github/tinygrad
1
1
Fork 0
mirror of https://github.com/tinygrad/tinygrad.git synced 2026-02-16 01:26:29 -05:00
Code Issues Packages Projects Releases Wiki Activity
2,952 Commits 227 Branches 14 Tags
86fbd413f3b9ac3fe01ccbceb0caa7eaa8f423fb
Commit Graph

2952 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Davi Silva
136dbd8b36 HIP CI that compiles (to RDNA3) but doesn't have to run (#2482) 
* hip amd compilation

* gate the test properly

* cleanup unused import

* remove superfluous numpy conversion

* add SpeedyNet tests (f32 [passes] & f16 [fails])

* make CI verbose (error log from hip compiler)

* test the real ops_hip

* Merge branch 'tinygrad:master' into ci/hip-compilation

* fix CI

* cleanup

* really fix CI

* Fix CI Three: the refixening

---------

Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com>
 2023-11-27 21:17:06 -08:00
George Hotz
756b01f46f why were these ever called buffer (#2483) 2023-11-27 21:02:07 -08:00
George Hotz
acbe6d1b53 Revert "HIP compilation on CI targeting RDNA3 (#2459)" (#2481) 
This reverts commit d275ff930a.
 2023-11-27 20:41:21 -08:00
qtkite
cb507a9389 Remove the toCPU copy (#2445) 
* Remove the rawbuffer copy in runtime/lib.py on line 44

* remove buffer view

* added metadata back, oops

* delayed cpu testcase

* whitespace

* whitespace

* buffer behavior as is

* Update test_jit.py
 2023-11-27 20:37:13 -08:00
Davi Silva
d275ff930a HIP compilation on CI targeting RDNA3 (#2459) 
* hip amd compilation

* gate the test properly

* cleanup unused import

* remove superfluous numpy conversion

* add SpeedyNet tests (f32 [passes] & f16 [fails])

* make CI verbose (error log from hip compiler)

* test the real ops_hip

* Merge branch 'tinygrad:master' into ci/hip-compilation

* fix CI

* cleanup

* really fix CI
 2023-11-27 20:33:11 -08:00
Yingbo Ma
d43485ae9e Fix graph_uops (#2457) 
* Load networkx when we need to graph uops

* Document GRAPHUOPS

* import nx in `graph_uops`
 2023-11-27 18:42:48 -08:00
Paul Gustafson
98cd9e8926 Add assertion to prevent nonsense mod values (#2474) 2023-11-27 18:37:44 -08:00
Davi Silva
186ac77ec3 Update hip_matmul.py (#2480) 2023-11-27 18:36:19 -08:00
chenyu
7f9a4c1285 fp16 and noshow flags for gpt2 (#2470) 2023-11-27 16:23:03 -05:00
qazal
e267a93124 reset seed on every run (#2468) 2023-11-27 12:55:54 -08:00
George Hotz
9e07824542 move device to device.py (#2466) 
* move device to device.py

* pylint test --disable R,C,W,E --enable E0611

* fix tests
 2023-11-27 11:34:37 -08:00
qazal
262cd26d28 Simplify openpilot kernel (#2460) 
* a conditional with the same results either way is a noop

* add unit test
 2023-11-27 10:02:27 -08:00
chenyu
61a80a0675 asserts LtNodes of SumNode with MulNode of Nodes (#2465) 2023-11-27 12:56:59 -05:00
chenyu
c4dfde761e remove the commented import (#2463) 2023-11-27 11:50:41 -05:00
Akshay Kashyap
a031afb2f6 Update display_name in resnet50 example (#2454) 2023-11-26 16:07:36 -08:00
Paul Gustafson
1d89c018fa Add isinstance check before gcd call in SumNode.__lt__ (#2450) 
* Add isinstance check before gcd call

* Delete blank lines

* Fix unit test typo

* Delete blank lines again

---------

Co-authored-by: Paul Gustafson <paul.gustafson@theambrusgroup.com>
 2023-11-26 13:05:04 -08:00
Paul Gustafson
58b1dd463e Add error code to type: ignore (#2451) 
Co-authored-by: Paul Gustafson <paul.gustafson@theambrusgroup.com>
 2023-11-26 13:04:10 -08:00
George Hotz
8e9cdef61f clean up the buffers (#2447) 
* clean up the buffers

* remove allocate_output

* functools.lru_cache is methodcache

* add TestShapeTrackerSize

* cache_clear

* no 0 sz buffer, add _ on functions that shouldn't be imported

* fix size

* if -> while
 2023-11-26 11:02:29 -08:00
George Hotz
f6f712e609 split out the three steps of exec_ast (#2446) 
* split out the three steps of exec_ast

* clean up extra args

* cleanups, bugfix

* allocate is a more normal name

* get_optimized_linearizer is better
 2023-11-26 09:07:37 -08:00
chenyu
511310737e test_linearizer_failures to run on all backends (#2443) 
* test_linearizer_failures to run on all backends

* test ubuntu and cuda

* failed only in CUDA CI

* move asserts
 2023-11-26 01:17:29 -05:00
George Hotz
c42d2c4731 strip whitespace 2023-11-25 14:09:06 -08:00
George Hotz
9eb2746d62 fix copy issue + add regression test (#2441) 2023-11-25 14:06:08 -08:00
andresgit
259a869fc1 Fix UnicodeDecodeError when debugging on Intel APU (#2421) 
* test DEBUG=5

* print prg if NVIDIA, fixes error on Intel APU
 2023-11-25 12:30:50 -08:00
George Hotz
7170a9a057 coder.py can write and run code (#2439) 
* wip mistral

* coder

* touchups

* cleanups

* mistral cleanups

* clean up cache create

* download the weights, fix tests

* fix llama loading

* global fixup

* clean up all

* move llama model

* cleanups

* Revert "cleanups"

This reverts commit a71c5d59eb.

* fine, leave it
 2023-11-25 12:27:54 -08:00
Davi Silva
df41a57e09 Fix: missing n_kv_heads for smaller models from huggingface (#2438) 
* fix: missing n_kv_heads for smaller models from huggingface

* a lil golfing
 2023-11-25 10:29:04 -08:00
George Hotz
96c12fdeab multibatch gpt2 (#2432) 
* support multibatch gpt-2

* multi output

* no default JIT in CI
 2023-11-24 18:10:10 -08:00
Tobias Fischer
5326bbc9a6 fix causal mask in Tensor class (#2425) 
* fixed causal mask in Tensor class

* added tests for scaled attention against pytorch

* cleaned up test formatting

* removed duplicate test
 2023-11-24 18:38:18 -05:00
chenyu
9a5d0e70de Device.DEFAULT instead of getenv to exclude tests (#2429) 2023-11-24 17:10:24 -05:00
chenyu
6223f8894d clean up ast_parse (#2428) 
* clean up ast_parse

* separate loops
 2023-11-24 16:43:32 -05:00
George Hotz
8ff2e13550 From teeny (#2426) 
* changes from teenygrad work

* support not supporting ImageDType/PtrDType

* fixups from teeny
 2023-11-24 12:50:56 -08:00
chenyu
9ae83fba04 flatten instead of reduce, improve type inference (#2423) 2023-11-24 13:19:22 -05:00
Francis Lata
7169de57e2 Update VITS to use fetch helper (#2422) 
* use fetch helper on vits

* remove duplicate weight loading
 2023-11-24 08:50:03 -08:00
George Hotz
857d440ea7 fail means fail (#2391) 
* flip order

* cleanup and comment out failing test
 2023-11-24 08:27:39 -08:00
George Hotz
1f4231a8f9 global pipefail 2023-11-24 08:03:49 -08:00
nimlgen
e68aebfff9 bring hip graph back (#2385) 
* bring hip graph back

* share with metal

* fix linter

* remove hasattrs

* Update ops_hip.py

* hip wrapper does not use _buf

---------

Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com>
 2023-11-24 07:53:44 -08:00
George Hotz
46b05daf7c simple display_name (#2416) 
* simple display_name

* name functions

* self.global_size [1]

* CompiledASTRunner display_name

* assert sizes are len 3

* 3 dims for GPU

* auto self.global_size
 2023-11-23 19:50:23 -08:00
George Hotz
12023b6824 onnx ops cleanup (#2413) 
* onnx ops cleanup

* revert those
 2023-11-23 18:39:49 -08:00
George Hotz
8f89e21fca torch and numpy don't share ops anymore (#2412) 
* torch and numpy don't share ops anymore

* that should be filtered out elsewhere

* still const

* graph + enet example cleanup

* hmm, we do still need it because of symbolic
 2023-11-23 16:58:10 -08:00
George Hotz
193be14b6c that had bugs, force an order (#2411) 2023-11-23 15:52:16 -08:00
George Hotz
65f4e6971b beautiful_mnist.py link 2023-11-23 14:58:22 -08:00
George Hotz
1b3b8de5e2 update readme examples 2023-11-23 14:54:52 -08:00
George Hotz
5bb720a777 Cocoa is no longer used 2023-11-23 14:31:21 -08:00
George Hotz
095e2ced61 add name support to fetch (#2407) 
* add name support

* use fetch in gpt2

* remove requests from main lib, networkx also optional

* umm, keep that assert

* updates to fetch

* i love the walrus so much

* stop bundling mnist with tinygrad

* err, https

* download cache names

* add DOWNLOAD_CACHE_VERSION

* need env.

* ugh, wrong path

* replace get_child
 2023-11-23 14:16:17 -08:00
nimlgen
397c093656 fix wait in jit (#2408) 2023-11-23 13:54:13 -08:00
qazal
b927942d58 Move HIP render logic to its dedicated place (#2394) 
* update HIP language

* vectorized render_cast with special treatment for hip only

* test coverage for all cases

---------

Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com>
 2023-11-23 13:03:29 -08:00
Francis Lata
6d672785db Update Whisper to use fetch helper (#2401) 
* update whisper to use new fetch helper

* simplify file opening

* update name

* update key name to "downloads-cache"
 2023-11-23 12:59:59 -08:00
George Hotz
0505c5ea50 remove force_wait, refactor to graph (#2405) 
* remove force_wait

* refactor

* get rid of stupid ASTRunner

* fix del in diskbuffer

* BufferOps.FROM_UNDERLYING

* put offset in the rawbuffer

* fix bugs

* use exec
 2023-11-23 12:46:07 -08:00
Ivan Beňovic
c5d585ea35 Fix Triton README broken link (#2406) 
* Remove triton from README

* Fix broken link
 2023-11-23 12:38:17 -08:00
chenyu
b27c845531 minor cleanup for View strides (#2404) 2023-11-23 13:40:01 -05:00
chenyu
64aa2f4156 clean up to_shape_strides (#2402) 2023-11-23 13:04:00 -05:00