Loaded the model but it wont respond and is stuck saying "typing" meanwhile gpu usage at 100%
#6
by
barncroft
- opened
Hey Metalx, I am currently using oobabooga and the model loads just fine but the model is just stuck while when I use the gpt4 x alpaca 13b model it barely uses any vram and has lightning quick responses. The 100% gpu usage, could that be because the 13b model has cuda optimizations because it says that in the file name but that still does explain why my model is just stuck on "typing" and never actually responding. But thank you for making this model I am looking forward to playing with it!
No idea, but I can infer that some kind of error happened and it's most likely visible in the console. Perhaps if you can provide a picture of the error in the console I can help further.