nimlgen/tinygrad
1
0
Fork 0
forked from tinygrad/tinygrad
Code Pull Requests Activity
14,223 Commits 334 Branches 15 Tags
1a90d06c602b83852eb5f87d22ea052d20eae7aa
Commit Graph
1230 Commits
Author SHA1 Message Date
nimlgenandGitHub 7c1ce50f63 hcq2: epoch (#17376) 
* hcq2: epoch

* x

* minor
 2026-08-03 16:37:25 +03:00
George HotzandGitHub 5a1c641f79 more arg -> val (#17349) 
* more arg -> val

* kimi

* more
 2026-07-31 22:44:24 -07:00
George HotzandGitHub 9082ecef5d use .val to access the value of Ops.CONST (#17347) 2026-07-31 19:51:42 -07:00
sirhcmandGitHub bfc9fc6e0a nicer TinyJit decorator (#17293) 2026-07-29 21:48:00 -04:00
George HotzandGitHub 57ae1bc7a7 rename MULTI to UNSHARD (#17267) 
* rename MULTI to UNSHARD

* comment updates (glm)

* rename method to unshard
 2026-07-28 16:51:41 -07:00
George HotzandGitHub 1757067fa6 add device range as src[1] to multi (kimi) (#17264) 
* add device range as src[1] to multi (kimi)

* cleanup

* fix tests
 2026-07-28 16:39:34 -07:00
George HotzandGitHub 7d48926293 switch _device_num to AxisType.DEVICE range (kimi) (#17252) 
* switch _device_num to AxisType.DEVICE range (kimi)

* improve

* don't exclude device for ranges

* current multi ends the range
 2026-07-28 13:52:47 -07:00
nimlgenandGitHub 8eaeede96d hcq2: do not cache beam (#17234) 2026-07-27 17:16:24 +03:00
qazalandGitHub 19c4d736f2 validate json output of viz.cli in CI (#17232) 
* validate viz.cli --json always prints valid JSON

* highest debug level

* jq empty we don't need a print

* gate that import
 2026-07-27 15:44:24 +09:00
nimlgenandGitHub 165f0626f8 speedy hcq2 (#17225) 
* faster

* x
 2026-07-27 01:31:56 +03:00
chenyuandGitHub 4b6760539b fix _prepare_jit_inputs for weak [pr] (#17222) 2026-07-26 14:03:27 -04:00
nimlgenandGitHub 3946df787d hcq2, cpu is hcq2-ish (#17197) 
* m

* i

* x

* x

* Df

* x

* x
 2026-07-26 02:38:15 +03:00
sirhcmandGitHub 9fdaa4bff2 standardize Program class (#17189) 2026-07-25 00:17:38 -04:00
nimlgenandGitHub 62273d50fc simpler hcq buffer (#17148) 
* simpleer hcq buffer

* x
 2026-07-23 13:28:51 +03:00
chenyuandGitHub c5b2b9242d python speed tweak for compile3 (#17147) 2026-07-23 02:02:44 -04:00
nimlgenandGitHub 4257939e50 remove copyin/copyout from Buffer (#17020) 
* remove copyin/copyout from Buffer

* x

* x

* x

* x
 2026-07-14 19:47:22 +03:00
nimlgenandGitHub e4bdc529c4 hcq2 ci (#17008) 
* hcq2 ci

* x
 2026-07-13 19:29:08 +03:00
George HotzandGitHub 40de90ab19 lil changes to cifar (#16972) 
* lil changes to cifar

* lil changes

* pool
 2026-07-10 14:19:46 -07:00
nimlgenandGitHub 52e060fd7a hcq2: link cache (#16965) 
* x

* x

* mini

* cleaner
 2026-07-10 23:47:16 +03:00
George HotzandGitHub 1df49a7bf5 dtype doesn't need to be explicit in most places now (glm) [PR] (#16951) 
* dtype doesn't need to be explicit in most places now (glm) [PR]

* fix amd pcode

* image idx long/int meh

* real image fix
 2026-07-09 14:41:34 -07:00
George HotzandGitHub 00d01d978d add dtype production rule checked with SPEC=2 (#16939) 
* add dtype production rule checked with SPEC=2

* no None support yet

* detach passes through

* work

* fix

* line len
 2026-07-08 18:04:34 -07:00
chenyuandGitHub 8085bd57ec remove old stale metadata codes [pr] (#16936) 2026-07-08 17:11:15 -04:00
George HotzandGitHub fdffc6c0c8 remove dtypes base (#16931) 
* remove dtypes base

* find/replace bug

* vcount is junk too
 2026-07-08 11:48:11 -07:00
George HotzandGitHub 2aebb6f6c4 move mop cleanup [pr] (#16915) 
* move mop cleanup [pr]

* those are mop cleanups

* more mop_cleanups

* revert

* movement.py
 2026-07-07 14:30:32 -07:00
chenyuandGitHub fb4a781fca fix a few imports [PR] (#16909) 2026-07-07 14:21:13 -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
nimlgenandGitHub 3732d6b43c jit prereqs for new hcq2 (#16807) 
* jit prereqs for new hcq2

* ty
 2026-06-30 23:00:10 +03:00
chenyuandGitHub 3a9b87e48f remove Ops.DEVICE from Ops.PROGRAM [PR] (#16797) 2026-06-29 17:41:31 -04:00
chenyuandGitHub 379b8d37fc read PROGRAM device from ParamArg [PR] (#16796) 
nothing reads from its Ops.DEVICE after this
 2026-06-29 16:06:37 -04:00
nimlgenandGitHub 4c3789661d hcq2: split hcq_schedule into hcq_compile/hcq_link for jit (#16784) 2026-06-28 10:19:48 +03:00
chenyuandGitHub 93a82d42bd Buffer arg -> max_numel [PR] (#16752) 
prep removing UNIQUE and use ParamArg for arg
 2026-06-25 14:50:20 -04:00
chenyuandGitHub 96c9a68205 imrpove some type casts (#16743) 2026-06-25 10:48:28 -04:00
chenyuandGitHub 9ae0a93d0e more const cleanups [PR] (#16682) 2026-06-20 20:41:43 -04:00
nimlgenandGitHub eda0a402d1 hcq2: fix multi (#16661) 2026-06-18 22:56:49 +03:00
chenyuandGitHub df015ad541 remove many type ignores [PR] (#16652) 2026-06-17 21:38:45 -04:00
nimlgenandGitHub 4dc51aff6e hcq2: jit (#16621) 
* hcq2: jit

* x

* x

* minor
 2026-06-15 06:35:35 +07:00
Philip SinitsinandGitHub 76c10cd635 jit: don't memplan buffers reachable from live tensors (#16588) 
The memory planner was suballocating BUFFERs created during JIT capture that are still referenced by external lazy tensor graphs, like the .grad tensors assigned by backward(). The replay then only writes the arena slices, so realizing such a tensor after the call reads freshly allocated memory and silently returns zeros. Hold every BUFFER reachable from a live Tensor instead of only the parameters of the return value; true internals are still planned. Fixes #16571.
 2026-06-12 17:51:54 +03:00
nimlgenandGitHub e5f498de3b hcq2: debug=2 info (#16569) 
* hcq2: debug=2 info

* t

* x

* hcq2: debug=2 info

* x
 2026-06-11 19:52:01 +03:00
qazalandGitHub a83710396c support mselect input to CALL, less kernels in allreduce (#16567) 
* support mselect input to CALL, less kernels in allreduce

* resolve mstack
 2026-06-11 18:10:47 +09:00
nimlgenandGitHub 2c9d2c0d31 jit: memplan before compile (#16560) 2026-06-10 15:05:15 +03:00
nimlgenandGitHub 99e37b1ee3 hcq2: deps (#16459) 
* start

* sin

* f
 2026-06-02 22:34:25 +03:00
George HotzandGitHub 1e7f1dcf49 add ParamArgs [pr] (#16421) 
* add ParamArgs

* fix export

* cleanups

* fixes

* simpler
 2026-05-28 19:17:17 -07:00
nimlgenandGitHub 0c385e31c6 hcq2 rewrite (#16375) 
* hcq2 rewrite

* fi

* x

* simpler
 2026-05-27 22:25:35 +03:00
George HotzandGitHub 156a4438d9 rename BUFFER_VIEW to SLICE (#16391) 
* rename BUFFER_VIEW to SLICE

* fix comments
 2026-05-26 18:15:00 -07:00
George HotzandGitHub 7f1b02854e bufferview offset is units of input dtype (#16378) 2026-05-26 08:49:31 -07:00
George HotzandGitHub 689ab6a49f move buffer view offset to src (#16364) 
* this work?

* failed
 2026-05-25 17:07:55 -07:00
nimlgenandGitHub 64095cf2e2 use get_buf in exec_kernel (#16356) 2026-05-25 15:13:40 +03:00
chenyuandGitHub e1715b3b92 extent jit const error to deviceless inputs (#16276) 2026-05-20 02:02:45 -04:00
chenyuandGitHub db639ebe3e deviceless const from UOp (#16243) 2026-05-18 14:14:12 -04:00
nimlgenandGitHub bef5f717bc fix nolocals and beam (#16232) 2026-05-16 18:09:19 +03:00