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

• bfbb8d3d0f fix ones, BS=2 stable diffusion, caching optimizer (#1312) George HotzandGitHub 2023-07-21 09:55:49 -07:00
• 9746f6d094 move hand coded optimizer (#1310) George HotzandGitHub 2023-07-21 07:53:12 -07:00
• d2c1e8409a Update arange to be (start, stop, step) (#1308) madt2709andGitHub 2023-07-20 21:27:23 -07:00
• f45013f0a3 stable diffusion: remove realizes we don't need geohot 2023-07-20 19:53:07 -07:00
• b58dd015e3 stable diffusion: remove import numpy as np geohot 2023-07-20 19:35:44 -07:00
• 35bc46289c stable diffusion: use new tinygrad primitives geohot 2023-07-20 19:25:12 -07:00
• 78a7a15753 Fix WSGL to render NaN and prevent shader compile error (#1268) Francis LamandGitHub 2023-07-20 18:00:33 -07:00
• 0a3d4f8103 Implementation of VITS TTS model (#1188) StanandGitHub 2023-07-21 02:37:14 +02:00
• d963024a13 optimizer small fix: return if there's nothing to optimize geohot 2023-07-20 16:57:30 -07:00
• 9dffc9ba23 Use nevergrad to optimize kernels (try 2) (#1301) George HotzandGitHub 2023-07-20 16:46:45 -07:00
• 8562b5a04f fixes error when trying to convert float4 -> half4 (#1300) DiogoandGitHub 2023-07-20 17:20:05 -04:00
• 50a399ffa3 real world test: relax memory geohot 2023-07-20 14:06:22 -07:00
• 17830e25da real world tests (#1297) George HotzandGitHub 2023-07-20 10:50:22 -07:00
• ca77d6cd72 bfloat16 in LLVM (enough for llama 2) (#1293) George HotzandGitHub 2023-07-19 20:18:32 -07:00
• 74e63fe4ee Added test_chunk and fixed (#1283) Umut ZenginandGitHub 2023-07-20 05:21:26 +03:00
• 3f2497160c strip whitespace geohot 2023-07-19 19:01:53 -07:00
• 65fe72f10b Cleanup loadop (#1291) George HotzandGitHub 2023-07-19 18:59:47 -07:00
• e3f717f614 fix CUDAProgram __init__ with DEBUG>=6 on Linux (#1288) Alexander SchlöglandGitHub 2023-07-19 23:36:58 +02:00
• f7b0320d8b add cifar training regression test (#1287) George HotzandGitHub 2023-07-19 14:17:09 -07:00
• 45ecae1ab3 Revert "Match Torch speed for sum reduction on M1 (#1187)" (#1286) George HotzandGitHub 2023-07-19 13:39:16 -07:00
• 120ae74008 Enable JIT test for size 1 tensor (#1285) chenyuandGitHub 2023-07-19 14:06:40 -04:00
• 940b6fd21a Revert "Fix constant folding for Tensor([3]) (#1227)" (#1274) chenyuandGitHub 2023-07-19 13:51:06 -04:00
• 0aed3f73da More JIT test cases (#1280) chenyuandGitHub 2023-07-19 13:45:43 -04:00
• 3db57d3118 Fix llama.py to load and concatenate 13B, 30B, and 65B models (#1275) Francis LamandGitHub 2023-07-19 10:22:33 -07:00
• d6637623e3 torch test touchup geohot 2023-07-19 09:37:23 -07:00
• 59af9b81c5 Match Torch speed for sum reduction on M1 (#1187) Alexander EdwardsandGitHub 2023-07-20 02:18:58 +10:00
• fde9f0e60d Slice migrated in Eye op (#1281) Umut ZenginandGitHub 2023-07-19 19:08:38 +03:00
• a5f5330d91 Add Fuzz Test symbolic / shapetracker to CI. (#1278) chenyuandGitHub 2023-07-19 12:05:45 -04:00
• 56ee97b37f dedup kernel args v2 (#1272) David HouandGitHub 2023-07-18 20:03:42 -07:00
• 37cc33269a cl fixes for multigpu (#1276) wozeparrotandGitHub 2023-07-18 22:59:30 -04:00
• fa0265b173 Fix: AssertionError Transpose/Permute when WHERE Op in LB (#1266) Umut ZenginandGitHub 2023-07-18 23:09:19 +03:00
• c96bf395df Enable JIT tests for supported devices, skip METAL and WEBGPU (#1265) chenyuandGitHub 2023-07-18 14:40:37 -04:00
• f8c539989e Re-open create cumsum speed test (#1255) Umut ZenginandGitHub 2023-07-18 04:59:36 +03:00
• ab3d281a6e Refactor MemOps (#1256) George HotzandGitHub 2023-07-17 16:36:33 -07:00
• ed472bffea Fix: negative axis in tensor.cumsum (#1261) StanandGitHub 2023-07-18 01:16:38 +02:00
• 64d39188ad Assembly ptx target current arch (#1250) OddityandGitHub 2023-07-17 10:45:43 -05:00
• 5a8ad57163 Add WHERE ternary (or trinary?) op (#1196) 3 people 2023-07-16 09:31:55 +02:00
• 91f797cd52 Moved mkdir in utils.download_file to diff line (#1249) StanandGitHub 2023-07-16 09:30:46 +02:00
• a8f2c16f8e add contiguous (#1246) Yixiang GaoandGitHub 2023-07-15 10:36:34 -05:00
• 872e2198fe Added nn.ConvTranspose1d (#1243) 3 people 2023-07-15 09:42:42 +02:00
• 7399f6dad7 display sass for both cuda code and ptx (#1240) OddityandGitHub 2023-07-15 02:36:04 -05:00
• 264d467f2b Added tensor.squeeze and support for testing exceptions (#1241) StanandGitHub 2023-07-15 09:33:24 +02:00
• a8f3b3f4ed Added test for nn.Conv1d (#1242) StanandGitHub 2023-07-15 09:30:50 +02:00
• 9c135c9450 add sqrt to ptx (#1236) David HouandGitHub 2023-07-13 07:26:11 -07:00
• 32be39554c Simplify symbolic.SumNode.__floordiv__ logic (#1220) chenyuandGitHub 2023-07-12 15:54:12 -04:00
• a9a1df785f Webgpu support (#1077) 4 people 2023-07-12 15:52:06 -04:00
• 613bcd945d Added Test Coverage to Int32 and Make Sure Tests Succeed (#1174) Yosef FrostandGitHub 2023-07-12 13:29:15 -04:00
• 8f2e2f5ee2 style: else-after-return (#1216) 3 people 2023-07-12 19:26:38 +02:00
• ab663c46e8 tensor cores: don't upcast if we can't. fix stable diffusion geohot 2023-07-12 10:21:02 -07:00
• 4f72eb823c Outdated repository URL (#1218) HeyandGitHub 2023-07-12 08:14:19 +02:00
• d0e21a7398 ci: don't install recommended packages for GPU (#1215) 3 people 2023-07-12 00:38:49 +02:00
• df86672bd4 Fix LazyBuffer SHUFFLE_PAD_OPS to prevent invalid pad movement (#1223) Francis LamandGitHub 2023-07-11 15:30:35 -07:00
• f75de602df fix typo in stable diffusion example (#1219) AN LongandGitHub 2023-07-12 06:26:40 +08:00
• ab645317c9 Fix constant folding for Tensor([3]) (#1227) chenyuandGitHub 2023-07-11 17:01:32 -04:00
• e2f6b09ffd [perf] optimize=True kwarg for np.einsum (#1213) Carson RadtkeandGitHub 2023-07-09 20:31:04 -05:00
• bb316a42af Fix pow to work with negative tensors (#1191) madt2709andGitHub 2023-07-09 17:33:04 -07:00
• 43385c7dbf remove contiguous on full (#1212) George HotzandGitHub 2023-07-09 17:31:15 -07:00
• 13a1abf9e7 remove tuple from type annotation in Tensor.__init__ (#1211) Carson RadtkeandGitHub 2023-07-09 18:27:07 -05:00
• e27f098946 View as namedtuple, cached methods (#1075) 3 people 2023-07-09 23:26:02 +02:00
• 1eb0e0cb3f implement common subexpression elimination (#1204) Carson RadtkeandGitHub 2023-07-09 16:22:53 -05:00
• beb4d3ab01 Tensor Cores 2: Local Buffers Edition (#1057) George HotzandGitHub 2023-07-09 09:06:00 -07:00
• 67e34b356a good stuff from tensor cores branch (#1199) George HotzandGitHub 2023-07-08 16:58:26 -07:00
• 7151382364 Refactor load/store before tensor cores (#1193) George HotzandGitHub 2023-07-08 15:54:58 -07:00
• ef1909500e remove superfluous parentheses (#1197) fluffy χατγιρλandGitHub 2023-07-09 00:11:02 +02:00
• 628ee46627 Fix bug where Tensor.randn returns inf (#1192) fluffy χατγιρλandGitHub 2023-07-08 21:03:46 +02:00
• d9c1d81e99 Revert "feat: cancel previous workflow runs on new commits (#1184)" (#1194) George HotzandGitHub 2023-07-08 11:26:13 -07:00
• 52600d532e add 20 minute timeout geohot 2023-07-07 23:02:28 -07:00
• d66a0c285d feat: cancel previous workflow runs on new commits (#1184) wozeparrotandGitHub 2023-07-08 01:55:35 -04:00
• e0c2ae8984 Update file paths (#1179) Jacky LeeandGitHub 2023-07-07 18:41:58 -07:00
• 0ad99038ef Revert "Revert "Fix ShapeTracker mismatch in LazyBuffer.fromCPU (#1156)" (#1181)" + add test geohot 2023-07-07 18:35:05 -07:00
• 2952b8e7a8 Fix up abstractions.py to include the Linearizer (#1177) George HotzandGitHub 2023-07-07 18:33:51 -07:00
• a374b62bfe Revert "Fix ShapeTracker mismatch in LazyBuffer.fromCPU (#1156)" (#1181) George HotzandGitHub 2023-07-07 18:29:05 -07:00
• 8ff7184b1b Fix ShapeTracker mismatch in LazyBuffer.fromCPU (#1156) fluffy χατγιρλandGitHub 2023-07-08 03:28:21 +02:00
• b8dfbba703 hip_matmul: f16 gemm 2048x2048 gets 36 TFLOPS geohot 2023-07-08 00:33:02 +00:00
• 69d33cab0d Fix: auto create parent dir when downloading file (#1173) StanandGitHub 2023-07-07 22:40:29 +02:00
• f40f8cd055 Initialise numpy arrays as float32 in DDPG (#1171) StanandGitHub 2023-07-07 21:05:31 +02:00
• 884b5965de ops_cuda fix race condition on cubin file read when testing with multiple cores (#1172) cloud11665andGitHub 2023-07-07 21:05:16 +02:00
• aa60feda48 Fix naming conflict with huggingface datasets (#1161) 3 people 2023-07-07 20:43:44 +03:00
• fd66d1ca00 fix Tensor.manual_seed() default to wrong type (#1168) Yahya LmallasandGitHub 2023-07-07 18:42:48 +01:00
• 9b6e57eccd helpers.py: improved test coverage + exception handling (#1165) StanandGitHub 2023-07-07 19:26:05 +02:00
• 8391648822 Over 90% on CIFAR with examples/hlb_cifar10.py (#1073) 3 people 2023-07-07 09:16:22 +05:30
• c5aea13a65 Fix evaluation stage in examples/transformer.py when using CUDA (#1150) BarathandGitHub 2023-07-07 09:07:10 +08:00
• 9975f24452 Fold expand preceding reduce if the reduction is on the same axis as the expansion (#1134) Rayan HatoutandGitHub 2023-07-06 21:41:05 +01:00
• f109af3cbb Don't save parents unless needed (#1142) cheeetooandGitHub 2023-07-05 20:11:57 -05:00
• 801564f31b Remove POW llop and add SQRT llop (#1104) 3 people 2023-07-05 20:07:58 -05:00
• b7369ffcff add ptx formatter + syntax highlighter (#1128) cloud11665andGitHub 2023-07-06 02:56:09 +02:00
• d1356cac27 Fix: Jacobian tests [WIP] (#1126) Reza RezvanandGitHub 2023-07-06 00:36:22 +02:00
• d363d25ee2 fix imports for examples/transformer.py (#1136) nimlgenandGitHub 2023-07-05 18:15:13 +03:00
• c3173ff281 Add return statement to the train function (#1135) Mehmet KuzucuandGitHub 2023-07-05 17:13:38 +02:00
• 981d4980c4 feat: reword contributing (#1131) wozeparrotandGitHub 2023-07-05 01:17:47 -04:00
• 793a670187 from tensor cores + lb touchup (#1127) George HotzandGitHub 2023-07-04 15:45:20 -07:00
• 2f968f8547 ignore cloudpickle type for local mypy geohot 2023-07-04 13:51:20 -07:00
• 87d21ea979 examples: simple conv bn geohot 2023-07-04 13:50:26 -07:00
• 535224ac20 Remove float64 (#1101) Reza RezvanandGitHub 2023-07-04 17:40:51 +02:00
• b4ce23e4b8 Make cross_process use cloudpickle (#1118) Daniel HipkeandGitHub 2023-07-04 00:47:34 -07:00
• c709dec8b5 gelu: weird test was broken for metal geohot 2023-07-04 00:43:54 -07:00
• daf8e1942f sigmoid: test large postive also and add note geohot 2023-07-04 00:18:31 -07:00
• 9e6067378f Broken Sigmoid backward: Add test and mlop for Sigmoid (#1113) 3 people 2023-07-04 12:44:22 +05:30
• d58a9603ab Create COCO data directory if it doesn't exist. (#1114) Daniel HipkeandGitHub 2023-07-03 18:15:53 -07:00
• a22aad7d32 Use generators instead of lists in anys and alls (#1111) 3 people 2023-07-04 01:06:06 +02:00