Search results

[tx] Implement mHC

tx

Status: Open.

#952 In NovaSky-AI/SkyRL;

· pcmoritz opened on Jan 25, 2026

Sticky session routing doc when inference refactoring lands

Status: Open.

#939 In NovaSky-AI/SkyRL;

· CharlieFRuan opened on Jan 24, 2026

[tx] Deflake qwen3 MoE test

tx

Status: Open.

#936 In NovaSky-AI/SkyRL;

· pcmoritz opened on Jan 24, 2026

[tx] Make it easy to run on a multi-node Ray cluster

tx

Status: Open.

#935 In NovaSky-AI/SkyRL;

· pcmoritz opened on Jan 24, 2026

On Policy Distillation checkpoints Loading

bugSomething isn't workingSomething isn't working

skyrl-train

Status: Open.

#895 In NovaSky-AI/SkyRL;

· White1973 opened on Jan 18, 2026

[tx] support long sequence

tx

Status: Open.

#891 In NovaSky-AI/SkyRL;

· raulchen opened on Jan 16, 2026

[skyrl-train][algorithm] Apply inference engine top-k mask to trainer

enhancementNew feature or requestNew feature or request

skyrl-train

Status: Open.

#877 In NovaSky-AI/SkyRL;

· erictang000 opened on Jan 14, 2026

[skyrl-train] Multi-modal support

skyrl-train

Status: Open.

#876 In NovaSky-AI/SkyRL;

· nithinvc opened on Jan 14, 2026

Supporting 120B MOE with FSDP2

skyrl-train

Status: Open.

#870 In NovaSky-AI/SkyRL;

· ashish230897 opened on Jan 13, 2026

[skyrl-train] consider updating grpo_norm_by_std default to false

skyrl-train

Status: Open.

#869 In NovaSky-AI/SkyRL;

· erictang000 opened on Jan 13, 2026

[skyrl-train][Harbor] Upstream Harbor training code to main skyrl-train

skyrl-train

Status: Open.

#866 In NovaSky-AI/SkyRL;

· CharlieFRuan opened on Jan 12, 2026

[tx] Implement DeepSeek model architecture

tx

Status: Open.

#865 In NovaSky-AI/SkyRL;

· pcmoritz opened on Jan 12, 2026