sirhcm/tinygrad
1
0
Fork 0
forked from tinygrad/tinygrad
Code Pull Requests 2 Actions 6 Activity
14,615 Commits 336 Branches 15 Tags
master
Commit Graph
63 Commits
Author SHA1 Message Date
chenyuandGitHub e3c4b3d9b2 fix logaddexp and asin at 0 (#17862) 
* fix logaddexp and asin at 0

* pi/2
 2026-08-31 09:51:03 -04:00
project224GitHubprojectai225Claude Opus 5
 		3123f947b0 fix erf gradient being zero at x=0 (#17861) 
erf used self.sign() * f(self.abs()). Both sign() and abs() have zero
gradient at exactly 0, so d/dx erf(0) came out as 0.0 instead of
2/sqrt(pi) ~= 1.1283792. Values on either side (including +/-1e-8) were
already correct, so the error was a single point discontinuity.

Derive the sign factor once with a where() and use it to form abs(x) as
s*x, so the gradient at 0 takes the right-hand limit instead of being
annihilated. Forward output is unchanged (erf(0) is still exactly 0.0 in
float32) and the line count is unchanged.


Claude-Session: https://claude.ai/code/session_01EXJpYja2w3bcjXzCaZwyWx

Co-authored-by: projectai225 <[email protected]>
Co-authored-by: Claude Opus 5 <[email protected]>
 2026-08-31 08:57:11 -04:00
Teddy TennantandGitHub fdb4109f5f fix logcumsumexp on fully masked prefixes (#17853) 2026-08-30 15:34:33 -04:00
Teddy TennantandGitHub 1c183e6de9 fix elu/celu/selu gradient being nan on large inputs (#17852) 
* fix elu/celu/selu gradient being nan on large inputs

* shrink shape for the elu/celu/selu overflow checks

* route selu and celu through elu
 2026-08-30 15:12:52 -04:00
Teddy TennantGitHubchenyu
 		0a1bfc9fad fix logsumexp on fully masked slices (#17822) 
Co-authored-by: chenyu <[email protected]>
 2026-08-29 11:48:10 -04:00
chenyuandGitHub 661440e672 fix bufferize_to_store for weak input (#17833) 2026-08-29 11:01:33 -04:00
chenyuandGitHub 6083de5cff fix nan compares (#17801) 
* fix nan compares

* fix
 2026-08-28 09:17:18 -04:00
Teddy TennantandGitHub 26c4b6319e fix interpolate with align_corners and an output size of 1 (#17790) 2026-08-27 19:17:05 -07:00
chenyuandGitHub fede358811 fix fancy indexing with uint8 index (#17787) 
* fix fancy indexing with uint8 index

* fix
 2026-08-27 19:59:27 -04:00
Teddy TennantandGitHub 5c3d044465 fix asinh gradient at zero (#17758) 2026-08-27 09:51:17 -04:00
chenyuandGitHub 6a3b297548 fix PTX NIR SPEC=2 for bool [pr] (#17753) 
* fix PTX NIR SPEC=2 for bool [pr]

storing bool with uint8 needs to pass SPEC

* the fix
 2026-08-26 14:07:29 -04:00
Teddy TennantandGitHub 4c20f1d357 fix asinh precision loss on negative inputs (#17749) 2026-08-26 10:08:04 -04:00
chenyuandGitHub 9f01775cf4 dtype_from_uop for THREEFRY and FDIV (#17727) 2026-08-25 08:43:59 -04:00
chenyuandGitHub 0e7ab863a0 x86 REX issue (#17580) 
* failing test

* fix
 2026-08-20 16:49:04 -04:00
chenyuandGitHub a8ecb73363 x64 imm uint64 (#17636) 
* x64 imm uint64

* fix
 2026-08-20 15:30:17 -04:00
BennettGitHubBennett
 		7064e76bc8 fix roll on zero-sized tensors (#17603) 
Signed-off-by: Bennett <[email protected]>
Co-authored-by: Bennett <[email protected]>
 2026-08-19 15:31:21 -04:00
chenyuandGitHub fae893753b no pm_fold_cast_const in get_kernel_graph [pr] (#17546) 
* no pm_fold_cast_const in get_kernel_graph [pr]

* maybe
 2026-08-15 09:57:05 -04:00
Robert JosephandGitHub 0c96cdc300 fix prod gradients at zero (#17404) 2026-08-07 09:56:52 -04:00
chenyuandGitHub d79772f057 fix pow on extreme inputs (#17397) 
* fix pow on extreme inputs

* WEBGPU
 2026-08-04 19:30:43 -04:00
Noah SchiroandGitHub f2c2f4456b Add softmin (#17292) 
* Add softmin

* Remove extra tests
 2026-07-30 20:05:11 -04:00
chenyuandGitHub acc2374b6f minor normalize cleanup [pr] (#17215) 
and add a few no-op clones in test_ops to prep for weak flip
 2026-07-26 00:39:53 -04:00
chenyuandGitHub a292dab8ab corner cases from weak const branch (#17173) 2026-07-24 00:09:56 -04:00
chenyuandGitHub f5d9c31d15 arange upcast to int64 with big N (#17167) 
simplifies _one_hot_along_dim
 2026-07-23 20:11:39 -04:00
chenyuandGitHub 3e031d6c4d AND OR XOR SHL SHR cannot have float operands [PR] (#17164) 
* AND OR XOR SHL SHR cannot have float operands [PR]

* dont
 2026-07-23 18:51:56 -04:00
chenyuandGitHub 2983987321 clean up some tests (#17154) 
duplicated or typo or slow
 2026-07-23 11:22:18 -04:00
sirhcmandGitHub c7027db715 lvp: skip test_masked_select (#17137) 2026-07-22 14:53:09 -04:00
chenyuandGitHub 92f9c850b4 fix pow(int, float) (#17126) 
* fix pow(int, float)

* onnx
 2026-07-21 18:48:24 -04:00
chenyuandGitHub f19a2ad771 single where mixin [pr] (#17118) 
* single where mixin [pr]

no shape broadcasting in ufix and _broadcasted anymore

* QCOM vectorized bool is broken
 2026-07-21 17:38:36 -04:00
chenyuandGitHub 9970a0aad0 fix Tensor << Tensor for x86 (#17082) 
* fix Tensor << Tensor for x86

* torch
 2026-07-19 12:31:05 -04:00
sirhcmandGitHub 353d8f1e13 use rusticl in ci (#16852) 2026-07-03 23:58:58 -04:00
chenyuandGitHub 59c7874724 fix isclose with scalar other (#16813) 2026-06-30 21:56:33 -04:00
Javier De JesusandGitHub 989f713c1b support negative pads in circular pad mode (#16448) 2026-05-31 09:28:45 -07:00
sirhcmandGitHub 8ddd1328df remove getenv(CI) (#16365) 
gone everywhere except test_interop, because torch MPS does not work in actions
 2026-05-25 20:23:33 -04:00
sirhcmandGitHub d8f86be613 webgpu: shader-f16 support in arch (#16370) 2026-05-25 19:20:59 -04:00
chenyuandGitHub 926d125a63 update test_stack (#16345) 
also skip COMPILE_ONLY, it was comparing 0==0
 2026-05-23 10:42:35 -04:00
sirhcmandGitHub 451f38155c start cleanup of the slowest tests (#16339) 2026-05-22 18:39:36 -04:00
sirhcmandGitHub c2d06570a5 remove getenv(CI) from core tinygrad (#16326) 2026-05-21 22:20:33 -04:00
sirhcmandGitHub 172f9493e1 move is_dtype_supported to renderer (#16226) 2026-05-20 21:19:37 -04:00
George HotzandGitHub 58d58c1659 remove DEVECTORIZE (#16290) 
* remove DEVECTORIZE

* fully remove DEVECTORIZE
 2026-05-20 13:25:49 -07:00
George HotzandGitHub a120709671 tighten shape spec for broadcasting (#16206) 
* tighten shape spec for broadcasting

* use IndexError, not ValueError

* needs size
 2026-05-18 22:12:04 -07:00
chenyuandGitHub 8631b6f17d remove use of requires_grad in test/ (#16237) 2026-05-16 17:21:07 -04:00
C TandGitHub 1b779a9058 add gelu approximate="none" (match pytorch) (#16162) 
* add gelu approximate="none" (match pytorch)

* lint

* pass through onnx Gelu approximate

* type annotate

* explicit math.sqrt

* keep tinygrad's gelu approximate="tanh" default
 2026-05-13 18:53:24 -07:00
chenyuandGitHub bdcdf1f1a1 jittable masked_select and nonzero (#16170) 
* jittable masked_select and nonzero

make jittable with `size=`, matches jax

* COMPILE_ONLY
 2026-05-12 16:39:36 -04:00
chenyuandGitHub 7c3e3fa154 fix empty input for masked_select and nonzero (#16168) 2026-05-12 15:36:51 -04:00
PawanandGitHub 4dd6ad3514 gradient: add TRUNC backward (#15925) 
* gradient: add TRUNC backward

* test: move round quantization gradient to test_ops
 2026-05-08 16:27:55 -07:00
chenyuandGitHub 072db9924c div to mixin (#16078) 
also deleted idiv method
 2026-05-07 12:52:37 -04:00
chenyuandGitHub 782d1ff80f Tensor.fmod (#16014) 
c-style mod matches torch
 2026-05-01 16:02:18 -04:00
chenyuandGitHub f911a63a6b don't allow negative num_classes in one_hot (#15859) 
no auto infer num_classes, matches jax
 2026-04-21 19:39:29 -04:00
sirhcmandGitHub 6adf4c3cd9 MOCKGPU interfaces (#15796) 2026-04-17 21:56:29 -04:00
chenyuandGitHub 0191cc73dc update arange range check (#15794) 
it was not checking negative steps correctly
 2026-04-17 16:07:50 -04:00