[GH-ISSUE #8226] ollama.com model quantization levels are not displayed correctly #51763

Open
opened 2026-04-28 20:55:03 -05:00 by GiteaMirror · 3 comments
Owner

Originally created by @CberYellowstone on GitHub (Dec 24, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/8226

What is the issue?

This issue is a continuation of #7816
The issue with incorrect local ollama quantization levels in #7816 has been resolved, but the same problem appears in the model cards of models uploaded to ollama.com.

example:
https://ollama.com/CBYellowstone/sakura-v1.0
image

it should be:
image

OS

Windows

GPU

Nvidia

CPU

Intel

Ollama version

0.5.4

Originally created by @CberYellowstone on GitHub (Dec 24, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/8226 ### What is the issue? This issue is a continuation of #7816 The issue with incorrect local ollama quantization levels in #7816 has been resolved, but the same problem appears in the model cards of models uploaded to ollama.com. example: https://ollama.com/CBYellowstone/sakura-v1.0 ![image](https://github.com/user-attachments/assets/8e01c271-4a30-4191-a3f6-6cedbca65ff0) it should be: ![image](https://github.com/user-attachments/assets/f31fda66-b63a-4827-a979-8d697f21610b) ### OS Windows ### GPU Nvidia ### CPU Intel ### Ollama version 0.5.4
GiteaMirror added the bugollama.com labels 2026-04-28 20:55:04 -05:00
Author
Owner

@bmizerany commented on GitHub (Jan 6, 2025):

Something looks off here. More than just the quantization are showing errors. Is it possible this model is a modified qwen2?

<!-- gh-comment-id:2572081460 --> @bmizerany commented on GitHub (Jan 6, 2025): Something looks off here. More than just the quantization are showing errors. Is it possible this model is a modified qwen2?
Author
Owner

@CberYellowstone commented on GitHub (Jan 6, 2025):

Something looks off here. More than just the quantization are showing errors. Is it possible this model is a modified qwen2?

Yes, its base model is qwen2.5.
This is its repository link SakuraLLM

<!-- gh-comment-id:2572329711 --> @CberYellowstone commented on GitHub (Jan 6, 2025): > Something looks off here. More than just the quantization are showing errors. Is it possible this model is a modified qwen2? Yes, its base model is qwen2.5. This is its repository link [SakuraLLM](https://github.com/SakuraLLM/SakuraLLM)
Author
Owner

@CberYellowstone commented on GitHub (Jan 6, 2025):

Something looks off here. More than just the quantization are showing errors. Is it possible this model is a modified qwen2?

In the specific example in the image, the model shown in the screenshot is this: Sakura-7B-Qwen2.5-v1.0-GGUF

<!-- gh-comment-id:2572332316 --> @CberYellowstone commented on GitHub (Jan 6, 2025): > Something looks off here. More than just the quantization are showing errors. Is it possible this model is a modified qwen2? In the specific example in the image, the model shown in the screenshot is this: [Sakura-7B-Qwen2.5-v1.0-GGUF](https://huggingface.co/SakuraLLM/Sakura-7B-Qwen2.5-v1.0-GGUF)
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#51763