Commit Graph

Select branches

Hide Pull Requests

geohot/updates

gh-pages

master

remove_anchors

#1

#2

#3

v0.10.0

v0.10.1

v0.10.2

v0.10.3

v0.11.0

v0.12.0

v0.13.0

v0.4.0

v0.5.0

v0.6.0

v0.7.0

v0.8.0

v0.9.0

v0.9.1

v0.9.2

Mono Color

• 5e1221845f refactor schedule edges to tuple[LazyBuffer, ...] [run_process_replay] (#6797) qazalandGitHub 2024-09-29 11:34:39 +08:00
• 68e59eb3f5 update mlperf-logging to 4.1.0-rc3 (#6796) chenyuandGitHub 2024-09-28 21:45:37 -04:00
• dab05ff070 match dataclass.replace in UOp.replace [run_process_replay] (#6792) qazalandGitHub 2024-09-28 16:28:49 +08:00
• 494b20e886 bert BS back to 54 (#6791) chenyuandGitHub 2024-09-27 22:16:05 -04:00
• 572d77d1d9 bert script delete eval data after eval (#6790) chenyuandGitHub 2024-09-27 20:54:00 -04:00
• f9c8e144ff chmod +x mlperf bert script for red (#6789) chenyuandGitHub 2024-09-27 11:27:32 -04:00
• d3a387be63 [MLPerf] Prepare openimages dataset script (#6747) Francis LataandGitHub 2024-09-27 11:13:56 -04:00
• bc82f8c5be use where in dropout (#6758) chenyuandGitHub 2024-09-27 11:11:43 -04:00
• 76b3c1e818 add all realized Buffers to schedule graph edges [run_process_replay] (#6786) qazalandGitHub 2024-09-27 19:25:51 +08:00
• 568c97f7a2 add UOp.define_global [run_process_replay] (#6787) qazalandGitHub 2024-09-27 19:24:03 +08:00
• b95f47784a qcom sleep when sync (#6785) nimlgenandGitHub 2024-09-27 19:14:10 +08:00
• fb3fe6f39b better VIZ (#6781) qazalandGitHub 2024-09-27 18:38:31 +08:00
• 2fc26890c9 default BS=9 in handcode_opt bert (#6783) chenyuandGitHub 2024-09-27 04:38:16 -04:00
• 9a3f6f392d llm.c tok/s geohot 2024-09-27 00:46:18 -07:00
• b0e70ab04f llm.c updates geohot 2024-09-27 15:25:59 +08:00
• eaa1e0eeeb rename constant_folder to sym [run_process_replay] (#6780) George HotzandGitHub 2024-09-27 14:54:54 +08:00
• 900b21ef0c viz delete const after fold (#6778) qazalandGitHub 2024-09-27 11:58:01 +08:00
• 94e43dc49a add Buffer.to_uop [run_process_replay] (#6777) qazalandGitHub 2024-09-27 11:41:23 +08:00
• 98a81b36e1 viz table view (#6743) qazalandGitHub 2024-09-27 10:26:46 +08:00
• bea7ed5986 add RUNMLPERF=1 to bert dev_run.sh (#6775) chenyuandGitHub 2024-09-26 11:00:49 -04:00
• c178dc1071 faster uops ci [run_process_replay] (#6774) George HotzandGitHub 2024-09-26 20:15:01 +08:00
• 249af24f18 metal bfloat as cast (#6773) George HotzandGitHub 2024-09-26 19:31:40 +08:00
• ed2f28388f render cast is rewrite rules [run_process_replay] (#6772) George HotzandGitHub 2024-09-26 19:03:31 +08:00
• 3c56aeee70 add Tensor.from_blob (#6765) nimlgenandGitHub 2024-09-26 18:33:19 +08:00
• 14ad47b515 rewrite to use uops if (#6764) George HotzandGitHub 2024-09-26 18:09:09 +08:00
• 7e7184bb13 cleaner ptx match rules [run_process_replay] (#6770) George HotzandGitHub 2024-09-26 17:44:10 +08:00
• 12de203a43 add IGNORE_JIT_FIRST_BEAM to bert scripts (#6769) chenyuandGitHub 2024-09-26 05:38:24 -04:00
• 15cd42cfb9 feat: support TRACEMETA=2 in handcode_opt (#6767) wozeparrotandGitHub 2024-09-26 16:58:29 +08:00
• 5a5fbfa1eb smaller bert script change (#6768) chenyuandGitHub 2024-09-26 04:54:28 -04:00
• abd484a9f7 fix: need numpy for docs and testing (#6766) wozeparrotandGitHub 2024-09-26 16:44:59 +08:00
• 2b899164c6 no numpy (#6751) wozeparrotandGitHub 2024-09-26 16:40:18 +08:00
• 7fca0bc912 use pattern matcher for image [run_process_replay] (#6762) George HotzandGitHub 2024-09-26 15:49:09 +08:00
• 197f8fd986 early uop globals with Buffer (#6753) qazalandGitHub 2024-09-26 15:34:21 +08:00
• e999281502 match_to_scalar (#6761) George HotzandGitHub 2024-09-26 14:50:47 +08:00
• 0c7d34ceb7 did vload do anything? [run_process_replay] (#6760) George HotzandGitHub 2024-09-26 14:46:16 +08:00
• ee4feedb77 delete test_variable_const [run_process_replay] (#6757) qazalandGitHub 2024-09-26 12:27:11 +08:00
• 0424c4967d fix handcode_opt.py for bert (#6756) chenyuandGitHub 2024-09-26 00:20:24 -04:00
• 396c96357b update mlperf bert scripts (#6755) chenyuandGitHub 2024-09-25 23:55:05 -04:00
• 717b394391 remove defaultdict from PatternMatcher [run_process_replay] (#6754) George HotzandGitHub 2024-09-26 11:25:01 +08:00
• 7e73c7b3cc hotfix: bump stable diffusion val distance geohot 2024-09-26 11:15:29 +08:00
• ff880f5be4 hotfix: force_transcendental to fix process replay geohot 2024-09-26 11:13:16 +08:00
• a6a70aa4bd add optional NEG and SUB (#6750) George HotzandGitHub 2024-09-26 10:50:53 +08:00
• 197dbbda0f add UnaryOps.NEG + BinaryOps.SUB so process replay can work geohot 2024-09-26 10:36:33 +08:00
• b199b699ed use shl everywhere (#6744) George HotzandGitHub 2024-09-26 09:59:36 +08:00
• 88160e59b2 gate engine.graph imports [run_process_replay] (#6748) qazalandGitHub 2024-09-26 09:13:49 +08:00
• 12e4a4900a hotfix: missing return in METAL dm benchmark (#6749) qazalandGitHub 2024-09-26 09:12:38 +08:00
• 8a15ccb414 start gc/mem usage tests for buffer schedule [run_process_replay] (#6737) qazalandGitHub 2024-09-26 08:26:31 +08:00
• b629a7998d early assert buffer count limit [run_process_replay] (#6746) qazalandGitHub 2024-09-26 08:24:26 +08:00
• 4ebc9589a6 feat: make buffer (#6745) wozeparrotandGitHub 2024-09-25 18:31:03 +08:00
• c100f3d406 default threefry (#6116) wozeparrotandGitHub 2024-09-25 17:45:13 +08:00
• 992cde05d7 Metal with CDLL instead of py-objc (#6545) 4 people 2024-09-25 17:43:01 +08:00
• cd534dee11 cstyle changes that don't pass process replay (#6734) George HotzandGitHub 2024-09-25 17:33:34 +08:00
• 232edcfd4f cast bool for type verify [run_process_replay] (#6742) George HotzandGitHub 2024-09-25 17:12:16 +08:00
• cb22ef379a truncate consts early (#6741) George HotzandGitHub 2024-09-25 16:49:51 +08:00
• e31552e2e0 qcom reinit queue on exec (#6728) nimlgenandGitHub 2024-09-25 16:08:50 +08:00
• 882339f729 remove parens from neg (#6738) George HotzandGitHub 2024-09-25 15:38:20 +08:00
• 5ad2f95d01 process replay diff stats (#6736) qazalandGitHub 2024-09-25 15:19:56 +08:00
• 56979aa3ed qcom ioctl log levels (#6735) nimlgenandGitHub 2024-09-25 14:59:27 +08:00
• 66af8bb54c use UOp.replace and UOp.define_var in validhack (#6730) chenyuandGitHub 2024-09-25 02:51:34 -04:00
• ff25bfb1b0 conv backward tests in test_simplify_valid_idx (#6727) chenyuandGitHub 2024-09-25 02:51:07 -04:00
• 6c69fec1ef viz more info for rewrite location (#6729) qazalandGitHub 2024-09-25 14:49:40 +08:00
• 39f78619ff cstyle replay [run_process_replay] (#6731) George HotzandGitHub 2024-09-25 14:26:05 +08:00
• e1caa24a92 qcom fix binded queue might be overwritten (#6712) nimlgenandGitHub 2024-09-25 12:45:23 +08:00
• dd575da7ee real minimum cstyle change (#6709) George HotzandGitHub 2024-09-25 12:40:46 +08:00
• e6a1b5aa8f more test_simplify_valid_idx cleanup (#6726) chenyuandGitHub 2024-09-24 23:47:42 -04:00
• 14524eeddc test_image_valid.py -> test_simplify_valid_idx.py (#6724) chenyuandGitHub 2024-09-24 23:32:27 -04:00
• e0d8685c99 test_masked_upcast_wino check device buf_max (#6723) qazalandGitHub 2024-09-25 11:26:53 +08:00
• f45d178a55 hotfix: support JIT_BATCH_SIZE=0, make that the default geohot 2024-09-25 10:36:04 +08:00
• 52e7f1c108 add new model CI geohot 2024-09-25 10:23:06 +08:00
• 76bd4c7d5f advanced setitem (#6262) 4 people 2024-09-25 03:14:59 +01:00
• 3bf25aae78 start work on global buffer count limit [run_process_replay] (#6722) qazalandGitHub 2024-09-25 09:51:56 +08:00
• b0ffe2452b bump line count to 9800 geohot 2024-09-25 09:13:56 +08:00
• 5c240c34aa split validhack into simplify idx and drop valids (#6719) chenyuandGitHub 2024-09-24 09:40:27 -04:00
• cefc3e9382 make all schedules immutable [run_process_replay] (#6718) qazalandGitHub 2024-09-24 21:08:16 +08:00
• 29330014ab give FUZZ_SCHEDULE views a base (#6717) qazalandGitHub 2024-09-24 19:20:37 +08:00
• f0019ad29c bump ci test timeout for test_speed_exec_time (#6715) nimlgenandGitHub 2024-09-24 18:44:09 +08:00
• 1c03fb69c9 viz dedup assert groupby ctx [run_process_replay] (#6714) qazalandGitHub 2024-09-24 18:17:21 +08:00
• 8d75326cb5 do not fold var with min==max (#6713) chenyuandGitHub 2024-09-24 06:16:34 -04:00
• 9e51879019 fix idx setup in image_valid test_openpilot_conv3 (#6710) chenyuandGitHub 2024-09-24 05:49:04 -04:00
• ae3f3fec38 refactor DEFINE_GLOBAL inputs to list [run_process_replay] (#6711) qazalandGitHub 2024-09-24 17:43:24 +08:00
• f932116e05 feat: small things from default_threefry (#6708) wozeparrotandGitHub 2024-09-24 17:00:47 +08:00
• f2700ac58a construct a candidate set to attempt valid idx rewrite (#6706) chenyuandGitHub 2024-09-24 04:12:21 -04:00
• 2be0b26a1f rand only supports single device (#6682) wozeparrotandGitHub 2024-09-24 16:07:44 +08:00
• 75b7627db7 qcom do not recreate memoryviews on updates (#6701) nimlgenandGitHub 2024-09-24 15:36:22 +08:00
• a6078c099f simpler idx rewrite structure in simplify_valid_image_load (#6704) chenyuandGitHub 2024-09-24 03:35:39 -04:00
• d3ed50c769 fix typo in 'Too many resources requested for launch' (#6705) nimlgenandGitHub 2024-09-24 15:33:01 +08:00
• ef7a74bfa0 feat: use /raid/downloads on tinybox (#6702) wozeparrotandGitHub 2024-09-24 15:26:31 +08:00
• ca66b11e07 qcom fix disasm (#6703) nimlgenandGitHub 2024-09-24 15:23:43 +08:00
• a473bf4ba9 do not always update float dims (#6699) nimlgenandGitHub 2024-09-24 14:40:45 +08:00
• 048483ee0b viz fold const nodes and UOp/float4 syntax highlight (#6695) qazalandGitHub 2024-09-24 14:36:59 +08:00
• 4bb1694f49 more tests about bounds of UOp divs (#6700) chenyuandGitHub 2024-09-24 00:41:43 -04:00
• 79aef64d70 update tests in test_image_valid (#6698) chenyuandGitHub 2024-09-24 00:04:21 -04:00
• 568757e087 fix model_eval.py in the mlperf folder searching for bert vocab in the wrong directory (#6649) Anurag LamsalandGitHub 2024-09-24 09:05:44 +05:45
• 4a2fa0b627 clean up apply OptOps.PADTO [run_process_replay] (#6694) chenyuandGitHub 2024-09-23 23:13:50 -04:00
• f703180356 hotfix missed cast in cstyle code_for_workitem (#6693) chenyuandGitHub 2024-09-23 22:18:18 -04:00
• 19c11792fd Flux.1 (#6334) 3 people 2024-09-24 03:08:04 +01:00
• 31b9c74c77 tiny import cleanup and fix typo (#6692) chenyuandGitHub 2024-09-23 21:48:23 -04:00
• 02c0c09fb9 VIZ syntax highlighting and new colors (#6686) qazalandGitHub 2024-09-24 09:41:07 +08:00
• 0ffbd75af8 Refactor TC [run_process_replay] (#6456) 3 people 2024-09-23 22:05:29 -03:00
• b9e6d42a1f Revert "gated native math in OpenCL (#6683)" (#6691) George HotzandGitHub 2024-09-24 08:48:10 +08:00