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

Quantization file name problem #1076

Open
Erimus-Koo opened this issue Jul 2, 2023 · 1 comment
Open

Quantization file name problem #1076

Erimus-Koo opened this issue Jul 2, 2023 · 1 comment

Comments

@Erimus-Koo
Copy link

On this site: https://ggml.ggerganov.com

ggml-model-whisper-large-q5_0.bin

This one looks large.en, not large.

Output like this:

[00:00:00.000 --> 00:00:03.920]   (speaking in foreign language)
[00:00:03.920 --> 00:00:07.840]   (speaking in foreign language)
[00:00:07.840 --> 00:00:11.760]   (speaking in foreign language)
[00:00:11.760 --> 00:00:15.680]   (speaking in foreign language)
[00:00:15.680 --> 00:00:19.600]   (speaking in foreign language)
[00:00:19.600 --> 00:00:23.520]   (speaking in foreign language)

I also tried the base-q5_1, that one is correct.

And what is the last number mean? q5_0 and q5_1, what is the difference?

@BackMountainDevil
Copy link

BackMountainDevil commented Dec 9, 2023

in #1396

arge-q5_0 == quantized large-v2

so I guess q means quantized. And readme#quantization says "quantize a model with Q5_0 method". so they are diff method. In issues/846, you can see one more q8_0

hg has large model and their sha1sum

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants