branch: master
Commits on master
- 9e958f2 Ptx simplify [pr] (#7877) 1 year ago
- e9c681c fix missing final rewrite in viz (#7883) 1 year ago
- a49a7c4 Improved mod folding (#7887) 1 year ago
- 5d92efb [BUGFIX] Tensor([]).data() (#7884) 1 year ago
- ac57d82 test_tiny on real NV/CUDA/AMD/HIP (#7886) 1 year ago
- 06a28d8 delete extra dtype check in uop const [pr] (#7880) 1 year ago
- 31337b4 cleanup Embedding call [pr] (#7869) 1 year ago
- ad9df26 add test for inconsistent behavior in float to int casting (#7870) 1 year ago
- 6b8a657 cleanup group_realizes [pr] (#7878) 1 year ago
- 5aee78a fix uop swizzle on BUFFER, new tests (#7875) 1 year ago
- 5d28a20 make tinychat local (#7871) 1 year ago
- 22d5def download llama3 70B (#7868) 1 year ago
- 6a8be3c don't change lazy state in schedule [pr] (#7867) 1 year ago
- 28e83e6 least controversial (#7863) 1 year ago
- 8c3d318 bottom up rewrite fixes substitute [pr] (#7862) 1 year ago
- 54d8f75 vectorized define_acc does not seem to get used (#7858) 1 year ago
- 40be917 move swizzle upats to ops, prereq for swizzle tc [pr] (#7861) 1 year ago
- 27a6cd7 cleanup swizzle upats [pr] (#7860) 1 year ago
- 5b2c03e defer realize folding to kernel splitting [pr] (#7849) 1 year ago
- 144e9f0 viz is local, new test, and new quantize [pr] (#7859) 1 year ago
- d43613e refactor image cast folding [pr] (#7852) 1 year ago
- c07daf4 move attention upcast (#7830) 1 year ago
- 5c5b1b9 less flaky benchmarks (#7855) 1 year ago
- 3b26e51 Tensor.cummax (#7854) 1 year ago
- fb10ea5 typedef bf16 amd (#7850) 1 year ago
- a352a69 simplify group_for_reduces in get_index [pr] (#7851) 1 year ago
- af5d77f move sint_to_uop from view.py to ops.py [pr] (#7848) 1 year ago
- f6d1201 variable_to_uop -> sint_to_uop [pr] (#7847) 1 year ago
- 40d7535 clean up DTYPES_DICT [pr] (#7845) 1 year ago
- 4453ab5 use ceildiv in View.stride [pr] (#7844) 1 year ago