nimlgen/tinygrad
1
0
Fork 0
forked from tinygrad/tinygrad
Code Pull Requests Activity
13,797 Commits 334 Branches 15 Tags
move_devec
Commit Graph
13797 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
geohot af0abe4032 fix flaky 2026-07-04 15:29:27 -07:00
geohot 964df8ec0e move devectorizer remains into coalese 2026-07-04 15:27:12 -07:00
George HotzandGitHub 43c1bbceb1 remove pm_render (#16868) 
* remove pm_render

* move merge_reduce_ends

* ext devectorize
 2026-07-04 15:18:43 -07:00
George HotzandGitHub a145fdce4b remove the .gep method (#16867) 2026-07-04 14:07:09 -07:00
wozeparrotandGitHub 971800b46a support none block divisible embed size (#16865) 2026-07-04 13:46:47 -07:00
George HotzandGitHub 3fd6f3d28c no more GEP (#16866) 2026-07-04 13:42:28 -07:00
George HotzandGitHub d76bcbaa02 deletions after new codegen (#16864) 
* deletions after new codegen

* delete the expander
 2026-07-04 13:12:11 -07:00
George HotzandGitHub c7e7687bd3 new codegen, try 3 (#16781) 
* new codegen, try 3 [pr]

* reduce

* diffs from cg2

* minor fixes

* mergable

* new local buffers

* fixes

* correct fix

* add barrier, and image is after that

* fix get

* wmma work

* devec wmma

* fixes for wmma

* slightly more flexible

* fix tensor cores

* fix linter

* tests

* remove invalid?

* remove extra

* wmma merge

* flip reshape/permute on wmma

* broadcast_binary None

* update symbolic for vecless

* x86 fix

* fix custom llama kernel

* is_ptr hack

* fix pre-commit

* fix wmma

* simpler

* fixes

* better hreduce

* move reduce axes

* all wmma tests pass

* allow stacked wmma

* permute cleanups

* late permute

* correct permute flip

* fix group for reduce

* fix wmma permuted

* all pass

* mop cleanup

* new test

* push permute/reshape

* revert that

* half
 2026-07-04 12:55:44 -07:00
nimlgenandGitHub fdf434062b hcq2: simpler (#16862) 
* x

* x

* x
 2026-07-04 22:07:37 +03:00
chenyuandGitHub 09922d2326 rework propagate_invalid [PR] (#16861) 2026-07-04 12:37:22 -04:00
qazalandGitHub 07f7383d29 llama: remove unused bf16 assembly gemm (#16859) 
* only hk bf16 gemm

* rm asm gemm

* more cleanup

* half isn't supported in asm gemm anymore

* more test edits

* unused

* remove TestMagicGu

* uop gemm is still tested

* minimal diff
 2026-07-04 18:41:54 +09:00
sirhcmandGitHub 353d8f1e13 use rusticl in ci (#16852) 2026-07-03 23:58:58 -04:00
chenyuandGitHub 709251e22d more generic fold_add_divmod_recombine [pr] (#16854) 2026-07-03 22:00:59 -04:00
George HotzandGitHub 62c951b2ce lil changes from codegen 6 (#16855) 2026-07-03 17:08:18 -07:00
George HotzandGitHub 41d6731bfd reduce removes ones (#16847) 
* reduce removes ones

* test changes

* lil clean
 2026-07-03 13:10:08 -07:00
nimlgenandGitHub 13d99388ab system: raise if unbind is unsuccessful (#16846) 2026-07-03 22:01:30 +03:00
chenyuandGitHub 5fb3cfb9bc remove early THREEFRY const folding [PR] (#16850) 
it folds automatically once decomposed
 2026-07-03 15:00:35 -04:00
George HotzandGitHub ed3dec4674 remove noop reshapes (#16849) 2026-07-03 11:39:57 -07:00
chenyuandGitHub ed7981f8f2 move divandmod next to symbolic [PR] (#16848) 2026-07-03 14:21:59 -04:00
chenyuandGitHub ee5e8cea68 no dtype.vec in select_dtype [PR] (#16845) 2026-07-03 13:19:12 -04:00
chenyuandGitHub 3dc44f3692 image vectorize -> _stack [PR] (#16844) 2026-07-03 12:51:35 -04:00
chenyuandGitHub 2f0690dcc2 minor gpudims cleanup [PR] (#16843) 2026-07-03 11:23:23 -04:00
chenyuandGitHub 6345c2883d simpler get_grouped_dims (#16842) 
works for more than 3D, also deleted get_contraction
 2026-07-03 10:29:11 -04:00
nimlgenandGitHub d73bca617b hcq2: fix deps tracking (#16841) 2026-07-03 15:51:25 +03:00
wozeparrotandGitHub 42be177ea8 gpt-oss-20b: model (#16837) 
* feat: gpt-oss model file

* clean: don't need that function

* feat: working model
 2026-07-03 01:29:56 -07:00
qazalandGitHub fd18dbad7b viz: do not check src[0] for INDEX/STAGE rendering (#16839) 
* viz: do not check src[0] for INDEX/STAGE rendering

* index too large
 2026-07-03 15:11:08 +09:00
geohot 0240998b06 hotfix: that test_ellipsis_with_int test was important 2026-07-02 19:19:38 -07:00
George HotzandGitHub e6fbede157 changes from new codegen (#16836) 2026-07-02 17:01:37 -07:00
sirhcmandGitHub e74c7042c3 ci: use tinymesa from pypi (#16835) 2026-07-02 19:25:10 -04:00
chenyuandGitHub bc115d44fd don't check isinstance ElementwiseMixin [PR] (#16834) 
won't work once Tensor no longer inherits mixin
 2026-07-02 18:28:52 -04:00
nimlgenandGitHub c2c784e028 hcq2: rewrite (#16775) 
* hcq2: rewrite

* rw

* mid

* jit

* tmp

* x

* x

* revert

* x

* x
 2026-07-03 00:00:08 +03:00
chenyuandGitHub 29b59962c9 ops and callify cleanups [PR] (#16833) 2026-07-02 16:51:44 -04:00
chenyuandGitHub e9dd2990b6 delete UOp.multibase [PR] (#16832) 2026-07-02 16:00:01 -04:00
chenyuandGitHub 682b098542 fix multi write after read in create_schedule (#16831) 2026-07-02 15:12:48 -04:00
chenyuandGitHub 48c2081378 clean up create_schedule [PR] (#16830) 2026-07-02 14:49:01 -04:00
chenyuandGitHub 3e73a2542b _apply_map_to_tensors always use walk rewrite (#16827) 
with assign it's wrong to recursively apply, fixed a random bug that counter does not update after round 2
 2026-07-02 10:43:25 -04:00
chenyuandGitHub b2be3c6c57 disk_copy_is_buffer maps each disk COPY once [PR] (#16826) 2026-07-02 09:58:34 -04:00
chenyuandGitHub b61285efa4 remove apply_after [PR] (#16825) 
tighten buffer_map spec and becomes_map is much smaller now, all intermediate afters are not needed
 2026-07-02 09:19:35 -04:00
qazalandGitHub 149fd91e22 viz: no tqdm noise in serve.py (#16824) 2026-07-02 15:37:26 +09:00
sirhcmandGitHub e0d7696ccd split benchmarks (#16812) 2026-07-01 22:28:15 -04:00
chenyuandGitHub 2e62dd308d move fs_load and fs_store to nn (#16821) 
similar to safe_load and safe_save
 2026-07-01 22:01:15 -04:00
chenyuandGitHub c8aed121cf fix WAR in create_schedule (#16823) 
* fix WAR in create_schedule

combined dep building, fixed a case that realize(x, y) != realize(y, x)

* fix
 2026-07-01 22:00:51 -04:00
chenyuandGitHub efd256b2a3 fix precompiled STORE+AFTER (#16822) 2026-07-01 17:07:37 -04:00
chenyuandGitHub 4b3de041e2 delete Ops.DEVICE (#16820) 
85 ops
 2026-07-01 12:20:05 -04:00
chenyuandGitHub fd7fd9cdca move GETADDR device to its arg (#16808) 2026-07-01 09:14:07 +03:00
George HotzandGitHub 3c07f31790 remove shape special case for Ops.CAST (#16817) 2026-06-30 22:37:07 -07:00
sirhcmandGitHub 4a4de1a966 mesa: use stdio autogen (#16815) 2026-06-30 23:33:55 -04:00
chenyuandGitHub 1261d719a8 tighter GETTUPLE spec (#16816) 2026-06-30 23:33:10 -04:00
chenyuandGitHub 59c7874724 fix isclose with scalar other (#16813) 2026-06-30 21:56:33 -04:00
chenyuandGitHub 372774fc41 empty to CreationMixin [PR] (#16811) 2026-06-30 20:42:42 -04:00