chenyu
03943cd1a0
use more _uop for cleanup [PR] ( #16531 )
...
`t.uop if isinstance(t, Tensor) else t` -> `t._uop`
2026-06-07 17:41:36 -04:00
chenyu
937aeaec60
remove device= from UPat.const [PR] ( #16530 )
2026-06-07 16:38:43 -04:00
George Hotz
eb1238436a
more prereqs for DL/DR -> BUFFER ( #16529 )
2026-06-07 12:25:11 -07:00
George Hotz
0336ba8eb1
buffer param arg + dsp fixups ( #16528 )
2026-06-07 12:07:00 -07:00
Dmitriy Strunin
75e903d533
remove unused device arg from _get_winograd_matcols ( #16527 )
2026-06-07 08:15:09 -04:00
chenyu
90b556ca48
move gradient to mixin [PR] ( #16526 )
2026-06-07 00:05:02 -04:00
chenyu
4e7c6260b0
clean up test_tesnor_uop_mixin ( #16525 )
...
most of those don't have UNIQUE anymore
2026-06-06 23:25:44 -04:00
George Hotz
2a2f81dd3d
remove ANON from addrspace, refactor marg ( #16523 )
...
* remove ANON from addrspace, refactor marg
* as_shape
* as_shape is cached
2026-06-06 09:49:09 -07:00
qazal
e69b4189b0
viz: hide STACK on PARAM by default ( #16522 )
2026-06-06 16:41:15 +09:00
Christopher Milan
857b1f5399
ci: more parallelism, less duplication ( #16509 )
2026-06-05 21:26:19 -04:00
wozeparrot
a1ec32cfd2
llama: current grad scaling ( #16518 )
2026-06-05 15:39:41 -07:00
Christopher Milan
8c0ba1da5c
cleanup more from test/backend ( #16521 )
2026-06-05 18:38:46 -04:00
chenyu
9982185b14
remove unused AFTER rules in pm_add_buffers[PR] ( #16519 )
2026-06-05 14:58:34 -04:00
nimlgen
5ebd44aa12
hcq2: merge queues ( #16514 )
...
* hcq2: mergw queues
* cleaner
2026-06-05 21:20:25 +03:00
chenyu
a51b5ba424
remove early fixup const copy [PR] ( #16516 )
2026-06-05 11:35:34 -04:00
Nueramarcos
8274140134
uop/ops: fix ~bool deprecation warning on Python 3.12+ (ORANGE Grok helped with the patch) ( #16512 )
2026-06-05 10:54:30 -04:00
chenyu
588c759a3d
remove unused GroupOp.Buffer [PR] ( #16515 )
2026-06-05 10:38:52 -04:00
qazal
79a13310b3
viz: kernel_graph.txt unique is per schedule ( #16511 )
2026-06-05 16:17:28 +09:00
Christopher Milan
9b0f75622c
many jit tests belong in unit ( #16508 )
2026-06-04 21:36:53 -04:00
chenyu
bb407d8b3c
fix transform_precompiled_call for MULTI ( #16510 )
...
based on my understanding for https://github.com/tinygrad/tinygrad/pull/16084
2026-06-04 20:09:58 -04:00
wozeparrot
f11f63007d
llama: immediate scaling on flag ( #16494 )
2026-06-04 10:30:00 -07:00
George Hotz
4fb8ce1831
update buffer in spec ( #16507 )
2026-06-04 10:12:31 -07:00
chenyu
4a8bf07a87
remove CONST(DEVICE) ( #16506 )
2026-06-04 11:29:46 -04:00
nimlgen
3838c8df1b
hcq2: move global sync ( #16504 )
2026-06-04 17:32:40 +03:00
chenyu
0faaf6df26
remove kwargs from arange and linspace [PR] ( #16505 )
...
it used to have requires_grad and device, now both are removed
2026-06-04 10:32:37 -04:00
qazal
3b1a5f9770
llama: a_bT and aT_b bf16 gemms ( #16487 )
...
* hk_bf16_gemm
* enable in 8b
* cleanups
* rename to USE_HK_BF16_GEMM
* work
* work
* work
* work
* change the gemms
* work
* work
* set as default
* work
* change
2026-06-04 23:30:21 +09:00
chenyu
5fad87252d
no device= into arange and eye ( #16503 )
2026-06-04 09:21:50 -04:00
nimlgen
11af81f96f
hcq2: cleaner ( #16502 )
2026-06-04 15:26:37 +03:00
chenyu
2c915c61ed
no CONST(DEVICE) in torch_backend ( #16499 )
2026-06-04 00:26:47 -04:00
wozeparrot
fd13080636
deviceless const skip axis check ( #16496 )
2026-06-03 19:13:20 -07:00
qazal
f7f03bd7e5
viz: better name for src id in kernel_graph.txt ( #16495 )
...
* viz: better name for src id in kernel_graph.txt
* better order
* cleanup
2026-06-04 11:09:29 +09:00
Christopher Milan
9dac781e45
ci: use uv ( #16492 )
2026-06-03 21:38:50 -04:00
George Hotz
9fdeaa402b
no anon addrspace, don't write hacks ( #16491 )
...
* no anon addrspace, don't write hacks
* revert that
* no reg there
2026-06-03 16:19:30 -07:00
chenyu
2f83d01ccf
fix deviceless materialize device ( #16493 )
...
symbolic arange currently does not fuse, which creates a deviceless UOp post rangeify that needs a device to bufferize
2026-06-03 19:13:21 -04:00
chenyu
19eb72ff60
remove use of full with buffer=False and non-None device= ( #16489 )
2026-06-03 16:21:24 -04:00
nimlgen
6f2a2857c8
hcq2: refactor deps ( #16490 )
2026-06-03 23:20:24 +03:00
chenyu
243446b44f
remove CONST(DEVICE) from const_like ( #16488 )
2026-06-03 14:04:51 -04:00
George Hotz
cee472a0ef
renderer Estimates uses maxel ( #16485 )
2026-06-03 10:55:00 -07:00
chenyu
8a4203638a
make full with buffer=False deviceless ( #16483 )
...
affects arange and eye
2026-06-03 12:35:59 -04:00
qazal
405866f2b7
viz: improve kernel_graph.py usability ( #16486 )
...
* better default
* always format kernel output
* also show ref
* sched num
2026-06-03 21:12:44 +09:00
Christopher Milan
f43cba5765
ci: native python where possible ( #16473 )
...
linters stays at 3.11
2026-06-02 22:40:12 -04:00
wozeparrot
7dcfd144b6
llama: columnwise fp8 scaling ( #16480 )
2026-06-02 18:55:45 -07:00
George Hotz
ffadd7a315
remove intel and amx support ( #16482 )
2026-06-02 18:53:05 -07:00
George Hotz
5f439e3b7c
refactor cstyle to avoid dtype [PR] ( #16478 )
...
* refactor cstyle to avoid dtype
* clean up rules
* add new style option
2026-06-02 18:27:12 -07:00
Christopher Milan
80eeb4dd21
mockgpu: use autogen.libc ( #16479 )
2026-06-02 19:59:36 -04:00
chenyu
a43b55d480
deviceless const folding schedule test ( #16477 )
2026-06-02 18:46:30 -04:00
George Hotz
14f843737b
renderer cleanups (pt 3) [PR] ( #16475 )
...
* renderer cleanups (pt 3)
* point refactors
* fix bugs
* fix PR
2026-06-02 14:24:24 -07:00
nimlgen
99e37b1ee3
hcq2: deps ( #16459 )
...
* start
* sin
* f
2026-06-02 22:34:25 +03:00
George Hotz
82f1c983d4
clean renderer migrations [pr] ( #16472 )
...
* clean renderer migrations
* minor webgpu
* use PARAM UOp as API
* make linter happy
2026-06-02 11:19:00 -07:00
Christopher Milan
9897658895
ci: fix ocelot compilation on macos ( #16471 )
2026-06-02 12:43:31 -04:00
