Commit Graph

78 Commits

Author SHA1 Message Date
YassineYousfi
a3f3e0c122 onnxmodel: fp16_to_fp32 (#30080)
* onnxmodel: force fp32

* rename

* rename this too
old-commit-hash: 04e239f7ed
2023-09-27 22:37:03 -07:00
Mitchell Goff
3e1903a6cb Converted onnx runner to python (#29856)
old-commit-hash: 9cc00c2101
2023-09-08 16:45:38 -07:00
Mitchell Goff
5732002296 Rewrite dmonitoringmodeld in python (#29740)
* Added dmonitoringmodeld.py

* Removed dmonitoringmodeld.cc

* Use ModelRunner helpers from runners/__init__.py

* Fixed DriverStateResult field ordering

* Some bug fixes

* Set calib input

* Look ma, no loop!

* Bump dmonitoringmodeld cpu usage in test_onroad

* Fixed memory leak caused by np.ctypes.data_as

* Formatting fixes

* chmod +x

* remove USE_ONNX_MODEL

* Realtime priority 1, formatting fixes
old-commit-hash: 503fa121ee
2023-09-07 19:46:43 -07:00
Mitchell Goff
cd7d73eaf8 Added ModelRunner helpers in modeld/runners/__init__.py (#29737)
* Added ModelRunner helper class

* import Runtime in __init__.py
old-commit-hash: d5a55136a1
2023-08-31 14:19:20 -07:00
mitchellgoffpc
aca5580900 Move Runtime into runmodel_pyx
old-commit-hash: f91e955a86
2023-08-31 13:51:50 -07:00
mitchellgoffpc
51066175f5 Set ADSP_LIBRARY_PATH in snpemodel_pyx
old-commit-hash: 18526c58b6
2023-08-31 13:11:39 -07:00
mitchellgoffpc
e9e98abe89 Remove USE_THNEED check from run.h
old-commit-hash: e0287fa93e
2023-08-25 19:05:49 -07:00
mitchellgoffpc
68e2c74cd3 Remove thneed logic from snpemodel
old-commit-hash: ef8e0fef55
2023-08-25 14:37:08 -07:00
Mitchell Goff
5033da5e27 Added cython bindings for model runners and commonmodel (#29607)
* Added cython bindings for model runners and commonmodel

* Removed cython language_level=3

* loop to set CXXFLAGS for both envs
old-commit-hash: e2e39d100b
2023-08-24 17:00:26 -07:00
Dean Lee
228d30d5a9 cpplint: add filter whitespace/tab (#29588)
old-commit-hash: 9c7bf50703
2023-08-24 08:06:58 -07:00
Dean Lee
0984736972 cpplint: add filter build/include_subdir (#29585)
old-commit-hash: 21da5a216f
2023-08-24 15:42:17 +01:00
Dean Lee
97e9d55588 cpplint: build/include_what_you_use (#29556)
* include_what_you_use

* remove comments

* include <memory>

---------

Co-authored-by: Adeeb Shihadeh <adeebshihadeh@gmail.com>
old-commit-hash: 2d99521e75
2023-08-23 13:25:17 -07:00
mitchellgoffpc
8a1e4d8575 Set paths for cl kernels and onnx_runner.py in modeld SConscript
old-commit-hash: 7699444816
2023-08-21 19:38:40 -07:00
mitchellgoffpc
b1a10e45d1 Move USE_RUNTIME flags into runmodel.h
old-commit-hash: 24090d2121
2023-08-21 19:26:48 -07:00
mitchellgoffpc
b69de80962 exit if onnxmodel exec fails
old-commit-hash: 8c478708ee
2023-08-21 19:23:23 -07:00
Justin Newberry
f33b4de80b Ruff: b905 (strict zip) (#29336)
* added mutable default args

* most of the Bs

* add comment about lrucache

* b905
old-commit-hash: e4ead4f183
2023-08-11 16:13:51 -07:00
Justin Newberry
3ee0786de3 CI: Pylint to ruff (#29294)
* pylint to riff

* pylint to riff

* pylint to riff

* revert more

* undo exclude removal

* exclude tinygrad

* set line length

* pylint exclusions to ruff

* same excludes as old linter

* fix tools QA

* remove unrequired check

* revert linting third_party

* ignore e402
old-commit-hash: 7fdd9fc37c
2023-08-09 13:26:54 -07:00
Justin Newberry
ff96892511 Revert "CI: pylint to ruff (#29276)" (#29284)
This reverts commit ffad7279cbad3c4f2ea786287b4624d3a3aa18d7.
old-commit-hash: a91764954b
2023-08-08 14:24:04 -07:00
Justin Newberry
99006365a3 CI: pylint to ruff (#29276)
* pylint to riff

* pylint to riff

* pylint to riff

* revert more

* undo exclude removal

* exclude tinygrad
old-commit-hash: cb31a53d8a
2023-08-08 13:41:30 -07:00
Dean Lee
b80325aa04 swaglog: enable type-checked against format string (#29096)
* check&fix fmt

* bump rednose

* correct format string

* bump rednose

---------

Co-authored-by: Adeeb Shihadeh <adeebshihadeh@gmail.com>
old-commit-hash: 940954c13e
2023-07-31 17:49:50 -07:00
Kacper Rączy
fe2fc9dd11 onnx_runner: force determinism (#29000)
old-commit-hash: 8ce72760e7
2023-07-18 05:52:07 +02:00
Adeeb Shihadeh
e83ec32c7b dmonitoringmodeld: add more logging around the init (#28920)
old-commit-hash: 8d13ddb943
2023-07-13 11:58:51 -07:00
Mitchell Goff
95051090a1 Refactor model runners (#28598)
* Started work on model runner refactor

* Fixed some compile errors

* everything compiles

* Fixed bug in SNPEModel

* updateInput -> setInputBuffer

* I understand nothing

* whoops lol

* use std::string instead of char*

* Move common logic into RunModel

* formatting fix
old-commit-hash: c9f00678af
2023-07-06 22:26:11 -07:00
YassineYousfi
52e1cf2e64 fix input order in onnxmodel (#28274)
old-commit-hash: d0a8b3780c
2023-05-23 16:11:51 -07:00
Mitchell Goff
22c5053bbc Nicki Minaj Model (#28218)
* New model: 2ff7490f-8a2f-4c0d-87a8-b0fa3a4a6a71

* Updated model_replay_ref_commit
old-commit-hash: 328b5d9d47
2023-05-22 15:11:04 -07:00
Harald Schäfer
f315819d47 Navmodel and driving style: update runner (#26762)
* Navmodel and driving style: update runner

* No driving style yet

* thneedrunner without extra inputs

* nav feature ref
old-commit-hash: 160c9ba1d2
2022-12-11 15:25:36 -08:00
HaraldSchafer
a8ec6f4cfd Rocket Launcher Model (#25963)
* 1456d261-d232-4654-8885-4d9fde883894/440 6b7d7cec-ead8-40f3-86cc-86d52c9b03fe/300

* compute only 9 tokens: 1456d261-d232-4654-8885-4d9fde883894/440 6b7d7cec-ead8-40f3-86cc-86d52c9b03fe/300

* tinygrad: cleanup gather

* 1456d261-d232-4654-8885-4d9fde883894/440 6b7d7cec-ead8-40f3-86cc-86d52c9b03fe/700

* empty commit for tests

* bump tinygrad

* dont use tinygrad matmul for now

* bump tinygrad

* 1456d261-d232-4654-8885-4d9fde883894/440 e63ab895-2222-4abd-a9a5-af86bb70e260/700

* float16 1456d261-d232-4654-8885-4d9fde883894/440 e63ab895-2222-4abd-a9a5-af86bb70e260/700

* increase steer rate cost

* Revert "increase steer rate cost"

This reverts commit 74ce9ab9be7ef17ecfec931f96851b12f37f2336.

* fork tinygrad

* empty commit for tests

* basics

* Kinda works

* new lat

* new tuning

* Move LATMPCN so scons compiles

* Update long weights

* Add tinygrad optim

* Update model ref

* update weights

* Update ref

* Try

* Error message for field ignore

* update model regf

* ref commit

* Fix onnx test

Co-authored-by: Yassine Yousfi <yyousfi1@binghamton.edu>
old-commit-hash: cb0b7375b7
2022-10-05 21:43:38 -07:00
George Hotz
273fab518b modeld: delete unused SNPE stuff after move to tinygrad (#25635)
* delete unused stuff

* remove CL interceptor from thneed since we don't use SNPE anymore

* remove dead files from release

* that's removed

* oops, didn't save
old-commit-hash: 6c39382d71
2022-09-01 18:28:20 -07:00
George Hotz
58f2a941da modeld: PC Thneed prereqs (#25615)
* pc thneed prereqs

* ugh, out of date

* that can stay private

* memcpy here is fine in SNPE variant

* release files

* thneed docs don't work anymore. they didn't look too useful

Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: b6e355a933
2022-08-30 17:57:14 -07:00
ZwX1616
cd2a98aa11 fullframe DM model (#24860)
* Revert "put cereal on master"

This reverts commit 4646c132bae7358079c9b2867725f8382906c1e5.

* Revert "Revert fullframe DM model (#24812)"

This reverts commit 59e8af4c3101785cead69a9880cc03e0a18081e1.

* revert revert cereal

* clip6

* 0.8 is fair

* Fiction compensation should be based on error

* Update refs

* Add deadzone

* not that

* good mg

* ref

* ref

* ee8f

* minor tweak

* ref

* recompile

* ref

* cereal

* match driverstatus

* new ref

* new ref

* pass token through jenkins credentials

* quote

* fix snpe dead weights

* final ref

Co-authored-by: Harald Schafer <harald.the.engineer@gmail.com>
Co-authored-by: Adeeb Shihadeh <adeebshihadeh@gmail.com>
old-commit-hash: 1f2f9ea9c9
2022-06-20 16:24:51 -07:00
ZwX1616
c73d4e5fc6 Revert fullframe DM model (#24812)
* Revert "fullframe DM: flip RHD yaw to use matching thresholds"

This reverts commit ce7daabc8847d18ba46e5d1879f5a6958d04ccc7.

* Revert "fullframe DM model (#24762)"

This reverts commit 817be81fb19004f4873881f6b29dcdfffbe7e3a8.

* revert cereal
old-commit-hash: c646eeee0a
2022-06-10 16:16:46 -07:00
ZwX1616
02f0f39b52 fullframe DM model (#24762)
* get log

* simplify two nonsense

* not needed

* libyuv is a joke

* clean up

* try small

* fast but not bad

* working

* clean up driverview

* simplified

* thats mirrored

* smol

* tweak

* ref is screen

* w/ ee

* update camera model

* no if TICI

* start

* update pose thresh

* less cpu more dsp

* new libyuv

* new snpe

* add files

* test

* should be fast

* update out len

* trigger test

* use master snpe

* add cereal

* update cereal

* refactor parsing

* missing ;

* get

* wrong type

* test model

* use driver data

* 10829278-72fe-4283-a118-2cef959ce174/1550

* no pf

* adapt driverview

* ;

* rhd learner

* update libyuv buildi x64

* ad4337ea

* remove blink slack

* test

* no

* use toggle

* b16

* fix for nv12

* 5b02cff5 both

* update test

* update cereal

* update cereal

* update cereal

* v2 packets

* revert libyuv

* no /

* update snpemodel

* ;

* memcpy

* fix test

* use toggle in driverview

* update power

* update replay

* Revert "update replay"

This reverts commit 1d0979ca59dbc89bc5890656e9501e83f0556d50.

* update model ref

* halve cpu

* fake 8bit onnx runner

* same thresh as report

* cereal master

Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: d6c07a6b15
2022-06-08 20:13:46 -07:00
Adeeb Shihadeh
3df7552b7b modeld: cleaner exit from onnx runner
old-commit-hash: 921a411da5
2022-06-02 20:16:21 -07:00
Willem Melching
f7e2813438 Small fixes for macOS build (#24620)
* Small fixes for macOS build

* remove gpu_libs
old-commit-hash: 1cc652602f
2022-05-23 13:46:48 +02:00
Adeeb Shihadeh
f4675fc260 Merge common/ and selfdrive/common (#24556)
* Merge common/ and selfdrive/common

* fix that

* fix version

* fix unit tests
old-commit-hash: cb8885cffb
2022-05-18 14:11:57 -07:00
George Hotz
1b18cef243 thneed: add flag to enable optimizer (#24568)
* improve the thneed compiler

* only init thneed if we are using the GPU

Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 0fc4b4df98
2022-05-17 17:39:18 -07:00
Shane Smiskol
b3e7498f27 onnxruntime: initialize CUDA provider before predictions (#24330)
* initialize CUDA runtime before predictions

* dmond pauses modeld execution...

* done in other PR
old-commit-hash: 60b7114024
2022-04-26 11:49:32 -07:00
Adeeb Shihadeh
5a9644261f remove comma two support (#24248)
* remove comma two support

* cleanup release files

* little more

* more libs

* no more gralloc

* add snpe back
old-commit-hash: 5c48e7bc86
2022-04-18 17:55:23 -07:00
ZwX1616
e8a8e37185 DM: more precision running on DSP + e2e outputs (#23900)
* update cereal

* run but not use

* log distraction type

* regression scaling

* clean up naming

* add calib buf

* add to header

* fake model

* no calib model

* adjust threshs

* 018a305f

* fix bn

* tweak1

* tweak2

* 0ff2/666

* tweak3

* t4

* t5

* fix out of bound

* skip when replaying old segments

* update ref

* fix onnxmodel

* get calib

* update model replay refs

* up ref
old-commit-hash: de4031c98e
2022-03-15 19:02:21 -07:00
George Hotz
7efe157848 thneed: a more sane way of doing record/debug (#23938)
Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 5c5a56c5e6
2022-03-10 16:33:35 -08:00
Willem Melching
ad5ccabce3 bigmodel (#23684)
* Added wide cam vipc client and bigmodel transform logic

* Added wide_frame to ModelState, should still work normally

* Refactored image input into addImage method, should still work normally

* Updated thneed/compile.cc

* Bigmodel, untested: 44f83118-b375-4d4c-ae12-2017124f0cf4/200

* Have to initialize extra buffer in SNPEModel

* Default paramater value in the wrong place I think

* Move USE_EXTRA to SConscript

* New model: 6c34d59a-acc3-4877-84bd-904c10745ba6/250

* move use extra check to runtime, not on C2

* this is always true

* more C2 checks

* log if frames are out of sync

* more logging on no frame

* store in pointer

* print sof

* add sync logic

* log based on sof difference as well

* keep both models

* less assumptions

* define above thneed

* typo

* simplify

* no need for second client is main is already wide

* more comments update

* no optional reference

* more logging to debug lags

* add to release files

* both defines

* New model: 6831a77f-2574-4bfb-8077-79b0972a2771/950

* Path offset no longer relevant

* Remove duplicate execute

* Moved bigmodel back to big_supercombo.dlc

* add wide vipc stream

* Tici must be tici

* Needs state too

* add wide cam support to model replay

* handle syncing better

* ugh, c2

* print that

* handle ecam lag

* skip first one

* so close

* update refs

Co-authored-by: mitchellgoffpc <mitchellgoffpc@gmail.com>
Co-authored-by: Harald Schafer <harald.the.engineer@gmail.com>
Co-authored-by: Adeeb Shihadeh <adeebshihadeh@gmail.com>
Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 85efde269d
2022-02-19 16:06:31 -08:00
redacid95
fac6a207f9 Snpemodel: Clean up after #23772 (#23780)
* #clean up snpemodel after #23772

* Leave benchmark remove loop
old-commit-hash: 1c9bf108af
2022-02-16 09:40:32 -08:00
George Hotz
37349c0fef add thneed optimizer (#23772)
* add thneed optimizer

* local work group opt

* kernels and final mods

* release files

* build system touchups

* fix kernel path, rand inputs for self test

* broken since extra is gone

* update model replay ref

Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 90beaebefb
2022-02-15 15:32:00 -08:00
jimw
ce83b94fa6 PC: fix onnxruntime excessive CPU (#23051)
* onnxruntime-gpu two threads

* relock pipfile

Co-authored-by: jwolffe <wolffja@gmail.com>
Co-authored-by: Willem Melching <willem.melching@gmail.com>
old-commit-hash: db9ea45cd5
2021-12-06 16:51:45 +01:00
Dean Lee
a80999fd4f moded/RunModel: fix possible memory leak without a virtual destructor (#22990)
old-commit-hash: d97ad1302a
2021-11-20 11:41:24 -08:00
George Hotz
b4adcd2e56 Use thneed directly on the loaded YUV data (#22236)
* completely untested

* it builds now

* bug fixes, save 1ms

* using a kernel to copy works

* more sane API to loadyuv

Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 83ff9ca331
2021-09-16 14:13:03 -07:00
Dean Lee
a02ecc02c4 ONNXModel: use the actual model file name (#22084)
* use the actual file name

* use macro

* fix typo

* continue

* Apply suggestions from code review

Co-authored-by: Adeeb Shihadeh <adeebshihadeh@gmail.com>
old-commit-hash: 7db04ab0c5
2021-09-02 14:21:15 -07:00
Dean Lee
f70a79b838 Use C++ header files instead of C header files. (#21192)
* use cstring instead of string.h

* use cstdio instead of stdio.h

* remove inttypes.h

* use cstdlib instead of stdlib.h

* use cstdint instead of stdint.h

* #include <cstddef>

* cstdlib

* use cmath

* remove stddef.h

* use cassert

* use csignal

* use ctime

* use cerror

* rebase master
old-commit-hash: c53cb5d570
2021-06-08 13:46:22 -07:00
Dean Lee
3a698faadc Use full include path (#21194)
old-commit-hash: b06b3ecb60
2021-06-08 12:09:30 -07:00
Dean Lee
e333e4f189 Cleanup selfdrive/ includes (#20822)
* cleanup include path

* continue

* format includes

* fix testraw.cc

* remove include path from SConstruct

* regroup

* rebase master

* almost done

* apply review

* rename FileReader.xx to filereader.xx

* rename Unlogger.x->unlogger.x

* rename FrameReader.xx -> framereader.xx

* apply reviews

* ui.h

* continue

* fix framebuffer.cc build error:mv util.h up

* full path to msm_media_info

* fix qcom2 camerad

Co-authored-by: Comma Device <device@comma.ai>
old-commit-hash: 7222d0f20d
2021-05-08 22:15:17 -07:00