Ahmad Fahadh Ilyas

> I think your code snippet is not the one that triggers the error message you're showing. You're missing the `function` keyword and also you have `override` there but the...

@cameel Here is my proof Here is how `vscode-solidity` telling me that writing `override` only is wrong...  Here is when I tried to compiled it... ![compiled...

> It's not that I don't believe you, I just don't know all the factors coming into play here and a broken snippet did not help :) > > First...

> i believe this is intentional ... the completion are the resulting outputs from the given prompts and are in fact concated together when being fed to the decoder-only model...

Hi, I also got this error. The weird thing is, if I'm running script from this git, the model train without any problem, but if I trained it with my...

> @LaaZa Btw, why is quantizing norm1, norm2, and router layers not helpful? I have little experience in the model layer code to infer the reason. Thanks. If we exclude...

does `inside_layer_modules` parameters has effect if you write the list format differently? Because if so, why don't you group all `w1` and `v1` inside one list and group all `w2`...

@Qubitium does `router.layer` must be included? Because mixtral did not include it (mixtral name it `gate`). And should normalization parameter be included?

> I think they shouldn't even if fixing padding allows for it. Transformers/Optimum would likely need to ignore both. Quantizing normalization is generally not done anyway. What about `outside_layer_modules`? What...

> @fahadh4ilyas something is broken in the beta packages. Downgrading to the last stable version resolved this for me -> version `2.1.22` I think the beta that might work as...