Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Do I have to set do_sample=True to generate different outputs with the same model? The official examples do not have this setting. #1575

Open
QygjW opened this issue Jan 24, 2025 · 1 comment
Open

Do I have to set do_sample=True to generate different outputs with the same model? The official examples do not have this setting. #1575

QygjW opened this issue Jan 24, 2025 · 1 comment

Comments

@QygjW
Copy link

QygjW commented Jan 24, 2025

When I use the same input, the model output is always exactly the same, even if I set the temperature parameter. Until I set de_sample=True, but the official examples have no hints or instructions about this.
@QygjW QygjW changed the title Do I have to set do_sample=True to output the same model? The official examples do not have this setting. Do I have to set do_sample=True to generate different outputs with the same model? The official examples do not have this setting. Jan 24, 2025
@QygjW
Copy link
Author

QygjW commented Jan 24, 2025 •
edited
Loading

File "/xxx/llamav32/lib/python3.10/site-packages/torch/nn/modules/module.py", line 1931, in getattr
raise AttributeError(
AttributeError: 'Qwen2Attention' object has no attribute 'temp_KV'. Did you mean: 'temp_O'?

What is wrong with this reasoning process?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@QygjW