branch: master
Commits on master
- a3869ff move gpuctypes in tree (#3253) 2 years ago
- bc92c4c onnx Einsum, CumSum, DepthToSpace, SpaceToDepth (#3252) 2 years ago
- e45ffdb cleanup onnx (#3249) 2 years ago
- 168b1f8 Fix hip_matmul gemm in extra (#3241) 2 years ago
- 7feeb11 hip launch speed (#3246) 2 years ago
- cb372b0 add device speed test (#3244) 2 years ago
- d0e116c fix maximum/where Scalar casting (#3194) 2 years ago
- 3628bea fix: big round even rounder round (#3242) 2 years ago
- da5e279 failed test cases for Tensor.round (#3240) 2 years ago
- b0b5eba fix _round in onnx_ops to look more like new Tensor.round (#3239) 2 years ago
- aa0d1b6 hotfix: don't use noqa: E702 that's just dumb 2 years ago
- b92945c hotfix: DEBUG >= 2 for kernels 2 years ago
- a8fbb03 minor hip cleanups (#3237) 2 years ago
- 3205fd8 fix cuda device var rewrite (#3233) 2 years ago
- ed8a327 hip mutex signal (#3234) 2 years ago
- 47f9887 hip events work (#3229) 2 years ago
- de7a3a5 save lines in llvm (#3231) 2 years ago
- 83d6142 reduce lines (#3230) 2 years ago
- afeadbe touch up Tensor.round and Tensor.neg (#3228) 2 years ago
- 0e103b4 implement Tensor.round (#3225) 2 years ago
- 8420538 fix neg logical_not inconsistencies (#3222) 2 years ago
- e2e4632 LoadOps SYNC (#3223) 2 years ago
- 2f4b3ab shard and to should preserve requires_grad (#3224) 2 years ago
- 23b084e add device name to device, all are constructed (#3221) 2 years ago
- 91a1b2b the runner does the build (#3220) 2 years ago
- 9e5409b cifar move GlobalCounters.reset() before shard (#3217) 2 years ago
- 595d05a test: fix test_linearizer to use the correct tc_dims (#3218) 2 years ago
- 3c179cc cifar only shuffle data at epoch start (#3216) 2 years ago
- 4a07ea3 buffer options should work (#3211) 2 years ago
- a06f34a remove dead lines from cstyle (#3212) 2 years ago