Full Code of xorbitsai/inference for AI

gitextract_u_nl6j7f/

├── .dockerignore
├── .gitattributes
├── .github/
│   ├── ISSUE_TEMPLATE/
│   │   ├── bug_report.yaml
│   │   └── feature_request.yaml
│   └── workflows/
│       ├── assign.yaml
│       ├── docker-cd.yaml
│       ├── issue.yaml
│       ├── pr_auto_run_gen_docs.yaml
│       ├── python.yaml
│       └── release.yaml
├── .gitignore
├── .pre-commit-config.yaml
├── .readthedocs.yaml
├── LICENSE
├── MANIFEST.in
├── README.md
├── README_ja_JP.md
├── README_zh_CN.md
├── benchmark/
│   ├── README.md
│   ├── benchmark_embedding.py
│   ├── benchmark_latency.py
│   ├── benchmark_long.py
│   ├── benchmark_rerank.py
│   ├── benchmark_runner.py
│   ├── benchmark_serving.py
│   └── utils.py
├── doc/
│   ├── Makefile
│   ├── source/
│   │   ├── _static/
│   │   │   └── switcher.json
│   │   ├── conf.py
│   │   ├── development/
│   │   │   ├── contributing_codebase.rst
│   │   │   ├── contributing_environment.rst
│   │   │   ├── index.rst
│   │   │   └── xinference_internals.rst
│   │   ├── examples/
│   │   │   ├── ai_podcast.rst
│   │   │   ├── chatbot.rst
│   │   │   ├── gradio_chatinterface.rst
│   │   │   ├── index.rst
│   │   │   ├── langchain_streamlit_doc_chat.rst
│   │   │   └── pdf_chatbot.rst
│   │   ├── gen_docs.py
│   │   ├── getting_started/
│   │   │   ├── environments.rst
│   │   │   ├── index.rst
│   │   │   ├── installation.rst
│   │   │   ├── installation_npu.rst
│   │   │   ├── logging.rst
│   │   │   ├── release_notes.rst
│   │   │   ├── troubleshooting.rst
│   │   │   ├── using_docker_image.rst
│   │   │   ├── using_kubernetes.rst
│   │   │   └── using_xinference.rst
│   │   ├── index.rst
│   │   ├── locale/
│   │   │   └── zh_CN/
│   │   │       └── LC_MESSAGES/
│   │   │           ├── development/
│   │   │           │   ├── contributing_codebase.po
│   │   │           │   ├── contributing_environment.po
│   │   │           │   ├── index.po
│   │   │           │   └── xinference_internals.po
│   │   │           ├── examples/
│   │   │           │   ├── ai_podcast.po
│   │   │           │   ├── chatbot.po
│   │   │           │   ├── gradio_chatinterface.po
│   │   │           │   ├── index.po
│   │   │           │   ├── langchain_streamlit_doc_chat.po
│   │   │           │   └── pdf_chatbot.po
│   │   │           ├── getting_started/
│   │   │           │   ├── environments.po
│   │   │           │   ├── index.po
│   │   │           │   ├── installation.po
│   │   │           │   ├── installation_npu.po
│   │   │           │   ├── logging.po
│   │   │           │   ├── release_notes.po
│   │   │           │   ├── troubleshooting.po
│   │   │           │   ├── using_docker_image.po
│   │   │           │   ├── using_kubernetes.po
│   │   │           │   └── using_xinference.po
│   │   │           ├── getting_started.po
│   │   │           ├── index.po
│   │   │           ├── models/
│   │   │           │   ├── builtin/
│   │   │           │   │   ├── audio/
│   │   │           │   │   │   └── index.po
│   │   │           │   │   ├── embedding/
│   │   │           │   │   │   ├── bge-base-en-v1.5.po
│   │   │           │   │   │   ├── bge-base-en.po
│   │   │           │   │   │   ├── bge-base-zh-v1.5.po
│   │   │           │   │   │   ├── bge-base-zh.po
│   │   │           │   │   │   ├── bge-large-en-v1.5.po
│   │   │           │   │   │   ├── bge-large-en.po
│   │   │           │   │   │   ├── bge-large-zh-noinstruct.po
│   │   │           │   │   │   ├── bge-large-zh-v1.5.po
│   │   │           │   │   │   ├── bge-large-zh.po
│   │   │           │   │   │   ├── bge-small-en-v1.5.po
│   │   │           │   │   │   ├── bge-small-zh-v1.5.po
│   │   │           │   │   │   ├── bge-small-zh.po
│   │   │           │   │   │   ├── e5-large-v2.po
│   │   │           │   │   │   ├── gte-base.po
│   │   │           │   │   │   ├── gte-large.po
│   │   │           │   │   │   ├── index.po
│   │   │           │   │   │   ├── jina-embeddings-v2-base-en.po
│   │   │           │   │   │   ├── jina-embeddings-v2-small-en.po
│   │   │           │   │   │   └── multilingual-e5-large.po
│   │   │           │   │   ├── image/
│   │   │           │   │   │   ├── flux.1-dev.po
│   │   │           │   │   │   ├── flux.1-schnell.po
│   │   │           │   │   │   ├── index.po
│   │   │           │   │   │   ├── kolors.po
│   │   │           │   │   │   ├── sd-turbo.po
│   │   │           │   │   │   ├── sd3-medium.po
│   │   │           │   │   │   ├── sdxl-turbo.po
│   │   │           │   │   │   ├── stable-diffusion-2-inpainting.po
│   │   │           │   │   │   ├── stable-diffusion-inpainting.po
│   │   │           │   │   │   ├── stable-diffusion-v1.5.po
│   │   │           │   │   │   ├── stable-diffusion-xl-base-1.0.po
│   │   │           │   │   │   └── stable-diffusion-xl-inpainting.po
│   │   │           │   │   ├── index.po
│   │   │           │   │   ├── llm/
│   │   │           │   │   │   ├── baichuan-2-chat.po
│   │   │           │   │   │   ├── baichuan-2.po
│   │   │           │   │   │   ├── baichuan-chat.po
│   │   │           │   │   │   ├── baichuan.po
│   │   │           │   │   │   ├── chatglm.po
│   │   │           │   │   │   ├── chatglm2-32k.po
│   │   │           │   │   │   ├── chatglm2.po
│   │   │           │   │   │   ├── chatglm3-32k.po
│   │   │           │   │   │   ├── chatglm3.po
│   │   │           │   │   │   ├── code-llama-instruct.po
│   │   │           │   │   │   ├── code-llama-python.po
│   │   │           │   │   │   ├── code-llama.po
│   │   │           │   │   │   ├── deepseek-chat.po
│   │   │           │   │   │   ├── deepseek-coder-instruct.po
│   │   │           │   │   │   ├── falcon-instruct.po
│   │   │           │   │   │   ├── falcon.po
│   │   │           │   │   │   ├── glaive-coder.po
│   │   │           │   │   │   ├── gorilla-openfunctions-v1.po
│   │   │           │   │   │   ├── gpt-2.po
│   │   │           │   │   │   ├── index.po
│   │   │           │   │   │   ├── internlm-20b.po
│   │   │           │   │   │   ├── internlm-7b.po
│   │   │           │   │   │   ├── internlm-chat-20b.po
│   │   │           │   │   │   ├── internlm-chat-7b.po
│   │   │           │   │   │   ├── llama-2-chat.po
│   │   │           │   │   │   ├── llama-2.po
│   │   │           │   │   │   ├── mistral-instruct-v0.1.po
│   │   │           │   │   │   ├── mistral-instruct-v0.2.po
│   │   │           │   │   │   ├── mistral-v0.1.po
│   │   │           │   │   │   ├── mixtral-instruct-v0.1.po
│   │   │           │   │   │   ├── mixtral-v0.1.po
│   │   │           │   │   │   ├── openbuddy.po
│   │   │           │   │   │   ├── openhermes-2.5.po
│   │   │           │   │   │   ├── opt.po
│   │   │           │   │   │   ├── orca.po
│   │   │           │   │   │   ├── qwen-chat.po
│   │   │           │   │   │   ├── starchat-beta.po
│   │   │           │   │   │   ├── starcoder.po
│   │   │           │   │   │   ├── starcoderplus.po
│   │   │           │   │   │   ├── tiny-llama.po
│   │   │           │   │   │   ├── vicuna-v1.3.po
│   │   │           │   │   │   ├── vicuna-v1.5-16k.po
│   │   │           │   │   │   ├── vicuna-v1.5.po
│   │   │           │   │   │   ├── wizardcoder-python-v1.0.po
│   │   │           │   │   │   ├── wizardlm-v1.0.po
│   │   │           │   │   │   ├── wizardmath-v1.0.po
│   │   │           │   │   │   ├── xverse-chat.po
│   │   │           │   │   │   ├── xverse.po
│   │   │           │   │   │   ├── yi-200k.po
│   │   │           │   │   │   ├── yi-chat.po
│   │   │           │   │   │   ├── yi.po
│   │   │           │   │   │   ├── zephyr-7b-alpha.po
│   │   │           │   │   │   └── zephyr-7b-beta.po
│   │   │           │   │   ├── rerank/
│   │   │           │   │   │   ├── bge-reranker-base.po
│   │   │           │   │   │   ├── bge-reranker-large.po
│   │   │           │   │   │   └── index.po
│   │   │           │   │   └── video/
│   │   │           │   │       ├── cogvideox-2b.po
│   │   │           │   │       └── index.po
│   │   │           │   ├── custom.po
│   │   │           │   ├── index.po
│   │   │           │   ├── lora.po
│   │   │           │   ├── model_abilities/
│   │   │           │   │   ├── audio.po
│   │   │           │   │   ├── chat.po
│   │   │           │   │   ├── embed.po
│   │   │           │   │   ├── flexible.po
│   │   │           │   │   ├── image.po
│   │   │           │   │   ├── index.po
│   │   │           │   │   ├── multimodal.po
│   │   │           │   │   ├── rerank.po
│   │   │           │   │   ├── tools.po
│   │   │           │   │   └── video.po
│   │   │           │   ├── model_memory.po
│   │   │           │   ├── model_update.po
│   │   │           │   ├── source/
│   │   │           │   │   └── source.po
│   │   │           │   ├── sources/
│   │   │           │   │   └── sources.po
│   │   │           │   ├── virtualenv.po
│   │   │           │   ├── xinference_model_hub.po
│   │   │           │   └── xinference_models_hub.po
│   │   │           ├── reference/
│   │   │           │   └── index.po
│   │   │           ├── reference.po
│   │   │           └── user_guide/
│   │   │               ├── auth_system.po
│   │   │               ├── backends.po
│   │   │               ├── cache_management.po
│   │   │               ├── client_api.po
│   │   │               ├── continuous_batching.po
│   │   │               ├── distributed_inference.po
│   │   │               ├── index.po
│   │   │               ├── launch.po
│   │   │               └── vllm_enhancement.po
│   │   ├── models/
│   │   │   ├── builtin/
│   │   │   │   ├── audio/
│   │   │   │   │   ├── belle-distilwhisper-large-v2-zh.rst
│   │   │   │   │   ├── belle-whisper-large-v2-zh.rst
│   │   │   │   │   ├── belle-whisper-large-v3-zh.rst
│   │   │   │   │   ├── chattts.rst
│   │   │   │   │   ├── cosyvoice-300m-instruct.rst
│   │   │   │   │   ├── cosyvoice-300m-sft.rst
│   │   │   │   │   ├── cosyvoice-300m.rst
│   │   │   │   │   ├── cosyvoice2-0.5b.rst
│   │   │   │   │   ├── f5-tts-mlx.rst
│   │   │   │   │   ├── f5-tts.rst
│   │   │   │   │   ├── fishspeech-1.5.rst
│   │   │   │   │   ├── fun-asr-mlt-nano-2512.rst
│   │   │   │   │   ├── fun-asr-nano-2512.rst
│   │   │   │   │   ├── index.rst
│   │   │   │   │   ├── indextts2.rst
│   │   │   │   │   ├── kokoro-82m-mlx.rst
│   │   │   │   │   ├── kokoro-82m-v1.1-zh.rst
│   │   │   │   │   ├── kokoro-82m.rst
│   │   │   │   │   ├── megatts3.rst
│   │   │   │   │   ├── melotts-chinese.rst
│   │   │   │   │   ├── melotts-english-v2.rst
│   │   │   │   │   ├── melotts-english-v3.rst
│   │   │   │   │   ├── melotts-english.rst
│   │   │   │   │   ├── melotts-french.rst
│   │   │   │   │   ├── melotts-japanese.rst
│   │   │   │   │   ├── melotts-korean.rst
│   │   │   │   │   ├── melotts-spanish.rst
│   │   │   │   │   ├── paraformer-zh-hotword.rst
│   │   │   │   │   ├── paraformer-zh-long.rst
│   │   │   │   │   ├── paraformer-zh-spk.rst
│   │   │   │   │   ├── paraformer-zh.rst
│   │   │   │   │   ├── qwen3-asr-0.6b.rst
│   │   │   │   │   ├── qwen3-asr-1.7b.rst
│   │   │   │   │   ├── seaco-paraformer-zh.rst
│   │   │   │   │   ├── sensevoicesmall.rst
│   │   │   │   │   ├── whisper-base-mlx.rst
│   │   │   │   │   ├── whisper-base.en-mlx.rst
│   │   │   │   │   ├── whisper-base.en.rst
│   │   │   │   │   ├── whisper-base.rst
│   │   │   │   │   ├── whisper-large-v3-mlx.rst
│   │   │   │   │   ├── whisper-large-v3-turbo-mlx.rst
│   │   │   │   │   ├── whisper-large-v3-turbo.rst
│   │   │   │   │   ├── whisper-large-v3.rst
│   │   │   │   │   ├── whisper-medium-mlx.rst
│   │   │   │   │   ├── whisper-medium.en-mlx.rst
│   │   │   │   │   ├── whisper-medium.en.rst
│   │   │   │   │   ├── whisper-medium.rst
│   │   │   │   │   ├── whisper-small-mlx.rst
│   │   │   │   │   ├── whisper-small.en-mlx.rst
│   │   │   │   │   ├── whisper-small.en.rst
│   │   │   │   │   ├── whisper-small.rst
│   │   │   │   │   ├── whisper-tiny-mlx.rst
│   │   │   │   │   ├── whisper-tiny.en-mlx.rst
│   │   │   │   │   ├── whisper-tiny.en.rst
│   │   │   │   │   └── whisper-tiny.rst
│   │   │   │   ├── embedding/
│   │   │   │   │   ├── bce-embedding-base_v1.rst
│   │   │   │   │   ├── bge-base-en-v1.5.rst
│   │   │   │   │   ├── bge-base-en.rst
│   │   │   │   │   ├── bge-base-zh-v1.5.rst
│   │   │   │   │   ├── bge-base-zh.rst
│   │   │   │   │   ├── bge-large-en-v1.5.rst
│   │   │   │   │   ├── bge-large-en.rst
│   │   │   │   │   ├── bge-large-zh-noinstruct.rst
│   │   │   │   │   ├── bge-large-zh-v1.5.rst
│   │   │   │   │   ├── bge-large-zh.rst
│   │   │   │   │   ├── bge-m3.rst
│   │   │   │   │   ├── bge-small-en-v1.5.rst
│   │   │   │   │   ├── bge-small-zh-v1.5.rst
│   │   │   │   │   ├── bge-small-zh.rst
│   │   │   │   │   ├── e5-large-v2.rst
│   │   │   │   │   ├── gme-qwen2-vl-2b-instruct.rst
│   │   │   │   │   ├── gme-qwen2-vl-7b-instruct.rst
│   │   │   │   │   ├── gte-base.rst
│   │   │   │   │   ├── gte-large.rst
│   │   │   │   │   ├── gte-qwen2.rst
│   │   │   │   │   ├── index.rst
│   │   │   │   │   ├── jina-clip-v2.rst
│   │   │   │   │   ├── jina-embeddings-v2-base-en.rst
│   │   │   │   │   ├── jina-embeddings-v2-base-zh.rst
│   │   │   │   │   ├── jina-embeddings-v2-small-en.rst
│   │   │   │   │   ├── jina-embeddings-v3.rst
│   │   │   │   │   ├── jina-embeddings-v4.rst
│   │   │   │   │   ├── m3e-base.rst
│   │   │   │   │   ├── m3e-large.rst
│   │   │   │   │   ├── m3e-small.rst
│   │   │   │   │   ├── multilingual-e5-large.rst
│   │   │   │   │   ├── qwen3-embedding-0.6b.rst
│   │   │   │   │   ├── qwen3-embedding-4b.rst
│   │   │   │   │   ├── qwen3-embedding-8b.rst
│   │   │   │   │   ├── qwen3-vl-embedding-2b.rst
│   │   │   │   │   ├── qwen3-vl-embedding-8b.rst
│   │   │   │   │   ├── text2vec-base-chinese-paraphrase.rst
│   │   │   │   │   ├── text2vec-base-chinese-sentence.rst
│   │   │   │   │   ├── text2vec-base-chinese.rst
│   │   │   │   │   ├── text2vec-base-multilingual.rst
│   │   │   │   │   └── text2vec-large-chinese.rst
│   │   │   │   ├── image/
│   │   │   │   │   ├── cogview4.rst
│   │   │   │   │   ├── deepseek-ocr.rst
│   │   │   │   │   ├── flux.1-dev.rst
│   │   │   │   │   ├── flux.1-kontext-dev.rst
│   │   │   │   │   ├── flux.1-schnell.rst
│   │   │   │   │   ├── flux.2-dev.rst
│   │   │   │   │   ├── flux.2-klein-4b.rst
│   │   │   │   │   ├── flux.2-klein-9b.rst
│   │   │   │   │   ├── got-ocr2_0.rst
│   │   │   │   │   ├── hunyuandit-v1.2-distilled.rst
│   │   │   │   │   ├── hunyuandit-v1.2.rst
│   │   │   │   │   ├── hunyuanocr.rst
│   │   │   │   │   ├── index.rst
│   │   │   │   │   ├── kolors.rst
│   │   │   │   │   ├── mineru2.5-2509-1.2b.rst
│   │   │   │   │   ├── paddleocr-vl.rst
│   │   │   │   │   ├── qwen-image-2512.rst
│   │   │   │   │   ├── qwen-image-edit-2509.rst
│   │   │   │   │   ├── qwen-image-edit-2511.rst
│   │   │   │   │   ├── qwen-image-edit.rst
│   │   │   │   │   ├── qwen-image-layered.rst
│   │   │   │   │   ├── qwen-image.rst
│   │   │   │   │   ├── sd-turbo.rst
│   │   │   │   │   ├── sd3-medium.rst
│   │   │   │   │   ├── sd3.5-large-turbo.rst
│   │   │   │   │   ├── sd3.5-large.rst
│   │   │   │   │   ├── sd3.5-medium.rst
│   │   │   │   │   ├── sdxl-turbo.rst
│   │   │   │   │   ├── stable-diffusion-2-inpainting.rst
│   │   │   │   │   ├── stable-diffusion-inpainting.rst
│   │   │   │   │   ├── stable-diffusion-v1.5.rst
│   │   │   │   │   ├── stable-diffusion-xl-base-1.0.rst
│   │   │   │   │   ├── stable-diffusion-xl-inpainting.rst
│   │   │   │   │   ├── z-image-turbo.rst
│   │   │   │   │   └── z-image.rst
│   │   │   │   ├── index.rst
│   │   │   │   ├── llm/
│   │   │   │   │   ├── baichuan-2-chat.rst
│   │   │   │   │   ├── baichuan-2.rst
│   │   │   │   │   ├── baichuan-m2.rst
│   │   │   │   │   ├── code-llama-instruct.rst
│   │   │   │   │   ├── code-llama-python.rst
│   │   │   │   │   ├── code-llama.rst
│   │   │   │   │   ├── codegeex4.rst
│   │   │   │   │   ├── codeqwen1.5-chat.rst
│   │   │   │   │   ├── codeqwen1.5.rst
│   │   │   │   │   ├── codeshell-chat.rst
│   │   │   │   │   ├── codeshell.rst
│   │   │   │   │   ├── codestral-v0.1.rst
│   │   │   │   │   ├── cogagent.rst
│   │   │   │   │   ├── deepseek-chat.rst
│   │   │   │   │   ├── deepseek-coder-instruct.rst
│   │   │   │   │   ├── deepseek-coder.rst
│   │   │   │   │   ├── deepseek-prover-v2.rst
│   │   │   │   │   ├── deepseek-r1-0528-qwen3.rst
│   │   │   │   │   ├── deepseek-r1-0528.rst
│   │   │   │   │   ├── deepseek-r1-distill-llama.rst
│   │   │   │   │   ├── deepseek-r1-distill-qwen.rst
│   │   │   │   │   ├── deepseek-r1.rst
│   │   │   │   │   ├── deepseek-v2-chat-0628.rst
│   │   │   │   │   ├── deepseek-v2-chat.rst
│   │   │   │   │   ├── deepseek-v2.5.rst
│   │   │   │   │   ├── deepseek-v3-0324.rst
│   │   │   │   │   ├── deepseek-v3.1.rst
│   │   │   │   │   ├── deepseek-v3.2-exp.rst
│   │   │   │   │   ├── deepseek-v3.2.rst
│   │   │   │   │   ├── deepseek-v3.rst
│   │   │   │   │   ├── deepseek-vl2.rst
│   │   │   │   │   ├── deepseek.rst
│   │   │   │   │   ├── dianjin-r1.rst
│   │   │   │   │   ├── ernie4.5.rst
│   │   │   │   │   ├── fin-r1.rst
│   │   │   │   │   ├── gemma-3-1b-it.rst
│   │   │   │   │   ├── gemma-3-it.rst
│   │   │   │   │   ├── glm-4.1v-thinking.rst
│   │   │   │   │   ├── glm-4.5.rst
│   │   │   │   │   ├── glm-4.5v.rst
│   │   │   │   │   ├── glm-4.6.rst
│   │   │   │   │   ├── glm-4.7-flash.rst
│   │   │   │   │   ├── glm-4.7.rst
│   │   │   │   │   ├── glm-4v.rst
│   │   │   │   │   ├── glm-5.rst
│   │   │   │   │   ├── glm-edge-chat.rst
│   │   │   │   │   ├── glm4-0414.rst
│   │   │   │   │   ├── glm4-chat-1m.rst
│   │   │   │   │   ├── glm4-chat.rst
│   │   │   │   │   ├── gorilla-openfunctions-v2.rst
│   │   │   │   │   ├── gpt-2.rst
│   │   │   │   │   ├── gpt-oss.rst
│   │   │   │   │   ├── huatuogpt-o1-llama-3.1.rst
│   │   │   │   │   ├── huatuogpt-o1-qwen2.5.rst
│   │   │   │   │   ├── index.rst
│   │   │   │   │   ├── internlm3-instruct.rst
│   │   │   │   │   ├── internvl3.rst
│   │   │   │   │   ├── kat-v1.rst
│   │   │   │   │   ├── kimi-k2.5.rst
│   │   │   │   │   ├── llama-2-chat.rst
│   │   │   │   │   ├── llama-2.rst
│   │   │   │   │   ├── llama-3-instruct.rst
│   │   │   │   │   ├── llama-3.1-instruct.rst
│   │   │   │   │   ├── llama-3.1.rst
│   │   │   │   │   ├── llama-3.2-vision-instruct.rst
│   │   │   │   │   ├── llama-3.2-vision.rst
│   │   │   │   │   ├── llama-3.3-instruct.rst
│   │   │   │   │   ├── llama-3.rst
│   │   │   │   │   ├── marco-o1.rst
│   │   │   │   │   ├── mineru2.5-2509-1.2b.rst
│   │   │   │   │   ├── minicpm-2b-dpo-bf16.rst
│   │   │   │   │   ├── minicpm-2b-dpo-fp16.rst
│   │   │   │   │   ├── minicpm-2b-dpo-fp32.rst
│   │   │   │   │   ├── minicpm-2b-sft-bf16.rst
│   │   │   │   │   ├── minicpm-2b-sft-fp32.rst
│   │   │   │   │   ├── minicpm-v-2.6.rst
│   │   │   │   │   ├── minicpm-v-4.5.rst
│   │   │   │   │   ├── minicpm3-4b.rst
│   │   │   │   │   ├── minicpm4.rst
│   │   │   │   │   ├── minimax-m2.5.rst
│   │   │   │   │   ├── minimax-m2.rst
│   │   │   │   │   ├── mistral-instruct-v0.1.rst
│   │   │   │   │   ├── mistral-instruct-v0.2.rst
│   │   │   │   │   ├── mistral-instruct-v0.3.rst
│   │   │   │   │   ├── mistral-large-instruct.rst
│   │   │   │   │   ├── mistral-nemo-instruct.rst
│   │   │   │   │   ├── mistral-v0.1.rst
│   │   │   │   │   ├── mixtral-8x22b-instruct-v0.1.rst
│   │   │   │   │   ├── mixtral-instruct-v0.1.rst
│   │   │   │   │   ├── mixtral-v0.1.rst
│   │   │   │   │   ├── moonlight-16b-a3b-instruct.rst
│   │   │   │   │   ├── openhermes-2.5.rst
│   │   │   │   │   ├── opt.rst
│   │   │   │   │   ├── orion-chat.rst
│   │   │   │   │   ├── ovis2.rst
│   │   │   │   │   ├── phi-2.rst
│   │   │   │   │   ├── phi-3-mini-128k-instruct.rst
│   │   │   │   │   ├── phi-3-mini-4k-instruct.rst
│   │   │   │   │   ├── qvq-72b-preview.rst
│   │   │   │   │   ├── qwen-chat.rst
│   │   │   │   │   ├── qwen1.5-chat.rst
│   │   │   │   │   ├── qwen1.5-moe-chat.rst
│   │   │   │   │   ├── qwen2-audio-instruct.rst
│   │   │   │   │   ├── qwen2-instruct.rst
│   │   │   │   │   ├── qwen2-moe-instruct.rst
│   │   │   │   │   ├── qwen2-vl-instruct.rst
│   │   │   │   │   ├── qwen2.5-coder-instruct.rst
│   │   │   │   │   ├── qwen2.5-coder.rst
│   │   │   │   │   ├── qwen2.5-instruct-1m.rst
│   │   │   │   │   ├── qwen2.5-instruct.rst
│   │   │   │   │   ├── qwen2.5-omni.rst
│   │   │   │   │   ├── qwen2.5-vl-instruct.rst
│   │   │   │   │   ├── qwen2.5.rst
│   │   │   │   │   ├── qwen3-coder.rst
│   │   │   │   │   ├── qwen3-instruct.rst
│   │   │   │   │   ├── qwen3-next-instruct.rst
│   │   │   │   │   ├── qwen3-next-thinking.rst
│   │   │   │   │   ├── qwen3-omni-instruct.rst
│   │   │   │   │   ├── qwen3-omni-thinking.rst
│   │   │   │   │   ├── qwen3-thinking.rst
│   │   │   │   │   ├── qwen3-vl-instruct.rst
│   │   │   │   │   ├── qwen3-vl-thinking.rst
│   │   │   │   │   ├── qwen3.5.rst
│   │   │   │   │   ├── qwen3.rst
│   │   │   │   │   ├── qwenlong-l1.rst
│   │   │   │   │   ├── qwq-32b-preview.rst
│   │   │   │   │   ├── qwq-32b.rst
│   │   │   │   │   ├── seallm_v2.5.rst
│   │   │   │   │   ├── seallm_v2.rst
│   │   │   │   │   ├── seallms-v3.rst
│   │   │   │   │   ├── seed-oss.rst
│   │   │   │   │   ├── skywork-math.rst
│   │   │   │   │   ├── skywork-or1-preview.rst
│   │   │   │   │   ├── skywork-or1.rst
│   │   │   │   │   ├── skywork.rst
│   │   │   │   │   ├── telechat.rst
│   │   │   │   │   ├── tiny-llama.rst
│   │   │   │   │   ├── wizardcoder-python-v1.0.rst
│   │   │   │   │   ├── wizardmath-v1.0.rst
│   │   │   │   │   ├── xiyansql-qwencoder-2504.rst
│   │   │   │   │   ├── xverse-chat.rst
│   │   │   │   │   ├── xverse.rst
│   │   │   │   │   ├── yi-1.5-chat-16k.rst
│   │   │   │   │   ├── yi-1.5-chat.rst
│   │   │   │   │   ├── yi-1.5.rst
│   │   │   │   │   ├── yi-200k.rst
│   │   │   │   │   ├── yi-chat.rst
│   │   │   │   │   └── yi.rst
│   │   │   │   ├── rerank/
│   │   │   │   │   ├── bce-reranker-base_v1.rst
│   │   │   │   │   ├── bge-reranker-base.rst
│   │   │   │   │   ├── bge-reranker-large.rst
│   │   │   │   │   ├── bge-reranker-v2-gemma.rst
│   │   │   │   │   ├── bge-reranker-v2-m3.rst
│   │   │   │   │   ├── bge-reranker-v2-minicpm-layerwise.rst
│   │   │   │   │   ├── index.rst
│   │   │   │   │   ├── jina-reranker-v2.rst
│   │   │   │   │   ├── jina-reranker-v3.rst
│   │   │   │   │   ├── minicpm-reranker.rst
│   │   │   │   │   ├── qwen3-reranker-0.6b.rst
│   │   │   │   │   ├── qwen3-reranker-4b.rst
│   │   │   │   │   ├── qwen3-reranker-8b.rst
│   │   │   │   │   ├── qwen3-vl-reranker-2b.rst
│   │   │   │   │   └── qwen3-vl-reranker-8b.rst
│   │   │   │   └── video/
│   │   │   │       ├── cogvideox-2b.rst
│   │   │   │       ├── cogvideox-5b.rst
│   │   │   │       ├── hunyuanvideo.rst
│   │   │   │       ├── index.rst
│   │   │   │       ├── wan2.1-1.3b.rst
│   │   │   │       ├── wan2.1-14b.rst
│   │   │   │       ├── wan2.1-flf2v-14b-720p.rst
│   │   │   │       ├── wan2.1-i2v-14b-480p.rst
│   │   │   │       ├── wan2.1-i2v-14b-720p.rst
│   │   │   │       ├── wan2.2-a14b.rst
│   │   │   │       ├── wan2.2-i2v-a14b.rst
│   │   │   │       └── wan2.2-ti2v-5b.rst
│   │   │   ├── custom.rst
│   │   │   ├── index.rst
│   │   │   ├── lora.rst
│   │   │   ├── model_abilities/
│   │   │   │   ├── audio.rst
│   │   │   │   ├── chat.rst
│   │   │   │   ├── embed.rst
│   │   │   │   ├── flexible.rst
│   │   │   │   ├── image.rst
│   │   │   │   ├── index.rst
│   │   │   │   ├── multimodal.rst
│   │   │   │   ├── rerank.rst
│   │   │   │   ├── tools.rst
│   │   │   │   └── video.rst
│   │   │   ├── model_memory.rst
│   │   │   ├── model_update.rst
│   │   │   ├── sources/
│   │   │   │   └── sources.rst
│   │   │   ├── virtualenv.rst
│   │   │   └── xinference_models_hub.rst
│   │   ├── norm_zh.py
│   │   ├── reference/
│   │   │   └── index.rst
│   │   └── user_guide/
│   │       ├── auth_system.rst
│   │       ├── backends.rst
│   │       ├── client_api.rst
│   │       ├── continuous_batching.rst
│   │       ├── distributed_inference.rst
│   │       ├── index.rst
│   │       ├── launch.rst
│   │       ├── metrics.rst
│   │       └── vllm_enhancement.rst
│   └── templates/
│       ├── audio.rst.jinja
│       ├── audio_index.rst.jinja
│       ├── embedding.rst.jinja
│       ├── embedding_index.rst.jinja
│       ├── image.rst.jinja
│       ├── image_index.rst.jinja
│       ├── llm.rst.jinja
│       ├── llm_index.rst.jinja
│       ├── metrics.jinja
│       ├── rerank.rst.jinja
│       ├── rerank_index.rst.jinja
│       ├── video.rst.jinja
│       └── video_index.rst.jinja
├── examples/
│   ├── AI_podcast.py
│   ├── AI_podcast_ZH.py
│   ├── AI_translate.py
│   ├── Custom_StableDiffusion_ControlNet.ipynb
│   ├── FunctionCall.ipynb
│   ├── LangChain_QA.ipynb
│   ├── LangChain_Streamlit_Doc_Chat.py
│   ├── StableDiffusionControlNet.ipynb
│   ├── Xinference_Quick_Start.ipynb
│   ├── audio_to_text.ipynb
│   ├── chat.py
│   ├── chat_vl.ipynb
│   └── gradio_chatinterface.py
├── pyproject.toml
├── setup.cfg
├── setup.py
├── versioneer.py
└── xinference/
    ├── __init__.py
    ├── _compat.py
    ├── _version.py
    ├── api/
    │   ├── __init__.py
    │   ├── dependencies.py
    │   ├── oauth2/
    │   │   ├── __init__.py
    │   │   ├── auth_service.py
    │   │   ├── types.py
    │   │   └── utils.py
    │   ├── responses.py
    │   ├── restful_api.py
    │   ├── routers/
    │   │   ├── __init__.py
    │   │   ├── admin.py
    │   │   ├── audio.py
    │   │   ├── embeddings.py
    │   │   ├── images.py
    │   │   ├── llm.py
    │   │   ├── models.py
    │   │   ├── rerank.py
    │   │   └── videos.py
    │   ├── schemas/
    │   │   ├── __init__.py
    │   │   └── requests.py
    │   ├── tests/
    │   │   ├── __init__.py
    │   │   ├── test_admin.py
    │   │   └── test_utils.py
    │   └── utils.py
    ├── client/
    │   ├── __init__.py
    │   ├── common.py
    │   ├── handlers.py
    │   ├── restful/
    │   │   ├── __init__.py
    │   │   ├── async_restful_client.py
    │   │   └── restful_client.py
    │   └── tests/
    │       ├── __init__.py
    │       ├── test_async_client.py
    │       ├── test_async_client_with_auth.py
    │       ├── test_client.py
    │       └── test_client_with_auth.py
    ├── conftest.py
    ├── constants.py
    ├── core/
    │   ├── __init__.py
    │   ├── cache_tracker.py
    │   ├── event.py
    │   ├── launch_strategy.py
    │   ├── metrics.py
    │   ├── model.py
    │   ├── otel.py
    │   ├── progress_tracker.py
    │   ├── resource.py
    │   ├── status_guard.py
    │   ├── supervisor.py
    │   ├── tests/
    │   │   ├── __init__.py
    │   │   ├── test_continuous_batching.py
    │   │   ├── test_launch_strategy.py
    │   │   ├── test_metrics.py
    │   │   ├── test_model.py
    │   │   ├── test_progressor.py
    │   │   ├── test_restful_api.py
    │   │   ├── test_types.py
    │   │   ├── test_utils.py
    │   │   └── test_worker.py
    │   ├── utils.py
    │   ├── virtual_env_manager.py
    │   └── worker.py
    ├── deploy/
    │   ├── __init__.py
    │   ├── cmdline.py
    │   ├── docker/
    │   │   ├── Dockerfile
    │   │   ├── Dockerfile.cpu
    │   │   ├── docker-compose-distributed.yml
    │   │   ├── docker-compose.yml
    │   │   ├── requirements/
    │   │   │   ├── requirements-base.txt
    │   │   │   ├── requirements-ml.txt
    │   │   │   └── requirements-models.txt
    │   │   └── requirements_cpu/
    │   │       ├── requirements_cpu-base.txt
    │   │       ├── requirements_cpu-ml.txt
    │   │       └── requirements_cpu-models.txt
    │   ├── local.py
    │   ├── supervisor.py
    │   ├── test/
    │   │   ├── __init__.py
    │   │   └── test_cmdline.py
    │   ├── utils.py
    │   └── worker.py
    ├── device_utils.py
    ├── fields.py
    ├── isolation.py
    ├── model/
    │   ├── __init__.py
    │   ├── audio/
    │   │   ├── __init__.py
    │   │   ├── chattts.py
    │   │   ├── core.py
    │   │   ├── cosyvoice.py
    │   │   ├── custom.py
    │   │   ├── f5tts.py
    │   │   ├── f5tts_mlx.py
    │   │   ├── fish_speech.py
    │   │   ├── funasr.py
    │   │   ├── indextts2.py
    │   │   ├── kokoro.py
    │   │   ├── kokoro_mlx.py
    │   │   ├── kokoro_zh.py
    │   │   ├── megatts.py
    │   │   ├── melotts.py
    │   │   ├── model_spec.json
    │   │   ├── qwen3_asr.py
    │   │   ├── tests/
    │   │   │   ├── __init__.py
    │   │   │   ├── bbc_news.npy
    │   │   │   ├── jfk.flac
    │   │   │   ├── test_chattts.py
    │   │   │   ├── test_cosyvoice.py
    │   │   │   ├── test_f5tts.py
    │   │   │   ├── test_f5tts_mlx.py
    │   │   │   ├── test_fish_speech.py
    │   │   │   ├── test_funasr.py
    │   │   │   ├── test_kokoro.py
    │   │   │   ├── test_megatts.py
    │   │   │   ├── test_melotts.py
    │   │   │   ├── test_whisper.py
    │   │   │   └── test_whisper_mlx.py
    │   │   ├── utils.py
    │   │   ├── whisper.py
    │   │   └── whisper_mlx.py
    │   ├── batch.py
    │   ├── cache_manager.py
    │   ├── core.py
    │   ├── custom.py
    │   ├── embedding/
    │   │   ├── __init__.py
    │   │   ├── cache_manager.py
    │   │   ├── core.py
    │   │   ├── custom.py
    │   │   ├── embed_family.py
    │   │   ├── flag/
    │   │   │   ├── __init__.py
    │   │   │   ├── core.py
    │   │   │   └── tests/
    │   │   │       ├── __init__.py
    │   │   │       └── test_flag.py
    │   │   ├── llama_cpp/
    │   │   │   ├── __init__.py
    │   │   │   ├── core.py
    │   │   │   └── tests/
    │   │   │       ├── __init__.py
    │   │   │       └── test_llama_cpp.py
    │   │   ├── model_spec.json
    │   │   ├── sentence_transformers/
    │   │   │   ├── __init__.py
    │   │   │   ├── core.py
    │   │   │   └── tests/
    │   │   │       ├── __init__.py
    │   │   │       └── test_sentence_transformers.py
    │   │   ├── tests/
    │   │   │   ├── __init__.py
    │   │   │   ├── test_embedding_models.py
    │   │   │   ├── test_integrated_embedding.py
    │   │   │   └── test_qwen3_vl_engine_params.py
    │   │   └── vllm/
    │   │       ├── __init__.py
    │   │       ├── core.py
    │   │       └── tests/
    │   │           ├── __init__.py
    │   │           └── test_vllm_embedding.py
    │   ├── flexible/
    │   │   ├── __init__.py
    │   │   ├── core.py
    │   │   ├── custom.py
    │   │   ├── launchers/
    │   │   │   ├── __init__.py
    │   │   │   ├── image_process_launcher.py
    │   │   │   ├── modelscope_launcher.py
    │   │   │   ├── transformers_launcher.py
    │   │   │   └── yolo_launcher.py
    │   │   ├── tests/
    │   │   │   ├── __init__.py
    │   │   │   └── test_flexible_models.py
    │   │   └── utils.py
    │   ├── image/
    │   │   ├── __init__.py
    │   │   ├── cache_manager.py
    │   │   ├── core.py
    │   │   ├── custom.py
    │   │   ├── engine.py
    │   │   ├── engine_family.py
    │   │   ├── model_spec.json
    │   │   ├── ocr/
    │   │   │   ├── __init__.py
    │   │   │   ├── deepseek_ocr.py
    │   │   │   ├── got_ocr2.py
    │   │   │   ├── hunyuan_ocr.py
    │   │   │   ├── mlx.py
    │   │   │   ├── ocr_family.py
    │   │   │   ├── paddleocr_vl.py
    │   │   │   └── vllm.py
    │   │   ├── scheduler/
    │   │   │   ├── __init__.py
    │   │   │   └── flux.py
    │   │   ├── sdapi.py
    │   │   ├── stable_diffusion/
    │   │   │   ├── __init__.py
    │   │   │   ├── core.py
    │   │   │   └── mlx.py
    │   │   ├── tests/
    │   │   │   ├── __init__.py
    │   │   │   ├── test_got_ocr2.py
    │   │   │   └── test_stable_diffusion.py
    │   │   └── utils.py
    │   ├── llm/
    │   │   ├── __init__.py
    │   │   ├── cache_manager.py
    │   │   ├── config_parser.py
    │   │   ├── core.py
    │   │   ├── custom.py
    │   │   ├── harmony.py
    │   │   ├── llama_cpp/
    │   │   │   ├── __init__.py
    │   │   │   ├── core.py
    │   │   │   └── tests/
    │   │   │       ├── __init__.py
    │   │   │       ├── test_gguf.py
    │   │   │       └── test_structured.py
    │   │   ├── llm_family.json
    │   │   ├── llm_family.py
    │   │   ├── lmdeploy/
    │   │   │   ├── __init__.py
    │   │   │   ├── core.py
    │   │   │   └── tests/
    │   │   │       └── __init__.py
    │   │   ├── memory.py
    │   │   ├── mlx/
    │   │   │   ├── __init__.py
    │   │   │   ├── core.py
    │   │   │   ├── distributed_models/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── core.py
    │   │   │   │   ├── deepseek_v3.py
    │   │   │   │   ├── qwen2.py
    │   │   │   │   ├── qwen3.py
    │   │   │   │   └── qwen3_moe.py
    │   │   │   └── tests/
    │   │   │       ├── __init__.py
    │   │   │       ├── test_distributed_model.py
    │   │   │       └── test_mlx.py
    │   │   ├── reasoning_parser.py
    │   │   ├── sglang/
    │   │   │   ├── __init__.py
    │   │   │   └── core.py
    │   │   ├── tests/
    │   │   │   ├── __init__.py
    │   │   │   ├── test_harmony.py
    │   │   │   ├── test_llm_family.py
    │   │   │   ├── test_llm_model.py
    │   │   │   ├── test_memory_estimate.py
    │   │   │   ├── test_multimodal.py
    │   │   │   ├── test_stream_options.py
    │   │   │   └── test_utils.py
    │   │   ├── tool_parsers/
    │   │   │   ├── __init__.py
    │   │   │   ├── abstract_tool_parser.py
    │   │   │   ├── deepseek_r1_tool_parser.py
    │   │   │   ├── deepseek_v3_1_tool_parser.py
    │   │   │   ├── deepseek_v3_tool_parser.py
    │   │   │   ├── glm4_tool_parser.py
    │   │   │   ├── llama3_tool_parser.py
    │   │   │   ├── minimax_tool_parser.py
    │   │   │   ├── qwen_tool_parser.py
    │   │   │   └── tests/
    │   │   │       ├── __init__.py
    │   │   │       ├── test_deepseek_r1_tool_parser.py
    │   │   │       ├── test_deepseek_v3_1_tool_parser.py
    │   │   │       ├── test_deepseek_v3_tool_parser.py
    │   │   │       ├── test_glm4_tool_parser.py
    │   │   │       ├── test_llama3_tool_parser.py
    │   │   │       └── test_qwen_tool_parser.py
    │   │   ├── transformers/
    │   │   │   ├── __init__.py
    │   │   │   ├── chatglm.py
    │   │   │   ├── core.py
    │   │   │   ├── deepseek_v2.py
    │   │   │   ├── gemma3.py
    │   │   │   ├── gpt_oss.py
    │   │   │   ├── multimodal/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── cogagent.py
    │   │   │   │   ├── core.py
    │   │   │   │   ├── deepseek_vl2.py
    │   │   │   │   ├── gemma3.py
    │   │   │   │   ├── glm4_1v.py
    │   │   │   │   ├── glm4v.py
    │   │   │   │   ├── intern_vl.py
    │   │   │   │   ├── minicpmv26.py
    │   │   │   │   ├── minicpmv45.py
    │   │   │   │   ├── ovis2.py
    │   │   │   │   ├── qwen-omni.py
    │   │   │   │   ├── qwen2_audio.py
    │   │   │   │   └── qwen2_vl.py
    │   │   │   ├── opt.py
    │   │   │   ├── tensorizer_utils.py
    │   │   │   ├── tests/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── test_opt.py
    │   │   │   │   └── test_tensorizer.py
    │   │   │   └── utils.py
    │   │   ├── utils.py
    │   │   └── vllm/
    │   │       ├── __init__.py
    │   │       ├── core.py
    │   │       ├── distributed_executor.py
    │   │       ├── distributed_executor_v1.py
    │   │       ├── tests/
    │   │       │   ├── __init__.py
    │   │       │   ├── test_core_chat_model.py
    │   │       │   └── test_distributed_executor.py
    │   │       ├── utils.py
    │   │       └── xavier/
    │   │           ├── __init__.py
    │   │           ├── allocator.py
    │   │           ├── block.py
    │   │           ├── block_manager.py
    │   │           ├── block_tracker.py
    │   │           ├── collective.py
    │   │           ├── collective_manager.py
    │   │           ├── engine.py
    │   │           ├── executor.py
    │   │           ├── scheduler.py
    │   │           ├── test/
    │   │           │   ├── __init__.py
    │   │           │   └── test_xavier.py
    │   │           ├── transfer.py
    │   │           └── utils.py
    │   ├── rerank/
    │   │   ├── __init__.py
    │   │   ├── cache_manager.py
    │   │   ├── core.py
    │   │   ├── custom.py
    │   │   ├── llama_cpp/
    │   │   │   ├── __init__.py
    │   │   │   ├── core.py
    │   │   │   └── tests/
    │   │   │       ├── __init__.py
    │   │   │       └── test_llama_cpp.py
    │   │   ├── model_spec.json
    │   │   ├── rerank_family.py
    │   │   ├── sentence_transformers/
    │   │   │   ├── __init__.py
    │   │   │   ├── core.py
    │   │   │   └── tests/
    │   │   │       ├── __init__.py
    │   │   │       └── test_sentence_transformers.py
    │   │   ├── tests/
    │   │   │   ├── __init__.py
    │   │   │   ├── test_qwen3_vl_reranker_virtualenv.py
    │   │   │   └── test_rerank.py
    │   │   ├── utils.py
    │   │   └── vllm/
    │   │       ├── __init__.py
    │   │       ├── core.py
    │   │       └── tests/
    │   │           ├── __init__.py
    │   │           └── test_vllm.py
    │   ├── scheduler/
    │   │   ├── __init__.py
    │   │   ├── batch.py
    │   │   ├── core.py
    │   │   └── request.py
    │   ├── tests/
    │   │   ├── __init__.py
    │   │   └── test_utils.py
    │   ├── utils.py
    │   └── video/
    │       ├── __init__.py
    │       ├── cache_manager.py
    │       ├── core.py
    │       ├── diffusers.py
    │       ├── model_spec.json
    │       └── tests/
    │           ├── __init__.py
    │           └── test_diffusers_video.py
    ├── thirdparty/
    │   ├── __init__.py
    │   ├── audiotools/
    │   │   ├── __init__.py
    │   │   ├── core/
    │   │   │   ├── __init__.py
    │   │   │   ├── audio_signal.py
    │   │   │   ├── display.py
    │   │   │   ├── dsp.py
    │   │   │   ├── effects.py
    │   │   │   ├── ffmpeg.py
    │   │   │   ├── loudness.py
    │   │   │   ├── playback.py
    │   │   │   ├── templates/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── headers.html
    │   │   │   │   ├── pandoc.css
    │   │   │   │   └── widget.html
    │   │   │   ├── util.py
    │   │   │   └── whisper.py
    │   │   ├── data/
    │   │   │   ├── __init__.py
    │   │   │   ├── datasets.py
    │   │   │   ├── preprocess.py
    │   │   │   └── transforms.py
    │   │   ├── metrics/
    │   │   │   ├── __init__.py
    │   │   │   ├── distance.py
    │   │   │   ├── quality.py
    │   │   │   └── spectral.py
    │   │   ├── ml/
    │   │   │   ├── __init__.py
    │   │   │   ├── accelerator.py
    │   │   │   ├── decorators.py
    │   │   │   ├── experiment.py
    │   │   │   └── layers/
    │   │   │       ├── __init__.py
    │   │   │       ├── base.py
    │   │   │       └── spectral_gate.py
    │   │   ├── post.py
    │   │   └── preference.py
    │   ├── cosyvoice/
    │   │   ├── __init__.py
    │   │   ├── bin/
    │   │   │   ├── average_model.py
    │   │   │   ├── export_jit.py
    │   │   │   ├── export_onnx.py
    │   │   │   ├── inference_deprecated.py
    │   │   │   ├── spk2info.pt
    │   │   │   └── train.py
    │   │   ├── cli/
    │   │   │   ├── __init__.py
    │   │   │   ├── cosyvoice.py
    │   │   │   ├── frontend.py
    │   │   │   └── model.py
    │   │   ├── dataset/
    │   │   │   ├── __init__.py
    │   │   │   ├── dataset.py
    │   │   │   └── processor.py
    │   │   ├── flow/
    │   │   │   ├── decoder.py
    │   │   │   ├── flow.py
    │   │   │   ├── flow_matching.py
    │   │   │   └── length_regulator.py
    │   │   ├── hifigan/
    │   │   │   ├── discriminator.py
    │   │   │   ├── f0_predictor.py
    │   │   │   ├── generator.py
    │   │   │   └── hifigan.py
    │   │   ├── llm/
    │   │   │   └── llm.py
    │   │   ├── tokenizer/
    │   │   │   ├── assets/
    │   │   │   │   └── multilingual_zh_ja_yue_char_del.tiktoken
    │   │   │   └── tokenizer.py
    │   │   ├── transformer/
    │   │   │   ├── __init__.py
    │   │   │   ├── activation.py
    │   │   │   ├── attention.py
    │   │   │   ├── convolution.py
    │   │   │   ├── decoder.py
    │   │   │   ├── decoder_layer.py
    │   │   │   ├── embedding.py
    │   │   │   ├── encoder.py
    │   │   │   ├── encoder_layer.py
    │   │   │   ├── label_smoothing_loss.py
    │   │   │   ├── positionwise_feed_forward.py
    │   │   │   ├── subsampling.py
    │   │   │   └── upsample_encoder.py
    │   │   ├── utils/
    │   │   │   ├── __init__.py
    │   │   │   ├── class_utils.py
    │   │   │   ├── common.py
    │   │   │   ├── executor.py
    │   │   │   ├── file_utils.py
    │   │   │   ├── frontend_utils.py
    │   │   │   ├── losses.py
    │   │   │   ├── mask.py
    │   │   │   ├── scheduler.py
    │   │   │   └── train_utils.py
    │   │   └── vllm/
    │   │       └── cosyvoice2.py
    │   ├── deepseek_vl/
    │   │   ├── __init__.py
    │   │   ├── models/
    │   │   │   ├── __init__.py
    │   │   │   ├── clip_encoder.py
    │   │   │   ├── image_processing_vlm.py
    │   │   │   ├── modeling_vlm.py
    │   │   │   ├── processing_vlm.py
    │   │   │   ├── projector.py
    │   │   │   ├── sam.py
    │   │   │   └── siglip_vit.py
    │   │   ├── serve/
    │   │   │   ├── __init__.py
    │   │   │   ├── app_deepseek.py
    │   │   │   ├── app_modules/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── gradio_utils.py
    │   │   │   │   ├── overwrites.py
    │   │   │   │   ├── presets.py
    │   │   │   │   └── utils.py
    │   │   │   ├── assets/
    │   │   │   │   ├── Kelpy-Codos.js
    │   │   │   │   ├── custom.css
    │   │   │   │   └── custom.js
    │   │   │   └── inference.py
    │   │   └── utils/
    │   │       ├── __init__.py
    │   │       ├── conversation.py
    │   │       └── io.py
    │   ├── deepseek_vl2/
    │   │   ├── __init__.py
    │   │   ├── models/
    │   │   │   ├── __init__.py
    │   │   │   ├── configuration_deepseek.py
    │   │   │   ├── conversation.py
    │   │   │   ├── modeling_deepseek.py
    │   │   │   ├── modeling_deepseek_vl_v2.py
    │   │   │   ├── processing_deepseek_vl_v2.py
    │   │   │   └── siglip_vit.py
    │   │   ├── serve/
    │   │   │   ├── __init__.py
    │   │   │   ├── app_modules/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── gradio_utils.py
    │   │   │   │   ├── overwrites.py
    │   │   │   │   ├── presets.py
    │   │   │   │   └── utils.py
    │   │   │   ├── assets/
    │   │   │   │   ├── Kelpy-Codos.js
    │   │   │   │   ├── custom.css
    │   │   │   │   ├── custom.js
    │   │   │   │   └── simsun.ttc
    │   │   │   └── inference.py
    │   │   └── utils/
    │   │       ├── __init__.py
    │   │       └── io.py
    │   ├── f5_tts/
    │   │   ├── __init__.py
    │   │   ├── api.py
    │   │   ├── configs/
    │   │   │   ├── E2TTS_Base_train.yaml
    │   │   │   ├── E2TTS_Small_train.yaml
    │   │   │   ├── F5TTS_Base_train.yaml
    │   │   │   └── F5TTS_Small_train.yaml
    │   │   ├── eval/
    │   │   │   ├── README.md
    │   │   │   ├── ecapa_tdnn.py
    │   │   │   ├── eval_infer_batch.py
    │   │   │   ├── eval_infer_batch.sh
    │   │   │   ├── eval_librispeech_test_clean.py
    │   │   │   ├── eval_seedtts_testset.py
    │   │   │   └── utils_eval.py
    │   │   ├── infer/
    │   │   │   ├── README.md
    │   │   │   ├── examples/
    │   │   │   │   ├── basic/
    │   │   │   │   │   └── basic.toml
    │   │   │   │   ├── multi/
    │   │   │   │   │   ├── country.flac
    │   │   │   │   │   ├── main.flac
    │   │   │   │   │   ├── story.toml
    │   │   │   │   │   ├── story.txt
    │   │   │   │   │   └── town.flac
    │   │   │   │   └── vocab.txt
    │   │   │   ├── infer_cli.py
    │   │   │   ├── infer_gradio.py
    │   │   │   ├── speech_edit.py
    │   │   │   └── utils_infer.py
    │   │   ├── model/
    │   │   │   ├── __init__.py
    │   │   │   ├── backbones/
    │   │   │   │   ├── README.md
    │   │   │   │   ├── dit.py
    │   │   │   │   ├── mmdit.py
    │   │   │   │   └── unett.py
    │   │   │   ├── cfm.py
    │   │   │   ├── dataset.py
    │   │   │   ├── modules.py
    │   │   │   ├── trainer.py
    │   │   │   └── utils.py
    │   │   ├── scripts/
    │   │   │   ├── count_max_epoch.py
    │   │   │   └── count_params_gflops.py
    │   │   ├── socket_server.py
    │   │   └── train/
    │   │       ├── README.md
    │   │       ├── datasets/
    │   │       │   ├── prepare_csv_wavs.py
    │   │       │   ├── prepare_emilia.py
    │   │       │   ├── prepare_libritts.py
    │   │       │   ├── prepare_ljspeech.py
    │   │       │   └── prepare_wenetspeech4tts.py
    │   │       ├── finetune_cli.py
    │   │       ├── finetune_gradio.py
    │   │       └── train.py
    │   ├── fish_speech/
    │   │   ├── __init__.py
    │   │   ├── fish_speech/
    │   │   │   ├── __init__.py
    │   │   │   ├── callbacks/
    │   │   │   │   ├── __init__.py
    │   │   │   │   └── grad_norm.py
    │   │   │   ├── configs/
    │   │   │   │   ├── base.yaml
    │   │   │   │   ├── firefly_gan_vq.yaml
    │   │   │   │   ├── lora/
    │   │   │   │   │   └── r_8_alpha_16.yaml
    │   │   │   │   └── text2semantic_finetune.yaml
    │   │   │   ├── conversation.py
    │   │   │   ├── datasets/
    │   │   │   │   ├── concat_repeat.py
    │   │   │   │   ├── protos/
    │   │   │   │   │   ├── text-data.proto
    │   │   │   │   │   ├── text_data_pb2.py
    │   │   │   │   │   └── text_data_stream.py
    │   │   │   │   ├── semantic.py
    │   │   │   │   └── vqgan.py
    │   │   │   ├── i18n/
    │   │   │   │   ├── README.md
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── core.py
    │   │   │   │   ├── locale/
    │   │   │   │   │   ├── en_US.json
    │   │   │   │   │   ├── es_ES.json
    │   │   │   │   │   ├── ja_JP.json
    │   │   │   │   │   ├── ko_KR.json
    │   │   │   │   │   ├── pt_BR.json
    │   │   │   │   │   └── zh_CN.json
    │   │   │   │   └── scan.py
    │   │   │   ├── models/
    │   │   │   │   ├── text2semantic/
    │   │   │   │   │   ├── __init__.py
    │   │   │   │   │   ├── lit_module.py
    │   │   │   │   │   ├── llama.py
    │   │   │   │   │   └── lora.py
    │   │   │   │   └── vqgan/
    │   │   │   │       ├── __init__.py
    │   │   │   │       ├── modules/
    │   │   │   │       │   ├── firefly.py
    │   │   │   │       │   └── fsq.py
    │   │   │   │       └── utils.py
    │   │   │   ├── scheduler.py
    │   │   │   ├── text/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── chn_text_norm/
    │   │   │   │   │   ├── .gitignore
    │   │   │   │   │   ├── README.md
    │   │   │   │   │   ├── __init__.py
    │   │   │   │   │   ├── basic_class.py
    │   │   │   │   │   ├── basic_constant.py
    │   │   │   │   │   ├── basic_util.py
    │   │   │   │   │   ├── cardinal.py
    │   │   │   │   │   ├── date.py
    │   │   │   │   │   ├── digit.py
    │   │   │   │   │   ├── fraction.py
    │   │   │   │   │   ├── money.py
    │   │   │   │   │   ├── percentage.py
    │   │   │   │   │   ├── telephone.py
    │   │   │   │   │   └── text.py
    │   │   │   │   ├── clean.py
    │   │   │   │   └── spliter.py
    │   │   │   ├── tokenizer.py
    │   │   │   ├── train.py
    │   │   │   ├── utils/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── braceexpand.py
    │   │   │   │   ├── context.py
    │   │   │   │   ├── file.py
    │   │   │   │   ├── instantiators.py
    │   │   │   │   ├── logger.py
    │   │   │   │   ├── logging_utils.py
    │   │   │   │   ├── rich_utils.py
    │   │   │   │   ├── spectrogram.py
    │   │   │   │   └── utils.py
    │   │   │   └── webui/
    │   │   │       ├── css/
    │   │   │       │   └── style.css
    │   │   │       ├── html/
    │   │   │       │   └── footer.html
    │   │   │       ├── js/
    │   │   │       │   └── animate.js
    │   │   │       ├── launch_utils.py
    │   │   │       └── manage.py
    │   │   └── tools/
    │   │       ├── api_client.py
    │   │       ├── api_server.py
    │   │       ├── download_models.py
    │   │       ├── e2e_webui.py
    │   │       ├── extract_model.py
    │   │       ├── file.py
    │   │       ├── fish_e2e.py
    │   │       ├── inference_engine/
    │   │       │   ├── __init__.py
    │   │       │   ├── reference_loader.py
    │   │       │   ├── utils.py
    │   │       │   └── vq_manager.py
    │   │       ├── llama/
    │   │       │   ├── build_dataset.py
    │   │       │   ├── eval_in_context.py
    │   │       │   ├── generate.py
    │   │       │   ├── merge_lora.py
    │   │       │   ├── quantize.py
    │   │       │   └── rebuild_tokenizer.py
    │   │       ├── run_webui.py
    │   │       ├── schema.py
    │   │       ├── sensevoice/
    │   │       │   ├── README.md
    │   │       │   ├── __init__.py
    │   │       │   ├── auto_model.py
    │   │       │   ├── fun_asr.py
    │   │       │   └── vad_utils.py
    │   │       ├── server/
    │   │       │   ├── agent/
    │   │       │   │   ├── __init__.py
    │   │       │   │   ├── generate.py
    │   │       │   │   ├── generation_utils.py
    │   │       │   │   └── pre_generation_utils.py
    │   │       │   ├── api_utils.py
    │   │       │   ├── exception_handler.py
    │   │       │   ├── inference.py
    │   │       │   ├── model_manager.py
    │   │       │   ├── model_utils.py
    │   │       │   └── views.py
    │   │       ├── smart_pad.py
    │   │       ├── vqgan/
    │   │       │   ├── create_train_split.py
    │   │       │   ├── extract_vq.py
    │   │       │   └── inference.py
    │   │       ├── webui/
    │   │       │   ├── __init__.py
    │   │       │   ├── inference.py
    │   │       │   └── variables.py
    │   │       └── whisper_asr.py
    │   ├── indextts/
    │   │   ├── BigVGAN/
    │   │   │   ├── ECAPA_TDNN.py
    │   │   │   ├── __init__.py
    │   │   │   ├── activations.py
    │   │   │   ├── alias_free_activation/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── cuda/
    │   │   │   │   │   ├── .gitignore
    │   │   │   │   │   ├── __init__.py
    │   │   │   │   │   ├── activation1d.py
    │   │   │   │   │   ├── anti_alias_activation.cpp
    │   │   │   │   │   ├── anti_alias_activation_cuda.cu
    │   │   │   │   │   ├── compat.h
    │   │   │   │   │   ├── load.py
    │   │   │   │   │   └── type_shim.h
    │   │   │   │   └── torch/
    │   │   │   │       ├── __init__.py
    │   │   │   │       ├── act.py
    │   │   │   │       ├── filter.py
    │   │   │   │       └── resample.py
    │   │   │   ├── alias_free_torch/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── act.py
    │   │   │   │   ├── filter.py
    │   │   │   │   └── resample.py
    │   │   │   ├── bigvgan.py
    │   │   │   ├── models.py
    │   │   │   ├── nnet/
    │   │   │   │   ├── CNN.py
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── linear.py
    │   │   │   │   └── normalization.py
    │   │   │   └── utils.py
    │   │   ├── __init__.py
    │   │   ├── cli.py
    │   │   ├── gpt/
    │   │   │   ├── __init__.py
    │   │   │   ├── conformer/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── attention.py
    │   │   │   │   ├── embedding.py
    │   │   │   │   └── subsampling.py
    │   │   │   ├── conformer_encoder.py
    │   │   │   ├── model.py
    │   │   │   ├── model_v2.py
    │   │   │   ├── perceiver.py
    │   │   │   ├── transformers_beam_search.py
    │   │   │   ├── transformers_generation_utils.py
    │   │   │   ├── transformers_gpt2.py
    │   │   │   └── transformers_modeling_utils.py
    │   │   ├── infer.py
    │   │   ├── infer_v2.py
    │   │   ├── s2mel/
    │   │   │   ├── dac/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── __main__.py
    │   │   │   │   ├── model/
    │   │   │   │   │   ├── __init__.py
    │   │   │   │   │   ├── base.py
    │   │   │   │   │   ├── dac.py
    │   │   │   │   │   ├── discriminator.py
    │   │   │   │   │   └── encodec.py
    │   │   │   │   ├── nn/
    │   │   │   │   │   ├── __init__.py
    │   │   │   │   │   ├── layers.py
    │   │   │   │   │   ├── loss.py
    │   │   │   │   │   └── quantize.py
    │   │   │   │   └── utils/
    │   │   │   │       ├── __init__.py
    │   │   │   │       ├── decode.py
    │   │   │   │       └── encode.py
    │   │   │   ├── hf_utils.py
    │   │   │   ├── modules/
    │   │   │   │   ├── alias_free_torch/
    │   │   │   │   │   ├── __init__.py
    │   │   │   │   │   ├── act.py
    │   │   │   │   │   ├── filter.py
    │   │   │   │   │   └── resample.py
    │   │   │   │   ├── audio.py
    │   │   │   │   ├── bigvgan/
    │   │   │   │   │   ├── activations.py
    │   │   │   │   │   ├── alias_free_activation/
    │   │   │   │   │   │   ├── cuda/
    │   │   │   │   │   │   │   ├── __init__.py
    │   │   │   │   │   │   │   ├── activation1d.py
    │   │   │   │   │   │   │   ├── anti_alias_activation.cpp
    │   │   │   │   │   │   │   ├── anti_alias_activation_cuda.cu
    │   │   │   │   │   │   │   ├── compat.h
    │   │   │   │   │   │   │   ├── load.py
    │   │   │   │   │   │   │   └── type_shim.h
    │   │   │   │   │   │   └── torch/
    │   │   │   │   │   │       ├── __init__.py
    │   │   │   │   │   │       ├── act.py
    │   │   │   │   │   │       ├── filter.py
    │   │   │   │   │   │       └── resample.py
    │   │   │   │   │   ├── bigvgan.py
    │   │   │   │   │   ├── config.json
    │   │   │   │   │   ├── env.py
    │   │   │   │   │   ├── meldataset.py
    │   │   │   │   │   └── utils.py
    │   │   │   │   ├── campplus/
    │   │   │   │   │   ├── DTDNN.py
    │   │   │   │   │   ├── classifier.py
    │   │   │   │   │   └── layers.py
    │   │   │   │   ├── commons.py
    │   │   │   │   ├── diffusion_transformer.py
    │   │   │   │   ├── encodec.py
    │   │   │   │   ├── flow_matching.py
    │   │   │   │   ├── gpt_fast/
    │   │   │   │   │   ├── generate.py
    │   │   │   │   │   ├── model.py
    │   │   │   │   │   └── quantize.py
    │   │   │   │   ├── hifigan/
    │   │   │   │   │   ├── f0_predictor.py
    │   │   │   │   │   └── generator.py
    │   │   │   │   ├── layers.py
    │   │   │   │   ├── length_regulator.py
    │   │   │   │   ├── openvoice/
    │   │   │   │   │   ├── __init__.py
    │   │   │   │   │   ├── api.py
    │   │   │   │   │   ├── attentions.py
    │   │   │   │   │   ├── checkpoints_v2/
    │   │   │   │   │   │   └── converter/
    │   │   │   │   │   │       └── config.json
    │   │   │   │   │   ├── commons.py
    │   │   │   │   │   ├── mel_processing.py
    │   │   │   │   │   ├── models.py
    │   │   │   │   │   ├── modules.py
    │   │   │   │   │   ├── openvoice_app.py
    │   │   │   │   │   ├── se_extractor.py
    │   │   │   │   │   ├── transforms.py
    │   │   │   │   │   └── utils.py
    │   │   │   │   ├── quantize.py
    │   │   │   │   ├── rmvpe.py
    │   │   │   │   ├── vocos/
    │   │   │   │   │   ├── __init__.py
    │   │   │   │   │   ├── heads.py
    │   │   │   │   │   ├── helpers.py
    │   │   │   │   │   ├── loss.py
    │   │   │   │   │   ├── models.py
    │   │   │   │   │   ├── modules.py
    │   │   │   │   │   ├── pretrained.py
    │   │   │   │   │   └── spectral_ops.py
    │   │   │   │   └── wavenet.py
    │   │   │   ├── optimizers.py
    │   │   │   └── wav2vecbert_extract.py
    │   │   ├── utils/
    │   │   │   ├── __init__.py
    │   │   │   ├── arch_util.py
    │   │   │   ├── checkpoint.py
    │   │   │   ├── common.py
    │   │   │   ├── feature_extractors.py
    │   │   │   ├── front.py
    │   │   │   ├── maskgct/
    │   │   │   │   └── models/
    │   │   │   │       ├── codec/
    │   │   │   │       │   ├── __init__.py
    │   │   │   │       │   ├── amphion_codec/
    │   │   │   │       │   │   ├── codec.py
    │   │   │   │       │   │   ├── quantize/
    │   │   │   │       │   │   │   ├── __init__.py
    │   │   │   │       │   │   │   ├── factorized_vector_quantize.py
    │   │   │   │       │   │   │   ├── lookup_free_quantize.py
    │   │   │   │       │   │   │   ├── residual_vq.py
    │   │   │   │       │   │   │   └── vector_quantize.py
    │   │   │   │       │   │   └── vocos.py
    │   │   │   │       │   ├── codec_dataset.py
    │   │   │   │       │   ├── codec_inference.py
    │   │   │   │       │   ├── codec_sampler.py
    │   │   │   │       │   ├── codec_trainer.py
    │   │   │   │       │   ├── facodec/
    │   │   │   │       │   │   ├── __init__.py
    │   │   │   │       │   │   ├── alias_free_torch/
    │   │   │   │       │   │   │   ├── __init__.py
    │   │   │   │       │   │   │   ├── act.py
    │   │   │   │       │   │   │   ├── filter.py
    │   │   │   │       │   │   │   └── resample.py
    │   │   │   │       │   │   ├── facodec_dataset.py
    │   │   │   │       │   │   ├── facodec_inference.py
    │   │   │   │       │   │   ├── facodec_trainer.py
    │   │   │   │       │   │   ├── modules/
    │   │   │   │       │   │   │   ├── JDC/
    │   │   │   │       │   │   │   │   ├── __init__.py
    │   │   │   │       │   │   │   │   ├── bst.t7
    │   │   │   │       │   │   │   │   └── model.py
    │   │   │   │       │   │   │   ├── attentions.py
    │   │   │   │       │   │   │   ├── commons.py
    │   │   │   │       │   │   │   ├── gradient_reversal.py
    │   │   │   │       │   │   │   ├── layers.py
    │   │   │   │       │   │   │   ├── quantize.py
    │   │   │   │       │   │   │   ├── style_encoder.py
    │   │   │   │       │   │   │   └── wavenet.py
    │   │   │   │       │   │   └── optimizer.py
    │   │   │   │       │   ├── kmeans/
    │   │   │   │       │   │   ├── repcodec_model.py
    │   │   │   │       │   │   └── vocos.py
    │   │   │   │       │   ├── melvqgan/
    │   │   │   │       │   │   └── melspec.py
    │   │   │   │       │   ├── ns3_codec/
    │   │   │   │       │   │   ├── README.md
    │   │   │   │       │   │   ├── __init__.py
    │   │   │   │       │   │   ├── alias_free_torch/
    │   │   │   │       │   │   │   ├── __init__.py
    │   │   │   │       │   │   │   ├── act.py
    │   │   │   │       │   │   │   ├── filter.py
    │   │   │   │       │   │   │   └── resample.py
    │   │   │   │       │   │   ├── facodec.py
    │   │   │   │       │   │   ├── gradient_reversal.py
    │   │   │   │       │   │   ├── melspec.py
    │   │   │   │       │   │   ├── quantize/
    │   │   │   │       │   │   │   ├── __init__.py
    │   │   │   │       │   │   │   ├── fvq.py
    │   │   │   │       │   │   │   └── rvq.py
    │   │   │   │       │   │   └── transformer.py
    │   │   │   │       │   ├── speechtokenizer/
    │   │   │   │       │   │   ├── model.py
    │   │   │   │       │   │   └── modules/
    │   │   │   │       │   │       ├── __init__.py
    │   │   │   │       │   │       ├── conv.py
    │   │   │   │       │   │       ├── lstm.py
    │   │   │   │       │   │       ├── norm.py
    │   │   │   │       │   │       ├── quantization/
    │   │   │   │       │   │       │   ├── __init__.py
    │   │   │   │       │   │       │   ├── ac.py
    │   │   │   │       │   │       │   ├── core_vq.py
    │   │   │   │       │   │       │   ├── distrib.py
    │   │   │   │       │   │       │   └── vq.py
    │   │   │   │       │   │       └── seanet.py
    │   │   │   │       │   └── vevo/
    │   │   │   │       │       └── vevo_repcodec.py
    │   │   │   │       └── tts/
    │   │   │   │           └── maskgct/
    │   │   │   │               ├── ckpt/
    │   │   │   │               │   └── wav2vec2bert_stats.pt
    │   │   │   │               ├── llama_nar.py
    │   │   │   │               └── maskgct_s2a.py
    │   │   │   ├── maskgct_utils.py
    │   │   │   ├── text_utils.py
    │   │   │   ├── typical_sampling.py
    │   │   │   ├── utils.py
    │   │   │   ├── webui_utils.py
    │   │   │   └── xtransformers.py
    │   │   └── vqvae/
    │   │       ├── __init__.py
    │   │       └── xtts_dvae.py
    │   ├── internvl/
    │   │   ├── __init__.py
    │   │   └── conversation.py
    │   ├── llava/
    │   │   ├── __init__.py
    │   │   ├── conversation.py
    │   │   ├── mm_utils.py
    │   │   └── model/
    │   │       ├── __init__.py
    │   │       ├── clip_encoder/
    │   │       │   ├── __init__.py
    │   │       │   ├── builder.py
    │   │       │   └── clip_encoder.py
    │   │       ├── constants.py
    │   │       ├── llava_arch.py
    │   │       ├── llava_llama.py
    │   │       └── multimodal_projector/
    │   │           ├── __init__.py
    │   │           └── builder.py
    │   ├── matcha/
    │   │   ├── VERSION
    │   │   ├── __init__.py
    │   │   ├── app.py
    │   │   ├── cli.py
    │   │   ├── data/
    │   │   │   ├── __init__.py
    │   │   │   ├── components/
    │   │   │   │   └── __init__.py
    │   │   │   └── text_mel_datamodule.py
    │   │   ├── hifigan/
    │   │   │   ├── LICENSE
    │   │   │   ├── README.md
    │   │   │   ├── __init__.py
    │   │   │   ├── config.py
    │   │   │   ├── denoiser.py
    │   │   │   ├── env.py
    │   │   │   ├── meldataset.py
    │   │   │   ├── models.py
    │   │   │   └── xutils.py
    │   │   ├── models/
    │   │   │   ├── __init__.py
    │   │   │   ├── baselightningmodule.py
    │   │   │   ├── components/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── decoder.py
    │   │   │   │   ├── flow_matching.py
    │   │   │   │   ├── text_encoder.py
    │   │   │   │   └── transformer.py
    │   │   │   └── matcha_tts.py
    │   │   ├── onnx/
    │   │   │   ├── __init__.py
    │   │   │   ├── export.py
    │   │   │   └── infer.py
    │   │   ├── text/
    │   │   │   ├── __init__.py
    │   │   │   ├── cleaners.py
    │   │   │   ├── numbers.py
    │   │   │   └── symbols.py
    │   │   ├── train.py
    │   │   └── utils/
    │   │       ├── __init__.py
    │   │       ├── audio.py
    │   │       ├── generate_data_statistics.py
    │   │       ├── get_durations_from_trained_model.py
    │   │       ├── instantiators.py
    │   │       ├── logging_utils.py
    │   │       ├── model.py
    │   │       ├── monotonic_align/
    │   │       │   ├── __init__.py
    │   │       │   ├── core.pyx
    │   │       │   └── setup.py
    │   │       ├── pylogger.py
    │   │       ├── rich_utils.py
    │   │       └── utils.py
    │   ├── megatts3/
    │   │   ├── __init__.py
    │   │   └── tts/
    │   │       ├── frontend_function.py
    │   │       ├── gradio_api.py
    │   │       ├── infer_cli.py
    │   │       ├── modules/
    │   │       │   ├── aligner/
    │   │       │   │   └── whisper_small.py
    │   │       │   ├── ar_dur/
    │   │       │   │   ├── ar_dur_predictor.py
    │   │       │   │   └── commons/
    │   │       │   │       ├── layers.py
    │   │       │   │       ├── nar_tts_modules.py
    │   │       │   │       ├── rel_transformer.py
    │   │       │   │       ├── rot_transformer.py
    │   │       │   │       ├── seq_utils.py
    │   │       │   │       └── transformer.py
    │   │       │   ├── llm_dit/
    │   │       │   │   ├── cfm.py
    │   │       │   │   ├── dit.py
    │   │       │   │   ├── time_embedding.py
    │   │       │   │   └── transformer.py
    │   │       │   └── wavvae/
    │   │       │       ├── decoder/
    │   │       │       │   ├── diag_gaussian.py
    │   │       │       │   ├── hifigan_modules.py
    │   │       │       │   ├── seanet_encoder.py
    │   │       │       │   └── wavvae_v3.py
    │   │       │       └── encoder/
    │   │       │           └── common_modules/
    │   │       │               ├── conv.py
    │   │       │               ├── lstm.py
    │   │       │               └── seanet.py
    │   │       └── utils/
    │   │           ├── audio_utils/
    │   │           │   ├── align.py
    │   │           │   ├── io.py
    │   │           │   └── plot.py
    │   │           ├── commons/
    │   │           │   ├── ckpt_utils.py
    │   │           │   └── hparams.py
    │   │           └── text_utils/
    │   │               ├── dict.json
    │   │               ├── ph_tone_convert.py
    │   │               ├── split_text.py
    │   │               └── text_encoder.py
    │   ├── melo/
    │   │   ├── __init__.py
    │   │   ├── api.py
    │   │   ├── app.py
    │   │   ├── attentions.py
    │   │   ├── commons.py
    │   │   ├── configs/
    │   │   │   └── config.json
    │   │   ├── data/
    │   │   │   └── example/
    │   │   │       └── metadata.list
    │   │   ├── data_utils.py
    │   │   ├── download_utils.py
    │   │   ├── infer.py
    │   │   ├── init_downloads.py
    │   │   ├── losses.py
    │   │   ├── main.py
    │   │   ├── mel_processing.py
    │   │   ├── models.py
    │   │   ├── modules.py
    │   │   ├── monotonic_align/
    │   │   │   ├── __init__.py
    │   │   │   └── core.py
    │   │   ├── preprocess_text.py
    │   │   ├── split_utils.py
    │   │   ├── text/
    │   │   │   ├── __init__.py
    │   │   │   ├── chinese.py
    │   │   │   ├── chinese_bert.py
    │   │   │   ├── chinese_mix.py
    │   │   │   ├── cleaner.py
    │   │   │   ├── cleaner_multiling.py
    │   │   │   ├── cmudict.rep
    │   │   │   ├── cmudict_cache.pickle
    │   │   │   ├── english.py
    │   │   │   ├── english_bert.py
    │   │   │   ├── english_utils/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── abbreviations.py
    │   │   │   │   ├── number_norm.py
    │   │   │   │   └── time_norm.py
    │   │   │   ├── es_phonemizer/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── base.py
    │   │   │   │   ├── cleaner.py
    │   │   │   │   ├── es_symbols.json
    │   │   │   │   ├── es_symbols.txt
    │   │   │   │   ├── es_symbols_v2.json
    │   │   │   │   ├── es_to_ipa.py
    │   │   │   │   ├── example_ipa.txt
    │   │   │   │   ├── gruut_wrapper.py
    │   │   │   │   ├── punctuation.py
    │   │   │   │   ├── spanish_symbols.txt
    │   │   │   │   └── test.ipynb
    │   │   │   ├── fr_phonemizer/
    │   │   │   │   ├── __init__.py
    │   │   │   │   ├── base.py
    │   │   │   │   ├── cleaner.py
    │   │   │   │   ├── en_symbols.json
    │   │   │   │   ├── example_ipa.txt
    │   │   │   │   ├── fr_symbols.json
    │   │   │   │   ├── fr_to_ipa.py
    │   │   │   │   ├── french_abbreviations.py
    │   │   │   │   ├── french_symbols.txt
    │   │   │   │   ├── gruut_wrapper.py
    │   │   │   │   └── punctuation.py
    │   │   │   ├── french.py
    │   │   │   ├── french_bert.py
    │   │   │   ├── japanese.py
    │   │   │   ├── japanese_bert.py
    │   │   │   ├── ko_dictionary.py
    │   │   │   ├── korean.py
    │   │   │   ├── opencpop-strict.txt
    │   │   │   ├── spanish.py
    │   │   │   ├── spanish_bert.py
    │   │   │   ├── symbols.py
    │   │   │   └── tone_sandhi.py
    │   │   ├── train.py
    │   │   ├── train.sh
    │   │   ├── transforms.py
    │   │   └── utils.py
    │   ├── mlx/
    │   │   ├── __init__.py
    │   │   └── flux/
    │   │       ├── __init__.py
    │   │       ├── autoencoder.py
    │   │       ├── clip.py
    │   │       ├── datasets.py
    │   │       ├── flux.py
    │   │       ├── layers.py
    │   │       ├── lora.py
    │   │       ├── model.py
    │   │       ├── sampler.py
    │   │       ├── t5.py
    │   │       ├── tokenizers.py
    │   │       ├── trainer.py
    │   │       └── utils.py
    │   └── whisper/
    │       ├── __init__.py
    │       ├── __main__.py
    │       ├── assets/
    │       │   ├── gpt2.tiktoken
    │       │   ├── mel_filters.npz
    │       │   └── multilingual.tiktoken
    │       ├── audio.py
    │       ├── decoding.py
    │       ├── model.py
    │       ├── normalizers/
    │       │   ├── __init__.py
    │       │   ├── basic.py
    │       │   ├── english.json
    │       │   └── english.py
    │       ├── timing.py
    │       ├── tokenizer.py
    │       ├── transcribe.py
    │       ├── triton_ops.py
    │       ├── utils.py
    │       └── version.py
    ├── types.py
    ├── ui/
    │   ├── __init__.py
    │   ├── gradio/
    │   │   ├── __init__.py
    │   │   ├── chat_interface.py
    │   │   ├── media_interface.py
    │   │   └── utils/
    │   │       ├── __init__.py
    │   │       └── latex.py
    │   └── web/
    │       └── ui/
    │           ├── .eslintignore
    │           ├── .eslintrc.yml
    │           ├── .gitignore
    │           ├── .prettierignore
    │           ├── .prettierrc.yml
    │           ├── package.json
    │           ├── public/
    │           │   └── index.html
    │           └── src/
    │               ├── App.js
    │               ├── components/
    │               │   ├── MenuSide.js
    │               │   ├── Title.js
    │               │   ├── alertComponent.js
    │               │   ├── apiContext.js
    │               │   ├── authAlertDialog.js
    │               │   ├── copyComponent.js
    │               │   ├── deleteDialog.js
    │               │   ├── errorMessageSnackBar.js
    │               │   ├── fetchWrapper.js
    │               │   ├── fetcher.js
    │               │   ├── hotkeyFocusTextField.js
    │               │   ├── successMessageSnackBar.js
    │               │   ├── tableTitle.js
    │               │   ├── themeButton.js
    │               │   ├── themeContext.js
    │               │   ├── titleTypography.js
    │               │   ├── translateButton.js
    │               │   ├── utils.js
    │               │   └── versionLabel.js
    │               ├── i18n.js
    │               ├── index.css
    │               ├── index.js
    │               ├── locales/
    │               │   ├── en.json
    │               │   ├── ja.json
    │               │   ├── ko.json
    │               │   └── zh.json
    │               ├── router/
    │               │   └── index.js
    │               ├── scenes/
    │               │   ├── _layout/
    │               │   │   └── index.js
    │               │   ├── cluster_info/
    │               │   │   ├── index.js
    │               │   │   ├── nodeInfo.js
    │               │   │   └── style.js
    │               │   ├── launch_model/
    │               │   │   ├── LaunchModel.js
    │               │   │   ├── components/
    │               │   │   │   ├── cachedListDialog.js
    │               │   │   │   ├── commandBuilder.js
    │               │   │   │   ├── dynamicFieldList.js
    │               │   │   │   ├── editCustomModelDialog.js
    │               │   │   │   ├── launchModelDrawer.js
    │               │   │   │   ├── modelFormConfig.js
    │               │   │   │   ├── pasteDialog.js
    │               │   │   │   ├── progress.js
    │               │   │   │   ├── selectField.js
    │               │   │   │   └── virtualenvListDialog.js
    │               │   │   ├── data/
    │               │   │   │   └── data.js
    │               │   │   ├── index.js
    │               │   │   ├── launchCustom.js
    │               │   │   ├── modelCard.js
    │               │   │   └── styles/
    │               │   │       └── modelCardStyle.css
    │               │   ├── login/
    │               │   │   ├── header.js
    │               │   │   └── login.js
    │               │   ├── register_model/
    │               │   │   ├── components/
    │               │   │   │   ├── addControlnet.js
    │               │   │   │   ├── addModelSpecs.js
    │               │   │   │   ├── addStop.js
    │               │   │   │   └── addVirtualenv.js
    │               │   │   ├── data/
    │               │   │   │   └── languages.js
    │               │   │   ├── index.js
    │               │   │   ├── registerModel.js
    │               │   │   └── styles/
    │               │   │       └── registerModelStyle.css
    │               │   └── running_models/
    │               │       └── index.js
    │               └── theme.js
    └── utils.py

SYMBOL INDEX (8657 symbols across 776 files)

FILE: benchmark/benchmark_embedding.py
  class EmbeddingBenchmarkRunner (line 31) | class EmbeddingBenchmarkRunner(ConcurrentBenchmarkRunner):
    method __init__ (line 32) | def __init__(
    method _run (line 52) | async def _run(self):
    method worker (line 59) | async def worker(self, i: int):
    method send_request (line 73) | async def send_request(self, request, warming_up: bool = False):
  function main (line 101) | def main(args: argparse.Namespace):

FILE: benchmark/benchmark_latency.py
  class LatencyBenchmarkRunner (line 29) | class LatencyBenchmarkRunner(BenchmarkRunner):
    method _run (line 30) | async def _run(self):
  function main (line 42) | def main(args: argparse.Namespace):

FILE: benchmark/benchmark_long.py
  class LongBenchmarkRunner (line 30) | class LongBenchmarkRunner(ConcurrentBenchmarkRunner):
    method _run (line 31) | async def _run(self):
    method worker (line 38) | async def worker(self, i: int):
  function main (line 53) | def main(args: argparse.Namespace):

FILE: benchmark/benchmark_rerank.py
  class RerankBenchmarkRunner (line 31) | class RerankBenchmarkRunner(ConcurrentBenchmarkRunner):
    method __init__ (line 32) | def __init__(
    method _run (line 54) | async def _run(self):
    method worker (line 61) | async def worker(self, i: int):
    method send_request (line 75) | async def send_request(self, request, warming_up: bool = False):
  function main (line 105) | def main(args: argparse.Namespace):

FILE: benchmark/benchmark_runner.py
  function remove_prefix (line 33) | def remove_prefix(text: str, prefix: str) -> str:
  class RequestOutput (line 40) | class RequestOutput:
  class BenchmarkRunner (line 50) | class BenchmarkRunner:
    method __init__ (line 51) | def __init__(
    method run (line 69) | async def run(self):
    method warm_up (line 76) | async def warm_up(self, num_requests: int = 5):
    method _run (line 83) | async def _run(self):
    method send_request (line 86) | async def send_request(self, request: tuple, warming_up: bool = False):
    method print_stats (line 180) | def print_stats(self):
  class ConcurrentBenchmarkRunner (line 378) | class ConcurrentBenchmarkRunner(BenchmarkRunner):
    method __init__ (line 379) | def __init__(
    method worker (line 400) | async def worker(self):

FILE: benchmark/benchmark_serving.py
  class ServingBenchmarkRunner (line 31) | class ServingBenchmarkRunner(ConcurrentBenchmarkRunner):
    method __init__ (line 32) | def __init__(
    method _run (line 55) | async def _run(self):
    method warm_up (line 63) | async def warm_up(self, num_requests: int = 5):
    method worker (line 72) | async def worker(self):
  function main (line 92) | def main(args: argparse.Namespace):

FILE: benchmark/utils.py
  function get_tokenizer (line 33) | def get_tokenizer(
  function sample_requests (line 94) | def sample_requests(
  function generate_sorting_prompts (line 143) | def generate_sorting_prompts(

FILE: doc/source/gen_docs.py
  function mock_engine_libraries (line 23) | def mock_engine_libraries():
  function mock_platform_checks (line 99) | def mock_platform_checks():
  function build_architecture_to_models (line 215) | def build_architecture_to_models(models):
  function get_metrics_from_url (line 223) | def get_metrics_from_url(metrics_url):
  function _can_use_transformers_legacy (line 238) | def _can_use_transformers_legacy(model, model_spec):
  function _extract_primary_model_src (line 244) | def _extract_primary_model_src(model):
  function main (line 251) | def main():

FILE: doc/source/norm_zh.py
  function _zh_len (line 24) | def _zh_len(s):
  function _zh_split (line 34) | def _zh_split(s):
  function _normalize (line 53) | def _normalize(string, prefix="", width=76):
  function main (line 112) | def main():

FILE: examples/AI_podcast.py
  function get_audio_devices (line 103) | def get_audio_devices() -> str:
  function callback (line 122) | def callback(indata, frames, time, status):
  function record_unlimited (line 129) | def record_unlimited() -> numpy.ndarray:
  function format_prompt (line 162) | def format_prompt(model, audio_input) -> str:
  function text_to_audio (line 168) | def text_to_audio(response, voice_id):
  function chat_with_bot (line 185) | def chat_with_bot(
  function check_word_order (line 337) | def check_word_order(string, first_word, second_word) -> int:

FILE: examples/AI_podcast_ZH.py
  function get_audio_devices (line 110) | def get_audio_devices() -> str:
  function callback (line 131) | def callback(indata, frames, time, status):
  function record_unlimited (line 139) | def record_unlimited() -> numpy.ndarray:
  function lanuch_model (line 176) | def lanuch_model(alice_or_bob, model_a, username, model_uid, system_prom...
  function format_prompt (line 213) | def format_prompt(model, audio_input) -> str:
  function text_to_audio (line 219) | def text_to_audio(response, voice_id):
  function construct_Baichuan_prompt (line 239) | def construct_Baichuan_prompt(
  function _base_sanitize_generate_config (line 262) | def _base_sanitize_generate_config() -> PytorchGenerateConfig:
  function baichuan_sanitize_generate_config (line 273) | def baichuan_sanitize_generate_config() -> PytorchGenerateConfig:
  function _convert_completion_to_chat (line 285) | def _convert_completion_to_chat(completion: Completion) -> ChatCompletion:
  function chat_with_bot (line 305) | def chat_with_bot(
  function check_word_order (line 490) | def check_word_order(string, first_word, second_word) -> int:

FILE: examples/AI_translate.py
  function _prompt (line 24) | def _prompt(text):

FILE: examples/LangChain_Streamlit_Doc_Chat.py
  function write_text_file (line 14) | def write_text_file(content, file_path):

FILE: examples/gradio_chatinterface.py
  function flatten (line 80) | def flatten(matrix: List[List[str]]) -> List[str]:
  function to_chat (line 86) | def to_chat(lst: List[str]) -> List[Dict[str, str]]:
  function generate_wrapper (line 98) | def generate_wrapper(message: str, history: List[List[str]]) -> str:

FILE: setup.py
  class ExtraCommandMixin (line 53) | class ExtraCommandMixin:
    method run (line 56) | def run(self):
    method register_pre_command (line 61) | def register_pre_command(cls, cmd):
  class CustomInstall (line 65) | class CustomInstall(ExtraCommandMixin, install):
  class CustomDevelop (line 69) | class CustomDevelop(ExtraCommandMixin, develop):
  class CustomSDist (line 73) | class CustomSDist(ExtraCommandMixin, sdist):
  class BuildWeb (line 76) | class BuildWeb(Command):
    method initialize_options (line 87) | def initialize_options(self):
    method finalize_options (line 90) | def finalize_options(self):
    method run (line 94) | def run(cls):
  function build_long_description (line 128) | def build_long_description():

FILE: versioneer.py
  class VersioneerConfig (line 331) | class VersioneerConfig:
  function get_root (line 335) | def get_root():
  function get_config_from_root (line 378) | def get_config_from_root(root):
  class NotThisMethod (line 416) | class NotThisMethod(Exception):
  function register_vcs_handler (line 425) | def register_vcs_handler(vcs, method):  # decorator
  function run_command (line 436) | def run_command(commands, args, cwd=None, verbose=False, hide_stderr=Fal...
  function git_get_keywords (line 1146) | def git_get_keywords(versionfile_abs):
  function git_versions_from_keywords (line 1174) | def git_versions_from_keywords(keywords, tag_prefix, verbose):
  function git_pieces_from_vcs (line 1245) | def git_pieces_from_vcs(tag_prefix, root, verbose, runner=run_command):
  function do_vcs_install (line 1385) | def do_vcs_install(versionfile_source, ipy):
  function versions_from_parentdir (line 1423) | def versions_from_parentdir(parentdir_prefix, root, verbose):
  function versions_from_file (line 1471) | def versions_from_file(filename):
  function write_to_version_file (line 1490) | def write_to_version_file(filename, versions):
  function plus_or_dot (line 1500) | def plus_or_dot(pieces):
  function render_pep440 (line 1507) | def render_pep440(pieces):
  function render_pep440_branch (line 1531) | def render_pep440_branch(pieces):
  function pep440_split_post (line 1560) | def pep440_split_post(ver):
  function render_pep440_pre (line 1570) | def render_pep440_pre(pieces):
  function render_pep440_post (line 1594) | def render_pep440_post(pieces):
  function render_pep440_post_branch (line 1621) | def render_pep440_post_branch(pieces):
  function render_pep440_old (line 1650) | def render_pep440_old(pieces):
  function render_git_describe (line 1672) | def render_git_describe(pieces):
  function render_git_describe_long (line 1692) | def render_git_describe_long(pieces):
  function render (line 1712) | def render(pieces, style):
  class VersioneerBadRootError (line 1754) | class VersioneerBadRootError(Exception):
  function get_versions (line 1758) | def get_versions(verbose=False):
  function get_version (line 1839) | def get_version():
  function get_cmdclass (line 1844) | def get_cmdclass(cmdclass=None):
  function do_setup (line 2160) | def do_setup():
  function scan_setup_py (line 2217) | def scan_setup_py():
  function setup_command (line 2254) | def setup_command():

FILE: xinference/__init__.py
  function _install (line 34) | def _install():

FILE: xinference/_compat.py
  class JSONSchema (line 78) | class JSONSchema(BaseModel):
  class ResponseFormatJSONSchema (line 85) | class ResponseFormatJSONSchema(BaseModel):
  class CreateChatCompletionOpenAI (line 95) | class CreateChatCompletionOpenAI(BaseModel):

FILE: xinference/_version.py
  function get_keywords (line 22) | def get_keywords():
  class VersioneerConfig (line 35) | class VersioneerConfig:
  function get_config (line 39) | def get_config():
  class NotThisMethod (line 53) | class NotThisMethod(Exception):
  function register_vcs_handler (line 61) | def register_vcs_handler(vcs, method):  # decorator
  function run_command (line 74) | def run_command(commands, args, cwd=None, verbose=False, hide_stderr=Fal...
  function versions_from_parentdir (line 120) | def versions_from_parentdir(parentdir_prefix, root, verbose):
  function git_get_keywords (line 151) | def git_get_keywords(versionfile_abs):
  function git_versions_from_keywords (line 179) | def git_versions_from_keywords(keywords, tag_prefix, verbose):
  function git_pieces_from_vcs (line 250) | def git_pieces_from_vcs(tag_prefix, root, verbose, runner=run_command):
  function plus_or_dot (line 390) | def plus_or_dot(pieces):
  function render_pep440 (line 397) | def render_pep440(pieces):
  function render_pep440_branch (line 421) | def render_pep440_branch(pieces):
  function pep440_split_post (line 450) | def pep440_split_post(ver):
  function render_pep440_pre (line 460) | def render_pep440_pre(pieces):
  function render_pep440_post (line 484) | def render_pep440_post(pieces):
  function render_pep440_post_branch (line 511) | def render_pep440_post_branch(pieces):
  function render_pep440_old (line 540) | def render_pep440_old(pieces):
  function render_git_describe (line 562) | def render_git_describe(pieces):
  function render_git_describe_long (line 582) | def render_git_describe_long(pieces):
  function render (line 602) | def render(pieces, style):
  function get_versions (line 644) | def get_versions():

FILE: xinference/api/dependencies.py
  function get_api (line 32) | def get_api(request: Request) -> "RESTfulAPI":

FILE: xinference/api/oauth2/auth_service.py
  class TokenData (line 30) | class TokenData(BaseModel):
  class AuthService (line 35) | class AuthService:
    method __init__ (line 36) | def __init__(self, auth_config_file: Optional[str]):
    method config (line 41) | def config(self):
    method is_legal_api_key (line 45) | def is_legal_api_key(key: str) -> bool:
    method init_auth_config (line 49) | def init_auth_config(self):
    method __call__ (line 70) | def __call__(
    method get_user (line 119) | def get_user(self, username: str) -> Optional[User]:
    method get_user_and_scopes_with_api_key (line 125) | def get_user_and_scopes_with_api_key(
    method authenticate_user (line 134) | def authenticate_user(self, username: str, password: str):
    method generate_token_for_user (line 142) | def generate_token_for_user(self, username: str, password: str):

FILE: xinference/api/oauth2/types.py
  class LoginUserForm (line 19) | class LoginUserForm(BaseModel):
  class User (line 24) | class User(LoginUserForm):
  class AuthConfig (line 29) | class AuthConfig(BaseModel):
  class AuthStartupConfig (line 35) | class AuthStartupConfig(BaseModel):

FILE: xinference/api/oauth2/utils.py
  function create_access_token (line 21) | def create_access_token(
  function verify_password (line 37) | def verify_password(plain_password, hashed_password):
  function get_password_hash (line 52) | def get_password_hash(password):

FILE: xinference/api/responses.py
  class JSONResponse (line 24) | class JSONResponse(StarletteJSONResponse):
    method render (line 27) | def render(self, content: Any) -> bytes:

FILE: xinference/api/restful_api.py
  class RESTfulAPI (line 91) | class RESTfulAPI(CancelMixin):
    method __init__ (line 95) | def __init__(
    method _init_allowed_ip_list (line 114) | def _init_allowed_ip_list(self):
    method _is_ip_allowed (line 138) | def _is_ip_allowed(self, ip: str) -> bool:
    method is_authenticated (line 151) | def is_authenticated(self):
    method handle_request_limit_error (line 155) | def handle_request_limit_error(e: Exception):
    method _set_trace_model (line 160) | def _set_trace_model(model_uid: Optional[str]) -> None:
    method _set_trace_model_type (line 176) | def _set_trace_model_type(model_type: Optional[str]) -> None:
    method _get_supervisor_ref (line 191) | async def _get_supervisor_ref(self) -> xo.ActorRefType[SupervisorActor]:
    method _get_event_collector_ref (line 198) | async def _get_event_collector_ref(self) -> xo.ActorRefType[EventColle...
    method _report_error_event (line 205) | async def _report_error_event(self, model_uid: Optional[str], content:...
    method serve (line 223) | def serve(self, logging_conf: Optional[dict] = None):
    method _get_builtin_prompts (line 357) | async def _get_builtin_prompts(self) -> JSONResponse:
    method _get_builtin_families (line 368) | async def _get_builtin_families(self) -> JSONResponse:
    method build_llm_registration_from_config (line 379) | async def build_llm_registration_from_config(
    method list_models (line 400) | async def list_models(self) -> JSONResponse:
    method anthropic_list_models (line 422) | async def anthropic_list_models(self) -> JSONResponse:
    method anthropic_get_model (line 449) | async def anthropic_get_model(self, model_id: str) -> JSONResponse:
    method describe_model (line 474) | async def describe_model(self, model_uid: str) -> JSONResponse:
    method launch_model (line 486) | async def launch_model(
    method get_instance_info (line 598) | async def get_instance_info(
    method get_model_replicas (line 612) | async def get_model_replicas(self, model_uid: str) -> JSONResponse:
    method get_launch_model_progress (line 626) | async def get_launch_model_progress(self, model_uid: str) -> JSONRespo...
    method cancel_launch_model (line 636) | async def cancel_launch_model(self, model_uid: str) -> JSONResponse:
    method launch_model_by_version (line 646) | async def launch_model_by_version(
    method get_model_versions (line 674) | async def get_model_versions(
    method build_gradio_interface (line 686) | async def build_gradio_interface(
    method build_gradio_media_interface (line 729) | async def build_gradio_media_interface(
    method terminate_model (line 769) | async def terminate_model(self, model_uid: str) -> JSONResponse:
    method _get_model_last_error (line 790) | async def _get_model_last_error(self, replica_model_uid: bytes, e: Exc...
    method create_completion (line 803) | async def create_completion(self, request: Request) -> Response:
    method create_message (line 878) | async def create_message(self, request: Request) -> Response:
    method create_embedding (line 1019) | async def create_embedding(self, request: Request) -> Response:
    method convert_ids_to_tokens (line 1048) | async def convert_ids_to_tokens(self, request: Request) -> Response:
    method rerank (line 1075) | async def rerank(self, request: Request) -> Response:
    method create_transcriptions (line 1108) | async def create_transcriptions(
    method create_translations (line 1152) | async def create_translations(
    method create_speech (line 1196) | async def create_speech(
    method create_images (line 1258) | async def create_images(self, request: Request) -> Response:
    method sdapi_options (line 1292) | async def sdapi_options(self, request: Request) -> Response:
    method sdapi_sd_models (line 1312) | async def sdapi_sd_models(self, request: Request) -> Response:
    method sdapi_samplers (line 1325) | async def sdapi_samplers(self, request: Request) -> Response:
    method sdapi_txt2img (line 1338) | async def sdapi_txt2img(self, request: Request) -> Response:
    method sdapi_img2img (line 1362) | async def sdapi_img2img(self, request: Request) -> Response:
    method create_variations (line 1386) | async def create_variations(
    method create_inpainting (line 1443) | async def create_inpainting(
    method create_ocr (line 1498) | async def create_ocr(
    method create_image_edits (line 1537) | async def create_image_edits(
    method _stream_image_edit (line 1701) | async def _stream_image_edit(
    method create_flexible_infer (line 1778) | async def create_flexible_infer(self, request: Request) -> Response:
    method create_videos (line 1803) | async def create_videos(self, request: Request) -> Response:
    method create_videos_from_images (line 1835) | async def create_videos_from_images(
    method create_videos_from_first_last_frame (line 1879) | async def create_videos_from_first_last_frame(
    method create_chat_completion (line 1925) | async def create_chat_completion(self, request: Request) -> Response:
    method query_engines_by_model_name (line 2077) | async def query_engines_by_model_name(
    method register_model (line 2100) | async def register_model(self, model_type: str, request: Request) -> J...
    method unregister_model (line 2118) | async def unregister_model(self, model_type: str, model_name: str) -> ...
    method update_model_type (line 2131) | async def update_model_type(self, request: Request) -> JSONResponse:
    method list_model_registrations (line 2167) | async def list_model_registrations(
    method get_model_registrations (line 2189) | async def get_model_registrations(
    method get_model_events (line 2204) | async def get_model_events(self, model_uid: str) -> JSONResponse:
    method abort_request (line 2216) | async def abort_request(
    method list_vllm_supported_model_families (line 2240) | async def list_vllm_supported_model_families(self) -> JSONResponse:
    method extract_guided_params (line 2257) | def extract_guided_params(raw_body: dict) -> dict:
    method _convert_openai_to_anthropic (line 2303) | def _convert_openai_to_anthropic(self, openai_response: dict, model: s...
  function run (line 2394) | def run(
  function run_in_subprocess (line 2429) | def run_in_subprocess(

FILE: xinference/api/routers/__init__.py
  function register_all_routes (line 17) | def register_all_routes(api: RESTfulAPI) -> None:

FILE: xinference/api/routers/admin.py
  function get_status (line 26) | async def get_status(api: "RESTfulAPI" = Depends(get_api)) -> JSONResponse:
  function get_address (line 36) | async def get_address(api: "RESTfulAPI" = Depends(get_api)) -> JSONRespo...
  function login_for_access_token (line 40) | async def login_for_access_token(
  function is_cluster_authenticated (line 50) | async def is_cluster_authenticated(
  function get_cluster_device_info (line 56) | async def get_cluster_device_info(
  function get_cluster_version (line 69) | async def get_cluster_version() -> JSONResponse:
  function get_devices_count (line 78) | async def get_devices_count(
  function get_workers_info (line 90) | async def get_workers_info(
  function get_supervisor_info (line 105) | async def get_supervisor_info(
  function abort_cluster (line 120) | async def abort_cluster(
  function list_cached_models (line 136) | async def list_cached_models(
  function list_model_files (line 154) | async def list_model_files(
  function confirm_and_remove_model (line 176) | async def confirm_and_remove_model(
  function list_virtual_envs (line 195) | async def list_virtual_envs(
  function remove_virtual_env (line 216) | async def remove_virtual_env(
  function get_progress (line 242) | async def get_progress(
  function register_routes (line 260) | def register_routes(api: "RESTfulAPI") -> None:

FILE: xinference/api/routers/audio.py
  function register_routes (line 13) | def register_routes(api: "RESTfulAPI") -> None:

FILE: xinference/api/routers/embeddings.py
  function register_routes (line 13) | def register_routes(api: "RESTfulAPI") -> None:

FILE: xinference/api/routers/images.py
  function register_routes (line 15) | def register_routes(api: "RESTfulAPI") -> None:

FILE: xinference/api/routers/llm.py
  function register_routes (line 15) | def register_routes(api: "RESTfulAPI") -> None:

FILE: xinference/api/routers/models.py
  function register_routes (line 13) | def register_routes(api: "RESTfulAPI") -> None:

FILE: xinference/api/routers/rerank.py
  function register_routes (line 13) | def register_routes(api: "RESTfulAPI") -> None:

FILE: xinference/api/routers/videos.py
  function register_routes (line 15) | def register_routes(api: "RESTfulAPI") -> None:

FILE: xinference/api/schemas/requests.py
  class CreateCompletionRequest (line 15) | class CreateCompletionRequest(CreateCompletion):
    class Config (line 16) | class Config:
  class CreateEmbeddingRequest (line 25) | class CreateEmbeddingRequest(BaseModel):
    class Config (line 32) | class Config:
  class RerankRequest (line 40) | class RerankRequest(BaseModel):
  class TextToImageRequest (line 51) | class TextToImageRequest(BaseModel):
  class SDAPIOptionsRequest (line 61) | class SDAPIOptionsRequest(BaseModel):
  class SDAPITxt2imgRequst (line 65) | class SDAPITxt2imgRequst(BaseModel):
  class SDAPIImg2imgRequst (line 81) | class SDAPIImg2imgRequst(BaseModel):
  class TextToVideoRequest (line 98) | class TextToVideoRequest(BaseModel):
  class SpeechRequest (line 106) | class SpeechRequest(BaseModel):
  class RegisterModelRequest (line 116) | class RegisterModelRequest(BaseModel):
  class AutoConfigLLMRequest (line 122) | class AutoConfigLLMRequest(BaseModel):
  class UpdateModelRequest (line 127) | class UpdateModelRequest(BaseModel):
  class BuildGradioInterfaceRequest (line 131) | class BuildGradioInterfaceRequest(BaseModel):
  class BuildGradioMediaInterfaceRequest (line 143) | class BuildGradioMediaInterfaceRequest(BaseModel):

FILE: xinference/api/tests/test_admin.py
  function _json_body (line 26) | def _json_body(response):
  function mock_supervisor (line 31) | def mock_supervisor():
  function mock_api (line 53) | def mock_api(mock_supervisor):
  function test_get_status_returns_200_and_data (line 66) | async def test_get_status_returns_200_and_data(mock_api, mock_supervisor):
  function test_get_status_raises_500_on_supervisor_error (line 74) | async def test_get_status_raises_500_on_supervisor_error(mock_api, mock_...
  function test_get_address_returns_supervisor_address (line 83) | async def test_get_address_returns_supervisor_address(mock_api):
  function test_get_cluster_version_returns_version (line 91) | async def test_get_cluster_version_returns_version():
  function test_is_cluster_authenticated_returns_auth_flag (line 99) | async def test_is_cluster_authenticated_returns_auth_flag(mock_api):
  function test_login_for_access_token_returns_token (line 111) | async def test_login_for_access_token_returns_token(mock_api):
  function test_get_cluster_device_info_returns_data (line 127) | async def test_get_cluster_device_info_returns_data(mock_api, mock_super...
  function test_get_devices_count_returns_data (line 140) | async def test_get_devices_count_returns_data(mock_api, mock_supervisor):
  function test_get_workers_info_returns_data (line 148) | async def test_get_workers_info_returns_data(mock_api, mock_supervisor):
  function test_get_supervisor_info_returns_data (line 158) | async def test_get_supervisor_info_returns_data(mock_api, mock_supervisor):
  function test_abort_cluster_returns_result_and_does_not_kill_in_test (line 166) | async def test_abort_cluster_returns_result_and_does_not_kill_in_test(
  function test_list_cached_models_returns_list (line 177) | async def test_list_cached_models_returns_list(mock_api, mock_supervisor):
  function test_list_model_files_returns_paths (line 188) | async def test_list_model_files_returns_paths(mock_api, mock_supervisor):
  function test_confirm_and_remove_model_returns_result (line 203) | async def test_confirm_and_remove_model_returns_result(mock_api, mock_su...
  function test_list_virtual_envs_returns_list (line 213) | async def test_list_virtual_envs_returns_list(mock_api, mock_supervisor):
  function test_remove_virtual_env_requires_model_name (line 226) | async def test_remove_virtual_env_requires_model_name(mock_api):
  function test_remove_virtual_env_returns_result (line 240) | async def test_remove_virtual_env_returns_result(mock_api, mock_supervis...
  function test_get_progress_returns_progress (line 254) | async def test_get_progress_returns_progress(mock_api, mock_supervisor):
  function test_get_progress_raises_400_on_key_error (line 263) | async def test_get_progress_raises_400_on_key_error(mock_api, mock_super...

FILE: xinference/api/tests/test_utils.py
  class DummyModel (line 21) | class DummyModel:
    method __init__ (line 24) | def __init__(self, uid: str):
  class DummySupervisor (line 28) | class DummySupervisor:
    method __init__ (line 31) | def __init__(self, models=None):
    method get_model (line 34) | async def get_model(self, model_uid: str):
  class TestRequireModel (line 40) | class TestRequireModel:
    method test_successful_get (line 44) | async def test_successful_get(self):
    method test_model_not_found_raises_400 (line 56) | async def test_model_not_found_raises_400(self):
    method test_unexpected_error_raises_500 (line 69) | async def test_unexpected_error_raises_500(self):
    method test_reports_error_event (line 81) | async def test_reports_error_event(self):
    method test_no_report_error_event_when_none (line 101) | async def test_no_report_error_event_when_none(self):

FILE: xinference/api/utils.py
  function require_model (line 27) | async def require_model(

FILE: xinference/client/common.py
  function convert_float_to_int_or_str (line 19) | def convert_float_to_int_or_str(model_size: float) -> Union[int, str]:
  function streaming_response_iterator (line 30) | def streaming_response_iterator(
  function async_streaming_response_iterator (line 66) | async def async_streaming_response_iterator(

FILE: xinference/client/restful/async_restful_client.py
  function _filter_params (line 35) | def _filter_params(params: Dict[Any, Any]) -> Dict[Any, Any]:
  function _get_error_string (line 43) | async def _get_error_string(response: aiohttp.ClientResponse) -> str:
  function _release_response (line 61) | async def _release_response(response: aiohttp.ClientResponse):
  class AsyncRESTfulModelHandle (line 67) | class AsyncRESTfulModelHandle:
    method __init__ (line 73) | def __init__(self, model_uid: str, base_url: str, auth_headers: Dict):
    method close (line 82) | async def close(self):
    method __del__ (line 88) | def __del__(self):
  class AsyncRESTfulEmbeddingModelHandle (line 94) | class AsyncRESTfulEmbeddingModelHandle(AsyncRESTfulModelHandle):
    method create_embedding (line 95) | async def create_embedding(
    method convert_ids_to_tokens (line 136) | async def convert_ids_to_tokens(
  class AsyncRESTfulRerankModelHandle (line 177) | class AsyncRESTfulRerankModelHandle(AsyncRESTfulModelHandle):
    method rerank (line 178) | async def rerank(
  class AsyncRESTfulImageModelHandle (line 239) | class AsyncRESTfulImageModelHandle(AsyncRESTfulModelHandle):
    method text_to_image (line 240) | async def text_to_image(
    method image_to_image (line 287) | async def image_to_image(
    method image_edit (line 370) | async def image_edit(
    method inpainting (line 521) | async def inpainting(
    method ocr (line 601) | async def ocr(self, image: Union[str, bytes], **kwargs):
  class AsyncRESTfulVideoModelHandle (line 623) | class AsyncRESTfulVideoModelHandle(AsyncRESTfulModelHandle):
    method text_to_video (line 624) | async def text_to_video(
    method image_to_video (line 663) | async def image_to_video(
    method flf_to_video (line 712) | async def flf_to_video(
  class AsyncRESTfulGenerateModelHandle (line 780) | class AsyncRESTfulGenerateModelHandle(AsyncRESTfulModelHandle):
    method generate (line 781) | async def generate(
  class AsyncRESTfulChatModelHandle (line 835) | class AsyncRESTfulChatModelHandle(AsyncRESTfulGenerateModelHandle):
    method chat (line 836) | async def chat(
  class AsyncRESTfulAudioModelHandle (line 917) | class AsyncRESTfulAudioModelHandle(AsyncRESTfulModelHandle):
    method transcriptions (line 918) | async def transcriptions(
    method translations (line 987) | async def translations(
    method speech (line 1054) | async def speech(
  class AsyncRESTfulFlexibleModelHandle (line 1142) | class AsyncRESTfulFlexibleModelHandle(AsyncRESTfulModelHandle):
    method infer (line 1143) | async def infer(
  class AsyncClient (line 1177) | class AsyncClient:
    method __init__ (line 1178) | def __init__(self, base_url, api_key: Optional[str] = None):
    method close (line 1190) | async def close(self):
    method __del__ (line 1196) | def __del__(self):
    method _set_token (line 1201) | def _set_token(self, token: Optional[str]):
    method _get_token (line 1206) | def _get_token(self) -> Optional[str]:
    method _check_cluster_authenticated (line 1213) | def _check_cluster_authenticated(self):
    method vllm_models (line 1231) | async def vllm_models(self) -> Dict[str, Any]:
    method login (line 1248) | async def login(self, username: str, password: str):
    method list_models (line 1269) | async def list_models(self) -> Dict[str, Dict[str, Any]]:
    method launch_model (line 1293) | async def launch_model(
    method terminate_model (line 1410) | async def terminate_model(self, model_uid: str):
    method get_launch_model_progress (line 1435) | async def get_launch_model_progress(self, model_uid: str) -> dict:
    method cancel_launch_model (line 1465) | async def cancel_launch_model(self, model_uid: str):
    method get_instance_info (line 1488) | async def get_instance_info(self, model_name: str, model_uid: str):
    method _get_supervisor_internal_address (line 1501) | async def _get_supervisor_internal_address(self):
    method get_model (line 1510) | async def get_model(self, model_uid: str) -> AsyncRESTfulModelHandle:
    method describe_model (line 1580) | async def describe_model(self, model_uid: str):
    method register_model (line 1632) | async def register_model(
    method unregister_model (line 1672) | async def unregister_model(self, model_type: str, model_name: str):
    method list_model_registrations (line 1699) | async def list_model_registrations(self, model_type: str) -> List[Dict...
    method list_cached_models (line 1730) | async def list_cached_models(
    method list_deletable_models (line 1769) | async def list_deletable_models(
    method confirm_and_remove_model (line 1801) | async def confirm_and_remove_model(
    method get_model_registration (line 1833) | async def get_model_registration(
    method query_engine_by_model_name (line 1862) | async def query_engine_by_model_name(
    method abort_request (line 1899) | async def abort_request(
    method get_workers_info (line 1934) | async def get_workers_info(self):
    method get_supervisor_info (line 1945) | async def get_supervisor_info(self):
    method get_progress (line 1956) | async def get_progress(self, request_id: str):
    method abort_cluster (line 1967) | async def abort_cluster(self):

FILE: xinference/client/restful/restful_client.py
  function _get_error_string (line 34) | def _get_error_string(response: requests.Response) -> str:
  class RESTfulModelHandle (line 47) | class RESTfulModelHandle:
    method __init__ (line 53) | def __init__(self, model_uid: str, base_url: str, auth_headers: Dict):
    method close (line 59) | def close(self):
    method __del__ (line 67) | def __del__(self):
  class RESTfulEmbeddingModelHandle (line 72) | class RESTfulEmbeddingModelHandle(RESTfulModelHandle):
    method create_embedding (line 73) | def create_embedding(self, input: Union[str, List[str]], **kwargs) -> ...
    method convert_ids_to_tokens (line 109) | def convert_ids_to_tokens(
  class RESTfulRerankModelHandle (line 147) | class RESTfulRerankModelHandle(RESTfulModelHandle):
    method rerank (line 148) | def rerank(
  class RESTfulImageModelHandle (line 206) | class RESTfulImageModelHandle(RESTfulModelHandle):
    method text_to_image (line 207) | def text_to_image(
    method image_to_image (line 251) | def image_to_image(
    method image_edit (line 332) | def image_edit(
    method inpainting (line 463) | def inpainting(
    method ocr (line 538) | def ocr(self, image: Union[str, bytes], **kwargs):
  class RESTfulVideoModelHandle (line 558) | class RESTfulVideoModelHandle(RESTfulModelHandle):
    method text_to_video (line 559) | def text_to_video(
    method image_to_video (line 595) | def image_to_video(
    method flf_to_video (line 642) | def flf_to_video(
  class RESTfulGenerateModelHandle (line 696) | class RESTfulGenerateModelHandle(RESTfulModelHandle):
    method generate (line 697) | def generate(
  class RESTfulChatModelHandle (line 751) | class RESTfulChatModelHandle(RESTfulGenerateModelHandle):
    method chat (line 752) | def chat(
  class RESTfulAudioModelHandle (line 832) | class RESTfulAudioModelHandle(RESTfulModelHandle):
    method transcriptions (line 833) | def transcriptions(
    method translations (line 898) | def translations(
    method speech (line 961) | def speech(
  class RESTfulFlexibleModelHandle (line 1042) | class RESTfulFlexibleModelHandle(RESTfulModelHandle):
    method infer (line 1043) | def infer(
  class Client (line 1079) | class Client:
    method __init__ (line 1080) | def __init__(self, base_url, api_key: Optional[str] = None):
    method close (line 1089) | def close(self):
    method __del__ (line 1097) | def __del__(self):
    method _set_token (line 1101) | def _set_token(self, token: Optional[str]):
    method _get_token (line 1106) | def _get_token(self) -> Optional[str]:
    method _check_cluster_authenticated (line 1113) | def _check_cluster_authenticated(self):
    method vllm_models (line 1127) | def vllm_models(self) -> Dict[str, Any]:
    method login (line 1140) | def login(self, username: str, password: str):
    method list_models (line 1156) | def list_models(self) -> Dict[str, Dict[str, Any]]:
    method launch_model (line 1179) | def launch_model(
    method terminate_model (line 1308) | def terminate_model(self, model_uid: str):
    method get_launch_model_progress (line 1332) | def get_launch_model_progress(self, model_uid: str) -> dict:
    method cancel_launch_model (line 1360) | def cancel_launch_model(self, model_uid: str):
    method get_instance_info (line 1382) | def get_instance_info(self, model_name: str, model_uid: str):
    method _get_supervisor_internal_address (line 1394) | def _get_supervisor_internal_address(self):
    method get_model (line 1402) | def get_model(self, model_uid: str) -> RESTfulModelHandle:
    method describe_model (line 1472) | def describe_model(self, model_uid: str):
    method register_model (line 1522) | def register_model(
    method unregister_model (line 1559) | def unregister_model(self, model_type: str, model_name: str):
    method list_model_registrations (line 1585) | def list_model_registrations(
    method list_cached_models (line 1619) | def list_cached_models(
    method list_deletable_models (line 1657) | def list_deletable_models(
    method confirm_and_remove_model (line 1687) | def confirm_and_remove_model(
    method get_model_registration (line 1717) | def get_model_registration(
    method query_engine_by_model_name (line 1745) | def query_engine_by_model_name(
    method abort_request (line 1781) | def abort_request(self, model_uid: str, request_id: str, block_duratio...
    method get_workers_info (line 1813) | def get_workers_info(self):
    method get_supervisor_info (line 1823) | def get_supervisor_info(self):
    method get_progress (line 1833) | def get_progress(self, request_id: str):
    method abort_cluster (line 1843) | def abort_cluster(self):

FILE: xinference/client/tests/test_async_client.py
  class _DummyAsyncResponse (line 30) | class _DummyAsyncResponse:
    method json (line 33) | async def json(self):
    method release (line 36) | def release(self):
    method wait_for_close (line 39) | async def wait_for_close(self):
  class _DummyAsyncSession (line 43) | class _DummyAsyncSession:
    method __init__ (line 44) | def __init__(self):
    method post (line 47) | async def post(self, url, json=None, headers=None):
    method close (line 51) | async def close(self):
  function test_async_RESTful_client (line 57) | async def test_async_RESTful_client(setup):
  function test_async_query_engines_by_name (line 184) | async def test_async_query_engines_by_name(setup):
  function test_async_list_cached_models (line 199) | async def test_async_list_cached_models(setup):
  function test_async_RESTful_client_for_embedding (line 208) | async def test_async_RESTful_client_for_embedding(setup):
  function test_async_RESTful_client_custom_model (line 229) | async def test_async_RESTful_client_custom_model(setup):
  function test_async_client_from_modelscope (line 370) | async def test_async_client_from_modelscope(setup):
  function test_async_client_custom_embedding_model (line 395) | async def test_async_client_custom_embedding_model(setup):
  function test_async_rerank (line 448) | async def test_async_rerank(setup):
  function set_auto_recover_limit (line 509) | def set_auto_recover_limit():
  function set_test_oom_error (line 516) | def set_test_oom_error():
  function setup_cluster (line 523) | def setup_cluster():
  function test_auto_recover (line 558) | async def test_auto_recover(set_auto_recover_limit, setup_cluster):
  function test_model_error (line 598) | async def test_model_error(set_test_oom_error, setup_cluster):
  function test_async_restful_chat_enable_thinking_injected (line 620) | async def test_async_restful_chat_enable_thinking_injected():

FILE: xinference/client/tests/test_async_client_with_auth.py
  function test_async_client_auth (line 22) | async def test_async_client_auth(setup_with_auth):

FILE: xinference/client/tests/test_client.py
  class _DummyResponse (line 33) | class _DummyResponse:
    method json (line 36) | def json(self):
  class _DummySession (line 40) | class _DummySession:
    method __init__ (line 41) | def __init__(self):
    method post (line 44) | def post(self, url, json=None, stream=None, headers=None):
    method close (line 48) | def close(self):
  function test_RESTful_client (line 53) | def test_RESTful_client(setup):
  function test_query_engines_by_name (line 192) | def test_query_engines_by_name(setup):
  function test_list_cached_models (line 202) | def test_list_cached_models(setup):
  function test_RESTful_client_for_embedding (line 209) | def test_RESTful_client_for_embedding(setup):
  function test_RESTful_client_custom_model (line 227) | def test_RESTful_client_custom_model(setup):
  function test_client_from_modelscope (line 360) | def test_client_from_modelscope(setup):
  function test_client_error (line 380) | def test_client_error():
  function test_client_custom_embedding_model (line 392) | def test_client_custom_embedding_model(setup):
  function set_auto_recover_limit (line 442) | def set_auto_recover_limit():
  function set_test_oom_error (line 449) | def set_test_oom_error():
  function setup_cluster (line 456) | def setup_cluster():
  function test_auto_recover (line 490) | def test_auto_recover(set_auto_recover_limit, setup_cluster):
  function test_model_error (line 528) | def test_model_error(set_test_oom_error, setup_cluster):
  function test_restful_chat_enable_thinking_injected (line 548) | def test_restful_chat_enable_thinking_injected():

FILE: xinference/client/tests/test_client_with_auth.py
  function test_client_auth (line 21) | def test_client_auth(setup_with_auth):

FILE: xinference/conftest.py
  function api_health_check (line 104) | def api_health_check(endpoint: str, max_attempts: int, sleep_interval: i...
  function _start_test_cluster (line 128) | async def _start_test_cluster(
  function run_test_cluster (line 154) | def run_test_cluster(address: str, logging_conf: Optional[Dict] = None):
  function run_test_cluster_in_subprocess (line 167) | def run_test_cluster_in_subprocess(
  function setup (line 179) | def setup():
  function setup_with_file_logging (line 211) | def setup_with_file_logging():
  function setup_with_auth (line 243) | def setup_with_auth():

FILE: xinference/constants.py
  function get_xinference_home (line 61) | def get_xinference_home() -> str:

FILE: xinference/core/cache_tracker.py
  class CacheTrackerActor (line 22) | class CacheTrackerActor(xo.Actor):
    method __init__ (line 23) | def __init__(self):
    method default_uid (line 28) | def default_uid(cls) -> str:
    method _map_address_to_file_location (line 32) | def _map_address_to_file_location(
    method _update_file_location (line 44) | def _update_file_location(data: Dict, origin_version_info: Dict):
    method record_model_version (line 51) | def record_model_version(self, version_info: Dict[str, List[Dict]], ad...
    method update_cache_status (line 72) | def update_cache_status(
    method unregister_model_version (line 91) | def unregister_model_version(self, model_name: str):
    method get_model_versions (line 94) | def get_model_versions(self, model_name: str) -> List[Dict]:
    method get_model_version_count (line 101) | def get_model_version_count(self, model_name: str) -> int:
    method list_cached_models (line 104) | def list_cached_models(
    method list_deletable_models (line 126) | def list_deletable_models(self, model_version: str, worker_ip: str) ->...
    method confirm_and_remove_model (line 140) | def confirm_and_remove_model(self, model_version: str, worker_ip: str):

FILE: xinference/core/event.py
  class EventType (line 24) | class EventType(Enum):
  class Event (line 30) | class Event(TypedDict):
  class EventCollectorActor (line 36) | class EventCollectorActor(xo.StatelessActor):
    method __init__ (line 37) | def __init__(self):
    method default_uid (line 44) | def default_uid(cls) -> str:
    method get_model_events (line 47) | def get_model_events(self, model_uid: str) -> List[Dict]:
    method report_event (line 54) | def report_event(self, model_uid: str, event: Event):

FILE: xinference/core/launch_strategy.py
  class LaunchStrategy (line 25) | class LaunchStrategy(ABC):
    method select_worker (line 31) | def select_worker(
  class IdleFirstLaunchStrategy (line 47) | class IdleFirstLaunchStrategy(LaunchStrategy):
    method __init__ (line 53) | def __init__(self, worker_status: Dict[str, "WorkerStatus"]):
    method _select_least_loaded_gpu (line 57) | def _select_least_loaded_gpu(
    method _reserve_slot (line 99) | def _reserve_slot(
    method select_worker (line 114) | def select_worker(

FILE: xinference/core/metrics.py
  function record_metrics (line 42) | def record_metrics(name, op, kwargs):
  function launch_metrics_export_server (line 47) | def launch_metrics_export_server(q, host=None, port=None):

FILE: xinference/core/model.py
  class _OutOfMemoryError (line 61) | class _OutOfMemoryError(Exception):
  function register_batching_multimodal_models (line 76) | def register_batching_multimodal_models(*model_names: str):
  function request_limit (line 86) | def request_limit(fn):
  function oom_check (line 122) | def oom_check(fn):
  class ModelActor (line 153) | class ModelActor(xo.StatelessActor, CancelMixin):
    method __pre_destroy__ (line 156) | async def __pre_destroy__(self):
    method __init__ (line 195) | def __init__(
    method __post_create__ (line 245) | async def __post_create__(self):
    method __repr__ (line 254) | def __repr__(self) -> str:
    method __getattr__ (line 257) | def __getattr__(self, attr: str):
    method decrease_serve_count (line 260) | def decrease_serve_count(self):
    method start_transfer_for_vllm (line 264) | async def start_transfer_for_vllm(self, rank_addresses: List[str]):
    method _record_completion_metrics (line 286) | async def _record_completion_metrics(
    method _get_worker_ref (line 323) | async def _get_worker_ref(self) -> xo.ActorRefType["WorkerActor"]:
    method _get_progress_tracker_ref (line 332) | async def _get_progress_tracker_ref(
    method _get_progressor (line 343) | async def _get_progressor(self, request_id: str):
    method is_vllm_backend (line 354) | def is_vllm_backend(self) -> bool:
    method is_sglang_backend (line 359) | def is_sglang_backend(self) -> bool:
    method load (line 364) | async def load(self):
    method wait_for_load (line 393) | async def wait_for_load(self):
    method need_create_pools (line 397) | def need_create_pools(self):
    method set_pool_addresses (line 400) | def set_pool_addresses(self, pool_addresses: List[str]):
    method get_pool_addresses (line 404) | def get_pool_addresses(self) -> Optional[List[str]]:
    method set_worker_addresses (line 409) | def set_worker_addresses(self, shard: int, worker_addresses: List[str]):
    method model_uid (line 413) | def model_uid(self):
    method get_driver_info (line 424) | def get_driver_info(self):
    method stop (line 430) | async def stop(self):
    method _handle_oom_error (line 436) | async def _handle_oom_error(self, ex):
    method _to_generator (line 447) | def _to_generator(self, output_type: str, gen: types.GeneratorType):
    method _to_async_gen (line 486) | async def _to_async_gen(self, output_type: str, gen: types.AsyncGenera...
    method _handle_pending_requests (line 527) | async def _handle_pending_requests(self):
    method _call_wrapper_json (line 568) | async def _call_wrapper_json(self, fn: Callable, *args, **kwargs):
    method _call_wrapper_binary (line 571) | async def _call_wrapper_binary(self, fn: Callable, *args, **kwargs):
    method _call_wrapper (line 575) | async def _call_wrapper(self, output_type: str, fn: Callable, *args, *...
    method generate (line 643) | async def generate(self, prompt: str, *args, **kwargs):
    method chat (line 669) | async def chat(self, messages: List[Dict], *args, **kwargs):
    method abort_request (line 715) | async def abort_request(
    method create_embedding (line 736) | async def create_embedding(self, input: Union[str, List[str]], *args, ...
    method convert_ids_to_tokens (line 749) | async def convert_ids_to_tokens(
    method rerank (line 762) | async def rerank(
    method transcriptions (line 790) | async def transcriptions(
    method translations (line 818) | async def translations(
    method speech (line 846) | async def speech(
    method text_to_image (line 872) | async def text_to_image(
    method txt2img (line 901) | async def txt2img(
    method image_to_image (line 920) | async def image_to_image(
    method img2img (line 953) | async def img2img(
    method inpainting (line 972) | async def inpainting(
    method ocr (line 1009) | async def ocr(
    method infer (line 1026) | async def infer(
    method text_to_video (line 1044) | async def text_to_video(
    method image_to_video (line 1069) | async def image_to_video(
    method flf_to_video (line 1098) | async def flf_to_video(
    method record_metrics (line 1127) | async def record_metrics(self, name, op, kwargs):
    method get_pending_requests_count (line 1131) | async def get_pending_requests_count(self):

FILE: xinference/core/otel.py
  function setup_otel (line 49) | def setup_otel(
  function _build_headers (line 150) | def _build_headers(api_key: str) -> Optional[dict]:
  function _setup_tracing (line 157) | def _setup_tracing(
  function _build_span_exporter (line 197) | def _build_span_exporter(
  function _setup_metrics (line 230) | def _setup_metrics(
  function _build_metric_exporter (line 260) | def _build_metric_exporter(
  function _instrument_fastapi (line 292) | def _instrument_fastapi(app) -> None:  # type: ignore[type-arg]
  class ClusterMetricsCollector (line 305) | class ClusterMetricsCollector:
    method __init__ (line 324) | def __init__(self) -> None:
    method update (line 328) | def update(self, worker_address: str, node_info: dict) -> None:
    method remove_worker (line 332) | def remove_worker(self, worker_address: str) -> None:
    method register (line 336) | def register(self) -> None:
    method _cpu_utilization_cb (line 393) | def _cpu_utilization_cb(self, options):  # type: ignore[no-untyped-def]
    method _cpu_count_cb (line 401) | def _cpu_count_cb(self, options):  # type: ignore[no-untyped-def]
    method _memory_used_cb (line 411) | def _memory_used_cb(self, options):  # type: ignore[no-untyped-def]
    method _memory_total_cb (line 419) | def _memory_total_cb(self, options):  # type: ignore[no-untyped-def]
    method _gpu_utilization_cb (line 429) | def _gpu_utilization_cb(self, options):  # type: ignore[no-untyped-def]
    method _gpu_mem_used_cb (line 445) | def _gpu_mem_used_cb(self, options):  # type: ignore[no-untyped-def]
    method _gpu_mem_total_cb (line 461) | def _gpu_mem_total_cb(self, options):  # type: ignore[no-untyped-def]
    method _gpu_mem_free_cb (line 477) | def _gpu_mem_free_cb(self, options):  # type: ignore[no-untyped-def]
  function get_cluster_metrics_collector (line 498) | def get_cluster_metrics_collector() -> Optional[ClusterMetricsCollector]:

FILE: xinference/core/progress_tracker.py
  class _ProgressInfo (line 39) | class _ProgressInfo:
  class ProgressTrackerActor (line 45) | class ProgressTrackerActor(xo.StatelessActor):
    method default_uid (line 49) | def default_uid(cls) -> str:
    method __init__ (line 52) | def __init__(
    method __post_create__ (line 64) | async def __post_create__(self):
    method __pre_destroy__ (line 67) | async def __pre_destroy__(self):
    method _clear_finished (line 71) | async def _clear_finished(self):
    method start (line 95) | def start(self, request_id: str, info: Optional[str] = None):
    method set_progress (line 100) | def set_progress(
    method get_progress (line 113) | def get_progress(self, request_id: str) -> float:
    method get_progress_info (line 116) | def get_progress_info(self, request_id: str) -> Tuple[float, Optional[...
  class Progressor (line 121) | class Progressor:
    method __init__ (line 124) | def __init__(
    method start (line 144) | async def start(self):
    method split_stages (line 148) | def split_stages(self, n_stage: int, stage_weight: Optional[List[float...
    method __enter__ (line 165) | def __enter__(self):
    method __exit__ (line 172) | def __exit__(self, exc_type, exc_val, exc_tb):
    method set_progress (line 180) | def set_progress(self, progress: float, info: Optional[str] = None):

FILE: xinference/core/resource.py
  class ResourceStatus (line 24) | class ResourceStatus:
  class GPUStatus (line 33) | class GPUStatus:
  function gather_node_info (line 42) | def gather_node_info() -> Dict[str, Union[ResourceStatus, GPUStatus]]:

FILE: xinference/core/status_guard.py
  class LaunchStatus (line 25) | class LaunchStatus(Enum):
  class ReplicaStatus (line 34) | class ReplicaStatus(BaseModel):
  class InstanceInfo (line 45) | class InstanceInfo(BaseModel):
    method update (line 56) | def update(self, **kwargs):
  class StatusGuardActor (line 61) | class StatusGuardActor(xo.StatelessActor):
    method __init__ (line 62) | def __init__(self):
    method default_uid (line 67) | def default_uid(cls) -> str:
    method _drop_terminated_info (line 71) | def _drop_terminated_info(instance_infos: List[InstanceInfo]) -> List[...
    method set_instance_info (line 78) | def set_instance_info(self, model_uid: str, info: InstanceInfo):
    method get_instance_info (line 81) | def get_instance_info(
    method get_instance_count (line 100) | def get_instance_count(self, model_name: str) -> int:
    method update_instance_info (line 103) | def update_instance_info(self, model_uid: str, info: Dict):
    method update_replica_status (line 106) | def update_replica_status(
    method get_replica_statuses (line 141) | def get_replica_statuses(self, model_uid: str) -> List[ReplicaStatus]:

FILE: xinference/core/supervisor.py
  function callback_for_async_launch (line 88) | def callback_for_async_launch(model_uid: str):
  class WorkerStatus (line 94) | class WorkerStatus:
  class ReplicaInfo (line 101) | class ReplicaInfo:
  class SupervisorActor (line 109) | class SupervisorActor(xo.StatelessActor):
    method __init__ (line 110) | def __init__(self):
    method default_uid (line 126) | def default_uid(cls) -> str:
    method _get_worker_ref_by_ip (line 129) | def _get_worker_ref_by_ip(
    method __post_create__ (line 138) | async def __post_create__(self):
    method get_cluster_device_info (line 307) | async def get_cluster_device_info(self, detailed: bool = False) -> List:
    method get_builtin_prompts (line 355) | async def get_builtin_prompts() -> Dict[str, Any]:
    method get_builtin_families (line 361) | async def get_builtin_families() -> Dict[str, List[str]]:
    method get_devices_count (line 385) | async def get_devices_count(self) -> int:
    method _choose_worker (line 415) | async def _choose_worker(
    method get_status (line 439) | def get_status(self) -> Dict:
    method _get_spec_dicts (line 445) | def _get_spec_dicts(
    method _to_llm_reg (line 465) | async def _to_llm_reg(
    method _to_embedding_model_reg (line 489) | async def _to_embedding_model_reg(
    method _to_rerank_model_reg (line 516) | async def _to_rerank_model_reg(
    method _to_image_model_reg (line 543) | async def _to_image_model_reg(
    method _to_audio_model_reg (line 578) | async def _to_audio_model_reg(
    method _to_video_model_reg (line 613) | async def _to_video_model_reg(
    method _to_flexible_model_reg (line 648) | async def _to_flexible_model_reg(
    method list_model_registrations (line 670) | async def list_model_registrations(
    method get_model_registration (line 695) | async def get_model_registration(self, model_type: str, model_name: st...
    method query_engines_by_model_name (line 706) | async def query_engines_by_model_name(
    method register_model (line 734) | async def register_model(
    method _sync_register_model (line 802) | async def _sync_register_model(
    method unregister_model (line 827) | async def unregister_model(self, model_type: str, model_name: str):
    method update_model_type (line 841) | async def update_model_type(self, model_type: str):
    method _gen_model_uid (line 876) | def _gen_model_uid(self, model_name: str) -> str:
    method get_model_versions (line 884) | async def get_model_versions(self, model_type: str, model_name: str) -...
    method get_model_version_count (line 887) | async def get_model_version_count(self, model_name: str) -> int:
    method launch_model_by_version (line 891) | async def launch_model_by_version(
    method _get_worker_refs_by_ip (line 923) | def _get_worker_refs_by_ip(self, ip: str) -> List[xo.ActorRefType["Wor...
    method launch_builtin_model (line 937) | async def launch_builtin_model(
    method _launch_builtin_sharded_model (line 1396) | async def _launch_builtin_sharded_model(
    method get_launch_builtin_model_progress (line 1557) | async def get_launch_builtin_model_progress(self, model_uid: str) -> f...
    method cancel_launch_builtin_model (line 1576) | async def cancel_launch_builtin_model(self, model_uid: str):
    method get_instance_info (line 1599) | async def get_instance_info(
    method get_replica_statuses (line 1607) | async def get_replica_statuses(self, model_uid: str) -> List[Dict]:
    method get_instance_count (line 1622) | async def get_instance_count(self, model_name: str) -> int:
    method _check_dead_nodes (line 1625) | async def _check_dead_nodes(self):
    method terminate_model (line 1679) | async def terminate_model(self, model_uid: str, suppress_exception=Fal...
    method get_model (line 1740) | async def get_model(self, model_uid: str) -> xo.ActorRefType["ModelAct...
    method get_model_status (line 1763) | async def get_model_status(self, replica_model_uid: str):
    method describe_model (line 1778) | async def describe_model(self, model_uid: str) -> Dict[str, Any]:
    method list_models (line 1801) | async def list_models(self) -> Dict[str, Dict[str, Any]]:
    method is_local_deployment (line 1813) | def is_local_deployment(self) -> bool:
    method list_cached_models (line 1821) | async def list_cached_models(
    method abort_request (line 1849) | async def abort_request(
    method add_worker (line 1886) | async def add_worker(self, worker_address: str):
    method remove_worker (line 1900) | async def remove_worker(self, worker_address: str):
    method report_worker_status (line 1928) | async def report_worker_status(
    method list_deletable_models (line 1957) | async def list_deletable_models(
    method confirm_and_remove_model (line 1982) | async def confirm_and_remove_model(
    method list_virtual_envs (line 2008) | async def list_virtual_envs(
    method list_virtual_env_packages (line 2043) | async def list_virtual_env_packages(
    method remove_virtual_env (line 2082) | async def remove_virtual_env(
    method get_workers_info (line 2135) | async def get_workers_info(self) -> List[Dict[str, Any]]:
    method get_supervisor_info (line 2141) | async def get_supervisor_info(self) -> Dict[str, Any]:
    method trigger_exit (line 2147) | async def trigger_exit(self) -> bool:
    method abort_cluster (line 2155) | async def abort_cluster(self) -> bool:
    method record_metrics (line 2164) | def record_metrics(name, op, kwargs):
    method get_progress (line 2167) | async def get_progress(self, request_id: str) -> float:
    method call_collective_manager (line 2170) | async def call_collective_manager(

FILE: xinference/core/tests/test_continuous_batching.py
  class BaseThread (line 27) | class BaseThread(threading.Thread):
    method __init__ (line 28) | def __init__(self):
    method run_internal (line 32) | def run_internal(self):
    method run (line 35) | def run(self):
    method join (line 41) | def join(self, timeout=None):
  class InferenceThread (line 47) | class InferenceThread(BaseThread):
    method __init__ (line 48) | def __init__(self, prompt, generate_config, client, model):
    method stream (line 57) | def stream(self):
    method run_internal (line 60) | def run_internal(self):
  class InferenceThreadWithError (line 79) | class InferenceThreadWithError(InferenceThread):
    method __init__ (line 80) | def __init__(self, prompt, generate_config, client, model, sleep=None):
    method run_internal (line 84) | def run_internal(self):
  class AbortThread (line 98) | class AbortThread(BaseThread):
    method __init__ (line 99) | def __init__(self, client, model_uid, request_id, expected_res, sleep=...
    method run_internal (line 107) | def run_internal(self):
  function test_continuous_batching (line 118) | def test_continuous_batching(setup):

FILE: xinference/core/tests/test_launch_strategy.py
  class DummyRef (line 23) | class DummyRef:
    method __init__ (line 24) | def __init__(self, address: str):
  class DummyWorkerRef (line 28) | class DummyWorkerRef:
    method __init__ (line 29) | def __init__(self, address: str, model_count: int, launched: list):
    method get_model_count (line 34) | async def get_model_count(self) -> int:
    method launch_builtin_model (line 37) | async def launch_builtin_model(self, *args, **kwargs):
    method wait_for_load (line 43) | async def wait_for_load(self, model_uid: str):
  class DummyStatusGuard (line 47) | class DummyStatusGuard:
    method __init__ (line 48) | def __init__(self):
    method set_instance_info (line 51) | async def set_instance_info(self, model_uid: str, instance_info):
    method update_instance_info (line 54) | async def update_instance_info(self, model_uid: str, updates: dict):
  function test_assign_replica_gpu_single_slot_reused (line 58) | def test_assign_replica_gpu_single_slot_reused():
  function test_assign_replica_gpu_slicing (line 64) | def test_assign_replica_gpu_slicing():
  function test_idle_first_prefers_empty_gpu (line 71) | def test_idle_first_prefers_empty_gpu(monkeypatch):
  function test_idle_first_balances_with_reserve (line 91) | def test_idle_first_balances_with_reserve(monkeypatch):
  function test_idle_first_fallback_to_count_when_no_alloc (line 113) | def test_idle_first_fallback_to_count_when_no_alloc():
  function test_multi_worker_multi_gpu_even_distribution (line 127) | def test_multi_worker_multi_gpu_even_distribution():
  function test_cpu_fallback_no_gpu_alloc (line 169) | def test_cpu_fallback_no_gpu_alloc():
  function test_idle_first_multi_gpu_single_worker (line 184) | def test_idle_first_multi_gpu_single_worker():
  function test_idle_first_multi_gpu_two_workers (line 204) | def test_idle_first_multi_gpu_two_workers():
  function test_distributed_launch_avoids_same_worker_for_shards (line 234) | async def test_distributed_launch_avoids_same_worker_for_shards():

FILE: xinference/core/tests/test_metrics.py
  function setup_cluster (line 21) | def setup_cluster():
  function test_metrics_exporter_server (line 57) | async def test_metrics_exporter_server(setup_cluster):
  function disable_metrics (line 95) | def disable_metrics():
  function test_disable_metrics_exporter_server (line 104) | async def test_disable_metrics_exporter_server(disable_metrics, setup_cl...
  function test_metrics_exporter_data (line 128) | async def test_metrics_exporter_data(setup_cluster):

FILE: xinference/core/tests/test_model.py
  class MockModelFamily (line 28) | class MockModelFamily:
    method to_description (line 29) | def to_description(self) -> dict:
  class MockModel (line 33) | class MockModel:
    method __init__ (line 34) | def __init__(self):
    method generate (line 37) | async def generate(self, prompt, **kwargs):
  class MockModelActor (line 46) | class MockModelActor(ModelActor):
    method __init__ (line 47) | def __init__(
    method __pre_destroy__ (line 61) | async def __pre_destroy__(self):
    method record_metrics (line 64) | async def record_metrics(self, name, op, kwargs):
  function setup_pool (line 69) | async def setup_pool():
  function test_concurrent_call (line 78) | async def test_concurrent_call(setup_pool):

FILE: xinference/core/tests/test_progressor.py
  function test_progressor (line 25) | async def test_progressor():

FILE: xinference/core/tests/test_restful_api.py
  class _DummyRequest (line 30) | class _DummyRequest:
    method __init__ (line 31) | def __init__(self, payload):
    method json (line 34) | async def json(self):
  function test_restful_api (line 39) | async def test_restful_api(setup):
  function test_restful_api_for_embedding (line 325) | def test_restful_api_for_embedding(setup):
  function _check_invalid_tool_calls (line 398) | def _check_invalid_tool_calls(endpoint, model_uid_res):
  function test_restful_api_for_tool_calls (line 449) | def test_restful_api_for_tool_calls(setup, model_format, quantization):
  function test_restful_api_for_llama3_tool_calls (line 628) | def test_restful_api_for_llama3_tool_calls(setup, model_format, quantiza...
  function test_restful_api_for_gorilla_openfunctions_tool_calls (line 719) | def test_restful_api_for_gorilla_openfunctions_tool_calls(
  function test_restful_api_for_qwen_tool_calls (line 820) | def test_restful_api_for_qwen_tool_calls(setup, model_format, quantizati...
  function test_restful_api_with_request_limits (line 1040) | def test_restful_api_with_request_limits(setup):
  function test_openai (line 1106) | async def test_openai(setup):
  function test_lang_chain (line 1169) | def test_lang_chain(setup):
  function test_chat_completion_enable_thinking_injected (line 1265) | async def test_chat_completion_enable_thinking_injected(payload, expected):
  function test_launch_model_async (line 1291) | def test_launch_model_async(setup):
  function test_cancel_launch_model (line 1333) | def test_cancel_launch_model(setup):
  function test_events (line 1372) | def test_events(setup):
  function test_launch_model_by_version (line 1412) | def test_launch_model_by_version(setup):
  function test_builtin_families (line 1448) | def test_builtin_families(setup):
  function anthropic_setup (line 1469) | def anthropic_setup():
  function test_convert_openai_to_anthropic_with_tools (line 1528) | def test_convert_openai_to_anthropic_with_tools(anthropic_setup):
  function test_convert_openai_to_anthropic_without_tools (line 1555) | def test_convert_openai_to_anthropic_without_tools(anthropic_setup):
  function test_convert_openai_to_anthropic_mixed_content (line 1579) | def test_convert_openai_to_anthropic_mixed_content(anthropic_setup):
  function test_convert_openai_to_anthropic_multiple_tools (line 1628) | def test_convert_openai_to_anthropic_multiple_tools(anthropic_setup):
  function test_convert_openai_to_anthropic_empty_response (line 1684) | def test_convert_openai_to_anthropic_empty_response(anthropic_setup):
  function test_convert_openai_to_anthropic_invalid_tool_arguments (line 1698) | def test_convert_openai_to_anthropic_invalid_tool_arguments(anthropic_se...
  function test_anthropic_tools_response_format (line 1737) | def test_anthropic_tools_response_format(anthropic_setup):
  function anthropic_api (line 1778) | def anthropic_api():
  function mock_supervisor (line 1786) | def mock_supervisor():
  function sample_models (line 1793) | def sample_models():
  function test_anthropic_list_models (line 1816) | async def test_anthropic_list_models(anthropic_api, mock_supervisor, sam...
  function test_anthropic_get_model_found (line 1838) | async def test_anthropic_get_model_found(anthropic_api, mock_supervisor,...
  function test_anthropic_models_format_compatibility (line 1860) | async def test_anthropic_models_format_compatibility(
  function test_anthropic_models_include_original_fields (line 1892) | async def test_anthropic_models_include_original_fields(

FILE: xinference/core/tests/test_types.py
  function check_fields (line 25) | def check_fields(a, b):
  function test_create_completion_types (line 48) | def test_create_completion_types():
  function test_create_chat_completion_types (line 68) | def test_create_chat_completion_types():
  function test_openai_requests (line 91) | def test_openai_requests():
  function test_create_message_with_tools (line 135) | def test_create_message_with_tools():
  function test_create_message_without_tools (line 176) | def test_create_message_without_tools():
  function test_create_message_with_tool_choice_function (line 197) | def test_create_message_with_tool_choice_function():
  function test_create_message_with_optional_fields (line 229) | def test_create_message_with_optional_fields():
  function test_model_and_messages_type (line 256) | def test_model_and_messages_type():
  function test_message_create_params_structure (line 275) | def test_message_create_params_structure():

FILE: xinference/core/tests/test_utils.py
  function test_replica_model_uid (line 23) | def test_replica_model_uid():
  class DummyVirtualEnvManager (line 35) | class DummyVirtualEnvManager:
    method __init__ (line 36) | def __init__(self, python_path: str):
    method get_python_path (line 39) | def get_python_path(self) -> str:
  function test_build_subpool_envs_for_virtual_env_disabled (line 43) | def test_build_subpool_envs_for_virtual_env_disabled():
  function test_build_subpool_envs_for_virtual_env_enabled (line 51) | def test_build_subpool_envs_for_virtual_env_enabled():

FILE: xinference/core/tests/test_worker.py
  class MockWorkerActor (line 26) | class MockWorkerActor(WorkerActor):
    method __init__ (line 27) | def __init__(
    method __post_create__ (line 35) | async def __post_create__(self):
    method __pre_destroy__ (line 38) | async def __pre_destroy__(self):
    method get_gpu_to_model_uid (line 41) | def get_gpu_to_model_uid(self):
    method get_user_specified_gpu_to_model_uids (line 44) | def get_user_specified_gpu_to_model_uids(self):
    method set_allow_multi_replica_per_gpu (line 47) | def set_allow_multi_replica_per_gpu(self, allow: bool):
    method is_model_vllm_backend (line 50) | async def is_model_vllm_backend(self, model_uid):
    method launch_builtin_model (line 62) | async def launch_builtin_model(
    method terminate_model (line 79) | async def terminate_model(self, model_uid: str):
  function setup_pool (line 88) | async def setup_pool():
  function test_allocate_cuda_devices (line 97) | async def test_allocate_cuda_devices(setup_pool):
  function test_terminate_model_flag (line 124) | async def test_terminate_model_flag(setup_pool):
  function test_merge_virtual_env_packages_override_and_append (line 173) | def test_merge_virtual_env_packages_override_and_append():
  class DummyVirtualEnvManager (line 191) | class DummyVirtualEnvManager:
    method __init__ (line 192) | def __init__(self):
    method install_packages (line 196) | def install_packages(self, packages, **kwargs):
  function test_prepare_virtual_env_injects_engine_vars (line 200) | def test_prepare_virtual_env_injects_engine_vars():
  function test_prepare_virtual_env_without_engine_vars (line 217) | def test_prepare_virtual_env_without_engine_vars():
  function test_prepare_virtual_env_inherit_pip_config (line 233) | def test_prepare_virtual_env_inherit_pip_config(monkeypatch):
  function test_prepare_virtual_env_keeps_system_markers (line 254) | def test_prepare_virtual_env_keeps_system_markers():
  function test_launch_embedding_model (line 280) | async def test_launch_embedding_model(setup_pool):
  function test_launch_model_with_gpu_idx (line 350) | async def test_launch_model_with_gpu_idx(setup_pool):

FILE: xinference/core/utils.py
  class AbortRequestMessage (line 37) | class AbortRequestMessage(Enum):
  function truncate_log_arg (line 43) | def truncate_log_arg(arg) -> str:
  function log_async (line 50) | def log_async(
  function log_sync (line 119) | def log_sync(logger, level=logging.DEBUG, log_exception=True):
  function iter_replica_model_uid (line 160) | def iter_replica_model_uid(model_uid: str, replica: int) -> Generator[st...
  function build_replica_model_uid (line 169) | def build_replica_model_uid(model_uid: str, rep_id: int) -> str:
  function parse_replica_model_uid (line 176) | def parse_replica_model_uid(replica_model_uid: str) -> Tuple[str, int]:
  function is_valid_model_uid (line 188) | def is_valid_model_uid(model_uid: str) -> bool:
  function gen_random_string (line 195) | def gen_random_string(length: int) -> str:
  function json_dumps (line 199) | def json_dumps(o):
  function purge_dir (line 208) | def purge_dir(d):
  function parse_model_version (line 223) | def parse_model_version(model_version: str, model_type: str) -> Tuple:
  function merge_virtual_env_packages (line 252) | def merge_virtual_env_packages(
  function build_subpool_envs_for_virtual_env (line 295) | def build_subpool_envs_for_virtual_env(
  function apply_engine_virtualenv_settings (line 318) | def apply_engine_virtualenv_settings(
  function filter_virtualenv_packages_by_markers (line 340) | def filter_virtualenv_packages_by_markers(
  function assign_replica_gpu (line 419) | def assign_replica_gpu(
  class CancelMixin (line 439) | class CancelMixin:
    method __init__ (line 442) | def __init__(self):
    method _add_running_task (line 447) | def _add_running_task(self, request_id: Optional[str]):
    method _cancel_running_task (line 462) | def _cancel_running_task(

FILE: xinference/core/virtual_env_manager.py
  function get_engine_virtualenv_packages (line 79) | def get_engine_virtualenv_packages(model_engine: Optional[str]) -> List[...
  function get_engine_virtualenv_extra_index_urls (line 85) | def get_engine_virtualenv_extra_index_urls(
  function get_engine_virtualenv_index_strategy (line 98) | def get_engine_virtualenv_index_strategy(model_engine: Optional[str]) ->...
  function resolve_virtualenv_python_path (line 104) | def resolve_virtualenv_python_path(virtual_env_manager: Any) -> Optional...
  function expand_engine_dependency_placeholders (line 129) | def expand_engine_dependency_placeholders(
  class VirtualEnvManager (line 157) | class VirtualEnvManager:
    method __init__ (line 165) | def __init__(self, worker_address: str):
    method list_virtual_envs (line 174) | def list_virtual_envs(
    method remove_virtual_env (line 231) | def remove_virtual_env(
    method check_virtual_env_exists (line 390) | def check_virtual_env_exists(self, model_name: str) -> Dict[str, Any]:
    method list_virtual_env_packages (line 403) | def list_virtual_env_packages(self, model_name: str) -> Dict[str, Any]:
    method _detect_python_version (line 421) | def _detect_python_version(self, env_path: str) -> str:
    method _is_valid_python_version (line 476) | def _is_valid_python_version(self, python_version: str) -> bool:

FILE: xinference/core/worker.py
  class ModelStatus (line 110) | class ModelStatus:
  class LaunchInfo (line 115) | class LaunchInfo:
  class WorkerActor (line 125) | class WorkerActor(xo.StatelessActor):
    method __init__ (line 126) | def __init__(
    method recover_sub_pool (line 206) | async def recover_sub_pool(self, address):
    method default_uid (line 262) | def default_uid(cls) -> str:
    method _get_spec_dicts_with_cache_status (line 265) | def _get_spec_dicts_with_cache_status(
    method _prefer_model_hub (line 287) | def _prefer_model_hub(self, model_family: Any, preferred_hub: str = "h...
    method __post_create__ (line 307) | async def __post_create__(self):
    method __pre_destroy__ (line 433) | async def __pre_destroy__(self):
    method trigger_exit (line 436) | async def trigger_exit(self) -> bool:
    method get_supervisor_ref (line 444) | async def get_supervisor_ref(self, add_worker: bool = True) -> xo.Acto...
    method get_devices_count (line 500) | def get_devices_count():
    method get_model_count (line 506) | def get_model_count(self) -> int:
    method is_model_vllm_backend (line 509) | async def is_model_vllm_backend(self, model_uid: str) -> bool:
    method allocate_devices (line 515) | def allocate_devices(self, model_uid: str, n_gpu: int) -> List[int]:
    method allocate_devices_with_gpu_idx (line 560) | async def allocate_devices_with_gpu_idx(
    method get_gpu_allocation_status (line 598) | async def get_gpu_allocation_status(self) -> Dict[str, Any]:
    method release_devices (line 610) | def release_devices(self, model_uid: str):
    method _create_subpool (line 628) | async def _create_subpool(
    method _check_model_is_valid (line 662) | def _check_model_is_valid(self, model_name: str, model_format: Optiona...
    method register_model (line 670) | async def register_model(self, model_type: str, model: str, persist: b...
    method unregister_model (line 694) | async def unregister_model(self, model_type: str, model_name: str):
    method update_model_type (line 703) | async def update_model_type(self, model_type: str):
    method _store_complete_model_configurations (line 794) | async def _store_complete_model_configurations(self, model_type: str, ...
    method list_model_registrations (line 833) | async def list_model_registrations(
    method get_model_registration (line 1154) | async def get_model_registration(self, model_type: str, model_name: st...
    method query_engines_by_model_name (line 1239) | async def query_engines_by_model_name(
    method _get_model_ability (line 1255) | async def _get_model_ability(self, model: Any, model_type: str) -> Lis...
    method update_cache_status (line 1271) | async def update_cache_status(self, model_name: str, version_info: Any):
    method _create_virtual_env_manager (line 1286) | def _create_virtual_env_manager(
    method _prepare_virtual_env (line 1373) | def _prepare_virtual_env(
    method _get_progressor (line 1460) | async def _get_progressor(self, request_id: str):
    method _upload_download_progress (line 1479) | def _upload_download_progress(
    method launch_builtin_model (line 1490) | async def launch_builtin_model(
    method wait_for_load (line 1803) | async def wait_for_load(self, model_uid: str):
    method cancel_launch_model (line 1808) | async def cancel_launch_model(self, model_uid: str):
    method terminate_model (line 1841) | async def terminate_model(self, model_uid: str, is_model_die=False):
    method get_model_launch_status (line 1948) | def get_model_launch_status(self, model_uid: str) -> Optional[str]:
    method list_models (line 1963) | async def list_models(self) -> Dict[str, Dict[str, Any]]:
    method get_model (line 1967) | def get_model(self, model_uid: str) -> xo.ActorRefType["ModelActor"]:
    method describe_model (line 1977) | def describe_model(self, model_uid: str) -> Dict[str, Any]:
    method report_status (line 1983) | async def report_status(self):
    method _periodical_report_status (line 1996) | async def _periodical_report_status(self):
    method list_cached_models (line 2016) | async def list_cached_models(
    method list_deletable_models (line 2040) | async def list_deletable_models(self, model_version: str) -> List[str]:
    method confirm_and_remove_model (line 2079) | async def confirm_and_remove_model(self, model_version: str) -> bool:
    method list_virtual_envs (line 2101) | async def list_virtual_envs(
    method list_virtual_env_packages (line 2127) | async def list_virtual_env_packages(self, model_name: str) -> Dict[str...
    method remove_virtual_env (line 2131) | async def remove_virtual_env(
    method get_workers_info (line 2142) | async def get_workers_info(self) -> Dict[str, Any]:
    method update_model_status (line 2149) | def update_model_status(self, model_uid: str, **kwargs):
    method get_model_status (line 2155) | def get_model_status(self, model_uid: str):
    method record_metrics (line 2159) | def record_metrics(name, op, kwargs):
    method start_transfer_for_vllm (line 2162) | async def start_transfer_for_vllm(
    method launch_rank0_model (line 2169) | async def launch_rank0_model(
    method recover_model (line 2202) | async def recover_model(self, launch_args: Dict[str, Any]):

FILE: xinference/deploy/cmdline.py
  function get_endpoint (line 59) | def get_endpoint(endpoint: Optional[str]) -> str:
  function get_hash_endpoint (line 71) | def get_hash_endpoint(endpoint: str) -> str:
  function get_stored_token (line 79) | def get_stored_token(
  function start_local_cluster (line 95) | def start_local_cluster(
  function cli (line 158) | def cli(
  function local (line 218) | def local(
  function supervisor (line 272) | def supervisor(
  function worker (line 334) | def worker(
  function register_model (line 394) | def register_model(
  function unregister_model (line 437) | def unregister_model(
  function list_model_registrations (line 475) | def list_model_registrations(
  function list_cached_models (line 624) | def list_cached_models(
  function remove_cache (line 680) | def remove_cache(
  function model_launch (line 893) | def model_launch(
  function model_list (line 1074) | def model_list(endpoint: Optional[str], api_key: Optional[str]):
  function model_terminate (line 1208) | def model_terminate(
  function model_generate (line 1246) | def model_generate(
  function model_chat (line 1349) | def model_chat(
  function vllm_models (line 1452) | def vllm_models(endpoint: Optional[str], api_key: Optional[str]):
  function cluster_login (line 1475) | def cluster_login(
  function query_engine_by_model_name (line 1536) | def query_engine_by_model_name(
  function cal_model_mem (line 1691) | def cal_model_mem(
  function stop_cluster (line 1752) | def stop_cluster(endpoint: str, api_key: Optional[str], check: bool):

FILE: xinference/deploy/local.py
  function _start_local_cluster (line 42) | async def _start_local_cluster(
  function run (line 82) | def run(
  function run_in_subprocess (line 119) | def run_in_subprocess(
  function main (line 149) | def main(

FILE: xinference/deploy/supervisor.py
  function _start_supervisor (line 35) | async def _start_supervisor(address: str, logging_conf: Optional[Dict] =...
  function run (line 52) | def run(address: str, logging_conf: Optional[Dict] = None):
  function run_in_subprocess (line 65) | def run_in_subprocess(
  function main (line 73) | def main(

FILE: xinference/deploy/test/test_cmdline.py
  function test_cmdline (line 37) | def test_cmdline(setup, stream, model_uid):
  function test_cmdline_model_path_error (line 150) | def test_cmdline_model_path_error(setup):
  function test_cmdline_of_custom_model (line 182) | def test_cmdline_of_custom_model(setup):
  function test_rotate_logs (line 275) | def test_rotate_logs(setup_with_file_logging):
  function test_list_cached_models (line 314) | def test_list_cached_models(setup):
  function test_remove_cache (line 331) | def test_remove_cache(setup):
  function test_launch_error_in_passing_parameters (line 345) | def test_launch_error_in_passing_parameters():

FILE: xinference/deploy/utils.py
  class LoggerNameFilter (line 34) | class LoggerNameFilter(logging.Filter):
    method filter (line 35) | def filter(self, record):
  function get_log_file (line 42) | def get_log_file(sub_dir: str):
  function get_config_dict (line 55) | def get_config_dict(
  function create_worker_actor_pool (line 141) | async def create_worker_actor_pool(
  function health_check (line 152) | def health_check(address: str, max_attempts: int, sleep_interval: int = ...
  function get_timestamp_ms (line 190) | def get_timestamp_ms():
  function handle_click_args_type (line 196) | def handle_click_args_type(arg: str) -> Any:
  function set_envs (line 218) | def set_envs(key: str, value: str):

FILE: xinference/deploy/worker.py
  function start_worker_components (line 30) | async def start_worker_components(
  function _start_worker (line 58) | async def _start_worker(
  function main (line 78) | def main(

FILE: xinference/device_utils.py
  function is_vacc_available (line 29) | def is_vacc_available() -> bool:
  function is_xpu_available (line 39) | def is_xpu_available() -> bool:
  function is_npu_available (line 43) | def is_npu_available() -> bool:
  function is_mlu_available (line 53) | def is_mlu_available() -> bool:
  function is_musa_available (line 63) | def is_musa_available() -> bool:
  function get_available_device (line 74) | def get_available_device() -> DeviceType:
  function is_device_available (line 92) | def is_device_available(device: str) -> bool:
  function move_model_to_available_device (line 113) | def move_model_to_available_device(model):
  function get_device_preferred_dtype (line 122) | def get_device_preferred_dtype(device: str) -> Union[torch.dtype, None]:
  function is_hf_accelerate_supported (line 140) | def is_hf_accelerate_supported(device: str) -> bool:
  function empty_cache (line 150) | def empty_cache():
  function get_available_device_env_name (line 177) | def get_available_device_env_name():
  function gpu_count (line 181) | def gpu_count():
  function _get_nvidia_gpu_mem_info (line 199) | def _get_nvidia_gpu_mem_info(gpu_id: int) -> Dict[str, float]:
  function get_nvidia_gpu_info (line 220) | def get_nvidia_gpu_info() -> Dict:

FILE: xinference/isolation.py
  class Isolation (line 20) | class Isolation:
    method __init__ (line 22) | def __init__(
    method _run (line 36) | def _run(self):
    method _cancel_all_tasks (line 43) | def _cancel_all_tasks(loop):
    method start (line 65) | def start(self):
    method call (line 73) | def call(self, coro: Coroutine) -> Any:
    method thread_ident (line 78) | def thread_ident(self):
    method loop (line 82) | def loop(self):
    method _stop (line 85) | async def _stop(self):
    method stop (line 88) | def stop(self):

FILE: xinference/model/__init__.py
  function _install (line 16) | def _install():

FILE: xinference/model/audio/__init__.py
  function register_custom_model (line 41) | def register_custom_model():
  function _need_filter (line 63) | def _need_filter(spec: dict):
  function _install (line 71) | def _install():
  function register_builtin_model (line 97) | def register_builtin_model():
  function has_downloaded_models (line 102) | def has_downloaded_models():
  function load_downloaded_models (line 109) | def load_downloaded_models():
  function load_model_family_from_json (line 124) | def load_model_family_from_json(json_filename, target_families):

FILE: xinference/model/audio/chattts.py
  class ChatTTSModel (line 28) | class ChatTTSModel:
    method __init__ (line 29) | def __init__(
    method model_ability (line 46) | def model_ability(self):
    method load (line 49) | def load(self):
    method speech (line 64) | def speech(

FILE: xinference/model/audio/core.py
  function get_audio_model_descriptions (line 42) | def get_audio_model_descriptions():
  class AudioModelFamilyV2 (line 48) | class AudioModelFamilyV2(CacheableModelSpec, ModelInstanceInfoMixin):
    class Config (line 62) | class Config:
    method to_description (line 65) | def to_description(self):
    method to_version_info (line 76) | def to_version_info(self):
  function generate_audio_description (line 88) | def generate_audio_description(
  function match_audio (line 96) | def match_audio(
  function create_audio_model_instance (line 136) | def create_audio_model_instance(

FILE: xinference/model/audio/cosyvoice.py
  class CosyVoiceModel (line 26) | class CosyVoiceModel:
    method __init__ (line 27) | def __init__(
    method model_ability (line 45) | def model_ability(self):
    method load (line 48) | def load(self):
    method _speech_handle (line 80) | def _speech_handle(
    method speech (line 154) | def speech(

FILE: xinference/model/audio/custom.py
  class CustomAudioModelFamilyV2 (line 32) | class CustomAudioModelFamilyV2(AudioModelFamilyV2):
    method parse_raw (line 39) | def parse_raw(
  class AudioModelRegistry (line 75) | class AudioModelRegistry(ModelRegistry):
    method __init__ (line 78) | def __init__(self):
  function get_user_defined_audios (line 86) | def get_user_defined_audios() -> List[CustomAudioModelFamilyV2]:
  function register_audio (line 93) | def register_audio(model_spec: CustomAudioModelFamilyV2, persist: bool):
  function unregister_audio (line 100) | def unregister_audio(model_name: str, raise_error: bool = True):

FILE: xinference/model/audio/f5tts.py
  class F5TTSModel (line 27) | class F5TTSModel:
    method __init__ (line 28) | def __init__(
    method model_ability (line 46) | def model_ability(self):
    method load (line 49) | def load(self):
    method _infer (line 92) | def _infer(self, ref_audio, ref_text, text_gen, model_obj, mel_spec_ty...
    method speech (line 152) | def speech(

FILE: xinference/model/audio/f5tts_mlx.py
  class F5TTSMLXModel (line 32) | class F5TTSMLXModel:
    method __init__ (line 33) | def __init__(
    method model_ability (line 51) | def model_ability(self):
    method load (line 54) | def load(self):
    method speech (line 127) | def speech(

FILE: xinference/model/audio/fish_speech.py
  function wav_chunk_header (line 31) | def wav_chunk_header(sample_rate=44100, bit_depth=16, channels=1):
  class FishSpeechModel (line 46) | class FishSpeechModel:
    method __init__ (line 47) | def __init__(
    method model_ability (line 66) | def model_ability(self):
    method load (line 69) | def load(self):
    method speech (line 115) | def speech(

FILE: xinference/model/audio/funasr.py
  class FunASRModel (line 27) | class FunASRModel:
    method __init__ (line 28) | def __init__(
    method model_ability (line 45) | def model_ability(self):
    method convert_to_openai_format (line 48) | def convert_to_openai_format(self, input_data):
    method load (line 86) | def load(self):
    method transcriptions (line 113) | def transcriptions(
    method translations (line 164) | def translations(

FILE: xinference/model/audio/indextts2.py
  class Indextts2 (line 27) | class Indextts2:
    method __init__ (line 28) | def __init__(
    method model_ability (line 45) | def model_ability(self):
    method load (line 48) | def load(self):
    method speech (line 80) | def speech(

FILE: xinference/model/audio/kokoro.py
  class KokoroModel (line 28) | class KokoroModel:
    method __init__ (line 29) | def __init__(
    method model_ability (line 46) | def model_ability(self):
    method load (line 49) | def load(self):
    method speech (line 88) | def speech(

FILE: xinference/model/audio/kokoro_mlx.py
  class KokoroMLXModel (line 28) | class KokoroMLXModel:
    method __init__ (line 29) | def __init__(
    method model_ability (line 46) | def model_ability(self):
    method load (line 49) | def load(self):
    method speech (line 68) | def speech(

FILE: xinference/model/audio/kokoro_zh.py
  class KokoroZHModel (line 30) | class KokoroZHModel:
    method __init__ (line 31) | def __init__(
    method _en_callable (line 48) | def _en_callable(self, text):
    method model_ability (line 60) | def model_ability(self):
    method load (line 63) | def load(self):
    method speech (line 89) | def speech(

FILE: xinference/model/audio/megatts.py
  class MegaTTSModel (line 25) | class MegaTTSModel:
    method __init__ (line 26) | def __init__(
    method model_ability (line 44) | def model_ability(self):
    method load (line 47) | def load(self):
    method speech (line 62) | def speech(

FILE: xinference/model/audio/melotts.py
  class MeloTTSModel (line 26) | class MeloTTSModel:
    method __init__ (line 27) | def __init__(
    method model_ability (line 44) | def model_ability(self):
    method load (line 47) | def load(self):
    method speech (line 76) | def speech(

FILE: xinference/model/audio/qwen3_asr.py
  class Qwen3ASRModel (line 31) | class Qwen3ASRModel:
    method __init__ (line 32) | def __init__(
    method model_ability (line 49) | def model_ability(self):
    method load (line 52) | def load(self):
    method _extract_text_and_language (line 87) | def _extract_text_and_language(self, result) -> Tuple[str, Optional[st...
    method transcriptions (line 105) | def transcriptions(
    method translations (line 145) | def translations(

FILE: xinference/model/audio/tests/test_chattts.py
  function test_chattts (line 21) | def test_chattts(setup):

FILE: xinference/model/audio/tests/test_cosyvoice.py
  function test_cosyvoice_sft (line 23) | def test_cosyvoice_sft(setup, model_name):
  function test_cosyvoice (line 76) | def test_cosyvoice(setup, model_name):
  function test_cosyvoice_instruct (line 127) | def test_cosyvoice_instruct(setup, model_name):

FILE: xinference/model/audio/tests/test_f5tts.py
  function test_f5tts (line 19) | def test_f5tts(setup):

FILE: xinference/model/audio/tests/test_f5tts_mlx.py
  function test_f5tts_mlx (line 19) | def test_f5tts_mlx(setup):

FILE: xinference/model/audio/tests/test_fish_speech.py
  function test_fish_speech (line 19) | def test_fish_speech(setup):

FILE: xinference/model/audio/tests/test_funasr.py
  function test_restful_api_for_funasr (line 20) | def test_restful_api_for_funasr(setup):
  function test_verbose_for_funasr (line 55) | def test_verbose_for_funasr(setup):

FILE: xinference/model/audio/tests/test_kokoro.py
  function test_kokoro (line 19) | def test_kokoro(setup):
  function test_kokoro_zh (line 55) | def test_kokoro_zh(setup):

FILE: xinference/model/audio/tests/test_megatts.py
  function test_megatts (line 18) | def test_megatts(setup):

FILE: xinference/model/audio/tests/test_melotts.py
  function test_melotts (line 19) | def test_melotts(setup):

FILE: xinference/model/audio/tests/test_whisper.py
  function test_restful_api_for_whisper (line 22) | def test_restful_api_for_whisper(setup):
  function test_transcriptions_for_whisper (line 82) | def test_transcriptions_for_whisper(setup):
  function test_register_custom_audio (line 141) | def test_register_custom_audio():
  function test_persistent_custom_audio (line 181) | def test_persistent_custom_audio():

FILE: xinference/model/audio/tests/test_whisper_mlx.py
  function test_restful_api_for_whisper (line 26) | def test_restful_api_for_whisper(setup):
  function test_transcriptions_for_whisper (line 80) | def test_transcriptions_for_whisper(setup):

FILE: xinference/model/audio/utils.py
  function _extract_pcm_from_wav_bytes (line 28) | def _extract_pcm_from_wav_bytes(wav_bytes):
  function ensure_sample_rate (line 35) | def ensure_sample_rate(
  function audio_stream_generator (line 62) | def audio_stream_generator(
  function audio_to_bytes (line 108) | def audio_to_bytes(response_format: str, sample_rate: int, tensor: "torc...

FILE: xinference/model/audio/whisper.py
  class WhisperModelConfig (line 34) | class WhisperModelConfig(TypedDict, total=False):
  class WhisperModel (line 41) | class WhisperModel:
    method __init__ (line 42) | def __init__(
    method _sanitize_model_config (line 62) | def _sanitize_model_config(
    method model_ability (line 74) | def model_ability(self):
    method load (line 77) | def load(self):
    method _call_model (line 112) | def _call_model(
    method transcriptions (line 201) | def transcriptions(
    method translations (line 226) | def translations(

FILE: xinference/model/audio/whisper_mlx.py
  class WhisperMLXModel (line 26) | class WhisperMLXModel:
    method __init__ (line 27) | def __init__(
    method model_ability (line 45) | def model_ability(self):
    method load (line 48) | def load(self):
    method transcriptions (line 93) | def transcriptions(
    method translations (line 112) | def translations(
    method _call (line 135) | def _call(

FILE: xinference/model/batch.py
  class BatchMixin (line 27) | class BatchMixin:
    method __init__ (line 32) | def __init__(self, func: _ExtensibleWrapper, **kwargs):
    method queue (line 48) | def queue(self):
    method _ensure_process_batch_running (line 53) | def _ensure_process_batch_running(self):
    method _get_batch_size (line 61) | def _get_batch_size(self, *args, **kwargs) -> int:
    method _process_batch (line 64) | async def _process_batch(self):
    method _wrap_method (line 107) | def _wrap_method(self):

FILE: xinference/model/cache_manager.py
  class CacheManager (line 12) | class CacheManager:
    method __init__ (line 15) | def __init__(self, model_family: "CacheableModelSpec"):
    method get_cache_dir (line 29) | def get_cache_dir(self):
    method get_cache_status (line 32) | def get_cache_status(self):
    method _cache_from_uri (line 36) | def _cache_from_uri(self, model_spec: "CacheableModelSpec") -> str:
    method _cache (line 60) | def _cache(self) -> str:
    method cache (line 117) | def cache(self) -> str:
    method register_custom_model (line 120) | def register_custom_model(self, model_type: str):
    method unregister_custom_model (line 130) | def unregister_custom_model(self, model_type: str):

FILE: xinference/model/core.py
  function create_model_instance (line 20) | def create_model_instance(
  class CacheableModelSpec (line 126) | class CacheableModelSpec(BaseModel):
  class VirtualEnvSettings (line 134) | class VirtualEnvSettings(BaseModel):

FILE: xinference/model/custom.py
  class ModelRegistry (line 29) | class ModelRegistry:
    method __init__ (line 32) | def __init__(self) -> None:
    method find_model (line 37) | def find_model(self, model_name: str):
    method get_custom_models (line 45) | def get_custom_models(self):
    method check_model_uri (line 49) | def check_model_uri(self, model_spec: "CacheableModelSpec"):
    method add_ud_model (line 56) | def add_ud_model(self, model_spec):
    method register (line 59) | def register(self, model_spec: "CacheableModelSpec", persist: bool):
    method remove_ud_model (line 83) | def remove_ud_model(self, model_spec):
    method remove_ud_model_files (line 86) | def remove_ud_model_files(self, model_spec):
    method unregister (line 92) | def unregister(
  class RegistryManager (line 110) | class RegistryManager:
    method get_registry (line 114) | def get_registry(cls, model_type: str) -> ModelRegistry:
  function migrate_from_v1_to_v2 (line 140) | def migrate_from_v1_to_v2(model_type: str, model_spec_cls: Type):

FILE: xinference/model/embedding/__init__.py
  function register_builtin_model (line 45) | def register_builtin_model():
  function register_custom_model (line 50) | def register_custom_model():
  function check_format_with_engine (line 72) | def check_format_with_engine(model_format, engine):
  function generate_engine_config_by_model_name (line 80) | def generate_engine_config_by_model_name(model_family: "EmbeddingModelFa...
  function has_downloaded_models (line 118) | def has_downloaded_models():
  function load_downloaded_models (line 127) | def load_downloaded_models():
  function load_model_family_from_json (line 144) | def load_model_family_from_json(json_filename, target_families):
  function load_downloaded_models_to_dict (line 168) | def load_downloaded_models_to_dict(target_dict):
  function _install (line 184) | def _install():

FILE: xinference/model/embedding/cache_manager.py
  class EmbeddingCacheManager (line 10) | class EmbeddingCacheManager(CacheManager):
    method __init__ (line 11) | def __init__(self, model_family: "EmbeddingModelFamilyV2"):
    method cache (line 25) | def cache(self) -> str:

FILE: xinference/model/embedding/core.py
  function get_embedding_model_descriptions (line 49) | def get_embedding_model_descriptions():
  class TransformersEmbeddingSpecV1 (line 55) | class TransformersEmbeddingSpecV1(BaseModel):
  class LlamaCppEmbeddingSpecV1 (line 64) | class LlamaCppEmbeddingSpecV1(BaseModel):
  class EmbeddingModelFamilyV2 (line 83) | class EmbeddingModelFamilyV2(BaseModel, ModelInstanceInfoMixin):
    class Config (line 93) | class Config:
    method to_description (line 96) | def to_description(self):
    method to_version_info (line 111) | def to_version_info(self):
  function get_model_version (line 125) | def get_model_version(embedding_model: EmbeddingModelFamilyV2) -> str:
  function generate_embedding_description (line 130) | def generate_embedding_description(
  class EmbeddingModel (line 142) | class EmbeddingModel(abc.ABC):
    method __init__ (line 143) | def __init__(
    method check_lib (line 166) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 171) | def match_json(
    method match (line 180) | def match(
    method load (line 196) | def load(self):
    method _fix_langchain_openai_inputs (line 201) | def _fix_langchain_openai_inputs(
    method _text_length (line 239) | def _text_length(text):
    method _create_embedding (line 250) | def _create_embedding(
    method create_embedding (line 271) | def create_embedding(
    method create_embedding (line 279) | def create_embedding(self, args_list, kwargs_list):
    method _extract_sentences_kwargs (line 329) | def _extract_sentences_kwargs(self, args, kwargs):
    method _get_batch_size (line 363) | def _get_batch_size(self, *args, **kwargs) -> int:
    method convert_ids_to_tokens (line 370) | def convert_ids_to_tokens(
    method _clean_cache_if_needed (line 400) | def _clean_cache_if_needed(self, all_token_nums: int):
  function create_embedding_model_instance (line 416) | def create_embedding_model_instance(

FILE: xinference/model/embedding/custom.py
  class CustomEmbeddingModelFamilyV2 (line 25) | class CustomEmbeddingModelFamilyV2(EmbeddingModelFamilyV2):
  class EmbeddingModelRegistry (line 32) | class EmbeddingModelRegistry(ModelRegistry):
    method __init__ (line 35) | def __init__(self):
    method add_ud_model (line 42) | def add_ud_model(self, model_spec):
    method check_model_uri (line 48) | def check_model_uri(self, model_family: "EmbeddingModelFamilyV2"):
    method remove_ud_model (line 56) | def remove_ud_model(self, model_family: "CustomEmbeddingModelFamilyV2"):
    method remove_ud_model_files (line 62) | def remove_ud_model_files(self, model_family: "CustomEmbeddingModelFam...
  function get_user_defined_embeddings (line 72) | def get_user_defined_embeddings() -> List[EmbeddingModelFamilyV2]:
  function register_embedding (line 79) | def register_embedding(model_family: CustomEmbeddingModelFamilyV2, persi...
  function unregister_embedding (line 86) | def unregister_embedding(model_name: str, raise_error: bool = True):

FILE: xinference/model/embedding/embed_family.py
  function match_embedding (line 31) | def match_embedding(
  function check_engine_by_model_name_and_engine (line 118) | def check_engine_by_model_name_and_engine(
  function check_engine_by_model_name_and_engine_with_virtual_env (line 147) | def check_engine_by_model_name_and_engine_with_virtual_env(

FILE: xinference/model/embedding/flag/core.py
  class FlagEmbeddingModel (line 40) | class FlagEmbeddingModel(EmbeddingModel, BatchMixin):
    method __init__ (line 41) | def __init__(
    method load (line 57) | def load(self):
    method _create_embedding (line 105) | def _create_embedding(
    method check_lib (line 285) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 292) | def match_json(

FILE: xinference/model/embedding/flag/tests/test_flag.py
  function test_embedding_model_with_flag (line 42) | async def test_embedding_model_with_flag():

FILE: xinference/model/embedding/llama_cpp/core.py
  class _Done (line 34) | class _Done:
  class _Error (line 38) | class _Error:
    method __init__ (line 39) | def __init__(self, msg):
  class XllamaCppEmbeddingModel (line 43) | class XllamaCppEmbeddingModel(EmbeddingModel, BatchMixin):
    method __init__ (line 44) | def __init__(self, *args, **kwargs) -> None:
    method _sanitize_model_config (line 52) | def _sanitize_model_config(self, llamacpp_model_config: Optional[dict]...
    method _is_darwin_and_apple_silicon (line 67) | def _is_darwin_and_apple_silicon(self):
    method _is_linux (line 70) | def _is_linux(self):
    method load (line 73) | def load(self):
    method _create_embedding (line 197) | def _create_embedding(
    method check_lib (line 232) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 239) | def match_json(

FILE: xinference/model/embedding/llama_cpp/tests/test_llama_cpp.py
  function test_embedding_model_with_xllamacpp (line 42) | async def test_embedding_model_with_xllamacpp():

FILE: xinference/model/embedding/sentence_transformers/core.py
  class SentenceTransformerEmbeddingModel (line 32) | class SentenceTransformerEmbeddingModel(EmbeddingModel, BatchMixin):
    method __init__ (line 33) | def __init__(self, *args, **kwargs) -> None:
    method load (line 38) | def load(self):
    method _create_embedding (line 158) | def _create_embedding(
    method _normalize_vl_inputs (line 455) | def _normalize_vl_inputs(
    method _create_qwen3_vl_embedding (line 471) | def _create_qwen3_vl_embedding(self, sentences, **kwargs):
    method check_lib (line 503) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 507) | def match_json(

FILE: xinference/model/embedding/sentence_transformers/tests/test_sentence_transformers.py
  function test_embedding_model_with_sentence_transformer (line 42) | async def test_embedding_model_with_sentence_transformer():
  function test_embedding_model_with_sentence_transformer_truncate_dim (line 81) | async def test_embedding_model_with_sentence_transformer_truncate_dim():

FILE: xinference/model/embedding/tests/test_embedding_models.py
  function test_engine_supported (line 76) | def test_engine_supported():
  function test_model_from_modelscope (line 83) | async def test_model_from_modelscope():
  function test_get_cache_status (line 103) | def test_get_cache_status():
  function test_from_local_uri (line 115) | def test_from_local_uri():
  function test_register_custom_embedding (line 143) | def test_register_custom_embedding():
  function test_register_fault_embedding (line 224) | def test_register_fault_embedding():
  function test_convert_ids_to_tokens (line 304) | def test_convert_ids_to_tokens():

FILE: xinference/model/embedding/tests/test_integrated_embedding.py
  function test_sparse_embedding (line 25) | def test_sparse_embedding(setup):
  function test_clip_embedding (line 48) | def test_clip_embedding(setup):
  function test_llama_cpp_embedding (line 83) | def test_llama_cpp_embedding(setup):

FILE: xinference/model/embedding/tests/test_qwen3_vl_engine_params.py
  function _assert_engine_params (line 35) | def _assert_engine_params(params, engine_name):
  function test_qwen3_vl_embedding_engine_params_with_virtualenv (line 44) | def test_qwen3_vl_embedding_engine_params_with_virtualenv():
  function _get_cached_model_path (line 54) | def _get_cached_model_path():
  function _get_virtualenv_site_packages (line 60) | def _get_virtualenv_site_packages(env_path: str) -> str:
  function _purge_modules (line 67) | def _purge_modules(prefixes):
  function _prepare_engine_virtualenv (line 73) | def _prepare_engine_virtualenv(engine_name: str, virtual_env_packages=No...
  function test_qwen3_vl_embedding_sentence_transformers_startup_virtualenv (line 120) | def test_qwen3_vl_embedding_sentence_transformers_startup_virtualenv():

FILE: xinference/model/embedding/vllm/core.py
  class VLLMEmbeddingModel (line 30) | class VLLMEmbeddingModel(EmbeddingModel, BatchMixin):
    method __init__ (line 31) | def __init__(self, *args, **kwargs):
    method load (line 36) | def load(self):
    method _get_detailed_instruct (line 85) | def _get_detailed_instruct(task_description: str, query: str) -> str:
    method _create_embedding (line 88) | def _create_embedding(
    method _embed_vl (line 172) | def _embed_vl(
    method check_lib (line 229) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 236) | def match_json(
    method wait_for_load (line 262) | def wait_for_load(self):
    method _set_context_length (line 266) | def _set_context_length(self):

FILE: xinference/model/embedding/vllm/tests/test_vllm_embedding.py
  function test_embedding_model_with_vllm (line 46) | async def test_embedding_model_with_vllm():
  function test_embedding_model_with_vllm_long_text (line 86) | async def test_embedding_model_with_vllm_long_text():
  function test_change_dim (line 161) | def test_change_dim(setup):

FILE: xinference/model/flexible/__init__.py
  function register_custom_model (line 37) | def register_custom_model():
  function _install (line 54) | def _install():

FILE: xinference/model/flexible/core.py
  class FlexibleModelSpec (line 28) | class FlexibleModelSpec(CacheableModelSpec, ModelInstanceInfoMixin):
    method parser_args (line 37) | def parser_args(self):
    class Config (line 40) | class Config:
    method to_description (line 43) | def to_description(self):
    method to_version_info (line 53) | def to_version_info(self):
  function generate_flexible_model_description (line 63) | def generate_flexible_model_description(
  function get_flexible_model_descriptions (line 75) | def get_flexible_model_descriptions():
  class FlexibleModel (line 81) | class FlexibleModel:
    method __init__ (line 82) | def __init__(
    method load (line 96) | def load(self):
    method infer (line 101) | def infer(self, *args, **kwargs):
    method model_uid (line 108) | def model_uid(self):
    method model_path (line 112) | def model_path(self):
    method device (line 116) | def device(self):
    method config (line 120) | def config(self):
  function match_flexible_model (line 124) | def match_flexible_model(model_name):
  function create_flexible_model_instance (line 133) | def create_flexible_model_instance(

FILE: xinference/model/flexible/custom.py
  class FlexibleModelRegistry (line 9) | class FlexibleModelRegistry(ModelRegistry):
    method __init__ (line 12) | def __init__(self):
    method register (line 19) | def register(self, model_spec: "FlexibleModelSpec", persist: bool):
  function get_flexible_models (line 51) | def get_flexible_models():
  function register_flexible_model (line 58) | def register_flexible_model(model_spec: "FlexibleModelSpec", persist: bo...
  function unregister_flexible_model (line 65) | def unregister_flexible_model(model_name: str, raise_error: bool = True):

FILE: xinference/model/flexible/launchers/image_process_launcher.py
  class ImageRemoveBackgroundModel (line 25) | class ImageRemoveBackgroundModel(FlexibleModel):
    method infer (line 26) | def infer(self, *args, **kwargs):
  function launcher (line 58) | def launcher(model_uid: str, model_spec: FlexibleModelSpec, **kwargs) ->...

FILE: xinference/model/flexible/launchers/modelscope_launcher.py
  class ModelScopePipelineModel (line 18) | class ModelScopePipelineModel(FlexibleModel):
    method load (line 19) | def load(self):
    method infer (line 32) | def infer(self, *args, **kwargs):
  function launcher (line 36) | def launcher(model_uid: str, model_spec: FlexibleModelSpec, **kwargs) ->...

FILE: xinference/model/flexible/launchers/transformers_launcher.py
  class MockModel (line 20) | class MockModel(FlexibleModel):
    method infer (line 21) | def infer(self, *args, **kwargs):
  class AutoModel (line 25) | class AutoModel(FlexibleModel):
    method load (line 26) | def load(self):
    method infer (line 30) | def infer(self, *args, **kwargs):
  class TransformersTextClassificationModel (line 34) | class TransformersTextClassificationModel(FlexibleModel):
    method load (line 35) | def load(self):
    method infer (line 40) | def infer(self, *args, **kwargs):
  function launcher (line 44) | def launcher(model_uid: str, model_spec: FlexibleModelSpec, **kwargs) ->...

FILE: xinference/model/flexible/launchers/yolo_launcher.py
  class UltralyticsModel (line 25) | class UltralyticsModel(FlexibleModel):
    method load (line 26) | def load(self):
    method infer (line 35) | def infer(self, *args, **kwargs):
  function launcher (line 53) | def launcher(model_uid: str, model_spec: FlexibleModelSpec, **kwargs) ->...

FILE: xinference/model/flexible/tests/test_flexible_models.py
  function test_register_flexible_model (line 20) | def test_register_flexible_model():
  function test_model (line 39) | def test_model():

FILE: xinference/model/flexible/utils.py
  function get_launcher (line 18) | def get_launcher(launcher_name: str):

FILE: xinference/model/image/__init__.py
  function register_custom_model (line 43) | def register_custom_model():
  function _install (line 65) | def _install():
  function register_builtin_model (line 102) | def register_builtin_model():
  function has_downloaded_models (line 107) | def has_downloaded_models():
  function load_downloaded_models (line 114) | def load_downloaded_models():
  function load_model_family_from_json (line 129) | def load_model_family_from_json(json_filename, target_families):

FILE: xinference/model/image/cache_manager.py
  class ImageCacheManager (line 7) | class ImageCacheManager(CacheManager):
    method __init__ (line 8) | def __init__(self, model_family):
    method cache_gguf (line 24) | def cache_gguf(self, quantization: Optional[str] = None):
    method cache_lightning (line 80) | def cache_lightning(self, lightning_version: Optional[str] = None):

FILE: xinference/model/image/core.py
  function get_image_model_descriptions (line 32) | def get_image_model_descriptions():
  class ImageModelFamilyV2 (line 38) | class ImageModelFamilyV2(CacheableModelSpec, ModelInstanceInfoMixin):
    class Config (line 58) | class Config:
    method to_description (line 61) | def to_description(self):
    method to_version_info (line 77) | def to_version_info(self):
  function generate_image_description (line 106) | def generate_image_description(
  function match_diffusion (line 114) | def match_diffusion(
  function create_ocr_model_instance (line 155) | def create_ocr_model_instance(
  function create_image_model_instance (line 206) | def create_image_model_instance(
  function _select_ocr_model_family (line 341) | def _select_ocr_model_family(

FILE: xinference/model/image/custom.py
  class CustomImageModelFamilyV2 (line 24) | class CustomImageModelFamilyV2(ImageModelFamilyV2):
  class ImageModelRegistry (line 35) | class ImageModelRegistry(ModelRegistry):
    method __init__ (line 38) | def __init__(self):
  function get_user_defined_images (line 46) | def get_user_defined_images() -> List[ImageModelFamilyV2]:
  function register_image (line 53) | def register_image(model_spec: CustomImageModelFamilyV2, persist: bool):
  function unregister_image (line 68) | def unregister_image(model_name: str, raise_error: bool = True):

FILE: xinference/model/image/engine.py
  class DiffusersImageModel (line 24) | class DiffusersImageModel(DiffusionModel, ImageEngineModel):
    method match (line 30) | def match(cls, model_family: "ImageModelFamilyV2") -> bool:
  class VLLMImageModel (line 34) | class VLLMImageModel(ImageEngineModel):
    method match (line 36) | def match(cls, model_family: "ImageModelFamilyV2") -> bool:
    method check_lib (line 41) | def check_lib(cls):
  class SGLangImageModel (line 48) | class SGLangImageModel(ImageEngineModel):
    method match (line 50) | def match(cls, model_family: "ImageModelFamilyV2") -> bool:
    method check_lib (line 55) | def check_lib(cls):
  function register_builtin_image_engines (line 62) | def register_builtin_image_engines() -> None:

FILE: xinference/model/image/engine_family.py
  class ImageEngineModel (line 25) | class ImageEngineModel:
    method __init__ (line 28) | def __init__(self, *args: Any, **kwargs: Any) -> None:
    method match (line 32) | def match(cls, model_family: "ImageModelFamilyV2") -> bool:
    method check_lib (line 36) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
  function check_engine_by_model_name_and_engine (line 48) | def check_engine_by_model_name_and_engine(
  function check_engine_by_model_name_and_engine_with_virtual_env (line 81) | def check_engine_by_model_name_and_engine_with_virtual_env(
  function generate_engine_config_by_model_name (line 134) | def generate_engine_config_by_model_name(model_family: "ImageModelFamily...

FILE: xinference/model/image/ocr/__init__.py
  function register_builtin_ocr_engines (line 36) | def register_builtin_ocr_engines() -> None:

FILE: xinference/model/image/ocr/deepseek_ocr.py
  class DeepSeekOCRModelSize (line 35) | class DeepSeekOCRModelSize:
    method __init__ (line 44) | def __init__(self, size_type: str):
    method from_string (line 64) | def from_string(cls, size_str: str) -> "DeepSeekOCRModelSize":
    method __str__ (line 68) | def __str__(self) -> str:
  function load_image (line 72) | def load_image(image_path: str) -> Optional[PIL.Image.Image]:
  function find_closest_aspect_ratio (line 87) | def find_closest_aspect_ratio(
  function dynamic_preprocess (line 112) | def dynamic_preprocess(
  function normalize_transform (line 166) | def normalize_transform(
  class BasicImageTransform (line 183) | class BasicImageTransform:
    method __init__ (line 186) | def __init__(
    method __call__ (line 206) | def __call__(self, x: PIL.Image.Image) -> torch.Tensor:
  function re_match (line 210) | def re_match(text: str) -> Tuple[List[Tuple], List[str], List[str]]:
  function extract_coordinates_and_label (line 225) | def extract_coordinates_and_label(
  function draw_bounding_boxes (line 239) | def draw_bounding_boxes(
  function process_image_with_refs (line 339) | def process_image_with_refs(
  function clean_ocr_annotations (line 347) | def clean_ocr_annotations(text: str) -> str:
  function extract_text_blocks (line 374) | def extract_text_blocks(text: str) -> List[Dict[str, Any]]:
  class DeepSeekOCRModel (line 416) | class DeepSeekOCRModel(OCRModel):
    method match (line 420) | def match(cls, model_family: "ImageModelFamilyV2") -> bool:
    method __init__ (line 424) | def __init__(
    method model_ability (line 445) | def model_ability(self):
    method load (line 448) | def load(self):
    method ocr (line 487) | def ocr(
    method visualize_ocr (line 583) | def visualize_ocr(
    method _visualize_single (line 653) | def _visualize_single(
    method _ocr_single (line 795) | def _ocr_single(
    method infer (line 940) | def infer(

FILE: xinference/model/image/ocr/got_ocr2.py
  class GotOCR2Model (line 28) | class GotOCR2Model(OCRModel):
    method match (line 32) | def match(cls, model_family: "ImageModelFamilyV2") -> bool:
    method __init__ (line 35) | def __init__(
    method model_ability (line 56) | def model_ability(self):
    method load (line 59) | def load(self):
    method ocr (line 75) | def ocr(

FILE: xinference/model/image/ocr/hunyuan_ocr.py
  class HunyuanOCRModel (line 29) | class HunyuanOCRModel(OCRModel):
    method match (line 33) | def match(cls, model_family: "ImageModelFamilyV2") -> bool:
    method __init__ (line 36) | def __init__(
    method model_ability (line 54) | def model_ability(self):
    method _load (line 57) | def _load(self):
    method load (line 82) | def load(self):
    method ocr (line 86) | def ocr(self, image: PIL.Image.Image, prompt: Optional[str] = None, **...

FILE: xinference/model/image/ocr/mlx.py
  class MLXDeepSeekOCRModel (line 30) | class MLXDeepSeekOCRModel(DeepSeekOCRModel):
    method __init__ (line 33) | def __init__(
    method match (line 45) | def match(cls, model_family) -> bool:
    method check_lib (line 50) | def check_lib(cls):
    method load (line 55) | def load(self):
    method ocr (line 245) | def ocr(
    method _ocr_single (line 257) | def _ocr_single(
    method _prepare_inputs (line 278) | def _prepare_inputs(self, image: PIL.Image.Image, prompt: str):
    method _generate_text (line 293) | def _generate_text(self, image: PIL.Image.Image, prompt: str, **kwargs...

FILE: xinference/model/image/ocr/ocr_family.py
  class OCRModel (line 25) | class OCRModel:
    method __init__ (line 28) | def __init__(self, *args: Any, **kwargs: Any) -> None:
    method match (line 32) | def match(cls, model_family: "ImageModelFamilyV2") -> bool:
    method check_lib (line 36) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
  function check_engine_by_model_name_and_engine (line 48) | def check_engine_by_model_name_and_engine(
  function check_engine_by_model_name_and_engine_with_virtual_env (line 77) | def check_engine_by_model_name_and_engine_with_virtual_env(
  function generate_engine_config_by_model_name (line 126) | def generate_engine_config_by_model_name(model_family: "ImageModelFamily...

FILE: xinference/model/image/ocr/paddleocr_vl.py
  class PaddleOCRVLModel (line 29) | class PaddleOCRVLModel(OCRModel):
    method match (line 35) | def match(cls, model_family: "ImageModelFamilyV2") -> bool:
    method __init__ (line 38) | def __init__(
    method model_ability (line 59) | def model_ability(self):
    method load (line 62) | def load(self):
    method ocr (line 99) | def ocr(
    method _process_single (line 173) | def _process_single(

FILE: xinference/model/image/ocr/vllm.py
  function _load_vllm_model (line 29) | def _load_vllm_model(model_path: str, model_kwargs: Dict[str, Any]):
  function _sanitize_vllm_kwargs (line 54) | def _sanitize_vllm_kwargs(kwargs: Dict[str, Any]) -> Dict[str, Any]:
  function _filter_engine_args (line 69) | def _filter_engine_args(model_kwargs: Dict[str, Any]) -> Dict[str, Any]:
  function _build_sampling_params (line 82) | def _build_sampling_params(kwargs: Dict[str, Any]):
  function _extract_text (line 107) | def _extract_text(outputs: List[Any]) -> List[str]:
  function _shutdown_vllm_model (line 117) | def _shutdown_vllm_model(model: Any) -> None:
  class VLLMDeepSeekOCRModel (line 146) | class VLLMDeepSeekOCRModel(DeepSeekOCRModel):
    method load (line 149) | def load(self):
    method stop (line 154) | def stop(self):
    method _prepare_inputs (line 159) | def _prepare_inputs(
    method ocr (line 167) | def ocr(
    method visualize_ocr (line 200) | def visualize_ocr(
  class VLLMGotOCR2Model (line 232) | class VLLMGotOCR2Model(GotOCR2Model):
  class VLLMHunyuanOCRModel (line 236) | class VLLMHunyuanOCRModel(HunyuanOCRModel):
    method load (line 239) | def load(self):
    method stop (line 249) | def stop(self):
    method _build_prompt (line 255) | def _build_prompt(self, image: PIL.Image.Image, prompt: str) -> str:
    method ocr (line 272) | def ocr(
  class VLLMPaddleOCRVLModel (line 307) | class VLLMPaddleOCRVLModel(PaddleOCRVLModel):

FILE: xinference/model/image/scheduler/flux.py
  class Text2ImageRequest (line 36) | class Text2ImageRequest:
    method __init__ (line 37) | def __init__(
    method _set_width_and_height (line 71) | def _set_width_and_height(self):
    method set_generate_kwargs (line 74) | def set_generate_kwargs(self, generate_kwargs: Dict):
    method prompt (line 78) | def prompt(self):
    method n (line 82) | def n(self):
    method size (line 86) | def size(self):
    method response_format (line 90) | def response_format(self):
    method kwargs (line 94) | def kwargs(self):
    method width (line 98) | def width(self):
    method height (line 102) | def height(self):
    method generate_kwargs (line 106) | def generate_kwargs(self):
    method request_id (line 110) | def request_id(self):
  class FluxBatchSchedulerActor (line 114) | class FluxBatchSchedulerActor(xo.StatelessActor):
    method gen_uid (line 116) | def gen_uid(cls, model_uid: str):
    method __init__ (line 119) | def __init__(self):
    method set_model (line 129) | def set_model(self, model):
    method __post_create__ (line 135) | async def __post_create__(self):
    method __pre_destroy__ (line 144) | async def __pre_destroy__(self):
    method add_request (line 154) | async def add_request(self, unique_id: str, future, *args, **kwargs):
    method abort_request (line 163) | async def abort_request(self, req_id: str) -> str:
    method _handle_request (line 178) | def _handle_request(
    method _empty_cache (line 226) | def _empty_cache():
    method step (line 231) | async def step(self):
    method run (line 263) | async def run(self):
  function _cat_tensors (line 275) | def _cat_tensors(infos: List[Dict]) -> Dict:
  function _batch_text_to_image_internal (line 286) | def _batch_text_to_image_internal(
  function _batch_text_to_image (line 511) | def _batch_text_to_image(

FILE: xinference/model/image/sdapi.py
  class SDAPIToDiffusersConverter (line 22) | class SDAPIToDiffusersConverter:
    method convert_to_diffusers (line 56) | def convert_to_diffusers(sd_type: str, params: dict) -> dict:
    method get_available_args (line 72) | def get_available_args(sd_type: str) -> set:
  class SDAPIDiffusionModelMixin (line 78) | class SDAPIDiffusionModelMixin:
    method _check_kwargs (line 80) | def _check_kwargs(sd_type: str, kwargs: dict):
    method txt2img (line 106) | def txt2img(self, **kwargs):
    method _decode_b64_img (line 118) | def _decode_b64_img(img_str: str) -> Image:
    method img2img (line 127) | def img2img(self, **kwargs):

FILE: xinference/model/image/stable_diffusion/core.py
  function model_accept_param (line 70) | def model_accept_param(params: Union[str, List[str]], model: Any) -> bool:
  class DiffusionModel (line 87) | class DiffusionModel(SDAPIDiffusionModelMixin):
    method __init__ (line 88) | def __init__(
    method model_ability (line 130) | def model_ability(self):
    method _is_flux2_model (line 133) | def _is_flux2_model(self) -> bool:
    method _get_pipeline_type (line 140) | def _get_pipeline_type(ability: str) -> type:
    method _get_controlnet_model (line 151) | def _get_controlnet_model(self, name: str, path: str):
    method _get_model (line 162) | def _get_model(
    method _apply_lora (line 221) | def _apply_lora(self):
    method _get_layer_cls (line 234) | def _get_layer_cls(self, layer: str):
    method load (line 242) | def load(self):
    method _should_use_batching (line 378) | def _should_use_batching(self) -> bool:
    method _get_quantize_config (line 384) | def _get_quantize_config(self, method: str, quantization: str, module:...
    method _quantize_text_encoder (line 436) | def _quantize_text_encoder(self, quantize_text_encoder: Optional[str]):
    method _quantize_transformer (line 475) | def _quantize_transformer(self):
    method _quantize_transformer_gguf (line 511) | def _quantize_transformer_gguf(self):
    method _process_lightning (line 527) | def _process_lightning(self, kwargs):
    method _load_to_device (line 564) | def _load_to_device(self, model):
    method get_max_num_images_for_batching (line 593) | def get_max_num_images_for_batching(self):
    method _get_scheduler (line 597) | def _get_scheduler(model: Any, sampler_name: str):
    method _need_set_scheduler (line 681) | def _need_set_scheduler(self, scheduler: Any) -> bool:
    method _reset_when_done (line 693) | def _reset_when_done(self, model: Any, sampler_name: str):
    method _release_after (line 708) | def _release_after():
    method _wrap_deepcache (line 718) | def _wrap_deepcache(self, model: Any):
    method _process_progressor (line 730) | def _process_progressor(kwargs: dict):
    method _call_model (line 749) | def _call_model(
    method _filter_kwargs (line 794) | def _filter_kwargs(cls, model, kwargs: dict):
    method text_to_image (line 805) | async def text_to_image(
    method _ensure_scheduler_started (line 832) | async def _ensure_scheduler_started(self):
    method _gen_config_for_lightning (line 837) | def _gen_config_for_lightning(self, kwargs):
    method _direct_text_to_image (line 849) | async def _direct_text_to_image(
    method abort_request (line 871) | async def abort_request(self, request_id: str) -> str:
    method pad_to_multiple (line 884) | def pad_to_multiple(image, multiple=8):
    method _model_expects_four_channel_input (line 892) | def _model_expects_four_channel_input(model: Any) -> bool:
    method _ensure_four_channel_image (line 898) | def _ensure_four_channel_image(image: Any, model: Any):
    method _ensure_three_channel_image (line 919) | def _ensure_three_channel_image(image: Any):
    method image_to_image (line 932) | def image_to_image(
    method inpainting (line 997) | def inpainting(

FILE: xinference/model/image/stable_diffusion/mlx.py
  function quantization_predicate (line 36) | def quantization_predicate(name: str, m) -> bool:
  function to_latent_size (line 40) | def to_latent_size(image_size: Tuple[int, int]):
  class MLXDiffusionModel (line 54) | class MLXDiffusionModel(SDAPIDiffusionModelMixin):
    method __init__ (line 55) | def __init__(
    method model_ability (line 81) | def model_ability(self):
    method support_model (line 85) | def support_model(model_name: str) -> bool:
    method load (line 88) | def load(self):
    method _apply_lora (line 118) | def _apply_lora(self):
    method _release_after (line 136) | def _release_after():
    method text_to_image (line 145) | def text_to_image(
    method image_to_image (line 217) | def image_to_image(self, **kwargs):
    method inpainting (line 220) | def inpainting(self, **kwargs):

FILE: xinference/model/image/tests/test_got_ocr2.py
  function test_got_ocr2 (line 30) | def test_got_ocr2(setup):

FILE: xinference/model/image/tests/test_stable_diffusion.py
  function test_model (line 48) | async def test_model():
  function test_progressor (line 73) | async def test_progressor():
  function test_restful_api_for_image_with_canny_controlnet (line 117) | def test_restful_api_for_image_with_canny_controlnet(setup):
  function test_restful_api_for_image_with_mlsd_controlnet (line 159) | def test_restful_api_for_image_with_mlsd_controlnet(setup):
  function test_restful_api_abort (line 205) | def test_restful_api_abort(setup, model_name):
  function test_restful_api_for_sd_turbo (line 263) | def test_restful_api_for_sd_turbo(setup, model_name):
  function test_restful_api_for_sd_image2image (line 306) | def test_restful_api_for_sd_image2image(setup):
  function test_restful_api_for_sd_inpainting (line 342) | def test_restful_api_for_sd_inpainting(setup):
  function test_get_cache_status (line 383) | def test_get_cache_status():
  function test_register_custom_image (line 395) | def test_register_custom_image():
  function test_persist_custom_image (line 418) | def test_persist_custom_image():
  function test_launch_custom_image (line 450) | def test_launch_custom_image(setup):
  function test_launch_custom_image_with_controlnet (line 511) | def test_launch_custom_image_with_controlnet(setup):

FILE: xinference/model/image/utils.py
  function get_model_version (line 30) | def get_model_version(
  function _flatten_images (line 40) | def _flatten_images(images):
  function _needs_png (line 52) | def _needs_png(image) -> bool:
  function handle_image_result (line 60) | def handle_image_result(response_format: str, images) -> ImageList:

FILE: xinference/model/llm/__init__.py
  function register_builtin_model (line 51) | def register_builtin_model():
  function check_format_with_engine (line 56) | def check_format_with_engine(model_format, engine):
  function generate_engine_config_by_model_family (line 65) | def generate_engine_config_by_model_family(model_family: "LLMFamilyV2"):
  function register_custom_model (line 122) | def register_custom_model():
  function has_downloaded_models (line 142) | def has_downloaded_models():
  function load_downloaded_models (line 151) | def load_downloaded_models():
  function load_model_family_from_json (line 168) | def load_model_family_from_json(json_filename, target_families):
  function _install (line 220) | def _install():

FILE: xinference/model/llm/cache_manager.py
  class LLMCacheManager (line 28) | class LLMCacheManager(CacheManager):
    method __init__ (line 29) | def __init__(
    method cache_uri (line 51) | def cache_uri(self) -> str:
    method cache_from_huggingface (line 75) | def cache_from_huggingface(self) -> str:
    method cache_from_modelscope (line 148) | def cache_from_modelscope(self) -> str:
    method cache_from_openmind_hub (line 223) | def cache_from_openmind_hub(self) -> str:
    method cache_from_csghub (line 252) | def cache_from_csghub(self) -> str:
    method cache (line 314) | def cache(self) -> str:

FILE: xinference/model/llm/config_parser.py
  function _resolve_config_and_dir (line 8) | def _resolve_config_and_dir(model_path: str) -> Tuple[str, str]:
  function _load_json_file (line 23) | def _load_json_file(path: str) -> Dict[str, Any]:
  function _load_tokenizer_config (line 28) | def _load_tokenizer_config(model_dir: str) -> Optional[Dict[str, Any]]:
  function _load_chat_template_file (line 35) | def _load_chat_template_file(model_dir: str) -> Optional[str]:
  function _get_first_value (line 44) | def _get_first_value(config: Dict[str, Any], *keys: str) -> Optional[Any]:
  function _infer_context_length (line 52) | def _infer_context_length(config: Dict[str, Any]) -> int:
  function _normalize_architectures (line 64) | def _normalize_architectures(config: Dict[str, Any]) -> List[str]:
  function _match_family_by_architectures (line 73) | def _match_family_by_architectures(
  function _load_builtin_families (line 90) | def _load_builtin_families() -> List[Dict[str, Any]]:
  function _infer_languages (line 102) | def _infer_languages(config: Dict[str, Any]) -> List[str]:
  function _format_size_in_billions (line 111) | def _format_size_in_billions(size_in_billions: float) -> Union[int, str]:
  function _extract_numeric_size (line 118) | def _extract_numeric_size(value: Any) -> Optional[float]:
  function _infer_model_size_in_billions (line 143) | def _infer_model_size_in_billions(config: Dict[str, Any]) -> Optional[Un...
  function _infer_quantization (line 206) | def _infer_quantization(config: Dict[str, Any], model_format: str) -> str:
  function _extract_chat_template (line 228) | def _extract_chat_template(tokenizer_config: Optional[Dict[str, Any]]) -...
  function _infer_model_format (line 237) | def _infer_model_format(config: Dict[str, Any]) -> str:
  function build_llm_registration_from_local_config (line 255) | def build_llm_registration_from_local_config(

FILE: xinference/model/llm/core.py
  function get_llm_version_infos (line 41) | def get_llm_version_infos():
  class LLM (line 47) | class LLM(abc.ABC):
    method __init__ (line 50) | def __init__(
    method check_lib (line 73) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method _is_darwin_and_apple_silicon (line 77) | def _is_darwin_and_apple_silicon():
    method _is_linux (line 81) | def _is_linux():
    method _has_cuda_device (line 86) | def _has_cuda_device():
    method _has_mlu_device (line 109) | def _has_mlu_device():
    method _has_vacc_device (line 126) | def _has_vacc_device():
    method _has_musa_device (line 140) | def _has_musa_device():
    method _get_cuda_count (line 166) | def _get_cuda_count():
    method load (line 184) | def load(self):
    method match (line 188) | def match(
    method match_json (line 199) | def match_json(
    method prepare_parse_reasoning_content (line 204) | def prepare_parse_reasoning_content(
    method prepare_parse_tool_calls (line 220) | def prepare_parse_tool_calls(self):
  function generate_llm_version_info (line 240) | def generate_llm_version_info(llm_family: "LLMFamilyV2") -> Dict[str, Li...
  function create_llm_model_instance (line 259) | def create_llm_model_instance(

FILE: xinference/model/llm/custom.py
  class LLMModelRegistry (line 30) | class LLMModelRegistry(ModelRegistry):
    method __init__ (line 33) | def __init__(self):
    method add_ud_model (line 40) | def add_ud_model(self, model_spec):
    method check_model_uri (line 46) | def check_model_uri(self, llm_family: "LLMFamilyV2"):
    method remove_ud_model (line 54) | def remove_ud_model(self, llm_family: "LLMFamilyV2"):
    method remove_ud_model_files (line 60) | def remove_ud_model_files(self, llm_family: "LLMFamilyV2"):
  function get_user_defined_llm_families (line 70) | def get_user_defined_llm_families():
  function register_llm (line 77) | def register_llm(llm_family: "LLMFamilyV2", persist: bool):
  function unregister_llm (line 84) | def unregister_llm(model_name: str, raise_error: bool = True):

FILE: xinference/model/llm/harmony.py
  class HarmonyStreamParser (line 22) | class HarmonyStreamParser:
    method __init__ (line 23) | def __init__(self):
    method feed (line 29) | def feed(self, text):
  function async_stream_harmony_chat_completion (line 123) | async def async_stream_harmony_chat_completion(

FILE: xinference/model/llm/llama_cpp/core.py
  function _schema_to_grammar (line 33) | def _schema_to_grammar(schema: Dict[str, Any]) -> Optional[str]:
  function _apply_response_format (line 46) | def _apply_response_format(generate_config: Dict[str, Any]) -> None:
  class _Done (line 63) | class _Done:
  class _Error (line 67) | class _Error:
    method __init__ (line 68) | def __init__(self, msg):
  class XllamaCppModel (line 72) | class XllamaCppModel(LLM, ChatModelMixin):
    method __init__ (line 75) | def __init__(
    method _sanitize_model_config (line 87) | def _sanitize_model_config(self, llamacpp_model_config: Optional[dict]...
    method check_lib (line 112) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 119) | def match_json(
    method load (line 131) | def load(self):
    method generate (line 273) | def generate(
    method chat (line 331) | def chat(

FILE: xinference/model/llm/llama_cpp/tests/test_gguf.py
  function test_gguf (line 22) | def test_gguf(setup):
  function surprise_image_base64 (line 42) | def surprise_image_base64():
  function test_gguf_multimodal (line 51) | def test_gguf_multimodal(setup, surprise_image_base64):

FILE: xinference/model/llm/llama_cpp/tests/test_structured.py
  class CarType (line 18) | class CarType(str, Enum):
  class CarDescription (line 25) | class CarDescription(BaseModel):
  function _load_json_from_message (line 31) | def _load_json_from_message(message: Any) -> Dict[str, Any]:
  function test_apply_response_format_sets_grammar (line 65) | def test_apply_response_format_sets_grammar(monkeypatch):
  function test_apply_response_format_handles_conversion_failure (line 89) | def test_apply_response_format_handles_conversion_failure(monkeypatch):
  function test_apply_response_format_ignores_non_schema (line 116) | def test_apply_response_format_ignores_non_schema(monkeypatch):
  function test_apply_response_format_uses_real_xllamacpp_if_available (line 123) | def test_apply_response_format_uses_real_xllamacpp_if_available():
  function test_llamacpp_qwen3_json_schema (line 151) | def test_llamacpp_qwen3_json_schema(setup):

FILE: xinference/model/llm/llm_family.py
  class LlamaCppLLMSpecV2 (line 52) | class LlamaCppLLMSpecV2(BaseModel):
    method validate_model_size_with_radix (line 69) | def validate_model_size_with_radix(cls, v: object) -> object:
  class PytorchLLMSpecV2 (line 80) | class PytorchLLMSpecV2(BaseModel):
    method validate_model_size_with_radix (line 93) | def validate_model_size_with_radix(cls, v: object) -> object:
  class MLXLLMSpecV2 (line 104) | class MLXLLMSpecV2(BaseModel):
    method validate_model_size_with_radix (line 117) | def validate_model_size_with_radix(cls, v: object) -> object:
  class LLMFamilyV2 (line 128) | class LLMFamilyV2(BaseModel, ModelInstanceInfoMixin):
    class Config (line 160) | class Config:
    method _resolve_architectures (line 163) | def _resolve_architectures(self) -> Optional[List[str]]:
    method has_architecture (line 173) | def has_architecture(self, *architectures: str) -> bool:
    method matches_supported_architectures (line 179) | def matches_supported_architectures(
    method to_description (line 187) | def to_description(self):
    method to_version_info (line 207) | def to_version_info(self):
  class CustomLLMFamilyV2 (line 235) | class CustomLLMFamilyV2(LLMFamilyV2):
    method parse_raw (line 237) | def parse_raw(
  function register_transformer (line 347) | def register_transformer(cls):
  function cache_model_tokenizer_and_config (line 365) | def cache_model_tokenizer_and_config(
  function cache_model_config (line 413) | def cache_model_config(llm_family: LLMFamilyV2):
  function _get_cache_dir_for_model_mem (line 438) | def _get_cache_dir_for_model_mem(
  function match_model_size (line 462) | def match_model_size(
  function convert_model_size_to_float (line 481) | def convert_model_size_to_float(
  function match_llm (line 495) | def match_llm(
  function check_engine_by_spec_parameters (line 593) | def check_engine_by_spec_parameters(
  function check_engine_by_spec_parameters_with_virtual_env (line 625) | def check_engine_by_spec_parameters_with_virtual_env(

FILE: xinference/model/llm/lmdeploy/core.py
  class LMDeployModelConfig (line 50) | class LMDeployModelConfig(TypedDict, total=False):
  class LMDeployGenerateConfig (line 68) | class LMDeployGenerateConfig(TypedDict, total=False):
  class LMDeployModel (line 84) | class LMDeployModel(LLM):
    method __init__ (line 87) | def __init__(
    method _sanitize_model_config (line 102) | def _sanitize_model_config(
    method load (line 112) | def load(self):
    method check_lib (line 126) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 133) | def match_json(
    method generate (line 138) | def generate(
  class LMDeployChatModel (line 146) | class LMDeployChatModel(LMDeployModel, ChatModelMixin):
    method load (line 147) | def load(self):
    method match_json (line 185) | def match_json(
    method async_chat (line 202) | async def async_chat(
    method _chat_stream (line 229) | async def _chat_stream(self, messages, include_usage):
    method _chat (line 279) | async def _chat(self, messages) -> ChatCompletion:
    method _generate (line 306) | async def _generate(
    method _get_prompt_input (line 478) | async def _get_prompt_input(

FILE: xinference/model/llm/memory.py
  class ModelLayersInfo (line 43) | class ModelLayersInfo:
  class ModelMemInfo (line 52) | class ModelMemInfo:
  function estimate_llm_gpu_memory (line 100) | def estimate_llm_gpu_memory(
  function estimate_llm_gpu_memory_details (line 130) | def estimate_llm_gpu_memory_details(
  function _load_item_from_json (line 182) | def _load_item_from_json(config_data: Any, *keys: str) -> str:
  function load_model_config_json (line 191) | def load_model_config_json(config_path: str) -> ModelLayersInfo:
  function get_model_layers_info (line 213) | def get_model_layers_info(
  function _get_default_layers_from_size (line 238) | def _get_default_layers_from_size(size_in_billion: float) -> ModelLayers...
  function _convert_to_mb_model_size (line 275) | def _convert_to_mb_model_size(model_size: float, quantization: Optional[...
  function _compute_inference_only_activation_memory (line 289) | def _compute_inference_only_activation_memory(
  function _compute_model_size_gguf (line 302) | def _compute_model_size_gguf(info: ModelLayersInfo, quantization: str) -...

FILE: xinference/model/llm/mlx/core.py
  class MLXBatchModel (line 64) | class MLXBatchModel:
    method __init__ (line 78) | def __init__(
    method _get_lock (line 92) | def _get_lock() -> asyncio.Lock:
    method _get_or_create_generator (line 98) | def _get_or_create_generator(self, temperature: float, top_p: float):
    method _ensure_background_worker (line 132) | def _ensure_background_worker(self, gen_dict):
    method _background_worker (line 138) | async def _background_worker(self, gen_dict):
    method generate_stream (line 188) | async def generate_stream(
    method generate (line 332) | async def generate(
  class MLXModelConfig (line 357) | class MLXModelConfig(TypedDict, total=False):
  class MLXGenerateConfig (line 371) | class MLXGenerateConfig(TypedDict, total=False):
  class PromptCache (line 387) | class PromptCache:
  function get_context_length (line 393) | def get_context_length(config: dict) -> int:
  class MLXModel (line 410) | class MLXModel(LLM):
    method __init__ (line 414) | def __init__(
    method set_loop (line 444) | def set_loop(self, loop: asyncio.AbstractEventLoop):
    method _cleanup_memory (line 449) | def _cleanup_memory(self):
    method driver_info (line 460) | def driver_info(self) -> Optional[dict]:
    method set_shard_info (line 463) | def set_shard_info(self, shard: int, address: str):
    method get_rank_addresses (line 471) | async def get_rank_addresses(self) -> Optional[Dict[int, str]]:
    method _sanitize_model_config (line 475) | def _sanitize_model_config(
    method _sanitize_generate_config (line 485) | def _sanitize_generate_config(
    method _load_model (line 505) | def _load_model(self, **kwargs):
    method _load_model_shard (line 552) | def _load_model_shard(self, **kwargs):
    method _get_classes (line 617) | def _get_classes(config: dict):
    method load (line 643) | def load(self):
    method wait_for_load (line 695) | def wait_for_load(self):
    method check_lib (line 733) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 740) | def match_json(
    method _get_prompt_cache (line 753) | def _get_prompt_cache(
    method _generate_stream_inner (line 778) | def _generate_stream_inner(self, **kwargs):
    method _prepare_inputs (line 812) | def _prepare_inputs(
    method _generate_stream (line 821) | def _generate_stream(
    method _run_non_drivers (line 928) | def _run_non_drivers(
    method async_generate (line 959) | async def async_generate(
  class MLXChatModel (line 1106) | class MLXChatModel(MLXModel, ChatModelMixin):
    method _sanitize_generate_config (line 1109) | def _sanitize_generate_config(
    method match_json (line 1125) | def match_json(
    method async_chat (line 1138) | async def async_chat(
  class MLXVisionModel (line 1193) | class MLXVisionModel(MLXModel, ChatModelMixin):
    method check_lib (line 1197) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 1204) | def match_json(
    method generate (line 1215) | def generate(
    method wait_for_load (line 1270) | def wait_for_load(self):
    method _load_model (line 1278) | def _load_model(self, **kwargs):
    method load (line 1294) | def load(self):
    method _generate_stream_inner (line 1317) | def _generate_stream_inner(self, **kwargs):
    method _prepare_inputs (line 1380) | def _prepare_inputs(
    method chat (line 1439) | def chat(

FILE: xinference/model/llm/mlx/distributed_models/core.py
  class ReceiverActor (line 32) | class ReceiverActor(xo.StatelessActor):
    method __init__ (line 33) | def __init__(self, *args, **kwargs):
    method gen_uid (line 39) | def gen_uid(cls, uid: str, rank: int):
    method send (line 42) | async def send(self, data: "mx.array"):
    method recv (line 49) | async def recv(self):
  class DistributedModelMixin (line 53) | class DistributedModelMixin:
    method __init__ (line 63) | def __init__(self):
    method prepare (line 73) | def prepare(self):
    method _send_stage_result (line 83) | def _send_stage_result(self, result: "mx.array"):
    method _wait_prev_stage_result (line 109) | def _wait_prev_stage_result(self):
    method _broadcast_result (line 122) | def _broadcast_result(self, result: "mx.array"):
    method _get_result (line 144) | def _get_result(self) -> "mx.array":
    method pipeline (line 154) | def pipeline(self):
  class SafeKVCache (line 167) | class SafeKVCache:
    method __init__ (line 174) | def __init__(self):
    method state (line 180) | def state(self):
    method state (line 193) | def state(self, v):
    method __getattr__ (line 203) | def __getattr__(self, name):

FILE: xinference/model/llm/mlx/distributed_models/deepseek_v3.py
  class DeepseekV3Model (line 27) | class DeepseekV3Model(_DeepseekV3Model, DistributedModelMixin):
    method __init__ (line 28) | def __init__(self, *args, **kwargs):
    method __call__ (line 32) | def __call__(
  class Model (line 69) | class Model(_Model):
    method __init__ (line 70) | def __init__(self, config: ModelArgs):

FILE: xinference/model/llm/mlx/distributed_models/qwen2.py
  class Qwen2Model (line 30) | class Qwen2Model(_Qwen2Model, DistributedModelMixin):
    method __init__ (line 31) | def __init__(self, *args, **kwargs):
    method __call__ (line 35) | def __call__(
  class Model (line 74) | class Model(_Model):
    method __init__ (line 75) | def __init__(self, args: ModelArgs):

FILE: xinference/model/llm/mlx/distributed_models/qwen3.py
  class Qwen3Model (line 30) | class Qwen3Model(_Qwen3Model, DistributedModelMixin):
    method __init__ (line 31) | def __init__(self, *args, **kwargs):
    method __call__ (line 35) | def __call__(
  class Model (line 75) | class Model(_Model):
    method __init__ (line 76) | def __init__(self, args: ModelArgs):

FILE: xinference/model/llm/mlx/distributed_models/qwen3_moe.py
  class Qwen3MoeModel (line 29) | class Qwen3MoeModel(_Qwen3MoeModel, DistributedModelMixin):
    method __init__ (line 30) | def __init__(self, *args, **kwargs):
    method __call__ (line 34) | def __call__(
  class Model (line 70) | class Model(_Model):
    method __init__ (line 71) | def __init__(self, args: ModelArgs):

FILE: xinference/model/llm/mlx/tests/test_distributed_model.py
  class ModelActor (line 25) | class ModelActor(xo.StatelessActor):
    method __init__ (line 26) | def __init__(self, rank: int, model_uid: str, model_path: str):
    method set_rank_addresses (line 33) | def set_rank_addresses(self, rank_addresses):
    method _load (line 36) | def _load(self):
    method load (line 71) | async def load(self):
    method _generate (line 74) | def _generate(self, prompt: str, **kwargs):
    method generate (line 84) | async def generate(self, prompt: str, **kwargs):
  function setup_pool (line 89) | async def setup_pool():
  function test_distributed (line 102) | async def test_distributed(setup_pool):

FILE: xinference/model/llm/mlx/tests/test_mlx.py
  class InferenceThread (line 26) | class InferenceThread(threading.Thread):
    method __init__ (line 29) | def __init__(self, prompt, generate_config, model):
    method run (line 37) | def run(self):
    method join (line 62) | def join(self, timeout=None):
  function test_load_mlx (line 73) | def test_load_mlx(setup):
  function test_load_mlx_vision (line 102) | def test_load_mlx_vision(setup):
  function test_mlx_parallel_inference (line 153) | def test_mlx_parallel_inference(setup):

FILE: xinference/model/llm/reasoning_parser.py
  class ReasoningParser (line 12) | class ReasoningParser:
    method __init__ (line 15) | def __init__(
    method extract_reasoning_content_streaming (line 32) | def extract_reasoning_content_streaming(
    method extract_reasoning_content (line 128) | def extract_reasoning_content(
    method check_content_parser (line 162) | def check_content_parser(self) -> bool:
    method _create_chat_completion_chunk (line 172) | def _create_chat_completion_chunk(
    method _create_completion_chunk (line 200) | def _create_completion_chunk(
    method is_enable_thinking (line 227) | def is_enable_thinking(self):
    method prepare_reasoning_content_streaming (line 233) | async def prepare_reasoning_content_streaming(
    method prepare_reasoning_content_sync (line 302) | def prepare_reasoning_content_sync(self, chunks: Iterator[CompletionCh...
    method prepare_reasoning_content (line 365) | def prepare_reasoning_content(self, completion):
    method prepare_first_reasoning_content_chunk (line 392) | def prepare_first_reasoning_content_chunk(

FILE: xinference/model/llm/sglang/core.py
  class SGLANGModelConfig (line 50) | class SGLANGModelConfig(TypedDict, total=False):
  class SGLANGGenerateConfig (line 66) | class SGLANGGenerateConfig(TypedDict, total=False):
  class SGLANGModel (line 116) | class SGLANGModel(LLM):
    method __init__ (line 119) | def __init__(
    method driver_info (line 137) | def driver_info(self) -> Optional[dict]:
    method load (line 140) | def load(self):
    method wait_for_load (line 241) | def wait_for_load(self):
    method stop (line 252) | def stop(self):
    method _sanitize_model_config (line 256) | def _sanitize_model_config(
    method _apply_fp4_config (line 286) | def _apply_fp4_config(self, model_config: SGLANGModelConfig) -> None:
    method _sanitize_generate_config (line 299) | def _sanitize_generate_config(
    method check_lib (line 334) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 341) | def match_json(
    method _convert_state_to_completion_chunk (line 376) | def _convert_state_to_completion_chunk(
    method _convert_state_to_completion (line 413) | def _convert_state_to_completion(
    method _filter_sampling_params (line 450) | def _filter_sampling_params(cls, sampling_params: dict):
    method _stream_generate (line 455) | async def _stream_generate(
    method _non_stream_generate (line 494) | async def _non_stream_generate(
    method async_generate (line 514) | async def async_generate(
  class SGLANGChatModel (line 657) | class SGLANGChatModel(SGLANGModel, ChatModelMixin):
    method match_json (line 659) | def match_json(
    method _sanitize_chat_config (line 689) | def _sanitize_chat_config(
    method is_tool_call_chunk_start (line 702) | def is_tool_call_chunk_start(chunk):
    method is_tool_call_chunk_end (line 706) | def is_tool_call_chunk_end(chunk):
    method async_chat (line 709) | async def async_chat(
  class SGLANGVisionModel (line 756) | class SGLANGVisionModel(SGLANGModel, ChatModelMixin):
    method match_json (line 758) | def match_json(
    method _sanitize_chat_config (line 794) | def _sanitize_chat_config(
    method async_chat (line 805) | async def async_chat(

FILE: xinference/model/llm/tests/test_harmony.py
  function test_streaming_parser_multiple_texts (line 20) | def test_streaming_parser_multiple_texts():
  function test_harmony_streaming_and_nonstreaming (line 59) | async def test_harmony_streaming_and_nonstreaming():
  function test_async_stream_chunks (line 443) | async def test_async_stream_chunks():

FILE: xinference/model/llm/tests/test_llm_family.py
  function test_deserialize_llm_family_v1 (line 36) | def test_deserialize_llm_family_v1():
  function test_cache_from_huggingface_pytorch (line 114) | def test_cache_from_huggingface_pytorch():
  function test_cache_from_huggingface_gguf (line 142) | def test_cache_from_huggingface_gguf():
  function test_cache_from_uri_local (line 176) | def test_cache_from_uri_local():
  function test_custom_llm (line 209) | def test_custom_llm():
  function test_persistent_custom_llm (line 240) | def test_persistent_custom_llm():
  function test_is_locale_chinese_simplified (line 278) | def test_is_locale_chinese_simplified():
  function test_match_llm (line 292) | def test_match_llm():
  function test_is_valid_file_uri (line 327) | def test_is_valid_file_uri():
  function test_get_cache_status_pytorch (line 333) | def test_get_cache_status_pytorch():
  function test_get_cache_status_gguf (line 371) | def test_get_cache_status_gguf():
  function test_parse_chat_template (line 407) | def test_parse_chat_template():
  function test_match_model_size (line 520) | def test_match_model_size():
  function test_convert_model_size_to_float (line 537) | def test_convert_model_size_to_float():
  function test_quert_engine_vLLM (line 548) | def test_quert_engine_vLLM():
  function test_quert_engine_SGLang (line 608) | def test_quert_engine_SGLang():
  function test_query_engine_general (line 665) | def test_query_engine_general():

FILE: xinference/model/llm/tests/test_llm_model.py
  function test_restful_api_for_deepseek_with_reasoning (line 42) | async def test_restful_api_for_deepseek_with_reasoning(
  function test_restful_api_for_deepseek_without_reasoning (line 103) | async def test_restful_api_for_deepseek_without_reasoning(
  function test_qwen3_with_thinking_params (line 171) | async def test_qwen3_with_thinking_params(
  function test_qwen3_with_tools (line 237) | async def test_qwen3_with_tools(setup):
  function setup_cluster (line 340) | def setup_cluster():
  function test_qwen3_enable_thinking (line 385) | async def test_qwen3_enable_thinking(

FILE: xinference/model/llm/tests/test_memory_estimate.py
  function test_llm_estimate_memory (line 18) | def test_llm_estimate_memory():

FILE: xinference/model/llm/tests/test_multimodal.py
  function test_restful_api_for_qwen_vl (line 24) | def test_restful_api_for_qwen_vl(setup, model_format, quantization):
  function test_restful_api_for_yi_vl (line 136) | def test_restful_api_for_yi_vl(setup, model_format, quantization):
  function test_restful_api_for_deepseek_vl (line 224) | def test_restful_api_for_deepseek_vl(setup, model_format, quantization):
  function test_restful_api_for_qwen_audio (line 326) | def test_restful_api_for_qwen_audio(setup):

FILE: xinference/model/llm/tests/test_stream_options.py
  function test_openai_stream_options_llamacpp_chatglm (line 27) | async def test_openai_stream_options_llamacpp_chatglm(setup):
  function test_openai_stream_options_llamacpp (line 115) | async def test_openai_stream_options_llamacpp(setup):
  function test_openai_stream_options_pytorch_chatglm (line 204) | async def test_openai_stream_options_pytorch_chatglm(setup):
  function test_openai_stream_options_pytorch (line 294) | async def test_openai_stream_options_pytorch(setup):
  function test_openai_stream_options_pytorch_deepseek_vl (line 384) | async def test_openai_stream_options_pytorch_deepseek_vl(setup):
  function test_openai_stream_options_pytorch_internlm2 (line 474) | async def test_openai_stream_options_pytorch_internlm2(setup):
  function test_openai_stream_options_pytorch_qwen_vl (line 564) | async def test_openai_stream_options_pytorch_qwen_vl(setup):
  function test_openai_stream_options_pytorch_yi_vl (line 654) | async def test_openai_stream_options_pytorch_yi_vl(setup):
  function test_openai_stream_options_sgalng (line 744) | async def test_openai_stream_options_sgalng(setup):
  function test_openai_stream_options_vllm (line 834) | async def test_openai_stream_options_vllm(setup):
  function test_openai_stream_tools_vllm (line 924) | async def test_openai_stream_tools_vllm(setup):

FILE: xinference/model/llm/tests/test_utils.py
  function test_is_valid_model_name (line 20) | def test_is_valid_model_name():
  function filter_ids_and_created (line 36) | def filter_ids_and_created(data):
  function test_post_process_completion_chunk_without_thinking (line 48) | def test_post_process_completion_chunk_without_thinking():
  function test_post_process_completion_chunk_with_thinking (line 448) | def test_post_process_completion_chunk_with_thinking():
  function test_post_process_completion_chunk_with_parser (line 1018) | def test_post_process_completion_chunk_with_parser():
  function test_post_process_completion_without_thinking (line 1522) | def test_post_process_completion_without_thinking():
  function test_post_process_completion_with_thinking (line 1580) | def test_post_process_completion_with_thinking():
  function test_post_process_completion_with_parser (line 1636) | def test_post_process_completion_with_parser():

FILE: xinference/model/llm/tool_parsers/__init__.py
  function register_tool_parser (line 8) | def register_tool_parser(name: str):

FILE: xinference/model/llm/tool_parsers/abstract_tool_parser.py
  class ToolParser (line 1) | class ToolParser:
    method extract_tool_calls (line 8) | def extract_tool_calls(self, model_output: str):
    method extract_tool_calls_streaming (line 20) | def extract_tool_calls_streaming(

FILE: xinference/model/llm/tool_parsers/deepseek_r1_tool_parser.py
  class DeepseekR1ToolParser (line 13) | class DeepseekR1ToolParser(ToolParser):
    method __init__ (line 21) | def __init__(self):
    method extract_tool_calls (line 47) | def extract_tool_calls(
    method _get_function_calls (line 137) | def _get_function_calls(self, model_output: str) -> List[str]:
    method extract_tool_calls_streaming (line 164) | def extract_tool_calls_streaming(

FILE: xinference/model/llm/tool_parsers/deepseek_v3_1_tool_parser.py
  class DeepseekV3_1ToolParser (line 13) | class DeepseekV3_1ToolParser(ToolParser):
    method __init__ (line 22) | def __init__(self):
    method extract_tool_calls (line 33) | def extract_tool_calls(
    method extract_tool_calls_streaming (line 104) | def extract_tool_calls_streaming(

FILE: xinference/model/llm/tool_parsers/deepseek_v3_tool_parser.py
  class DeepseekV3ToolParser (line 13) | class DeepseekV3ToolParser(ToolParser):
    method __init__ (line 22) | def __init__(self):
    method _parse_json_function_call (line 30) | def _parse_json_function_call(
    method extract_tool_calls (line 50) | def extract_tool_calls(
    method extract_tool_calls_streaming (line 123) | def extract_tool_calls_streaming(

FILE: xinference/model/llm/tool_parsers/glm4_tool_parser.py
  class Glm4ToolParser (line 12) | class Glm4ToolParser(ToolParser):
    method __init__ (line 21) | def __init__(self):
    method _parse_json_function_call (line 29) | def _parse_json_function_call(
    method extract_tool_calls (line 49) | def extract_tool_calls(
    method extract_tool_calls_streaming (line 94) | def extract_tool_calls_streaming(

FILE: xinference/model/llm/tool_parsers/llama3_tool_parser.py
  class Llama3ToolParser (line 11) | class Llama3ToolParser(ToolParser):
    method __init__ (line 20) | def __init__(self):
    method extract_tool_calls (line 26) | def extract_tool_calls(
    method extract_tool_calls_streaming (line 51) | def extract_tool_calls_streaming(

FILE: xinference/model/llm/tool_parsers/minimax_tool_parser.py
  class MiniMaxToolParser (line 13) | class MiniMaxToolParser(ToolParser):
    method __init__ (line 21) | def __init__(self):
    method _parse_param_value (line 46) | def _parse_param_value(self, value: str) -> Any:
    method _parse_invoke_calls (line 55) | def _parse_invoke_calls(self, tool_block: str) -> List[Tuple[str, Dict...
    method _get_function_calls (line 64) | def _get_function_calls(self, model_output: str) -> List[str]:
    method _get_function_calls_streaming (line 76) | def _get_function_calls_streaming(self, model_output: str) -> List[str]:
    method is_contain_think (line 80) | def is_contain_think(self, model_output: str) -> bool:
    method _has_unclosed_tool_call (line 83) | def _has_unclosed_tool_call(self, text: str) -> bool:
    method extract_tool_calls (line 90) | def extract_tool_calls(
    method extract_tool_calls_streaming (line 124) | def extract_tool_calls_streaming(

FILE: xinference/model/llm/tool_parsers/qwen_tool_parser.py
  class QwenToolParser (line 13) | class QwenToolParser(ToolParser):
    method __init__ (line 22) | def __init__(self):
    method _parse_json_function_call (line 47) | def _parse_json_function_call(
    method _parse_json_function_call_stream (line 85) | def _parse_json_function_call_stream(
    method is_contain_think_end_token (line 105) | def is_contain_think_end_token(self, model_output: str) -> bool:
    method is_contain_think (line 117) | def is_contain_think(self, model_output: str) -> bool:
    method is_contain_tool_call (line 129) | def is_contain_tool_call(self, model_output: str) -> bool:
    method is_contain_tool_call_start_token (line 141) | def is_contain_tool_call_start_token(self, model_output: str) -> bool:
    method is_contain_tool_call_end_token (line 153) | def is_contain_tool_call_end_token(self, model_output: str) -> bool:
    method _get_function_calls (line 165) | def _get_function_calls(self, model_output: str) -> List[str]:
    method _get_function_calls_streaming (line 192) | def _get_function_calls_streaming(self, model_output: str) -> List[str]:
    method extract_tool_calls (line 207) | def extract_tool_calls(
    method _has_unclosed_tool_call (line 275) | def _has_unclosed_tool_call(self, text: str) -> bool:
    method extract_tool_calls_streaming (line 294) | def extract_tool_calls_streaming(

FILE: xinference/model/llm/tool_parsers/tests/test_deepseek_r1_tool_parser.py
  function test_tool_parser_extract_calls_without_thinking (line 4) | def test_tool_parser_extract_calls_without_thinking():

FILE: xinference/model/llm/tool_parsers/tests/test_deepseek_v3_1_tool_parser.py
  function test_extract_tool_calls_single_call (line 4) | def test_extract_tool_calls_single_call():
  function test_extract_tool_calls_multiple_calls (line 17) | def test_extract_tool_calls_multiple_calls():
  function test_extract_tool_calls_no_tool_call (line 36) | def test_extract_tool_calls_no_tool_call():
  function test_extract_tool_calls_streaming_full_sequence (line 217) | def test_extract_tool_calls_streaming_full_sequence():
  function test_extract_tool_calls_streaming_multi_sequence (line 242) | def test_extract_tool_calls_streaming_multi_sequence():
  function test_extract_tool_calls_streaming_split_token (line 268) | def test_extract_tool_calls_streaming_split_token():
  function test_extract_tool_calls_invalid_json (line 301) | def test_extract_tool_calls_invalid_json():

FILE: xinference/model/llm/tool_parsers/tests/test_deepseek_v3_tool_parser.py
  function test_tool_parser_extract_calls_without_thinking (line 4) | def test_tool_parser_extract_calls_without_thinking():

FILE: xinference/model/llm/tool_parsers/tests/test_glm4_tool_parser.py
  function test_tool_parser_extract_calls (line 4) | def test_tool_parser_extract_calls():
  function test_tool_parser_extract_calls_streaming (line 16) | def test_tool_parser_extract_calls_streaming():

FILE: xinference/model/llm/tool_parsers/tests/test_qwen_tool_parser.py
  function test_tool_parser_extract_calls_streaming_without_thinking_multi (line 4) | def test_tool_parser_extract_calls_streaming_without_thinking_multi():
  function test_tool_parser_extract_calls_streaming_without_thinking (line 287) | def test_tool_parser_extract_calls_streaming_without_thinking():
  function test_tool_parser_extract_calls_streaming_with_thinking (line 410) | def test_tool_parser_extract_calls_streaming_with_thinking():
  function test_tool_parser_extract_calls_streaming_with_parser (line 583) | def test_tool_parser_extract_calls_streaming_with_parser():
  function test_tool_parser_extract_calls_without_thinking_multi (line 714) | def test_tool_parser_extract_calls_without_thinking_multi():
  function test_tool_parser_extract_calls_without_thinking (line 729) | def test_tool_parser_extract_calls_without_thinking():
  function test_tool_parser_extract_calls_with_thinking (line 741) | def test_tool_parser_extract_calls_with_thinking():
  function test_tool_parser_extract_calls_with_parser (line 757) | def test_tool_parser_extract_calls_with_parser():

FILE: xinference/model/llm/transformers/__init__.py
  function import_submodules (line 22) | def import_submodules(package_path: str, package_name: str, globals_dict...

FILE: xinference/model/llm/transformers/chatglm.py
  class ChatglmPytorchChatModel (line 39) | class ChatglmPytorchChatModel(PytorchChatModel):
    method __init__ (line 42) | def __init__(
    method _get_model_class (line 58) | def _get_model_class(self):
    method _load_model (line 63) | def _load_model(self, **kwargs):
    method match_json (line 88) | def match_json(
    method _handle_tools (line 102) | def _handle_tools(self, messages, generate_config):
    method _process_messages (line 120) | def _process_messages(messages, tools=None, tool_choice="none"):
    method _process_response_non_streaming (line 208) | def _process_response_non_streaming(
    method _process_response_streaming (line 268) | def _process_response_streaming(output, tools, end=False):
    method _stream_chat (line 299) | def _stream_chat(self, inputs, tools, **kwargs):
    method _get_generate_kwargs (line 327) | def _get_generate_kwargs(generate_config):
    method chat (line 352) | def chat(  # type: ignore
    method prepare_sanitize_generate_config (line 445) | def prepare_sanitize_generate_config(self, req: InferenceRequest):
    method prepare_batch_inference (line 459) | def prepare_batch_inference(self, req_list: List[InferenceRequest]):
    method handle_chat_result_non_streaming (line 495) | def handle_chat_result_non_streaming(self, req: InferenceRequest):
    method handle_chat_result_streaming (line 509) | def handle_chat_result_streaming(self, req: InferenceRequest):

FILE: xinference/model/llm/transformers/core.py
  function register_non_default_model (line 63) | def register_non_default_model(*architectures: str):
  class PytorchModel (line 92) | class PytorchModel(LLM):
    method __init__ (line 95) | def __init__(
    method _sanitize_model_config (line 111) | def _sanitize_model_config(
    method _sanitize_generate_config (line 129) | def _sanitize_generate_config(
    method _check_tensorizer_integrity (line 145) | def _check_tensorizer_integrity(self):
    method _load_tensorizer (line 158) | def _load_tensorizer(self, **kwargs):
    method _save_tensorizer (line 172) | def _save_tensorizer(self, **kwargs):
    method _get_model_class (line 180) | def _get_model_class(self):
    method _get_components (line 185) | def _get_components(self, **kwargs):
    method _load_model (line 202) | def _load_model(self, **kwargs):
    method _apply_lora (line 227) | def _apply_lora(self):
    method apply_bnb_quantization (line 251) | def apply_bnb_quantization(
    method apply_fp_quantization (line 277) | def apply_fp_quantization(
    method apply_quantization_config (line 313) | def apply_quantization_config(
    method load (line 322) | def load(self):
    method _should_use_batching (line 410) | def _should_use_batching(self) -> bool:
    method _ensure_scheduler_started (line 439) | async def _ensure_scheduler_started(self):
    method generate (line 444) | async def generate(self, prompt: str, generate_config: Optional[dict] ...
    method _direct_generate (line 475) | async def _direct_generate(
    method _queue_to_async_generator (line 481) | async def _queue_to_async_generator(self, queue):
    method abort_request (line 502) | async def abort_request(self, request_id: str) -> Optional[str]:
    method stop_scheduler (line 510) | async def stop_scheduler(self):
    method stop (line 515) | def stop(self):
    method check_lib (line 537) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 544) | def match_json(
    method build_prefill_attention_mask (line 561) | def build_prefill_attention_mask(
    method build_decode_attention_mask (line 594) | def build_decode_attention_mask(
    method build_prefill_position_ids (line 631) | def build_prefill_position_ids(
    method build_decode_position_ids (line 654) | def build_decode_position_ids(
    method build_prefill_token_type_ids (line 668) | def build_prefill_token_type_ids(
    method build_decode_token_type_ids (line 677) | def build_decode_token_type_ids(
    method build_prefill_inputs (line 686) | def build_prefill_inputs(self, prompts: List, req_list: List[Inference...
    method build_prefill_kwargs (line 698) | def build_prefill_kwargs(self, prompts: List, req_list: List[Inference...
    method build_decode_kwargs (line 720) | def build_decode_kwargs(
    method get_batch_size_and_seq_len_indexes_from_kv (line 743) | def get_batch_size_and_seq_len_indexes_from_kv() -> Tuple[int, int]:
    method get_dtype (line 751) | def get_dtype(self):
    method get_context_len (line 755) | def get_context_len(self):
    method get_max_num_seqs (line 759) | def get_max_num_seqs(self) -> int:
    method prepare_sanitize_generate_config (line 762) | def prepare_sanitize_generate_config(self, req: InferenceRequest):
    method merge_kv_cache (line 765) | def merge_kv_cache(self, past_cache, new_cache):
    method prepare_batch_inference (line 870) | def prepare_batch_inference(self, req_list: List[InferenceRequest]):
    method get_builtin_stop_token_ids (line 904) | def get_builtin_stop_token_ids(self) -> Tuple:
    method handle_batch_inference_results (line 921) | def handle_batch_inference_results(self, req_list: List[InferenceReque...
    method batch_inference (line 956) | def batch_inference(self, req_list: List[InferenceRequest]):
    method build_reduced_kv_cache (line 965) | def build_reduced_kv_cache(self, cache, skipped_indexes: Set[int]):
  class PytorchChatModel (line 976) | class PytorchChatModel(PytorchModel, ChatModelMixin):
    method __init__ (line 977) | def __init__(
    method _sanitize_generate_config (line 993) | def _sanitize_generate_config(
    method match_json (line 1009) | def match_json(
    method chat (line 1026) | async def chat(
    method _direct_chat (line 1061) | async def _direct_chat(
    method load (line 1069) | def load(self):
    method _get_full_prompt (line 1072) | def _get_full_prompt(self, messages: List[Dict], tools, generate_confi...
    method prepare_batch_inference (line 1098) | def prepare_batch_inference(self, req_list: List[InferenceRequest]):
    method handle_chat_result_non_streaming (line 1114) | def handle_chat_result_non_streaming(self, req: InferenceRequest):
    method handle_chat_result_streaming (line 1126) | def handle_chat_result_streaming(self, req: InferenceRequest):
    method handle_batch_inference_results (line 1151) | def handle_batch_inference_results(self, req_list: List[InferenceReque...

FILE: xinference/model/llm/transformers/deepseek_v2.py
  class DeepSeekV2PytorchChatModel (line 27) | class DeepSeekV2PytorchChatModel(PytorchChatModel):
    method _load_model (line 30) | def _load_model(self, **kwargs):
    method match_json (line 64) | def match_json(

FILE: xinference/model/llm/transformers/gemma3.py
  class Gemma3TextChatModel (line 26) | class Gemma3TextChatModel(PytorchChatModel):
    method match_json (line 30) | def match_json(
    method _load_model (line 45) | def _load_model(self, **kwargs):
    method _get_full_prompt (line 62) | def _get_full_prompt(self, messages: List[Dict], tools, generate_confi...
    method build_prefill_kwargs (line 65) | def build_prefill_kwargs(self, prompts: List, req_list: List[Inference...
    method merge_kv_cache (line 94) | def merge_kv_cache(self, past_cache, new_cache):
    method build_decode_attention_mask (line 128) | def build_decode_attention_mask(
    method build_decode_position_ids (line 136) | def build_decode_position_ids(
    method build_reduced_kv_cache (line 144) | def build_reduced_kv_cache(self, cache, skipped_indexes: Set[int]):

FILE: xinference/model/llm/transformers/gpt_oss.py
  class GPTOSSPytorchChatModel (line 33) | class GPTOSSPytorchChatModel(PytorchChatModel):
    method _sanitize_model_config (line 36) | def _sanitize_model_config(
    method match_json (line 44) | def match_json(
    method chat (line 61) | async def chat(  # type:ignore

FILE: xinference/model/llm/transformers/multimodal/cogagent.py
  class CogAgentChatModel (line 34) | class CogAgentChatModel(PytorchMultiModalModel):
    method __init__ (line 37) | def __init__(self, *args, **kws):
    method match_json (line 51) | def match_json(
    method decide_device (line 63) | def decide_device(self):
    method load_processor (line 67) | def load_processor(self):
    method load_multimodal_model (line 74) | def load_multimodal_model(self):
    method _message_content_to_cogagent (line 86) | def _message_content_to_cogagent(self, content):
    method _history_content_to_cogagent (line 115) | def _history_content_to_cogagent(self, chat_history: List[Dict]):
    method _get_query_and_history (line 156) | def _get_query_and_history(
    method build_inputs_from_messages (line 184) | def build_inputs_from_messages(
    method build_generate_kwargs (line 216) | def build_generate_kwargs(
    method build_streaming_iter (line 229) | def build_streaming_iter(

FILE: xinference/model/llm/transformers/multimodal/core.py
  class PytorchMultiModalModel (line 29) | class PytorchMultiModalModel(PytorchChatModel):
    method __init__ (line 30) | def __init__(self, *args, **kwargs):
    method decide_device (line 38) | def decide_device(self):
    method load_processor (line 45) | def load_processor(self):
    method load_multimodal_model (line 52) | def load_multimodal_model(self):
    method load (line 58) | def load(self):
    method build_inputs_from_messages (line 70) | def build_inputs_from_messages(
    method build_generate_kwargs (line 83) | def build_generate_kwargs(
    method build_streaming_iter (line 94) | def build_streaming_iter(
    method get_stop_strs (line 106) | def get_stop_strs(self) -> List[str]:
    method check_conditions (line 109) | def check_conditions(self, new_text: str) -> Tuple[str, bool]:
    method generate_non_streaming (line 117) | def generate_non_streaming(
    method generate_streaming (line 149) | def generate_streaming(
    method chat (line 271) | def chat(

FILE: xinference/model/llm/transformers/multimodal/deepseek_vl2.py
  class DeepSeekVL2ChatModel (line 35) | class DeepSeekVL2ChatModel(PytorchMultiModalModel):
    method __init__ (line 38) | def __init__(self, *args, **kwargs):
    method match_json (line 43) | def match_json(
    method decide_device (line 55) | def decide_device(self):
    method load_processor (line 60) | def load_processor(self):
    method load_multimodal_model (line 69) | def load_multimodal_model(self):
    method _message_content_to_deepseek (line 85) | def _message_content_to_deepseek(content) -> Tuple[str, List[str]]:
    method get_stop_strs (line 148) | def get_stop_strs(self) -> List[str]:
    method build_generate_kwargs (line 153) | def build_generate_kwargs(self, generate_config: Dict):
    method build_inputs_from_messages (line 157) | def build_inputs_from_messages(
    method build_streaming_iter (line 213) | def build_streaming_iter(
    method check_conditions (line 228) | def check_conditions(self, new_text: str) -> Tuple[str, bool]:

FILE: xinference/model/llm/transformers/multimodal/gemma3.py
  class Gemma3ChatModel (line 29) | class Gemma3ChatModel(PytorchMultiModalModel):
    method match_json (line 33) | def match_json(
    method _sanitize_model_config (line 48) | def _sanitize_model_config(
    method decide_device (line 57) | def decide_device(self):
    method load_processor (line 62) | def load_processor(self):
    method load_multimodal_model (line 74) | def load_multimodal_model(self):
    method build_inputs_from_messages (line 82) | def build_inputs_from_messages(
    method build_generate_kwargs (line 97) | def build_generate_kwargs(
    method build_streaming_iter (line 106) | def build_streaming_iter(

FILE: xinference/model/llm/transformers/multimodal/glm4_1v.py
  class Glm4_1VModel (line 35) | class Glm4_1VModel(PytorchMultiModalModel):
    method match_json (line 42) | def match_json(
    method decide_device (line 54) | def decide_device(self):
    method load_processor (line 58) | def load_processor(self):
    method load_multimodal_model (line 64) | def load_multimodal_model(self):
    method _get_processed_msgs (line 79) | def _get_processed_msgs(messages: List[Dict]) -> List[Dict]:
    method build_inputs_from_messages (line 120) | def build_inputs_from_messages(
    method get_stop_strs (line 146) | def get_stop_strs(self) -> List[str]:
    method get_builtin_stop_token_ids (line 149) | def get_builtin_stop_token_ids(self) -> Tuple:
    method build_generate_kwargs (line 154) | def build_generate_kwargs(
    method build_streaming_iter (line 166) | def build_streaming_iter(

FILE: xinference/model/llm/transformers/multimodal/glm4v.py
  class Glm4VModel (line 37) | class Glm4VModel(PytorchMultiModalModel):
    method match_json (line 44) | def match_json(
    method decide_device (line 56) | def decide_device(self):
    method load_processor (line 60) | def load_processor(self):
    method load_multimodal_model (line 67) | def load_multimodal_model(self):
    method _get_processed_msgs (line 83) | def _get_processed_msgs(messages: List[Dict]) -> List[Dict]:
    method build_inputs_from_messages (line 118) | def build_inputs_from_messages(
    method build_generate_kwargs (line 134) | def build_generate_kwargs(
    method get_stop_strs (line 145) | def get_stop_strs(self) -> List[str]:
    method build_streaming_iter (line 148) | def build_streaming_iter(
    method _get_full_prompt (line 172) | def _get_full_prompt(self, messages, tools, generate_config: dict):
    method prepare_sanitize_generate_config (line 186) | def prepare_sanitize_generate_config(self, req: InferenceRequest):
    method build_prefill_inputs (line 199) | def build_prefill_inputs(self, prompts: List, req_list: List[Inference...
    method is_empty (line 236) | def is_empty(images_list: Optional[List[List[torch.Tensor]]]):
    method get_full_attention_mask (line 248) | def get_full_attention_mask(
    method build_prefill_kwargs (line 286) | def build_prefill_kwargs(self, prompts: List, req_list: List[Inference...
    method build_decode_attention_mask (line 304) | def build_decode_attention_mask(

FILE: xinference/model/llm/transformers/multimodal/intern_vl.py
  class InternVLChatModel (line 32) | class InternVLChatModel(PytorchMultiModalModel):
    method match_json (line 39) | def match_json(
    method decide_device (line 51) | def decide_device(self):
    method load_processor (line 82) | def load_processor(self):
    method load_multimodal_model (line 89) | def load_multimodal_model(self):
    method _build_transform (line 106) | def _build_transform(self, input_size=448):
    method _get_index (line 125) | def _get_index(bound, fps, max_frame, first_idx=0, num_segments=32):
    method _find_closest_aspect_ratio (line 143) | def _find_closest_aspect_ratio(
    method _dynamic_preprocess (line 160) | def _dynamic_preprocess(
    method _load_video (line 205) | def _load_video(
    method _message_content_to_intern (line 233) | def _message_content_to_intern(self, content, image_cnt):
    method _get_prompt_and_chat_history (line 270) | def _get_prompt_and_chat_history(
    method _load_image (line 295) | def _load_image(self, image_file, input_size=448, max_num=12):
    method build_inputs_from_messages (line 305) | def build_inputs_from_messages(
    method build_generate_kwargs (line 383) | def build_generate_kwargs(
    method build_streaming_iter (line 393) | def build_streaming_iter(
    method check_conditions (line 415) | def check_conditions(self, new_text: str) -> Tuple[str, bool]:

FILE: xinference/model/llm/transformers/multimodal/minicpmv26.py
  class MiniCPMV26Model (line 36) | class MiniCPMV26Model(PytorchMultiModalModel):
    method match_json (line 40) | def match_json(
    method _sanitize_model_config (line 52) | def _sanitize_model_config(
    method decide_device (line 61) | def decide_device(self):
    method load_processor (line 70) | def load_processor(self):
    method load_multimodal_model (line 86) | def load_multimodal_model(self):
    method _message_content_to_chat (line 109) | def _message_content_to_chat(self, content):
    method _convert_to_specific_style (line 167) | def _convert_to_specific_style(self, messages: List[Dict]) -> Tuple:
    method build_inputs_from_messages (line 203) | def build_inputs_from_messages(
    method build_generate_kwargs (line 215) | def build_generate_kwargs(
    method build_streaming_iter (line 221) | def build_streaming_iter(
    method prepare_sanitize_generate_config (line 234) | def prepare_sanitize_generate_config(self, req: InferenceRequest):
    method _handle_input_ids_and_images (line 253) | def _handle_input_ids_and_images(self, msgs: List[Dict]) -> Dict:
    method _get_full_prompt (line 286) | def _get_full_prompt(self, messages: List[Dict], tools, generate_confi...
    method build_prefill_kwargs (line 294) | def build_prefill_kwargs(self, prompts: List, req_list: List[Inference...
    method build_decode_position_ids (line 327) | def build_decode_position_ids(
    method batch_inference (line 332) | def batch_inference(self, req_list: List[InferenceRequest]):

FILE: xinference/model/llm/transformers/multimodal/minicpmv45.py
  class MiniCPMV45Model (line 36) | class MiniCPMV45Model(PytorchMultiModalModel):
    method match_json (line 40) | def match_json(
    method _sanitize_model_config (line 52) | def _sanitize_model_config(
    method decide_device (line 62) | def decide_device(self):
    method load_processor (line 71) | def load_processor(self):
    method load_multimodal_model (line 87) | def load_multimodal_model(self):
    method _message_content_to_chat (line 110) | def _message_content_to_chat(self, content):
    method _convert_to_specific_style (line 168) | def _convert_to_specific_style(self, messages: List[Dict]) -> Tuple:
    method build_inputs_from_messages (line 204) | def build_inputs_from_messages(
    method build_generate_kwargs (line 216) | def build_generate_kwargs(
    method build_streaming_iter (line 222) | def build_streaming_iter(
    method prepare_sanitize_generate_config (line 235) | def prepare_sanitize_generate_config(self, req: InferenceRequest):
    method _handle_input_ids_and_images (line 254) | def _handle_input_ids_and_images(self, msgs: List[Dict]) -> Dict:
    method _get_full_prompt (line 288) | def _get_full_prompt(self, messages: List[Dict], tools, generate_confi...
    method build_prefill_kwargs (line 296) | def build_prefill_kwargs(self, prompts: List, req_list: List[Inference...
    method build_decode_position_ids (line 329) | def build_decode_position_ids(
    method batch_inference (line 334) | def batch_inference(self, req_list: List[InferenceRequest]):

FILE: xinference/model/llm/transformers/multimodal/ovis2.py
  class Ovis2ChatModel (line 30) | class Ovis2ChatModel(PytorchMultiModalModel):
    method __init__ (line 33) | def __init__(self, *args, **kws):
    method match_json (line 39) | def match_json(
    method decide_device (line 56) | def decide_device(self):
    method load_processor (line 59) | def load_processor(self):
    method load_multimodal_model (line 62) | def load_multimodal_model(self):
    method _parse_messages_ovis (line 77) | def _parse_messages_ovis(messages: List[Dict]) -> List[Dict]:
    method _convert_video_tensors_to_pil (line 96) | def _convert_video_tensors_to_pil(video_inputs: List) -> List[Image.Im...
    method _generate_chat_data (line 137) | def _generate_chat_data(self, messages: List[Dict]):
    method build_generate_kwargs (line 194) | def build_generate_kwargs(
    method build_inputs_from_messages (line 210) | def build_inputs_from_messages(
    method build_streaming_iter (line 232) | def build_streaming_iter(

FILE: xinference/model/llm/transformers/multimodal/qwen-omni.py
  class QwenOmniChatModel (line 43) | class QwenOmniChatModel(PytorchMultiModalModel):
    method __init__ (line 53) | def __init__(self, *args, **kwargs):
    method match_json (line 62) | def match_json(
    method decide_device (line 87) | def decide_device(self):
    method load_processor (line 92) | def load_processor(self):
    method load_multimodal_model (line 103) | def load_multimodal_model(self):
    method _transform_messages (line 132) | def _transform_messages(
    method build_inputs_from_messages (line 150) | def build_inputs_from_messages(
    method build_generate_kwargs (line 181) | def build_generate_kwargs(
    method build_streaming_iter (line 192) | def build_streaming_iter(
    method generate_non_streaming (line 211) | def generate_non_streaming(

FILE: xinference/model/llm/transformers/multimodal/qwen2_audio.py
  class Qwen2AudioChatModel (line 32) | class Qwen2AudioChatModel(PytorchMultiModalModel):
    method match_json (line 36) | def match_json(
    method decide_device (line 48) | def decide_device(self):
    method load_processor (line 52) | def load_processor(self):
    method load_multimodal_model (line 64) | def load_multimodal_model(self):
    method _transform_messages (line 76) | def _transform_messages(
    method build_inputs_from_messages (line 99) | def build_inputs_from_messages(
    method build_generate_kwargs (line 113) | def build_generate_kwargs(
    method build_streaming_iter (line 119) | def build_streaming_iter(

FILE: xinference/model/llm/transformers/multimodal/qwen2_vl.py
  class Qwen2VLChatModel (line 44) | class Qwen2VLChatModel(PytorchMultiModalModel):
    method _sanitize_model_config (line 52) | def _sanitize_model_config(
    method match_json (line 62) | def match_json(
    method decide_device (line 86) | def decide_device(self):
    method load_processor (line 92) | def load_processor(self):
    method load_multimodal_model (line 105) | def load_multimodal_model(self):
    method build_inputs_from_messages (line 176) | def build_inputs_from_messages(
    method build_generate_kwargs (line 199) | def build_generate_kwargs(self, generate_config: Dict) -> Dict[str, Any]:
    method build_streaming_iter (line 204) | def build_streaming_iter(
    method prepare_sanitize_generate_config (line 232) | def prepare_sanitize_generate_config(self, req: InferenceRequest):
    method _get_full_prompt (line 244) | def _get_full_prompt(self, messages: List[Dict], tools, generate_confi...
    method build_prefill_kwargs (line 247) | def build_prefill_kwargs(self, prompts: List, req_list: List[Inference...

FILE: xinference/model/llm/transformers/opt.py
  class OptPytorchModel (line 25) | class OptPytorchModel(PytorchModel):
    method __init__ (line 28) | def __init__(
    method match_json (line 45) | def match_json(
    method build_prefill_position_ids (line 57) | def build_prefill_position_ids(
    method build_decode_position_ids (line 67) | def build_decode_position_ids(

FILE: xinference/model/llm/transformers/tensorizer_utils.py
  function _filter_kwargs (line 41) | def _filter_kwargs(kwargs):
  function _file_is_non_empty (line 48) | def _file_is_non_empty(
  function get_tensorizer_dir (line 57) | def get_tensorizer_dir(model_path: str) -> str:
  function check_tensorizer_integrity (line 62) | def check_tensorizer_integrity(
  function load_from_tensorizer (line 79) | def load_from_tensorizer(
  function _load_pretrained_from_tensorizer (line 131) | def _load_pretrained_from_tensorizer(
  function _load_model_from_tensorizer (line 168) | def _load_model_from_tensorizer(
  function save_to_tensorizer (line 260) | def save_to_tensorizer(
  function _tensorizer_serialize_model (line 276) | def _tensorizer_serialize_model(
  function _tensorizer_serialize_pretrained (line 310) | def _tensorizer_serialize_pretrained(

FILE: xinference/model/llm/transformers/tests/test_opt.py
  function test_opt_pytorch_model (line 25) | async def test_opt_pytorch_model(setup, quantization):
  function test_opt_fp4_model (line 76) | async def test_opt_fp4_model(setup):

FILE: xinference/model/llm/transformers/tests/test_tensorizer.py
  class TestTensorizerSerializeModel (line 19) | class TestTensorizerSerializeModel:
    method setup_and_teardown (line 21) | def setup_and_teardown(self):
    method _cleanup_directory (line 64) | def _cleanup_directory(self, directory):
    method test_tensor_file_exists (line 69) | def test_tensor_file_exists(self):
  function mock_environment (line 95) | def mock_environment(tmp_path):
  function test_tensorizer_serialize_model_cache_exists (line 108) | def test_tensorizer_serialize_model_cache_exists(

FILE: xinference/model/llm/transformers/utils.py
  function get_context_length (line 47) | def get_context_length(config) -> int:
  function prepare_logits_processor (line 70) | def prepare_logits_processor(
  function _get_token_from_logits (line 86) | def _get_token_from_logits(
  function _pad_to_max_length (line 115) | def _pad_to_max_length(x: List[int], max_len: int, pad: int) -> List[int]:
  function _pad_seqs_inplace (line 120) | def _pad_seqs_inplace(seqs: List[List[int]], reqs: List[InferenceRequest...
  function get_max_src_len (line 132) | def get_max_src_len(context_len: int, r: InferenceRequest) -> int:
  function pad_prefill_tokens (line 140) | def pad_prefill_tokens(
  function _get_completion (line 153) | def _get_completion(
  function _get_pad_param (line 188) | def _get_pad_param(seq_len_idx: int, pad_len: int) -> Tuple:
  function get_batch_size_and_seq_len_from_kv_cache (line 194) | def get_batch_size_and_seq_len_from_kv_cache(kv, xinf_model_obj: "Pytorc...
  function convert_to_cache_cls (line 216) | def convert_to_cache_cls(cache) -> DynamicCache:
  function _batch_inference_one_step_internal (line 226) | def _batch_inference_one_step_internal(
  function batch_inference_one_step (line 484) | def batch_inference_one_step(

FILE: xinference/model/llm/utils.py
  class ChatModelMixin (line 110) | class ChatModelMixin:
    method __init__ (line 111) | def __init__(self):
    method _compile_jinja_template (line 119) | def _compile_jinja_template(chat_template):
    method _build_from_raw_template (line 136) | def _build_from_raw_template(
    method get_full_context (line 145) | def get_full_context(
    method _get_chat_template_kwargs_from_generate_config (line 179) | def _get_chat_template_kwargs_from_generate_config(
    method convert_messages_with_content_list_to_str_conversion (line 205) | def convert_messages_with_content_list_to_str_conversion(
    method get_specific_prompt (line 224) | def get_specific_prompt(model_family: str, messages: List[ChatCompleti...
    method _to_chat_completion_chunk (line 294) | def _to_chat_completion_chunk(
    method _get_first_chat_completion_chunk (line 386) | def _get_first_chat_completion_chunk(
    method _get_final_chat_completion_chunk (line 418) | def _get_final_chat_completion_chunk(
    method _to_chat_completion_chunks (line 438) | def _to_chat_completion_chunks(
    method _tools_to_messages_for_deepseek (line 476) | def _tools_to_messages_for_deepseek(
    method _async_to_chat_completion_chunks (line 505) | async def _async_to_chat_completion_chunks(
    method _to_chat_completion (line 543) | def _to_chat_completion(
    method _eval_glm_chat_arguments (line 600) | def _eval_glm_chat_arguments(c) -> List[Tuple]:
    method _handle_qwen_tool_result (line 617) | def _handle_qwen_tool_result(cls, text: str) -> List[Tuple]:
    method _eval_qwen_chat_arguments (line 686) | def _eval_qwen_chat_arguments(
    method _eval_llama3_chat_arguments (line 695) | def _eval_llama3_chat_arguments(cls, c) -> List[Tuple]:
    method _eval_deepseek_chat_arguments (line 704) | def _eval_deepseek_chat_arguments(cls, c) -> List[Tuple]:
    method _eval_deepseek_r1_arguments (line 772) | def _eval_deepseek_r1_arguments(cls, c) -> List[Tuple]:
    method _eval_tool_arguments (line 818) | def _eval_tool_arguments(
    method _post_process_completion_chunk (line 840) | def _post_process_completion_chunk(
    method _post_process_completion (line 922) | def _post_process_completion(
    method _transform_messages (line 1013) | def _transform_messages(
    method _async_to_tool_completion_chunks (line 1050) | async def _async_to_tool_completion_chunks(
  function get_model_version (line 1093) | def get_model_version(
  function _decode_image (line 1102) | def _decode_image(_url):
  function _decode_image_without_rgb (line 1121) | def _decode_image_without_rgb(_url):
  function generate_completion_chunk (line 1141) | def generate_completion_chunk(
  function generate_completion (line 1175) | def generate_completion(
  function generate_chat_completion (line 1201) | def generate_chat_completion(
  function get_stop_token_ids_from_config_file (line 1230) | def get_stop_token_ids_from_config_file(model_path: str) -> Optional[Lis...
  function normalize_response_format (line 1252) | def normalize_response_format(
  function parse_messages (line 1278) | def parse_messages(messages: List[Dict]) -> Tuple:

FILE: xinference/model/llm/vllm/core.py
  class VLLMModelConfig (line 88) | class VLLMModelConfig(TypedDict, total=False):
  class VLLMGenerateConfig (line 117) | class VLLMGenerateConfig(TypedDict, total=False):
  function _get_effective_vllm_version (line 165) | def _get_effective_vllm_version() -> version.Version:
  function _virtual_env_allows_missing_vllm (line 177) | def _virtual_env_allows_missing_vllm() -> bool:
  function _append_unique (line 185) | def _append_unique(target: List[str], *items: str) -> None:
  function _update_vllm_supported_lists (line 209) | def _update_vllm_supported_lists() -> None:
  class VLLMModel (line 325) | class VLLMModel(LLM):
    method __init__ (line 328) | def __init__(
    method set_xavier_config (line 358) | def set_xavier_config(self, value: Optional[Dict]):
    method set_worker_addresses (line 361) | def set_worker_addresses(self, shard: int, worker_addresses: List[str]):
    method driver_info (line 371) | def driver_info(self) -> Optional[dict]:
    method need_create_pools (line 375) | def need_create_pools(self):
    method set_pool_addresses (line 378) | def set_pool_addresses(self, pool_addresses: List[str]):
    method get_pool_addresses (line 381) | def get_pool_addresses(self) -> Optional[List[str]]:
    method set_loop (line 384) | def set_loop(self, loop: asyncio.AbstractEventLoop):
    method _is_vllm_v1 (line 389) | def _is_vllm_v1(self) -> bool:
    method load (line 405) | def load(self):
    method wait_for_load (line 628) | def wait_for_load(self):
    method _set_context_length (line 640) | def _set_context_length(self):
    method _enable_v1_if_supported (line 652) | def _enable_v1_if_supported(self, engine_args: "vllm.AsyncEngineArgs"):
    method _preprocess_load_gguf (line 696) | def _preprocess_load_gguf(self):
    method stop (line 740) | def stop(self):
    method init_xavier (line 760) | async def init_xavier(self):
    method _check_healthy (line 763) | async def _check_healthy(self, interval: int = 30):
    method parse_str_field_to_dict (line 783) | def parse_str_field_to_dict(
    method _sanitize_model_config (line 828) | def _sanitize_model_config(
    method _sanitize_generate_config (line 891) | def _sanitize_generate_config(
    method check_lib (line 979) | def check_lib(cls) -> Union[bool, Tuple[bool, str]]:
    method match_json (line 989) | def match_json(
    method _convert_request_output_to_completion_chunk (line 1031) | def _convert_request_output_to_completion_chunk(
    method _convert_request_output_to_completion (line 1058) | def _convert_request_output_to_completion(
    method _get_tokenizer (line 1090) | async def _get_tokenizer(self, lora_request: Any) -> Any:
    method _tokenize (line 1124) | def _tokenize(self, tokenizer: Any, prompt: str, config: dict) -> List...
    method _gen_tokens_prompt (line 1148) | async def _gen_tokens_prompt(
    method async_generate (line 1162) | async def async_generate(
  class VLLMChatModel (line 1532) | class VLLMChatModel(VLLMModel, ChatModelMixin):
    method match_json (line 1534) | def match_json(
    method _sanitize_chat_config (line 1580) | def _sanitize_chat_config(
    method is_tool_call_chunk_start (line 1616) | def is_tool_call_chunk_start(chunk):
    method is_tool_call_chunk_end (line 1620) | def is_tool_call_chunk_end(chunk):
    method prefill_messages (line 1624) | def prefill_messages(messages: List[Dict]) -> List[Dict]:
    method async_chat (line 1650) | async def async_chat(
  class VLLMMultiModel (line 1718) | class VLLMMultiModel(VLLMModel, ChatModelMixin):
    method match_json (line 1720) | def match_json(
    method _attach_video_metadata (line 1777) | def _attach_video_metadata(
    method _sanitize_model_config (line 1801) | def _sanitize_model_config(
    method _sanitize_chat_config (line 1842) | def _sanitize_chat_config(
    method _gen_tokens_prompt (line 1861) | async def _gen_tokens_prompt(
    method _handle_base64_images (line 1882) | def _handle_base64_images(self, messages, temp_files):
    method async_chat (line 1930) | async def async_chat(

FILE: xinference/model/llm/vllm/distributed_executor.py
  class WorkerActor (line 45) | class WorkerActor(xo.StatelessActor):
    method __init__ (line 46) | def __init__(self, vllm_config: "VllmConfig", rpc_rank: int = 0, **kwa...
    method __post_create__ (line 50) | async def __post_create__(self):
    method __getattr__ (line 59) | def __getattr__(self, item):
    method gen_uid (line 63) | def gen_uid(cls, rank):
    method execute_method (line 66) | def execute_method(self, method: Union[str, Callable], *args, **kwargs):
  class WorkerWrapper (line 82) | class WorkerWrapper:
    method __init__ (line 83) | def __init__(
    method execute_method (line 91) | def execute_method(self, method: Union[str, Callable], *args, **kwargs):
    method execute_method_async (line 95) | async def execute_method_async(self, method: Union[str, Callable], *ar...
    method kill (line 98) | def kill(self):
  class XinferenceDistr

[
  {
    "path": ".dockerignore",
    "chars": 106,
    "preview": "doc/\n.idea/\n.github/\nbuild/\nxinference.egg-info/\nxinference/web/ui/build/\nxinference/web/ui/node_modules/\n"
  },
  {
    "path": ".gitattributes",
    "chars": 36,
    "preview": "xinference/_version.py export-subst\n"
  },
  {
    "path": ".github/ISSUE_TEMPLATE/bug_report.yaml",
    "chars": 3204,
    "preview": "name: \"Bug Report\"\ndescription: Submit a bug report to help us improve Xinference. You should provide useful information"
  },
  {
    "path": ".github/ISSUE_TEMPLATE/feature_request.yaml",
    "chars": 932,
    "preview": "name: \"Feature request\"\ndescription: Submit a request for a new Xinference feature / 提交一个新的 Xinference 的功能建议\nlabels: [ \""
  },
  {
    "path": ".github/workflows/assign.yaml",
    "chars": 606,
    "preview": "name: Assign\non:\n  issue_comment:\n    types: created\n\npermissions:\n  contents: read\n\njobs:\n  issue_assign:\n    permissio"
  },
  {
    "path": ".github/workflows/docker-cd.yaml",
    "chars": 2690,
    "preview": "name: Xinference CD for DockerHub\n\non:\n  schedule:\n    - cron: '0 18 * * *'\n  push:\n    tags:\n      - '*'\n  workflow_dis"
  },
  {
    "path": ".github/workflows/issue.yaml",
    "chars": 760,
    "preview": "name: Close inactive issues\non:\n  schedule:\n    - cron: \"0 19 * * *\"\n  workflow_dispatch:\n\njobs:\n  close-issues:\n    run"
  },
  {
    "path": ".github/workflows/pr_auto_run_gen_docs.yaml",
    "chars": 6561,
    "preview": "name: Auto run gen_docs.py and commit changes to PR\n\non:\n  pull_request_target:\n    types: [opened, synchronize]\n\npermis"
  },
  {
    "path": ".github/workflows/python.yaml",
    "chars": 16842,
    "preview": "name: Python CI\n\non:\n  push:\n    branches:\n      - '*'\n  pull_request:\n    types: ['opened', 'reopened', 'synchronize']\n"
  },
  {
    "path": ".github/workflows/release.yaml",
    "chars": 1389,
    "preview": "name: Build and upload to PyPI\n\non:\n  push:\n    tags:\n      - '*'\n\nconcurrency:\n  group: ${{ github.workflow }}-${{ gith"
  },
  {
    "path": ".gitignore",
    "chars": 2078,
    "preview": "# Byte-compiled / optimized / DLL files\n__pycache__/\n*.py[cod]\n*$py.class\n\n# C extensions\n*.so\n\n# Distribution / packagi"
  },
  {
    "path": ".pre-commit-config.yaml",
    "chars": 1102,
    "preview": "files: xinference\nrepos:\n  - repo: https://github.com/psf/black\n    rev: 25.1.0\n    hooks:\n      - id: black\n        exc"
  },
  {
    "path": ".readthedocs.yaml",
    "chars": 307,
    "preview": "version: 2\n\n# Build documentation in the docs/ directory with Sphinx\nsphinx:\n   configuration: doc/source/conf.py\n\nbuild"
  },
  {
    "path": "LICENSE",
    "chars": 11356,
    "preview": "                                 Apache License\n                           Version 2.0, January 2004\n                   "
  },
  {
    "path": "MANIFEST.in",
    "chars": 449,
    "preview": "global-include *.pyx\nglobal-include *.pxd\nglobal-include xinference/**/*.json\nglobal-exclude *.c\nglobal-exclude *.cpp\nin"
  },
  {
    "path": "README.md",
    "chars": 12557,
    "preview": "<div align=\"center\">\n<img src=\"./assets/xorbits-logo.png\" width=\"180px\" alt=\"xorbits\" />\n\n# Xorbits Inference: Model Ser"
  },
  {
    "path": "README_ja_JP.md",
    "chars": 6235,
    "preview": "<div align=\"center\">\n<img src=\"./assets/xorbits-logo.png\" width=\"180px\" alt=\"xorbits\" />\n\n# Xorbits Inference: モデルサービングを"
  },
  {
    "path": "README_zh_CN.md",
    "chars": 9724,
    "preview": "<div align=\"center\">\n<img src=\"./assets/xorbits-logo.png\" width=\"180px\" alt=\"xorbits\" />\n\n# Xorbits Inference：模型推理， 轻而易举"
  },
  {
    "path": "benchmark/README.md",
    "chars": 1916,
    "preview": "# Benchmarking Xinference\n\n## Downloading the ShareGPT dataset\n\nYou can download the dataset by running:\n```bash\nwget ht"
  },
  {
    "path": "benchmark/benchmark_embedding.py",
    "chars": 5820,
    "preview": "# Copyright 2022-2025 XProbe Inc.\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not use "
  },
  {
    "path": "benchmark/benchmark_latency.py",
    "chars": 3314,
    "preview": "# Copyright 2022-2023 XProbe Inc.\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not use "
  },
  {
    "path": "benchmark/benchmark_long.py",
    "chars": 4302,
    "preview": "# Copyright 2022-2023 XProbe Inc.\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not use "
  },
  {
    "path": "benchmark/benchmark_rerank.py",
    "chars": 6023,
    "preview": "# Copyright 2022-2023 XProbe Inc.\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not use "
  },
  {
    "path": "benchmark/benchmark_runner.py",
    "chars": 15640,
    "preview": "# Copyright 2022-2023 XProbe Inc.\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not use "
  },
  {
    "path": "benchmark/benchmark_serving.py",
    "chars": 5916,
    "preview": "# Copyright 2022-2023 XProbe Inc.\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not use "
  },
  {
    "path": "benchmark/utils.py",
    "chars": 6166,
    "preview": "# Copyright 2022-2023 XProbe Inc.\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not use "
  },
  {
    "path": "doc/Makefile",
    "chars": 1157,
    "preview": "# Minimal makefile for Sphinx documentation\n#\n\n# You can set these variables from the command line, and also\n# from the "
  },
  {
    "path": "doc/source/_static/switcher.json",
    "chars": 257,
    "preview": "[\n  {\n    \"name\": \"简体中文(Chinese)\",\n    \"version\": \"zh-cn\",\n    \"url\": \"https://inference.readthedocs.io/zh-cn/latest/\"\n "
  },
  {
    "path": "doc/source/conf.py",
    "chars": 4217,
    "preview": "# Configuration file for the Sphinx documentation builder.\n#\n# This file only contains a selection of the most common op"
  },
  {
    "path": "doc/source/development/contributing_codebase.rst",
    "chars": 4167,
    "preview": "=============================\nContributing to the code base\n=============================\n\n.. contents:: Table of conten"
  },
  {
    "path": "doc/source/development/contributing_environment.rst",
    "chars": 3966,
    "preview": "==================================\nCreating a development environment\n==================================\n\n.. contents:: "
  },
  {
    "path": "doc/source/development/index.rst",
    "chars": 172,
    "preview": ".. _development_index:\n\n===========\nDevelopment\n===========\n\n.. toctree::\n    :maxdepth: 2\n\n    contributing_environment"
  },
  {
    "path": "doc/source/development/xinference_internals.rst",
    "chars": 15258,
    "preview": "===========================\nThe internals of Xinference\n===========================\n\n.. contents:: Table of contents:\n  "
  },
  {
    "path": "doc/source/examples/ai_podcast.rst",
    "chars": 3269,
    "preview": ".. _examples_ai_podcast:\n\n======================\nExample: AI Podcast 🎙\n======================\n\n**Description**:\n\n🎙️AI Po"
  },
  {
    "path": "doc/source/examples/chatbot.rst",
    "chars": 1275,
    "preview": ".. _examples_chatbot:\n\n========================\nExample: CLI chatbot 🤖️\n========================\n\n**Description**:\n\nDemo"
  },
  {
    "path": "doc/source/examples/gradio_chatinterface.rst",
    "chars": 1376,
    "preview": ".. _examples_gradio_chatinterface:\n\n===============================\nExample: Gradio ChatInterface🤗\n====================="
  },
  {
    "path": "doc/source/examples/index.rst",
    "chars": 2180,
    "preview": ".. _examples_index:\n\n========\nExamples\n========\n\n.. toctree::\n   :maxdepth: 2\n   :hidden:\n\n   ai_podcast\n   chatbot\n   g"
  },
  {
    "path": "doc/source/examples/langchain_streamlit_doc_chat.rst",
    "chars": 1390,
    "preview": ".. _examples_langchain_streamlit_doc_chat:\n\n=======================================\nExample: LangChain Streamlit Doc Cha"
  },
  {
    "path": "doc/source/examples/pdf_chatbot.rst",
    "chars": 1062,
    "preview": ".. _examples_pdf_chatbot:\n\n======================\nExample: PDF Chatbot📚\n======================\n\n**Description**:\n\nThis e"
  },
  {
    "path": "doc/source/gen_docs.py",
    "chars": 27256,
    "preview": "# Copyright 2022-2023 XProbe Inc.\n#\n# Licensed under the Apache License, Version 2.0 (the \"License\");\n# you may not use "
  },
  {
    "path": "doc/source/getting_started/environments.rst",
    "chars": 3543,
    "preview": ".. _environments:\n\n======================\nEnvironments Variables\n======================\n\nXINFERENCE_ENDPOINT\n~~~~~~~~~~~"
  },
  {
    "path": "doc/source/getting_started/index.rst",
    "chars": 249,
    "preview": ".. _getting_started_index:\n\n===============\nGetting Started\n===============\n\n\n.. toctree::\n   :maxdepth: 2\n\n   installat"
  },
  {
    "path": "doc/source/getting_started/installation.rst",
    "chars": 6816,
    "preview": ".. _installation:\n\n============\nInstallation\n============\nXinference can be installed with ``pip`` on Linux, Windows, an"
  },
  {
    "path": "doc/source/getting_started/installation_npu.rst",
    "chars": 1661,
    "preview": ".. _installation_npu:\n\n\n=================================\nInstallation Guide for Ascend NPU\n============================"
  },
  {
    "path": "doc/source/getting_started/logging.rst",
    "chars": 1993,
    "preview": ".. _logging:\n\n=====================\nLogging in Xinference\n=====================\n\nConfigure Log Level\n###################"
  },
  {
    "path": "doc/source/getting_started/release_notes.rst",
    "chars": 3043,
    "preview": ".. _release_ntoes:\n\nRelease Notes\n=============\n\nThis page provides a version-by-version index of Xinference release not"
  },
  {
    "path": "doc/source/getting_started/troubleshooting.rst",
    "chars": 11927,
    "preview": ".. _troubleshooting:\n\n===============\nTroubleshooting\n===============\n\n\nNo huggingface repo access\n====================="
  },
  {
    "path": "doc/source/getting_started/using_docker_image.rst",
    "chars": 5390,
    "preview": ".. _using_docker_image:\n\n=======================\nXinference Docker Image\n=======================\n\nXinference provides of"
  },
  {
    "path": "doc/source/getting_started/using_kubernetes.rst",
    "chars": 2899,
    "preview": ".. _using_kubernetes:\n\n########################\nXinference on Kubernetes\n########################\n\n************\nHelm Sup"
  },
  {
    "path": "doc/source/getting_started/using_xinference.rst",
    "chars": 16025,
    "preview": ".. _using_xinference:\n\n================\nUsing Xinference\n================\n\n\nRun Xinference Locally\n====================="
  },
  {
    "path": "doc/source/index.rst",
    "chars": 6227,
    "preview": ".. _index:\n\n======================\nWelcome to Xinference!\n======================\n\n.. toctree::\n   :maxdepth: 2\n   :hidde"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/development/contributing_codebase.po",
    "chars": 8084,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/development/contributing_environment.po",
    "chars": 8155,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/development/index.po",
    "chars": 711,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/development/xinference_internals.po",
    "chars": 30084,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/examples/ai_podcast.po",
    "chars": 6656,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/examples/chatbot.po",
    "chars": 3258,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/examples/gradio_chatinterface.po",
    "chars": 3456,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/examples/index.po",
    "chars": 4598,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/examples/langchain_streamlit_doc_chat.po",
    "chars": 3667,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/examples/pdf_chatbot.po",
    "chars": 3049,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/getting_started/environments.po",
    "chars": 8166,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/getting_started/index.po",
    "chars": 719,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/getting_started/installation.po",
    "chars": 13125,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/getting_started/installation_npu.po",
    "chars": 3148,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/getting_started/logging.po",
    "chars": 3731,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/getting_started/release_notes.po",
    "chars": 4841,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2025, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/getting_started/troubleshooting.po",
    "chars": 20625,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/getting_started/using_docker_image.po",
    "chars": 9767,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/getting_started/using_kubernetes.po",
    "chars": 5066,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/getting_started/using_xinference.po",
    "chars": 20502,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/getting_started.po",
    "chars": 719,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/index.po",
    "chars": 4952,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/audio/index.po",
    "chars": 879,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-base-en-v1.5.po",
    "chars": 1806,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-base-en.po",
    "chars": 1728,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-base-zh-v1.5.po",
    "chars": 1806,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-base-zh.po",
    "chars": 1728,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-large-en-v1.5.po",
    "chars": 1822,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-large-en.po",
    "chars": 1744,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-large-zh-noinstruct.po",
    "chars": 1912,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-large-zh-v1.5.po",
    "chars": 1822,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-large-zh.po",
    "chars": 1744,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-small-en-v1.5.po",
    "chars": 1821,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-small-zh-v1.5.po",
    "chars": 1821,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/bge-small-zh.po",
    "chars": 1743,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/e5-large-v2.po",
    "chars": 1740,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/gte-base.po",
    "chars": 1691,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/gte-large.po",
    "chars": 1707,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/index.po",
    "chars": 838,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/jina-embeddings-v2-base-en.po",
    "chars": 1961,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/jina-embeddings-v2-small-en.po",
    "chars": 1976,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/embedding/multilingual-e5-large.po",
    "chars": 1890,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/flux.1-dev.po",
    "chars": 1464,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/flux.1-schnell.po",
    "chars": 1508,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/index.po",
    "chars": 879,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/kolors.po",
    "chars": 1437,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/sd-turbo.po",
    "chars": 1436,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/sd3-medium.po",
    "chars": 1496,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/sdxl-turbo.po",
    "chars": 1458,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/stable-diffusion-2-inpainting.po",
    "chars": 1667,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/stable-diffusion-inpainting.po",
    "chars": 1642,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/stable-diffusion-v1.5.po",
    "chars": 1656,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/stable-diffusion-xl-base-1.0.po",
    "chars": 1696,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/image/stable-diffusion-xl-inpainting.po",
    "chars": 1684,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/index.po",
    "chars": 716,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/baichuan-2-chat.po",
    "chars": 3184,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/baichuan-2.po",
    "chars": 3087,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/baichuan-chat.po",
    "chars": 2300,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/baichuan.po",
    "chars": 3639,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/chatglm.po",
    "chars": 2867,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/chatglm2-32k.po",
    "chars": 2351,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/chatglm2.po",
    "chars": 3041,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/chatglm3-32k.po",
    "chars": 2357,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/chatglm3.po",
    "chars": 3013,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/code-llama-instruct.po",
    "chars": 6565,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/code-llama-python.po",
    "chars": 6280,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/code-llama.po",
    "chars": 5754,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/deepseek-chat.po",
    "chars": 4681,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/deepseek-coder-instruct.po",
    "chars": 6644,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/falcon-instruct.po",
    "chars": 3074,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/falcon.po",
    "chars": 2750,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/glaive-coder.po",
    "chars": 2332,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/gorilla-openfunctions-v1.po",
    "chars": 3462,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/gpt-2.po",
    "chars": 2155,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/index.po",
    "chars": 23867,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/internlm-20b.po",
    "chars": 2360,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/internlm-7b.po",
    "chars": 2368,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/internlm-chat-20b.po",
    "chars": 2508,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/internlm-chat-7b.po",
    "chars": 2425,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/llama-2-chat.po",
    "chars": 5882,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/llama-2.po",
    "chars": 5344,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/mistral-instruct-v0.1.po",
    "chars": 3406,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/mistral-instruct-v0.2.po",
    "chars": 3515,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/mistral-v0.1.po",
    "chars": 3223,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/mixtral-instruct-v0.1.po",
    "chars": 3395,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/mixtral-v0.1.po",
    "chars": 3148,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/openbuddy.po",
    "chars": 2417,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/openhermes-2.5.po",
    "chars": 3144,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/opt.po",
    "chars": 2124,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/orca.po",
    "chars": 3360,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/qwen-chat.po",
    "chars": 7922,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/starchat-beta.po",
    "chars": 2304,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/starcoder.po",
    "chars": 2254,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/starcoderplus.po",
    "chars": 2310,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/tiny-llama.po",
    "chars": 2425,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/vicuna-v1.3.po",
    "chars": 5534,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/vicuna-v1.5-16k.po",
    "chars": 3014,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/vicuna-v1.5.po",
    "chars": 2892,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/wizardcoder-python-v1.0.po",
    "chars": 6395,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/wizardlm-v1.0.po",
    "chars": 3092,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/wizardmath-v1.0.po",
    "chars": 3788,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/xverse-chat.po",
    "chars": 2999,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/xverse.po",
    "chars": 3575,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/yi-200k.po",
    "chars": 2920,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/yi-chat.po",
    "chars": 2994,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/yi.po",
    "chars": 3472,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/zephyr-7b-alpha.po",
    "chars": 2421,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/llm/zephyr-7b-beta.po",
    "chars": 2400,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/rerank/bge-reranker-base.po",
    "chars": 1349,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/rerank/bge-reranker-large.po",
    "chars": 1359,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/rerank/index.po",
    "chars": 828,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/video/cogvideox-2b.po",
    "chars": 1360,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/builtin/video/index.po",
    "chars": 879,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/custom.po",
    "chars": 16197,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/index.po",
    "chars": 6347,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/lora.po",
    "chars": 3132,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_abilities/audio.po",
    "chars": 30230,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_abilities/chat.po",
    "chars": 7978,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_abilities/embed.po",
    "chars": 4768,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_abilities/flexible.po",
    "chars": 10050,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_abilities/image.po",
    "chars": 17890,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_abilities/index.po",
    "chars": 726,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_abilities/multimodal.po",
    "chars": 8161,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_abilities/rerank.po",
    "chars": 1666,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_abilities/tools.po",
    "chars": 6717,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_abilities/video.po",
    "chars": 9698,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_memory.po",
    "chars": 3869,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/model_update.po",
    "chars": 2185,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2025, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/source/source.po",
    "chars": 4139,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/sources/sources.po",
    "chars": 2669,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/virtualenv.po",
    "chars": 17514,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/xinference_model_hub.po",
    "chars": 9088,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2025, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/models/xinference_models_hub.po",
    "chars": 10329,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2025, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/reference/index.po",
    "chars": 14246,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/reference.po",
    "chars": 713,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/user_guide/auth_system.po",
    "chars": 9379,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/user_guide/backends.po",
    "chars": 15128,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/user_guide/cache_management.po",
    "chars": 726,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/user_guide/client_api.po",
    "chars": 6017,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/user_guide/continuous_batching.po",
    "chars": 6243,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/user_guide/distributed_inference.po",
    "chars": 4823,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/user_guide/index.po",
    "chars": 709,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/user_guide/launch.po",
    "chars": 7851,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2025, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/locale/zh_CN/LC_MESSAGES/user_guide/vllm_enhancement.po",
    "chars": 2561,
    "preview": "# SOME DESCRIPTIVE TITLE.\n# Copyright (C) 2023, Xorbits Inc.\n# This file is distributed under the same license as the Xi"
  },
  {
    "path": "doc/source/models/builtin/audio/belle-distilwhisper-large-v2-zh.rst",
    "chars": 512,
    "preview": ".. _models_builtin_belle-distilwhisper-large-v2-zh:\n\n===============================\nBelle-distilwhisper-large-v2-zh\n==="
  },
  {
    "path": "doc/source/models/builtin/audio/belle-whisper-large-v2-zh.rst",
    "chars": 470,
    "preview": ".. _models_builtin_belle-whisper-large-v2-zh:\n\n=========================\nBelle-whisper-large-v2-zh\n====================="
  },
  {
    "path": "doc/source/models/builtin/audio/belle-whisper-large-v3-zh.rst",
    "chars": 470,
    "preview": ".. _models_builtin_belle-whisper-large-v3-zh:\n\n=========================\nBelle-whisper-large-v3-zh\n====================="
  },
  {
    "path": "doc/source/models/builtin/audio/chattts.rst",
    "chars": 366,
    "preview": ".. _models_builtin_chattts:\n\n=======\nChatTTS\n=======\n\n- **Model Name:** ChatTTS\n- **Model Family:** ChatTTS\n- **Abilitie"
  },
  {
    "path": "doc/source/models/builtin/audio/cosyvoice-300m-instruct.rst",
    "chars": 485,
    "preview": ".. _models_builtin_cosyvoice-300m-instruct:\n\n=======================\nCosyVoice-300M-Instruct\n=======================\n\n- "
  }
]

// ... and 1435 more files (download for full content)