Search results

loss mask question

Status: Open.

#1477 In THUDM/slime;

· glennccc opened on Jan 21, 2026

请求支持 rollout 粒度的 sglang 请求路由提高多轮 rollout 的吞吐量

Status: Open.

#1469 In THUDM/slime;

· zchuz opened on Jan 20, 2026

Do not support Gemma3 in convert_hf_to_torch_dist

Status: Open.

#1468 In THUDM/slime;

· Zhaoyi-Yan opened on Jan 20, 2026

cannot import name 'GPU_MEMORY_TYPE_CUDA_GRAPH' from 'sglang.srt.constants'

Status: Open.

#1462 In THUDM/slime;

· CHRIS123540 opened on Jan 20, 2026

The Sglang version in docker image

Status: Open.

#1455 In THUDM/slime;

· AIR-hl opened on Jan 19, 2026

Bug: the generate_rollout in sft_rollout.py has not truncation

Status: Open.

#1450 In THUDM/slime;

· MikaStars39 opened on Jan 18, 2026

on_policy_distillation, advantage_estimator

Status: Open.

#1449 In THUDM/slime;

· EthanChen1234 opened on Jan 18, 2026

An error occurred while converting Llama3.2-1B-Instruct to Megatron ckpt.

Status: Open.

Bug

#1430 In THUDM/slime;

· SmileShaun opened on Jan 16, 2026

How to implement multi-teacher distillation?

Status: Open.

#1417 In THUDM/slime;

· whq-pku1 opened on Jan 14, 2026

tools/convert_fsdp_to_hf.py Don't support VLM model

Status: Open.

#1416 In THUDM/slime;

· Cerberous opened on Jan 14, 2026

Bug: Group normalization in _post_process_rewards normalizes across entire batch when sample counts are unequal

Status: Open.

#1414 In THUDM/slime;

· ccggddmm opened on Jan 14, 2026

rollout_log_probs如果是None，会影响GRPO的训练吗

Status: Open.

#1413 In THUDM/slime;

· glennccc opened on Jan 14, 2026