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.

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

ggml : allow using the default ggml_type value in the name function #9458

Merged
merged 1 commit into from
Sep 14, 2024

Conversation

ykhrustalev
Copy link
Contributor

When running on Windows, the quantization utility attempts to print the types that are not set which leads to a crash.
@github-actions github-actions bot added the ggml changes relating to the ggml tensor library for machine learning label Sep 13, 2024
@ggerganov
Copy link
Owner

When running on Windows, the quantization utility attempts to print the types that are not set which leads to a crash.

There should be no types without a name. What command do you use to trigger this?

@ykhrustalev
Copy link
Contributor Author

ykhrustalev commented Sep 13, 2024

that is manually added extra logging that is added to the quantize.cpp

    fprintf(stderr, "  output_tensor_type       : %s\n", ggml_type_name(params.output_tensor_type));
    fprintf(stderr, "  token_embedding_type     : %s\n", ggml_type_name(params.token_embedding_type));
    fprintf(stderr, "  special_layer_quant_type : %s\n", ggml_type_name(params.special_layer_quant_type));

where params.output_tensor_type is a default value

@ggerganov ggerganov merged commit 822b632 into ggerganov:master Sep 14, 2024
1 check passed
dsx1986 pushed a commit to dsx1986/llama.cpp that referenced this pull request Oct 29, 2024
When running on Windows, the quantization utility attempts to print the types that are not set which leads to a crash.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
ggml changes relating to the ggml tensor library for machine learning
Projects
None yet
Development

Successfully merging this pull request may close these issues.

2 participants