f902af4f0b
increase metal ci test timeout to 20 minutes ( #4920 )
...
make it less annoying for now
2024-06-11 18:45:51 -04:00
7f3d9e6d94
revert hsa autogen removal ( #4914 )
...
* Revert "only install comgr in AMD CI (#4909 )"
This reverts commit 7f03420d05
2024-06-11 12:55:45 -04:00
7f03420d05
only install comgr in AMD CI ( #4909 )
...
* test
* delete hsa autogen
2024-06-11 06:19:33 -04:00
8b5bcf309a
process replay in all of CI ( #4884 )
2024-06-10 14:49:29 -04:00
f42183ba28
hotfix: relax cifar to 93.2
2024-06-09 13:09:21 +02:00
654a8b9ef7
retire hsa ( #4885 )
...
* retire hsa
* EMULATE_AMD
2024-06-09 11:33:03 +03:00
6327b50e51
amd in benchmarks ( #4861 )
...
* amd in benchmarks
* remove all hsa
2024-06-08 23:24:46 +03:00
66dfd5e7bf
faster codegen process replay ( #4858 )
...
* faster codegen process replay
* use self.copy
* regenerate
* delete copy
* test a real error [run_process_replay]
* revert the error change
2024-06-07 16:20:57 +03:00
0db9674dea
skip process replay on master ( #4808 )
2024-06-03 12:29:28 +03:00
f64fa51a64
process replay for test/* ( #4799 )
...
* add input to unit tests [run_process_replay]
* add setup [run_process_replay]
* run tests [run_process_replay]
* add cuda and amd [run_process_replay]
* run everything but BEAM=2 [run_process_replay]
* skip export_model [run_process_replay]
* fix amd CI
* add concurrency back
2024-06-03 12:01:58 +03:00
240d6b5bc0
process replay benchmarks ( #4668 )
2024-06-01 14:36:21 +03:00
bd2e7c8b31
amd registers from file ( #4778 )
...
* amd registers from file
* remove commentes
* linetr
* no off
2024-05-31 18:48:57 +03:00
a4de81e9a6
Update ocelot version ( #4715 )
2024-05-24 14:32:53 -04:00
38bc38cdff
fix llama example quantize ( #4699 )
...
* fix llama example quantize
import quantize layers from new example llama3
add to mac benchmark
* fix that
* save the files
2024-05-23 15:35:26 -04:00
72560e30fe
add CACHELEVEL=0 to tinybox green GEMM BEAM ( #4693 )
...
* add CACHELEVEL=0 to tinybox green GEMM BEAM
* BEAM=4 is more stable
2024-05-22 23:59:50 -04:00
af56f0e68a
fix HSA/KFD load for system-wide installation ( #4218 )
...
Co-authored-by: wozeparrot <wozeparrot@gmail.com>
2024-05-22 20:33:21 -07:00
12339f6564
disable cuda test in ci ( #4630 )
...
Co-authored-by: chenyu <chenyu@fastmail.com>
2024-05-22 23:23:32 -04:00
498cf3e7e0
fuzzer path search for DEFINE_ACC ( #4656 )
...
* insert acc
* add test_ops
* find toposorts
* todo - not yet ready
* remove the import
* atol and childless children
2024-05-23 00:50:01 +03:00
458a3961eb
catch compile errors in uops tests ( #4672 )
...
* use helper and compile
* llama beam=2
* ast length
* skip float4, fix hsa
* use empty tensors
2024-05-21 12:20:35 +03:00
00432496d7
feat: tinyboxgreen ( #4366 )
...
* feat: tinyboxgreen
* feat: tinyboxgreenv2
* fix symlink weights
* fix: remove llama 2 70b for now
* feat: naming
* fix: remove extra cifar steps
* feat: disable mixtral on nvidia
2024-05-20 22:39:34 -04:00
8a0d1ca7bb
CI test timeout 20 min -> 10 min ( #4645 )
...
if it takes more than 10 usually setup fails anyway. also updated matmul_kfd -> matmul_amd in benchmark
2024-05-18 13:58:28 -04:00
b74cc1d01a
uops cleanup ( #4634 )
...
* def add cleanup
* minor speedup
* add back ptx speed
* a little faster
* merge that
* only linearize once for ptx
* two graph rewrites for ptx, bug?
2024-05-17 20:02:38 -07:00
07b350a8f4
new uops is an actual graph ( #4560 )
...
* new uops is an actual graph
* it's way slower
* simpler
* fix define acc
* render_loop unique
* ops test pass
* add pattern matcher back, there's bugs
* rewrite
* use priority queue
* recursive children
* fix tests
* fix tests with SINK
* fix abstractions
* fix assembly
* simpler
* link define_acc
* fix DEFINE_ACC placement
* type verify
* full cmp
* fix cmp
* ACCESS_ACC
* insert DEFINE_ACC
* fix PHI
* recursive rewrite
* fix many tests
* sum collapse
* more patterns
* correct change
* fold arange
* fix that lin test
* space
* big folding rule works
* close
* has more maxes, meh
* cached node replace
* set changed
* simplest folding yet
* works
* works
* DIV
* all tests pass
* del
* fuzz linearizer fails
* sum_collapse
* test depth 2 cf
* fix lin test 14
* fix clang depth
* disable that
* failure 14 is fixed
* fix ptx
* failure 27 is fixed
* fix llama
* run_cnt
* Revert "Optimize PTX gated loads index calculation (#4304 )"
This reverts commit d97d5a7689
2024-05-17 18:00:18 -07:00
ca1df20fa9
benchmark name fix - resnet eval is on eval data ( #4628 )
2024-05-17 12:56:12 -04:00
e5d4e6a8aa
BEAM=2 in green CI for 100 TFLOPS ( #4624 )
2024-05-16 23:28:28 -04:00
eb9689336e
nv mockgpu ( #4600 )
...
* mockgpu nv
* works
* comment that out
* fix merge
* setup gpuocelot
* install packages
* not run all of them
* passes
* fix ci
* almost
* should pass
* linter
* linter 2
* try this?
* ugn, not supported
* ci
* remove ticket from description
* better descs
2024-05-15 23:46:08 +03:00
5ba611787d
move image into tensor.py. delete features ( #4603 )
...
* move image into tensor.py
* change setup.py
* openpilot tests need pythonpath now
2024-05-15 10:50:25 -07:00
afa9753d39
ruff cleanup ( #4594 )
...
* check editor config
* no editorconfig, it doesn't work
* ruff cleanups
2024-05-14 21:16:14 -07:00
9425973bc7
docs cleanup and move ( #4593 )
...
* cleanup and move
* docs-legacy is gone
* don't update setup.py
2024-05-14 20:44:59 -07:00
fd02ab1e8b
move disassemblers and openpilot ( #4592 )
...
* move disassemblers and openpilot
* delete junk
* put that in pre-commit
* fixup readme
2024-05-14 19:30:02 -07:00
9b02aef45a
remove rhip ( #4579 )
...
* remove rhip
* remove hip runner
2024-05-14 17:58:19 +03:00
2131556c2c
amd mockgpu ( #4535 )
...
* start mock amd gpu
* virt files
* cleaner
* init ci
* small fixes
* linter
* better?
* ugh
* linter
* fix
* diable some
* run shorter
* fixes
* add hcq test
* fix
* fix cmd revert
2024-05-14 14:28:04 +03:00
5de4a46f10
re-enable gpt2 half/beam mac benchmark ( #4496 )
...
* re-enable gpt2 half/beam mac benchmark
from fuzzer it seems to be flaky due to numerical issue, not kernel bug. we used to have half in splitted reduce.
run this in M1 Max for 20 loops and it's fine
* that should be jitted
2024-05-09 19:15:32 -04:00
c508eb7425
revert the removal of CAST_BEFORE_VIEW ( #4471 )
...
this brings most of the memory gain for resnet back.
2024-05-08 00:14:29 -04:00
760776c59d
merge EfficientNet to C with clang job ( #4426 )
...
* merge ImageNet to C with linters
* add to clang
* delete from linter
2024-05-05 20:33:12 +03:00
d4062cb6fc
NV tensor_cores in kernel.py ( #4399 )
2024-05-02 22:33:08 -04:00
dce7ac0160
NOCLANG=1 for tinybox green ci. ( #4378 )
...
CLANG was disabled for tinybox red for speed
2024-05-01 13:31:01 -04:00
4a26718ca9
feat: tinyboxgreen ( #4365 )
2024-04-30 19:05:37 -04:00
fdc8fabae5
disable flaky mac gpt2 beam benchmark and add back cifar mac with JIT=2 ( #4358 )
...
* debug flaky mac gpt2 beam run
* disable for now
2024-04-30 10:41:37 -04:00
bb849a57d1
[MLPerf] UNet3D dataloader ( #4343 )
...
* add support for train/val datasets for kits19
* split dataset into train and val sets
* add tests for kits19 dataloader
* add MLPerf dataset tests to CI
* update unet3d model_eval script
* fix linting
* add nibabel
* fix how mock dataset gets created
* update ref implementation with permalink and no edits
* clean up test and update rand_flip implementation
* cleanups
2024-04-28 22:34:18 -04:00
3ec4b745d6
JIT=2 for mac cifar benchmark ( #4300 )
...
also double BS for resnet training benchmark to match submission target
2024-04-25 18:33:40 -04:00
c1fbacb182
resnet benchmarks use DEFAULT_FLOAT=HALF ( #4285 )
...
also update LR default to scaled based on 1536 (the BS we are submitting)
2024-04-24 12:10:57 -04:00
002a14088e
Ptx store gate cast to bool ( #4284 )
...
* Cast gate to bool
* Update
* Add PTX fuzzing to benchmark
2024-04-24 11:43:44 -04:00
dbe3e1d548
or true fixes ci ( #4283 )
...
* or true fixes ci
* all with two pipes
2024-04-24 20:48:26 +08:00
759b4f41c3
few more KFD -> AMD ( #4262 )
...
benchmark gemm and default_parallel
2024-04-23 10:15:37 -04:00
3f6c7ca8bf
test: fix test_tensor_core_padded on CUDA and add to benchmarks ( #4258 )
...
* test: fix test_tensor_core_padded on CUDA and add to benchmarks
* fix linter
* run both tests in one call
2024-04-22 23:22:11 -04:00
bbb0ad4800
wmma: widen TC usage in search by using PADTO on TC axes when possible ( #4216 )
...
* wmma: widen TC usage in search by using PADTO on TC axes when possible
* test: start tests for the new padding TC behavior
* search: upgrade padded TC search to TC_OPT >= 2
* test: add behavior and correctness test for padded TC
added optional argument to apply_tensor_core to set TC_OPT level
* linearizer: add tests for the PADTO behvaior and docs
2024-04-22 16:50:31 -04:00
9e53d6cffa
hotfix: 8000 lines
2024-04-22 20:58:16 +04:00
e6227bdb15
nv driver ( #4044 )
...
* start
* fix err 93
* gpu
* ioctl mappings
* alloc like cuda
* semaphores
* wait for semaphores value
* start ops_nv
* very simple kernels work
* init several gpus
* qmd dumper
* dirty, but most of kernels work
* always all test_ops
* progress, more tests, stable
* test_ops passes, gpt2 works
but wth big fifo, wrap of fifo doesn't work, i think it's something coherency releated
* need better sync
* fix sync
* alloc2
* all tests pass!
* cleanup 1
* cleanup
* multigpu, simple transfer
* fix sync
* correct init
* nv_gpu autogen + sync bug fix
* clean extra/nv_gpu_driver
* p2p
* clean up
* remove old gen
* small fixes
* cleanup
* cleanup 2
* small fixes
* bigger queue size
* cleanups
* wait
* fixed signals for devs
* fix hang + parallel beam
* small fixes
* detect when local memory is big in kernel
* correct assert
* small fixes
* correct tls size est
* one va space
* less lines
* shorter
* save 2 lines
* save some lines
* remove type ignores
---------
Co-authored-by: George Hotz <72895+geohot@users.noreply.github.com>
2024-04-22 19:50:20 +04:00
f1d9d0a151
cleanup external_test_opt ( #4234 )
...
no more OPT=2 or OPT=3, check strict number of kernels, enabled tests that fusion works now
2024-04-20 04:00:08 -04:00
chenyu
qazal
George Hotz
nimlgen
Szymon Ożóg
Yury Zhuravlev
wozeparrot
George Hotz
Francis Lata
Francis Lam