ARahim3 / kaggle-tpu-lab
Frontier-class open models on a free Kaggle TPU v5e-8: GLM-5.3-Flash 320B MoE (~165 tok/s, our own JAX engine) and Qwen3.8-27B bf16 (~130 tok/s), 262k context, prefix caching. Works with Claude Code, Codex, opencode and pi.
Why this job: topics: llm-inference, vllm, llm-serving
What it does
Each model has its own folder with a run-all Kaggle notebook, the kernel script behind it, and a write-up of how it works and what we measured.
Get started
No single install command. Read the steps in the README (opens in new tab)
- Maintenance
- Active
- Last commit 34 days ago
- Stars
- 782
- +24 this week
- Licence
- MIT
- Allows business use
- Downloads for
- None listed
- No release files
- Latest version
- None yet
- No releases
- Built with
- Python
- 55% of the code
Is it looked after?
- Contributors
- 1
- Last 90 days
- Issues
- 16% closed
- 16 still open
- Pull requests
- 0 merged
- 6 waiting
- Releases
- None
- Never published
- Contributing guide (opens in new tab)
- No security policy
- No code of conduct
Similar repos
| Repo | Stars | Last commit | Licence | Downloads for |
|---|---|---|---|---|
| vllm-project / vllm | 96,588 | today | Apache-2.0 | Mac, Linux |
| simplexai-labs / PolyStrata | 938 | 2 days ago | MIT | Windows |
| mw00 / project-maya | 2,712 | 2 days ago | MIT | No downloads |
More: languages, topics and project record
Languages
- Python 55%
- Jupyter Notebook 45%
- Shell <1%
Topics
- anthropic-api
- claude-code
- coding-agents
- free-gpu
- glm
- glm-5-3-flash
- jax
- kaggle
- llm-inference
- llm-serving
- long-context
- mixture-of-experts
Project record
- JobLocal LLMs
- StatusListed
- Commits25
- Forks105
- Watching7
- Stars in 28 days+178
- CreatedAug 31, 2026
- First seen by usNov 7, 2026
- CheckedNov 7, 2026