Vasily Shamporov
Vasily Shamporov
@raymondlo84 this is basically the same as #211 so I am hesitant to do anything about this notebook until #211 gets to a state when the CI passes so that...
> Cell 12: Timeout of 2000 seconds exceeded while executing cell. Failed to interrupt kernel in 5 seconds, so failing without traceback. > > Looks like it got timeout... Would...
@helena-intel could you please assist with interpreting the CI failures?
@AlexanderDokuchaev @mkaglins
Greetings, @jylansesuifeng! NNCF currently does not provide a user-facing interface to put fake quantizers into arbitrary locations. The reasoning for this is that the best-performing model actually has to have...
@mkaglins @evgeniya-egupova
Greetings, @dqawami ! Thanks for reporting this. Please provide an exact reproducer - the NNCF config .json file or part thereof that exhibits incorrect behaviour.
No, in fact, the storage scope comparisons are limited to determining the train-only modules and referencing their scopes against the storage scope that is being tracked during the one-time replacement...
@dongkwan-kim01 as you can see, the model's expected accuracy value is 67.55, which means that this very same model was specifically measured to reach 67.55 accuracy in the past. If...
The improvement looks great, but now, just to be on the safe side, we would like to understand what changed exactly between the situation where you posted the screenshots with...