gpt-oss-20b on Google Colab

[yishangupenn]

Copied from upstream: openai/openai-cookbook#2003
Original author: @pcuenca
Originally opened: 2025-08-06

---

Summary

This PR adds a guide that shows how to run OpenAI gpt-oss-20b on a 16 GB Google Colab GPU.

Motivation

Shortly after the release yesterday, transformers added mxfp4 support for previous cuda architectures (Ada, Ampere, Tesla), not just Hopper. This is of great interest to the community, but it requires specific instructions to install transformers from the main branch and update the dependencies so the Triton kernels work. The new guide demonstrates this process, and can run directly on a Google Colab notebook.

---

For new content

When contributing new content, read through our contribution guidelines, and mark the following action items as completed:

• I have added a new entry in registry.yaml (and, optionally, in authors.yaml) so that my content renders on the cookbook website.
• I have conducted a self-review of my content based on the contribution guidelines:
  • Relevance: This content is related to building with OpenAI technologies and is useful to others.
  • Uniqueness: I have searched for related examples in the OpenAI Cookbook, and verified that my content offers new insights or unique information compared to existing documentation.
  • Spelling and Grammar: I have checked for spelling or grammatical mistakes.
  • Clarity: I have done a final read-through and verified that my submission is well-organized and easy to understand.
  • Correctness: The information I include is correct and all of my code executes successfully.
  • Completeness: I have explained everything fully, including all necessary references and citations.

We will rate each of these areas on a scale from 1 to 4, and will only accept contributions that score 3 or higher on all areas. Refer to our contribution guidelines for more details.