[GH-ISSUE #8111] Would like to get Qwen2.5-Coder-32B, which model should I pull #51697

Closed
opened 2026-04-28 20:46:02 -05:00 by GiteaMirror · 4 comments
Owner

Originally created by @jiusi9 on GitHub (Dec 16, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/8111

I want to use Qwen2.5-Coder-32B, not Qwen2.5-Coder-32B-instruct.
Which model should I pull using ollama. qwen2.5-coder:32b-base is right?

I want to know what the difference is between qwen2.5-coder:32b-base and qwen2.5-coder:32b-instruct.

Originally created by @jiusi9 on GitHub (Dec 16, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/8111 I want to use Qwen2.5-Coder-32B, not Qwen2.5-Coder-32B-instruct. Which model should I pull using ollama. **qwen2.5-coder:32b-base** is right? I want to know what the difference is between qwen2.5-coder:32b-base and qwen2.5-coder:32b-instruct.
GiteaMirror added the model label 2026-04-28 20:46:02 -05:00
Author
Owner

@Rakhsan commented on GitHub (Dec 16, 2024):

hey you should have probably asked chatgpt. SO qwen instruct is made for instruction following and the base is just good at writing code not following the instruction. I recommend just using instruct model

<!-- gh-comment-id:2544647270 --> @Rakhsan commented on GitHub (Dec 16, 2024): hey you should have probably asked chatgpt. SO qwen instruct is made for instruction following and the base is just good at writing code not following the instruction. I recommend just using instruct model
Author
Owner

@jiusi9 commented on GitHub (Dec 16, 2024):

hey you should have probably asked chatgpt. SO qwen instruct is made for instruction following and the base is just good at writing code not following the instruction. I recommend just using instruct model

Thanks a lot.

I‘m looking for some different abilities models are good at chat or generate.
Could you tell me about which model is like Qwen2.5-Coder-32B of huggingface, which model is like Qwen2.5-Coder-32B-instruct ?

<!-- gh-comment-id:2544861686 --> @jiusi9 commented on GitHub (Dec 16, 2024): > hey you should have probably asked chatgpt. SO qwen instruct is made for instruction following and the base is just good at writing code not following the instruction. I recommend just using instruct model Thanks a lot. I‘m looking for some different abilities models are good at chat or generate. Could you tell me about which model is like Qwen2.5-Coder-32B of huggingface, which model is like Qwen2.5-Coder-32B-instruct ?
Author
Owner

@rick-github commented on GitHub (Dec 16, 2024):

Qwen2.5-Coder-32B -> qwen2.5-coder:32b-base-fp16
Qwen2.5-Coder-32B-instruct -> qwen2.5-coder:32b-instruct-fp16

<!-- gh-comment-id:2545772726 --> @rick-github commented on GitHub (Dec 16, 2024): Qwen2.5-Coder-32B -> [qwen2.5-coder:32b-base-fp16](https://ollama.com/library/qwen2.5-coder:32b-base-fp16) Qwen2.5-Coder-32B-instruct -> [qwen2.5-coder:32b-instruct-fp16](https://ollama.com/library/qwen2.5-coder:32b-instruct-fp16)
Author
Owner

@jiusi9 commented on GitHub (Dec 17, 2024):

Thanks for your answer again.

<!-- gh-comment-id:2547302683 --> @jiusi9 commented on GitHub (Dec 17, 2024): Thanks for your answer again.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#51697