Update to genai usage #2714
Update to genai usage #2714
Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
|
View / edit / reply to this conversation on ReviewNB eaidova commented on 2025-02-05T04:49:15Z Please add info about genai usage also in general notebook description too.
image tokens is not correct wording, probably you mean tensor/array (tokens associated with tokenization process that encode string to numeric representation, reverse process is detokenization, back from indices to string).
IMHO, "just run" is your value judgement - it is better to avoid such constructions and give people make own decision (for you it maybe something simple, but somebody may have different opinion) that highlight simplicity of actions, but if it is simple, why you need highlight that? :)
I think this text can be simplified:
Pipeline |
|
View / edit / reply to this conversation on ReviewNB eaidova commented on 2025-02-05T04:49:15Z Line #2. !optimum-cli export openvino -m stabilityai/stable-diffusion-xl-base-1.0 --weight-format int8 {model_dir}
can you please also move it to use optimum_cli helper? |
|
View / edit / reply to this conversation on ReviewNB eaidova commented on 2025-02-05T04:53:43Z Line #3. num_steps = 30 generated images looks worse than previous. I believe it may be connected with too large number of steps for generation. Model that you use now is lcm, it usually requires 4 steps only to get results.
|
|
tiny-sd needs to be excluded for awhile |
CVS-161647