbe Try to open llvm first when opening metal (#8949) 1 year ago
- 67b70e4 Fix incorrect __del__ (#8950) 1 year ago
- 9ed2d0d refactor into subactions (#8946) 1 year ago
- 133caca Autogen webgpu dawn, removing wgpu-py dependency (f16 support part 1) (#8646) 1 year ago
- dbda72f hotfix: raise line limit to 11200 for new webgpu backend 1 year ago
- b1e1319 ci speed on the enterprise plan [pr] (#8942) 1 year ago
- 3b67712 [bounty] Fix LLVM=1 NO_DEVECTORIZE=1 python3 test/test_ops.py TestOps.test_strided_conv2d_simple (#8937) 1 year ago
- f542428 failing test for the devectorize [pr] (#8940) 1 year ago
- ee1a0fb am_smi: print device name (#8939) 1 year ago
- a092b63 Tuple -> tuple, List -> list [pr] (#8936) 1 year ago
- d5183e1 remove unneeded annotation import (#8934) 1 year ago
- 00d72a5 setitem isinstance cleanup [pr] (#8932) 1 year ago
- 81e2411 hotfix: save 1 line (#8931) 1 year ago
- eb1144b hotfix: only check current graph when excluding nodes in viz (#8930) 1 year ago
- 3cc0508 llvm no devectorize, the right way (#8901) 1 year ago
- 8b16c65 add compile3 benchmark [pr] (#8929) 1 year ago
- 79fb5c6 hotfix: test_shard_no_recompile shouldn't rely on schedule order [pr] (#8928) 1 year ago
- 1249e8d objc fast msg, try 2 [pr] (#8927) 1 year ago
- 86feb98 am: add support for 7600 (#8910) 1 year ago
- ae45826 hotfix: GRAPH_ONE_KERNEL + fix timing 1 year ago
- 1c53e8b Revert "objc fast msg (#8922)" (#8926) 1 year ago
- c3f99a7 objc fast msg (#8922) 1 year ago
- a2e7e49 prepickle scheduler process replay [pr] (#8924) 1 year ago
- 89d7480 hotfix: don't sink views [pr] (#8923) 1 year ago
- 0cbb7d7 hotfix: metal has known sync issue 1 year ago
- a8e54df benchmark single kernel launch (#8921) 1 year ago
- 3e082d4 add float4 support to LLVM (#8920) 1 year ago
- b05c536 cleanup some llvm stuff [pr] (#8919) 1 year ago
- 44e0eab Fix AttributeError occurring after ValueError in _apply_uop (#8905) 1 year ago
- 30695da remove Tensor._to_const_val (#8917) 1 year ago