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

• 396ce6cfc9 clean up graph dedup function [run_process_replay] (#5169) George HotzandGitHub 2024-06-26 15:07:34 -07:00
• 3a04e518ec print_tree UPat +fix (#5132) kormannandGitHub 2024-06-27 00:02:19 +02:00
• 0ba093dea0 hotfix: only validate stable diffusion when using threefry (#5166) chenyuandGitHub 2024-06-26 16:50:38 -04:00
• e4a5870b36 validate stable_diffusion output (#5163) chenyuandGitHub 2024-06-26 16:42:21 -04:00
• 21b225ac45 llama3 download works (#5160) nimlgenandGitHub 2024-06-26 22:45:13 +03:00
• c91b3c4079 shard llama3 on 0 sometimes (#5157) wozeparrotandGitHub 2024-06-26 18:50:57 +00:00
• 294bd1a9ff refactor: name check [run_process_replay] (#5158) Roelof van DijkandGitHub 2024-06-26 20:39:41 +02:00
• 2c80583e14 perf: cache const UOp creation [run_process_replay] (#5156) Roelof van DijkandGitHub 2024-06-26 20:13:14 +02:00
• eda2824cd8 freeze uop [run_process_replay] (#5155) George HotzandGitHub 2024-06-26 10:18:15 -07:00
• e267f3161d Add MLLogger (#5125) Elias WahlandGitHub 2024-06-26 18:23:56 +02:00
• 16405b973a fix hcq sync (#5062) nimlgenandGitHub 2024-06-26 17:50:37 +03:00
• 3604642847 Llama shard axis 0 sometimes (#5123) 4 people 2024-06-26 07:35:25 -07:00
• fd27f19e92 graph tests (#5153) nimlgenandGitHub 2024-06-26 16:31:20 +03:00
• 7b709c3ccd switch tensorcoreoptions to tuple [run_process_replay] (#5143) 3 people 2024-06-26 04:12:53 -07:00
• 6ca7b13ed1 limit pickled objects [run_process_replay] (#5154) qazalandGitHub 2024-06-26 13:51:32 +03:00
• ee4f080a14 rewrite div const [run_process_replay] [no_assert] (#5151) George HotzandGitHub 2024-06-25 20:23:14 -07:00
• 666a9c1448 don't view origin buffer when sharding (#5122) David HouandGitHub 2024-06-25 20:19:09 -07:00
• 89e106686a simpler unmatch [run_process_replay] (#5149) George HotzandGitHub 2024-06-25 19:57:40 -07:00
• c98ca23cb9 test pickle variable (#5150) George HotzandGitHub 2024-06-25 19:49:21 -07:00
• 8fcc41582f make buffer view optional with a flag (#5120) David HouandGitHub 2024-06-25 19:13:20 -07:00
• 63ba2d05d1 uops dfs cleanup (#5147) George HotzandGitHub 2024-06-25 18:51:42 -07:00
• 6841ea3baf don't allow duplicate variables (#5148) George HotzandGitHub 2024-06-25 18:47:29 -07:00
• cc7fafcd8b sink folding rule [run_process_replay] (#5145) George HotzandGitHub 2024-06-25 18:34:44 -07:00
• fa78755f19 Add new patterns to unfold division (#5139) Jhenner TigrerosandGitHub 2024-06-25 20:07:47 -05:00
• c4fdb9c725 second iteration on verify_lazyop (#5140) qazalandGitHub 2024-06-25 09:44:32 +03:00
• dade7677cf validate llama3 output only with model "LLaMA-3/8B-SF-DPO" (#5138) chenyuandGitHub 2024-06-24 20:58:25 -04:00
• 981afb114f safely fold NEG in lazy.py (#5135) qazalandGitHub 2024-06-25 02:40:37 +03:00
• 7948b05738 fix uneven shard with shrink and pad args on sharded axis (#5131) chenyuandGitHub 2024-06-24 16:55:50 -04:00
• 18e70deec3 verify_lazyop (#5124) qazalandGitHub 2024-06-24 23:45:35 +03:00
• fe707bc968 hotfix: don't use is for comparing dtype (#5128) qazalandGitHub 2024-06-24 21:12:34 +03:00
• dfa562dbc1 DEFINE_ACC takes UOps.CONST in vin instead of arg (#4975) 3 people 2024-06-24 11:25:33 -05:00
• d84beaa6dd tiny profiler cleanups (#5126) nimlgenandGitHub 2024-06-24 17:02:31 +03:00
• 4a7d403777 cleanup test_multitensor (#5118) chenyuandGitHub 2024-06-23 20:54:22 -04:00
• c0ba5e0dfb multi copy_to_device return the copy on same device if possible (#5117) chenyuandGitHub 2024-06-23 20:25:56 -04:00
• b563cd52ed linearizer: change globals to merge into left axis/gridDims.x first (#5033) Francis LamandGitHub 2024-06-23 15:53:15 -07:00
• 69f116a7e1 nv/amd profiler (#4718) nimlgenandGitHub 2024-06-23 17:10:12 +03:00
• 64a3b7931e simplify render_ops ctx [run_process_replay] (#5116) qazalandGitHub 2024-06-23 16:56:32 +03:00
• 28bf8d86d8 test_linearizer with multi output ASTs (#5115) qazalandGitHub 2024-06-23 15:41:24 +03:00
• ee0c6dfc15 build Tensor._tri with movements only (#5110) chenyuandGitHub 2024-06-23 00:07:36 -04:00
• 20fabd8a5b update Tensor.triu and Tensor.tril (#5109) chenyuandGitHub 2024-06-22 21:59:50 -04:00
• 8f6ae84e4a minor cleanup of conv_transpose2d (#5108) chenyuandGitHub 2024-06-22 21:31:47 -04:00
• 33211f356b fix desc in tqdm (#5107) chenyuandGitHub 2024-06-22 19:00:38 -04:00
• 055e616302 cleanup mnist data load in beautiful_mnist (#5106) chenyuandGitHub 2024-06-22 18:31:51 -04:00
• 5516b790ad hotfix append colon space to tqdm set_description (#5105) chenyuandGitHub 2024-06-22 18:09:14 -04:00
• e356807696 tinytqdm.set_description and tinytrange (#5101) chenyuandGitHub 2024-06-22 14:45:06 -04:00
• 8080298739 s/tinytqdm/tqdm (#5103) chenyuandGitHub 2024-06-22 14:18:26 -04:00
• 9f875123b6 small changes from lowerer. [run_process_replay] [no_assert] (#5102) George HotzandGitHub 2024-06-22 11:09:35 -07:00
• e468601226 update llama attention casting (#5096) chenyuandGitHub 2024-06-22 10:57:17 -04:00
• ca021229e4 fix attention to always return in the same dtype as input (#5100) chenyuandGitHub 2024-06-22 10:34:57 -04:00
• 2dcef5a0d7 hcq spec (#5081) nimlgenandGitHub 2024-06-22 15:32:12 +03:00
• 8bd6cb9511 update llama model RMSNorm casting (#5095) chenyuandGitHub 2024-06-21 23:02:04 -04:00
• 0c857ae2d6 some onnx_ops cleanups (#5094) chenyuandGitHub 2024-06-21 22:01:32 -04:00
• f4a041af16 Simplify graph_dedup [run_process_replay] (#5084) kormannandGitHub 2024-06-21 21:12:30 +02:00
• 00593d6095 clean the long lines in avg_pool2d and max_pool2d (#5091) chenyuandGitHub 2024-06-21 14:46:56 -04:00
• a971dc6218 argmax(axis=None) is argmax.flatten().argmax(0) (#5090) chenyuandGitHub 2024-06-21 14:17:10 -04:00
• 166a2b19b5 fix reduce axis of 0d tensors (#5089) chenyuandGitHub 2024-06-21 13:51:40 -04:00
• 3ff048b68c type annotate reduce axis in tensor.py (#5088) chenyuandGitHub 2024-06-21 13:06:10 -04:00
• 36b4a492a1 explicitly check getitem indices can have at most one ellipsis (#5087) chenyuandGitHub 2024-06-21 12:33:18 -04:00
• f1e758bacb graph fuzzer (#5082) nimlgenandGitHub 2024-06-21 18:47:23 +03:00
• 5717a54b28 don't use Tensor.empty in kernel opts tests (#5086) qazalandGitHub 2024-06-21 18:41:03 +03:00
• 8aa786232d docs for running process replay locally (#5083) qazalandGitHub 2024-06-21 16:55:08 +03:00
• fb1bf48cfe io_uring for copies from disk (#5035) nimlgenandGitHub 2024-06-21 11:36:51 +03:00
• b69afc67d8 tinybox docs typo geohot 2024-06-20 17:58:40 -07:00
• 6bc5e5f41c start tinybox docs geohot 2024-06-20 17:04:45 -07:00
• f6d6760f71 don't cast tuple to list before creating Tensor (#5071) chenyuandGitHub 2024-06-20 13:32:56 -04:00
• 97f1347dd9 fix check_process_replay for special characters (#5072) qazalandGitHub 2024-06-20 20:23:29 +03:00
• 6f6b3b10c9 import from uops, not linearizer (#5064) George HotzandGitHub 2024-06-20 08:08:44 -07:00
• 50700171ef minor cleanup to reshape arg handling (#5070) chenyuandGitHub 2024-06-20 10:27:27 -04:00
• f4355d0f1b check Tensor.permute input arg is a valid permutation (#5069) chenyuandGitHub 2024-06-20 10:01:28 -04:00
• 24c89a2a33 move assert_equiv_uops to helpers + use == for dtypes (#5067) qazalandGitHub 2024-06-20 16:39:34 +03:00
• e8f39fcaaa check arg to Tensor.flip can appear only once (#5068) chenyuandGitHub 2024-06-20 09:33:42 -04:00
• 55e02cdd84 generic gate folding (#5061) qazalandGitHub 2024-06-20 16:10:08 +03:00
• bdca2da2be typannos (#5059) kormannandGitHub 2024-06-20 15:02:31 +02:00
• 5f7edc7a46 minor cleanup getting output shape of _pool (#5065) chenyuandGitHub 2024-06-20 09:00:48 -04:00
• a6a5dba637 Revert "UPat for has_valid in load/store (#5052)" (#5056) qazalandGitHub 2024-06-19 20:53:36 +03:00
• ee01e464e3 use process replay as a diff creator (#4903) qazalandGitHub 2024-06-19 18:17:31 +03:00
• 99fc275c27 UPat line savings [run_process_replay] (#5053) qazalandGitHub 2024-06-19 12:43:20 +03:00
• 71194df1da UPat for has_valid in load/store [run_process_replay] (#5052) qazalandGitHub 2024-06-19 12:20:22 +03:00
• 996788358d minor change to gelu (#5048) chenyuandGitHub 2024-06-18 22:26:56 -04:00
• 4c7e316ded update pylint for ops_python (#5046) chenyuandGitHub 2024-06-18 20:15:34 -04:00
• acb715c64c fix: llama3 special tokens (#5045) wozeparrotandGitHub 2024-06-19 00:08:44 +00:00
• a8e9307e0b pylint runtime/ and shape/ (#5044) chenyuandGitHub 2024-06-18 19:48:18 -04:00
• cc2be9064f fix out of bound python list into numpy array (#5043) chenyuandGitHub 2024-06-18 18:05:21 -04:00
• 4e5add4d01 move test_tqdm to test/unit/ (#5042) chenyuandGitHub 2024-06-18 17:41:39 -04:00
• 2b2488f2e2 revert creating Tensor from a list without numpy (#5041) chenyuandGitHub 2024-06-18 17:31:22 -04:00
• e2c5054bdd update resnet.load_from_pretrained (#5040) chenyuandGitHub 2024-06-18 16:29:22 -04:00
• a3ed4176c8 use tinytqdm in active tests and examples (#5038) chenyuandGitHub 2024-06-18 16:01:19 -04:00
• fe332464d2 src->vin [run_process_replay] (#5036) kormannandGitHub 2024-06-18 21:23:49 +02:00
• f171006ded Should this symbolic test fail? (#4501) 4 people 2024-06-18 15:21:26 -04:00
• 7c3b877216 rename uop [run_process_replay] (#5031) kormannandGitHub 2024-06-18 20:34:05 +02:00
• dc942bf1f6 jit sampling functionn in test_randomness.test_multinomial (#5034) chenyuandGitHub 2024-06-18 14:21:05 -04:00
• f31ef11537 Better default hparams for large BS (#5030) Elias WahlandGitHub 2024-06-18 17:13:06 +02:00
• 8d33998e0d [run_process_replay] linearizer: fix get_grouping_dims to respect global/local max (#4855) Francis LamandGitHub 2024-06-18 06:51:27 -07:00
• 7842559952 simplification of exp2 (#5023) joeshmoe0112358andGitHub 2024-06-18 06:51:16 -07:00
• acc8f5e30e print_tree for uops (#5028) kormannandGitHub 2024-06-18 15:36:14 +02:00
• c8cd6e725c Remove BinaryOps.SUB. Replace SUB by ADD and NEG in all tests. Regenerate dataset (#4977) 3 people 2024-06-18 06:06:13 -07:00
• 620fa6e5a2 check Tensor.reshape can have at most one -1 (#5026) chenyuandGitHub 2024-06-18 08:17:12 -04:00
• 7bfa9101c0 Float in scaled dot product attention (#4985) Elias WahlandGitHub 2024-06-18 14:16:41 +02:00
• 194a168630 hcq signal scheduler (#5016) nimlgenandGitHub 2024-06-18 14:02:21 +03:00
• e9c6a36894 remove CACHELEVEL=0 in llama3 benchmark (#5025) chenyuandGitHub 2024-06-17 22:43:16 -04:00