help wanted
Repository-Metriken
- Stars
- (17.706 Stars)
- PR-Merge-Metriken
- (Durchschn. Merge 6T 16h) (10 gemergte PRs in 30 T)
Beschreibung
Description
I'm currently implementing XDeepFM but can't get training on the gpu. I followed the instructions in the GPU support section here.
When I trigger the training, I see the GPU memory in use but utilization stays @ 0. Sometimes, it changes to 1% and then 6% but the majority of the time, it stays @ 0%.
A single cpu stays at 100% however. I do not believe the GPU is being utilized. Any ideas?