Search results

How to train on multi host tpu pod

Status: Open.

#560 In google/tunix;

· siddagra opened on Oct 12, 2025

add sglang-jax inference framework for sampler, similar to vllm

enhancementNew feature or requestNew feature or request

Status: Open.

#554 In google/tunix;

· pathfinder-pf opened on Oct 11, 2025

VLM support for DPO

enhancementNew feature or requestNew feature or request

Status: Open.

#510 In google/tunix;

· immuntasir opened on Oct 6, 2025

Add larger Qwen 2.5 models

enhancementNew feature or requestNew feature or request

Status: Open.

#509 In google/tunix;

· abheesht17 opened on Oct 6, 2025

Failed build: Tunix Package Tests

build failed

Status: Open.

#489 In google/tunix;

· github-actions opened on Oct 3, 2025

RL Bug/Issue Tracker

bugSomething isn't workingSomething isn't working

Status: Open.

#390 In google/tunix;

· abheesht17 opened on Sep 26, 2025

Question: how to run tunix with vLLM?

Status: Open.

#256 In google/tunix;

· OhadRubin opened on Sep 5, 2025

Question: Where to implement rejection sampling for GRPO?

Status: Open.

#216 In google/tunix;

· OhadRubin opened on Aug 25, 2025

Wandb

Status: Open.

#137 In google/tunix;

· shivajid opened on Jul 27, 2025

Multiple rollouts per prompt

Status: Open.

#122 In google/tunix;

· yhtang opened on Jul 21, 2025

Qwen Example relies on HF Transformer and Accelerate, which is removed

Status: Open.

#94 In google/tunix;

· shivajid opened on Jul 5, 2025

Adding linting configuration file

Status: Open.

#93 In google/tunix;

· Practicinginhell opened on Jul 4, 2025