Commit Graph
10489 Commits
Author SHA1 Message Date
geohot 1a780f2851 error 2025-10-08 19:37:28 +08:00
geohot 42210f5487 failing 2025-10-08 19:36:39 +08:00
George HotzandGitHub 23783681ca Merge branch 'master' into rtoposort 2025-10-08 19:15:38 +08:00
geohot 9759899cb7 actually make the file readable 2025-10-08 19:14:02 +08:00
qazalandGitHub c5a1f9f5f9 no ShapeTrackers in multi.py (#12521)
* switch multi to all movement ops

* inline dvars
2025-10-08 14:04:05 +03:00
George HotzandGitHub 29509a1a57 Merge branch 'master' into rtoposort 2025-10-08 18:48:17 +08:00
geohot 8c3242bb45 disable failing test 2025-10-08 18:47:27 +08:00
geohot b264772ae4 touch up tests + add comments 2025-10-08 18:38:48 +08:00
chenyuandGitHub ee0382ad99 remove ShapeTracker.invert (#12520) 2025-10-08 18:37:34 +08:00
geohot 3f927e7bb3 move to indexing.py 2025-10-08 18:32:21 +08:00
geohot e131a5ef0f mstack/mselect no hacks 2025-10-08 18:20:36 +08:00
chenyuandGitHub d5058427ea remove ShapeTracker.real_size (#12519) 2025-10-08 06:15:29 -04:00
geohot fc01a7cc15 tests 2025-10-08 18:02:58 +08:00
qazalandGitHub 6f26603f06 delete swizzler.py (#12518)
* delete swizzler

* remove merge_views tests

* don't need rewrites_for_views

* apply_rewrites
2025-10-08 13:02:34 +03:00
geohot 04b7b68242 ending ranges fix 2025-10-08 18:00:21 +08:00
geohot f09f46d028 put in rangeify map 2025-10-08 17:32:40 +08:00
qazalandGitHub 7e0b14243e delete grouper and kernelize (#12517)
* delete grouper and kernelize

* +sys.setrecursionlimit
2025-10-08 12:27:26 +03:00
geohot 9e884403e9 don't delete those shape ops 2025-10-08 17:20:10 +08:00
chenyuandGitHub 942022c309 smaller LLAMA_LAYER in Test llama 3 training (#12516)
very slow now
2025-10-08 05:10:51 -04:00
geohot cd8942cc77 tests passing 2025-10-08 17:06:43 +08:00
chenyuandGitHub e701106a64 remove FUSE_ARANGE (#12511)
it was the default already
2025-10-08 04:54:07 -04:00
qazalandGitHub 291a19650b move Kernel dataclass to rangeify (#12510) 2025-10-08 11:30:06 +03:00
qazalandGitHub ad49f8148b switch process_replay to rangeify (#12509) 2025-10-08 11:26:43 +03:00
chenyuandGitHub da1f46ff3f remove RANGEIFY specific test jobs (#12507) 2025-10-08 04:12:04 -04:00
George HotzandGitHub e3f073b957 Merge branch 'master' into rtoposort 2025-10-08 16:04:43 +08:00
geohot 91d60d83e0 FAST 2025-10-08 16:04:33 +08:00
1e567a5cf8 make RANGEIFY=1 the default (#12161)
Co-authored-by: chenyu <[email protected]>
Co-authored-by: Sieds Lykles <[email protected]>
Co-authored-by: qazal <[email protected]>
2025-10-08 03:46:09 -04:00
geohot c9ab0a0e26 pad fix 2025-10-08 15:31:40 +08:00
nimlgenandGitHub 9e7103647d amd: rename cmd_id to sqtt_next_cmd_id (#12503)
* amd: rename cmd_id to sqtt_next_cmd_id

* and typo
2025-10-08 15:16:19 +08:00
George HotzandGitHub b4eb5fcac5 Merge branch 'master' into rtoposort 2025-10-08 15:14:43 +08:00
geohot 985c0c20a9 progress 2025-10-08 14:33:06 +08:00
nimlgenandGitHub 4a756a37d8 amd: support rocm7 (#12502)
* amd: support rocm7

* mock
2025-10-08 14:30:39 +08:00
geohot 5042e0b5af should work 2025-10-08 14:06:51 +08:00
geohot 53eb2af4ce fast rangeify works for mnist 2025-10-08 12:58:40 +08:00
qazalandGitHub 60b6dca5ba update some tests instead of expect_rangeify_fails (#12500)
* update test_clone_doesnt_dedup to use base

* new_flat_buffer passes

* fix test_reorder_expand

* remove the view stuff

* remove that test, we don't want this view const behavior

* test_setitem_becomes_subbuffer is good
2025-10-08 07:42:31 +03:00
geohot e35152a004 work 2025-10-08 12:34:07 +08:00
geohot d823a6d9d5 fast rangeify 2025-10-08 12:19:43 +08:00
qazalandGitHub 84597ed53c early assert for device mistmatched asts in rangeify (#12499)
* early assert for device mistmatched asts in rangeify

* alt also passes
2025-10-08 07:19:36 +03:00
qazalandGitHub 2e19354c1c viz: reorder timeline graphs (#12498)
* viz: reorder timeline graphs

* update test_viz with the new order
2025-10-08 07:10:23 +03:00
George HotzandGitHub d06226b575 fix SPEC and all_tensors iterator (#12496) 2025-10-07 23:18:17 -04:00
qazalandGitHub a7cb80bfab use recursive_property in UOp device (#12477)
* simple failing test with RecursionError

* switch to @recursive_property

* merge 2

* diff
2025-10-08 06:15:05 +03:00
geohot d23324779c rtoposort is fast, can replace rangeify with this 2025-10-08 11:06:12 +08:00
George HotzandGitHub a6d59a0b45 backward_slice to get srcs recursively (#12494)
* change name to backward_slice

* faster check

* clean up comments and names

* comment
2025-10-08 10:31:42 +08:00
chenyuandGitHub eb3bc277b3 remove ASSERT_MIN_STEP_TIME in external_benchmark_openpilot (#12495)
should add for compile3 and compile 3 only
2025-10-07 22:13:42 -04:00
qazalandGitHub 239f9a3029 update viz to not use children [pr] (#12493) 2025-10-08 04:35:01 +03:00
Sieds LyklesandGitHub b465c17b56 Revert "UOp.factor and add chain sorting (#12413)" (#12492)
This reverts commit e74be4a140.
2025-10-08 03:20:23 +02:00
George HotzandGitHub 945cc46475 delete children tracking from uop (#12491)
* delete children tracking from uop

* uop children no longer exists

* no tracked children

* that test is flaky too
2025-10-08 09:04:14 +08:00
nimlgenandGitHub 648e5bb223 hcq: do not raise when fini (#12487)
* hcq: do not raise when fini

* Revert "hcq: do not raise when fini"

This reverts commit 44af5f7d05.

* this way

* runtime is fine

* nn
2025-10-07 23:27:03 +08:00
George HotzandGitHub a2345787b9 parents is faster than sparents (#12490) 2025-10-07 21:31:50 +08:00
George HotzandGitHub 12c4963489 add more rangeify pm tests (#12488) 2025-10-07 05:45:38 -04:00