Release v0.9.2

What's Changed

• [pre-commit.ci] pre-commit suggestions by @pre-commit-ci[bot] in #1167
• Fix: AgentInstance bug by @Freder-chen in #1169
• refactor: replace --bf16 with --data_type (bf16/fp16/fp32) by @LYMDLUT in #1170
• feat: Add multi-stage wake_up support for vLLM engine by @LYMDLUT in #1171
• feat: add save_hf_ckpt and disable_ds_ckpt options to training scripts by @LYMDLUT in #1173
• fix: change if to elif for eval_task handling in batch inference by @LYMDLUT in #1174
• Add PromptDataset collate_fn wiring by @Freder-chen in #1177
• add log logprobs_diff by @richardodliu in #1156
• docs: add Bilibili video link to README files by @LYMDLUT in #1180
• fix: pass ring_attn_group to reward model forward in rm_trainer by @yurekami in #1182
• feat: support prorlv2 recipes and math verify @xiaoxigua999

New Contributors

• @yurekami made their first contribution in #1182

Full Changelog: v0.9.1...v0.9.2