Skip to content

[WIP] Kimi K2.5 NVFP4 recipe #267

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

Draft
kedarpotdar-nv wants to merge 1 commit into
base: main
Choose a base branch
from
+3 −0
Draft

[WIP] Kimi K2.5 NVFP4 recipe #267

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
3 changes: 3 additions & 0 deletions moonshotai/Kimi-K2.5.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -87,4 +87,7 @@ print(f"Response costs: {time.time() - start:.2f}s")
print(f"Generated text: {response.choices[0].message.content}")
```

## Running Kimi-K2.5-NVFP4
The [Kimi-K2.5-NVFP4](https://huggingface.co/nvidia/Kimi-K2.5-NVFP4) checkpoint is an NVFP4-quantized version of Kimi-K2.5, quantized using [NVIDIA Model Optimizer](https://github.com/NVIDIA/TensorRT-Model-Optimizer). Only the weights and activations of the linear operators within transformer blocks in MoE are quantized to FP4, enabling efficient deployment on fewer GPUs.
Comment on lines +90 to +91
Copy link
Contributor

@gemini-code-assist gemini-code-assist bot Mar 4, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

medium

The section ## Running Kimi-K2.5-NVFP4 describes the model but doesn't include the command or 'recipe' to run it. Considering the PR title and the section title, it would be very helpful to add an example command, similar to the one provided for the base Kimi-K2.5 model.


For more usage examples, check out the [vLLM user guide for multimodal models](https://docs.vllm.ai/en/latest/features/multimodal_inputs.html) and the [official Kimi-K2.5 Hugging Face page](https://huggingface.co/moonshotai/Kimi-K2.5)!