branch: master
Commits on master
- 63ba2d0 uops dfs cleanup (#5147) 1 year ago
- 6841ea3 don't allow duplicate variables (#5148) 1 year ago
- cc7fafc sink folding rule [run_process_replay] (#5145) 1 year ago
- fa78755 Add new patterns to unfold division (#5139) 1 year ago
- c4fdb9c second iteration on verify_lazyop (#5140) 1 year ago
- dade767 validate llama3 output only with model "LLaMA-3/8B-SF-DPO" (#5138) 1 year ago
- 981afb1 safely fold NEG in lazy.py (#5135) 1 year ago
- 7948b05 fix uneven shard with shrink and pad args on sharded axis (#5131) 1 year ago
- 18e70de verify_lazyop (#5124) 1 year ago
- fe707bc hotfix: don't use is for comparing dtype (#5128) 1 year ago
- dfa562d DEFINE_ACC takes UOps.CONST in vin instead of arg (#4975) 1 year ago
- d84beaa tiny profiler cleanups (#5126) 1 year ago
- 4a7d403 cleanup test_multitensor (#5118) 1 year ago
- c0ba5e0 multi copy_to_device return the copy on same device if possible (#5117) 1 year ago
- b563cd5 linearizer: change globals to merge into left axis/gridDims.x first (#5033) 1 year ago
- 69f116a nv/amd profiler (#4718) 1 year ago
- 64a3b79 simplify render_ops ctx [run_process_replay] (#5116) 1 year ago
- 28bf8d8 test_linearizer with multi output ASTs (#5115) 1 year ago
- ee0c6df build Tensor._tri with movements only (#5110) 1 year ago
- 20fabd8 update Tensor.triu and Tensor.tril (#5109) 1 year ago
- 8f6ae84 minor cleanup of conv_transpose2d (#5108) 1 year ago
- 33211f3 fix desc in tqdm (#5107) 1 year ago
- 055e616 cleanup mnist data load in beautiful_mnist (#5106) 1 year ago
- 5516b79 hotfix append colon space to tqdm set_description (#5105) 1 year ago
- e356807 tinytqdm.set_description and tinytrange (#5101) 1 year ago
- 8080298 s/tinytqdm/tqdm (#5103) 1 year ago
- 9f87512 small changes from lowerer. [run_process_replay] [no_assert] (#5102) 1 year ago
- e468601 update llama attention casting (#5096) 1 year ago
- ca02122 fix attention to always return in the same dtype as input (#5100) 1 year ago
- 2dcef5a hcq spec (#5081) 1 year ago