on_device_model: Add vram_mb to model descriptor. [chromium/src : main]
Alan Ding (Gerrit)
Alan Ding voted Commit-Queue+1![]( "Open in Gerrit")
| Commit-Queue | +1 |
Related details
- Chi Yo Tsai
- Nathan Memmott
Code-Coverage
Code-Owners
Code-Review
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Nathan Memmott (Gerrit)
Nathan Memmott added 1 comment![]( "Open in Gerrit")
uint64_t vram_mb = 0;Placing this here is causing the optimization guide internal bots to fail. You can move it to the end to make them happy.
Related details
- Alan Ding
- Chi Yo Tsai
Code-Coverage
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Nathan Memmott (Gerrit)
Nathan Memmott added 1 comment![]( "Open in Gerrit")
ml::GetDeviceAndPerformanceInfo(*chrome_ml_);It looks like this will trigger a benchmark to run, but all we need vram_mb?
Chi Yo Tsai (Gerrit)
Chi Yo Tsai added 1 comment![]( "Open in Gerrit")
ml::GetDeviceAndPerformanceInfo(*chrome_ml_);It looks like this will trigger a benchmark to run, but all we need vram_mb?
Currently our VRAM fetching is tied to our micro benchmark. Probably something that we should fix. Otherwise overriding the performance class blocks anything that depends on a VRAM check
Related details
- Alan Ding
Code-Coverage
Code-Owners
Code-Review
No-Unresolved-Comments
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Nathan Memmott (Gerrit)
Nathan Memmott added 1 comment![]( "Open in Gerrit")
ml::GetDeviceAndPerformanceInfo(*chrome_ml_);Chi Yo TsaiIt looks like this will trigger a benchmark to run, but all we need vram_mb?
Currently our VRAM fetching is tied to our micro benchmark. Probably something that we should fix. Otherwise overriding the performance class blocks anything that depends on a VRAM check
Could we pass in the already calculated vram from the browser process? It's saved in components/optimization_guide/core/model_execution/performance_class.cc to the local state preference model_execution::prefs::localstate::kOnDeviceVramMb. We could retrieve it from the `PerformanceClassifier` and pass it through the `LoadModelParams`.
Alan Ding (Gerrit)
Alan Ding added 2 comments![]( "Open in Gerrit")
Placing this here is causing the optimization guide internal bots to fail. You can move it to the end to make them happy.
Done
Chi Yo TsaiIt looks like this will trigger a benchmark to run, but all we need vram_mb?
Nathan MemmottCurrently our VRAM fetching is tied to our micro benchmark. Probably something that we should fix. Otherwise overriding the performance class blocks anything that depends on a VRAM check
Could we pass in the already calculated vram from the browser process? It's saved in components/optimization_guide/core/model_execution/performance_class.cc to the local state preference model_execution::prefs::localstate::kOnDeviceVramMb. We could retrieve it from the `PerformanceClassifier` and pass it through the `LoadModelParams`.
- Chi Yo Tsai
- Nathan Memmott
Code-Coverage
Code-Owners
Code-Review
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Nathan Memmott (Gerrit)
Nathan Memmott voted and added 1 comment![]( "Open in Gerrit")
Votes added by Nathan Memmott
| Code-Review | +1 |
1 comment
Related details
- Alan Ding
- Chi Yo Tsai
Code-Coverage
Code-Owners
Code-Review
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
gwsq (Gerrit)
Message from gwsq![]( "Open in Gerrit")
From googleclient/chrome/chromium_gwsq/ipc/config.gwsq:
IPC: dch...@chromium.org
📎 It looks like you’re making a possibly security-sensitive change! 📎 IPC security review isn’t a rubberstamp, so your friendly security reviewer will need a fair amount of context to review your CL effectively. Please review your CL description and code comments to make sure they provide context for someone unfamiliar with your project/area. Pay special attention to where data comes from and which processes it flows between (and their privilege levels). Feel free to point your security reviewer at design docs, bugs, or other links if you can’t reasonably make a self-contained CL description. (Also see https://cbea.ms/git-commit/).
IPC reviewer(s): dch...@chromium.org
Reviewer source(s):
dch...@chromium.org is from context(googleclient/chrome/chromium_gwsq/ipc/config.gwsq)
Related details
- Chi Yo Tsai
- Daniel Cheng
- Steven Holte
Code-Coverage
Code-Owners
Code-Review
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Daniel Cheng (Gerrit)
Daniel Cheng voted and added 1 comment![]( "Open in Gerrit")
Votes added by Daniel Cheng
| Code-Review | +1 |
1 comment
- Alan Ding
- Chi Yo Tsai
- Steven Holte
Steven Holte (Gerrit)
Steven Holte voted Code-Review+1![]( "Open in Gerrit")
| Code-Review | +1 |
Related details
- Alan Ding
- Chi Yo Tsai
Code-Coverage
Code-Owners
Code-Review
Review-Enforcement
| Inspect html for hidden footers to help with email filtering. To unsubscribe visit settings. |
Alan Ding (Gerrit)
Alan Ding voted and added 1 comment![]( "Open in Gerrit")
Votes added by Alan Ding
| Commit-Queue | +2 |
1 comment
- Chi Yo Tsai
Alan Ding (Gerrit)
Alan Ding voted Commit-Queue+2![]( "Open in Gerrit")
| Commit-Queue | +2 |