Align text-only training accuracy between AutoModel (NVIDIA NeMo Automodel) and XTuner after model onboarding. Use to validate same-input loss and gradients, investigate cross-framework numerical discrepancies, or verify a fix. Read the pitfalls first, freeze…
Verify that a new change to XTuner does not regress RL training accuracy, by running the GRPO (Qwen3-8B gsm8k) and OPD (Qwen3-VL-2B multi-teacher MOPD) end-to-end trainings and comparing their tracker metrics against fixed baselines. Use when the user asks to…
Add or audit an XTuner chat-template implementation and its loss mask from an official Hugging Face model repository or local model directory. Use when supporting a new model under xtuner/v1/data_proto/messages, checking an existing implementation against…
Implement a missing XTuner Hugging Face config export from the official HF model when possible, then validate it against the installed Transformers round-trip and versioned inference-engine field contracts. Use when adding a model, when hf_config is missing…
Use when the user wants to instrument, inspect, or debug XTuner traces to reconstruct a sample/request execution path, follow cross-process call chains, or identify latency hotspots, bottleneck stages, and abnormal paths from span data.
Use when creating, modifying, or reviewing XTuner RL judgers under xtuner/v1/rl/judger or configs that build custom judgers. Enforces the preferred Judger payload contract and the BaseJudger/ComposedJudger constraints.
Add support for a HuggingFace model in XTuner so it can be trained with the full set of parallelism strategies and optimization switches. The user can provide a HuggingFace Hub repo id, a local model directory, or a model family already supported by…
Merge a base HF model dir with an extra HF model dir by index diff — take tensors only present in extra, append them to a new output dir in HF-standard shard layout. Base wins on overlap.
Use when `make html` or `sphinx-build` fails with `Extension error (sphinx.ext.autosummary)`, `ImportExceptionGroup`, or other Sphinx autosummary / autodoc import failures. Provides automated pdb diagnosis for `autodoc_mock_imports` issues.
Synchronize xtuner's supported model documentation (docs/en/pretrain_sft/advanced_tutorial/model.md and docs/zh_cn/pretrain_sft/advanced_tutorial/model.md) with the actual Config classes defined under xtuner/v1/model/. Use when (1) new TransformerConfig,…