* add param that requests keyframe
* set keyframe request to false on cleanup
* timing and errors
* video_enabled option and ability to add video later
* remove loggerd changes from this PR
* video enabled default to on, clean identifier
* request keyframe on start regardless of video enabled or not
* move error handling additions to different PR
* add param that requests keyframe
* set keyframe request to false on cleanup
* clean
* timing and errors
* remove other changes
* clean diff
* clean diff
* clean up live adaptation params and functions
* better function names
* fix syntax
* remove unused var
* clean diff
* improve lat maneuver report
* rm torque jerk
* add jitter maneuver
* add to 30mph
* add jerk
* color
* more state
* override
* abort reason
* abort
* rm text
* pad for small text
- patch `aioice` to only get host address of active network interface (wlan0 when on wifi, ppp0 when not)
- use common webrtc session between `startStream` and `addIceCandidates`
- first time connection speed ups
- move imports to top of file to move delay to webrtcd server start up not first connection
- warm up webrtc stack by creating a webrtc connection object to import all submodule classes
op.sh: don't eval CMD, run argv directly so special chars work
Previously op_run_command joined "$@" into a single string and eval'd
it, which broke on inputs containing shell metacharacters. For example,
'op esim --download "LPA:1$[rsp.truphone.com]($url)$QRF-SPEEDTEST" name'
would fail with a syntax error on the unquoted parens. Run "$@" instead
and only use $* for the printed display string. Also quote the unquoted
$@ in the other op_run_command callers.
* modeld: RL driving model with 3-file split
Split the driving model into vision + off_policy + on_policy ONNX
files and wire up the RL policy:
- 3-file model split (vision / off_policy / on_policy), replacing the
combined big_driving_policy/vision models
- compiler updates for the split models
- actually consume the policy action in modeld
- add desire state to the driving model
- model iterations (smoothness, off/on-policy weight updates)
* modeld: update driving model
* 1e72cf5a-785f-45ea-888f-28cdb14785de/100
* tinygrad hack
* fix parsing
* looser timing
* big
* Remove unnecessary modeld rebase changes
* Tighten modeld split cleanup
---------
Co-authored-by: Comma Device <device@comma.ai>
Co-authored-by: Armandpl <adpl33@gmail.com>
* bitrate controller
* ability to set the quality so a certain level, make message handler a switch case
* distinct bitrate levels and exponential backoff of raising bitrate
* add med level threshold
* fix
* fix
* simplify
* clean
Robbie Wadley