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

feat: qwen2.5 coder Proposal #801

Merged
merged 2 commits into from
Jan 3, 2025
Merged

feat: qwen2.5 coder Proposal #801

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
d153dbe
feat: add qwen2.5 coder 7b
zhuangqh Jan 2, 2025
ba0bdd5
fix
zhuangqh Jan 2, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
50 changes: 50 additions & 0 deletions docs/proposals/20250103-qwen2.5-coder.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,50 @@
---
title: Proposal for new model support
authors:
- "Qinghui Zhuang"
reviewers:
- "Kaito contributor"
creation-date: 2025-01-03
last-updated: 2025-01-03
status: provisional
---

# Title
Add Qwen 2.5 Coder to Kaito supported model list

## Glossary
N/A

## Summary
- **Model description**: Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). It brings significantly improvements in code generation, code reasoning and code fixing with long-context support up to 128K tokens. For more information, refer to the [Qwen2.5 Documentation](https://qwenlm.github.io/blog/qwen2.5/) and access the model on [Hugging Face](https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct).
- **Model usage statistics**: In the past month, Qwen2.5-Coder-7B-Instruct has garnered 118,568 downloads on Hugging Face, reflecting its widespread popularity. Google Trends data shows a high level of search interest in <!-- markdown-link-check-disable --> ["qwen2.5"](https://trends.google.com/trends/explore?q=qwen2.5), <!-- markdown-link-check-enable --> indicating strong market curiosity.
- **Model license**: Qwen2.5-Coder series is distributed under the Apache 2.0 license, ensuring broad usability and modification rights.

## Requirements

The following table describes the basic model characteristics and the resource requirements of running it.

| Field | Notes|
|----|----|
| Family name| Qwen 2.5 Coder|
| Type| `text-generation` |
| Download site| https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct |
| Version| 0eb6b1ed2d0c4306bc637d09ecef51e59d3dfe05 |
| Storage size| 100GB |
| GPU count| 1 GPU |
| Total GPU memory| 24 GB |
| Per GPU memory | `N/A` |

## Runtimes

This section describes how to configure the runtime framework to support the inference calls.

| Options | Notes|
|----|----|
| Runtime | Huggingface Transformer |
| Distributed Inference| False |
| Custom configurations| Precision: BF16. Can run on one machine with 24 GB of GPU memory. |

# History

- [x] 03/01/2025: Open proposal PR.
Loading