Skip to content

Conversation

@YassineYousfi
Copy link
Contributor

No description provided.

@github-actions
Copy link
Contributor

github-actions bot commented Oct 7, 2025

@commaci-public
Copy link
Contributor

ref for commit 83b81b8: https://raw.githubusercontent.com/commaai/ci-artifacts/refs/heads/model_replay_nnv-driving/8494c69d3c710e81|000001d4--2648a9a404_model_tici_83b81b83cdf76a577946a1567f4644e1a018443c.zst

Model Replay Differences
All Model Replay Plots

@YassineYousfi YassineYousfi merged commit 4d08542 into master Oct 9, 2025
13 checks passed
@YassineYousfi YassineYousfi deleted the nnv-driving branch October 9, 2025 19:58
@benmarten
Copy link

ref for commit 83b81b8: https://raw.githubusercontent.com/commaai/ci-artifacts/refs/heads/model_replay_nnv-driving/8494c69d3c710e81|000001d4--2648a9a404_model_tici_83b81b83cdf76a577946a1567f4644e1a018443c.zst
Model Replay Differences



All Model Replay Plots

Just curious what are the x/y-axis here representative of?

@ChristopherRabotin
Copy link
Contributor

Hi there,

Are these plots for the experimental mode? I just tried out the experimental mode today, and the vehicle no longer accelerates on high ways, regardless of whether there is a vehicle in front of it or not. Setting the max speed to 65 mph when at 65 leads the vehicle to slow down over several minutes. The desired acceleration plot here could explain this new behavior.

I will try the standard mode tomorrow.

Long time user (2020) and engineer.

@YassineYousfi
Copy link
Contributor Author

Just curious what are the x/y-axis here representative of?

x = frame index
y = dependent on the output we're plotting (everything is in SI units, desire is a probability)

@YassineYousfi
Copy link
Contributor Author

Hi there,

Are these plots for the experimental mode? I just tried out the experimental mode today, and the vehicle no longer accelerates on high ways, regardless of whether there is a vehicle in front of it or not. Setting the max speed to 65 mph when at 65 leads the vehicle to slow down over several minutes. The desired acceleration plot here could explain this new behavior.

I will try the standard mode tomorrow.

Long time user (2020) and engineer.

These plots are just sanity checks, not tests to verify model performance.
They just replay 60 frames from a test segment and pass them through the model.
Please send us a segment of your issues with experimental mode in the Feedback section of our discord.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants