-
Notifications
You must be signed in to change notification settings - Fork 1.3k
All issues
Issue creation is restricted in this repository
Issues
is:issue state:open
is:issue state:open
Search results
- Status: Open.#2409 In THUDM/slime;
- Status: Open.#2402 In THUDM/slime;
- Status: Open.#2400 In THUDM/slime;
- Status: Open.#2399 In THUDM/slime;
[Bug] Qwen-VL patch bypasses SGLang token-ID preservation and breaks multi-turn TITO
bugSomething isn't workingSomething isn't workingStatus: Open.#2388 In THUDM/slime;[Question] Would an adaptive in-reward KL controller fit Slime's PPO/RLHF scope?
questionFurther information is requestedFurther information is requestedStatus: Open.#2387 In THUDM/slime;- Status: Open.#2386 In THUDM/slime;
[Bug] Broadcasting full rollout_routed_experts data to every Megatron training rank will cause host OOM
bugSomething isn't workingSomething isn't workingStatus: Open.#2384 In THUDM/slime;- Status: Open.#2374 In THUDM/slime;
- Status: Open.#2370 In THUDM/slime;
[Bug] Fully-async multi-agent rollout crashes on nested groups
bugSomething isn't workingSomething isn't workingStatus: Open.#2344 In THUDM/slime;[Discussion] Support background evaluation while fully-async training rollout is running
questionFurther information is requestedFurther information is requestedStatus: Open.#2339 In THUDM/slime;