diff --git a/doc/source/conf.py b/doc/source/conf.py index 0c5441f239..001e744e3e 100644 --- a/doc/source/conf.py +++ b/doc/source/conf.py @@ -121,7 +121,7 @@ "type": "fontawesome", }]) html_theme_options["external_links"] = [ - {"name": "产品官网", "url": "https://xorbits.cn/inference"}, + {"name": "产品官网", "url": "https://xorbits.cn"}, ] html_favicon = "_static/favicon.svg" diff --git a/xinference/constants.py b/xinference/constants.py index 66e9983a93..bcb815b2fe 100644 --- a/xinference/constants.py +++ b/xinference/constants.py @@ -39,6 +39,7 @@ def get_xinference_home() -> str: # if user has already set `XINFERENCE_HOME` env, change huggingface and modelscope default download path os.environ["HUGGINGFACE_HUB_CACHE"] = os.path.join(home_path, "huggingface") os.environ["MODELSCOPE_CACHE"] = os.path.join(home_path, "modelscope") + os.environ["XDG_CACHE_HOME"] = os.path.join(home_path, "openmind_hub") # In multi-tenant mode, # gradio's temporary files are stored in their respective home directories, # to prevent insufficient permissions diff --git a/xinference/core/worker.py b/xinference/core/worker.py index 567bcf14f0..ca156c7354 100644 --- a/xinference/core/worker.py +++ b/xinference/core/worker.py @@ -785,7 +785,9 @@ async def launch_builtin_model( peft_model_config: Optional[PeftModelConfig] = None, request_limits: Optional[int] = None, gpu_idx: Optional[Union[int, List[int]]] = None, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, model_path: Optional[str] = None, **kwargs, ): diff --git a/xinference/model/audio/core.py b/xinference/model/audio/core.py index 5672b216d4..fe2daaeb1f 100644 --- a/xinference/model/audio/core.py +++ b/xinference/model/audio/core.py @@ -100,7 +100,9 @@ def generate_audio_description( def match_audio( model_name: str, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, ) -> AudioModelFamilyV1: from ..utils import download_from_modelscope from . import BUILTIN_AUDIO_MODELS, MODELSCOPE_AUDIO_MODELS @@ -152,7 +154,9 @@ def create_audio_model_instance( devices: List[str], model_uid: str, model_name: str, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, model_path: Optional[str] = None, **kwargs, ) -> Tuple[ diff --git a/xinference/model/core.py b/xinference/model/core.py index 4591d255b0..1cccbb6f2d 100644 --- a/xinference/model/core.py +++ b/xinference/model/core.py @@ -55,7 +55,9 @@ def create_model_instance( model_size_in_billions: Optional[Union[int, str]] = None, quantization: Optional[str] = None, peft_model_config: Optional[PeftModelConfig] = None, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, model_path: Optional[str] = None, **kwargs, ) -> Tuple[Any, ModelDescription]: diff --git a/xinference/model/embedding/core.py b/xinference/model/embedding/core.py index 8bb6c05626..ae66b945b2 100644 --- a/xinference/model/embedding/core.py +++ b/xinference/model/embedding/core.py @@ -433,7 +433,9 @@ def encode( def match_embedding( model_name: str, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, ) -> EmbeddingModelSpec: from ..utils import download_from_modelscope from . import BUILTIN_EMBEDDING_MODELS, MODELSCOPE_EMBEDDING_MODELS @@ -469,7 +471,9 @@ def create_embedding_model_instance( devices: List[str], model_uid: str, model_name: str, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, model_path: Optional[str] = None, **kwargs, ) -> Tuple[EmbeddingModel, EmbeddingModelDescription]: diff --git a/xinference/model/image/core.py b/xinference/model/image/core.py index 581358b789..432a70c1a4 100644 --- a/xinference/model/image/core.py +++ b/xinference/model/image/core.py @@ -125,7 +125,9 @@ def generate_image_description( def match_diffusion( model_name: str, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, ) -> ImageModelFamilyV1: from ..utils import download_from_modelscope from . import BUILTIN_IMAGE_MODELS, MODELSCOPE_IMAGE_MODELS @@ -213,7 +215,9 @@ def create_image_model_instance( model_uid: str, model_name: str, peft_model_config: Optional[PeftModelConfig] = None, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, model_path: Optional[str] = None, **kwargs, ) -> Tuple[ diff --git a/xinference/model/llm/__init__.py b/xinference/model/llm/__init__.py index 88b0e49651..cc7b452694 100644 --- a/xinference/model/llm/__init__.py +++ b/xinference/model/llm/__init__.py @@ -32,6 +32,7 @@ BUILTIN_LLM_MODEL_TOOL_CALL_FAMILIES, BUILTIN_LLM_PROMPT_STYLE, BUILTIN_MODELSCOPE_LLM_FAMILIES, + BUILTIN_OPENMIND_HUB_LLM_FAMILIES, LLAMA_CLASSES, LLM_ENGINES, LMDEPLOY_CLASSES, @@ -258,6 +259,36 @@ def _install(): if "tools" in model_spec.model_ability: BUILTIN_LLM_MODEL_TOOL_CALL_FAMILIES.add(model_spec.model_name) + openmind_hub_json_path = os.path.join( + os.path.dirname(os.path.abspath(__file__)), "llm_family_openmind_hub.json" + ) + for json_obj in json.load( + codecs.open(openmind_hub_json_path, "r", encoding="utf-8") + ): + model_spec = LLMFamilyV1.parse_obj(json_obj) + BUILTIN_OPENMIND_HUB_LLM_FAMILIES.append(model_spec) + + # register prompt style, in case that we have something missed + # if duplicated with huggingface json, keep it as the huggingface style + + if ( + "chat" in model_spec.model_ability + and isinstance(model_spec.chat_template, str) + and model_spec.model_name not in BUILTIN_LLM_PROMPT_STYLE + ): + BUILTIN_LLM_PROMPT_STYLE[model_spec.model_name] = { + "chat_template": model_spec.chat_template, + "stop_token_ids": model_spec.stop_token_ids, + "stop": model_spec.stop, + } + # register model family + if "chat" in model_spec.model_ability: + BUILTIN_LLM_MODEL_CHAT_FAMILIES.add(model_spec.model_name) + else: + BUILTIN_LLM_MODEL_GENERATE_FAMILIES.add(model_spec.model_name) + if "tools" in model_spec.model_ability: + BUILTIN_LLM_MODEL_TOOL_CALL_FAMILIES.add(model_spec.model_name) + csghub_json_path = os.path.join( os.path.dirname(os.path.abspath(__file__)), "llm_family_csghub.json" ) @@ -288,6 +319,7 @@ def _install(): for llm_specs in [ BUILTIN_LLM_FAMILIES, BUILTIN_MODELSCOPE_LLM_FAMILIES, + BUILTIN_OPENMIND_HUB_LLM_FAMILIES, BUILTIN_CSGHUB_LLM_FAMILIES, ]: for llm_spec in llm_specs: @@ -298,6 +330,7 @@ def _install(): for families in [ BUILTIN_LLM_FAMILIES, BUILTIN_MODELSCOPE_LLM_FAMILIES, + BUILTIN_OPENMIND_HUB_LLM_FAMILIES, BUILTIN_CSGHUB_LLM_FAMILIES, ]: for family in families: diff --git a/xinference/model/llm/core.py b/xinference/model/llm/core.py index d4f1ed32c1..9141dc1106 100644 --- a/xinference/model/llm/core.py +++ b/xinference/model/llm/core.py @@ -193,7 +193,9 @@ def create_llm_model_instance( model_size_in_billions: Optional[Union[int, str]] = None, quantization: Optional[str] = None, peft_model_config: Optional[PeftModelConfig] = None, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, model_path: Optional[str] = None, **kwargs, ) -> Tuple[LLM, LLMDescription]: diff --git a/xinference/model/llm/llm_family.py b/xinference/model/llm/llm_family.py index 413b4229ae..513a13c5f8 100644 --- a/xinference/model/llm/llm_family.py +++ b/xinference/model/llm/llm_family.py @@ -41,6 +41,7 @@ create_symlink, download_from_csghub, download_from_modelscope, + download_from_openmind_hub, is_valid_model_uri, parse_uri, retry_download, @@ -239,6 +240,7 @@ def parse_raw( BUILTIN_LLM_FAMILIES: List["LLMFamilyV1"] = [] BUILTIN_MODELSCOPE_LLM_FAMILIES: List["LLMFamilyV1"] = [] +BUILTIN_OPENMIND_HUB_LLM_FAMILIES: List["LLMFamilyV1"] = [] BUILTIN_CSGHUB_LLM_FAMILIES: List["LLMFamilyV1"] = [] SGLANG_CLASSES: List[Type[LLM]] = [] @@ -301,6 +303,9 @@ def cache( elif llm_spec.model_hub == "modelscope": logger.info(f"Caching from Modelscope: {llm_spec.model_id}") return cache_from_modelscope(llm_family, llm_spec, quantization) + elif llm_spec.model_hub == "openmind_hub": + logger.info(f"Caching from openmind_hub: {llm_spec.model_id}") + return cache_from_openmind_hub(llm_family, llm_spec, quantization) elif llm_spec.model_hub == "csghub": logger.info(f"Caching from CSGHub: {llm_spec.model_id}") return cache_from_csghub(llm_family, llm_spec, quantization) @@ -474,7 +479,7 @@ def _skip_download( model_revision: Optional[str], quantization: Optional[str] = None, ) -> bool: - if model_format == "pytorch": + if model_format in ["pytorch", "mindspore"]: model_hub_to_meta_path = { "huggingface": _get_meta_path( cache_dir, model_format, "huggingface", quantization @@ -482,6 +487,9 @@ def _skip_download( "modelscope": _get_meta_path( cache_dir, model_format, "modelscope", quantization ), + "openmind_hub": _get_meta_path( + cache_dir, model_format, "openmind_hub", quantization + ), "csghub": _get_meta_path(cache_dir, model_format, "csghub", quantization), } if valid_model_revision(model_hub_to_meta_path[model_hub], model_revision): @@ -702,6 +710,50 @@ def cache_from_modelscope( return cache_dir +def cache_from_openmind_hub( + llm_family: LLMFamilyV1, + llm_spec: "LLMSpecV1", + quantization: Optional[str] = None, +) -> str: + """ + Cache model from openmind_hub. Return the cache directory. + """ + from openmind_hub import snapshot_download + + cache_dir = _get_cache_dir(llm_family, llm_spec) + if _skip_download( + cache_dir, + llm_spec.model_format, + llm_spec.model_hub, + llm_spec.model_revision, + quantization, + ): + return cache_dir + + if llm_spec.model_format in ["pytorch", "mindspore"]: + download_dir = retry_download( + snapshot_download, + llm_family.model_name, + { + "model_size": llm_spec.model_size_in_billions, + "model_format": llm_spec.model_format, + }, + llm_spec.model_id, + revision=llm_spec.model_revision, + ) + create_symlink(download_dir, cache_dir) + + else: + raise ValueError(f"Unsupported format: {llm_spec.model_format}") + + meta_path = _get_meta_path( + cache_dir, llm_spec.model_format, llm_spec.model_hub, quantization + ) + _generate_meta_file(meta_path, llm_family, llm_spec, quantization) + + return cache_dir + + def cache_from_huggingface( llm_family: LLMFamilyV1, llm_spec: "LLMSpecV1", @@ -893,7 +945,9 @@ def match_llm( model_format: Optional[str] = None, model_size_in_billions: Optional[Union[int, str]] = None, quantization: Optional[str] = None, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, ) -> Optional[Tuple[LLMFamilyV1, LLMSpecV1, str]]: """ Find an LLM family, spec, and quantization that satisfy given criteria. @@ -924,6 +978,12 @@ def _apply_format_to_model_id(spec: LLMSpecV1, q: str) -> LLMSpecV1: + BUILTIN_LLM_FAMILIES + user_defined_llm_families ) + elif download_hub == "openmind_hub": + all_families = ( + BUILTIN_OPENMIND_HUB_LLM_FAMILIES + + BUILTIN_LLM_FAMILIES + + user_defined_llm_families + ) elif download_hub == "csghub": all_families = ( BUILTIN_CSGHUB_LLM_FAMILIES @@ -938,6 +998,12 @@ def _apply_format_to_model_id(spec: LLMSpecV1, q: str) -> LLMSpecV1: + BUILTIN_LLM_FAMILIES + user_defined_llm_families ) + elif download_from_openmind_hub(): + all_families = ( + BUILTIN_OPENMIND_HUB_LLM_FAMILIES + + BUILTIN_LLM_FAMILIES + + user_defined_llm_families + ) elif download_from_csghub(): all_families = ( BUILTIN_CSGHUB_LLM_FAMILIES diff --git a/xinference/model/llm/llm_family_openmind_hub.json b/xinference/model/llm/llm_family_openmind_hub.json new file mode 100644 index 0000000000..60ade3db51 --- /dev/null +++ b/xinference/model/llm/llm_family_openmind_hub.json @@ -0,0 +1,1359 @@ +[ + { + "version": 1, + "context_length": 32768, + "model_name": "internlm2-chat", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat" + ], + "model_description": "The second generation of the InternLM model, InternLM2.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "none" + ], + "model_id": "PyTorch-NPU/internlm2_chat_7b", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 20, + "quantizations": [ + "none" + ], + "model_id": "AI-Research/internlm2-chat-20b", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{{ '' }}{% for message in messages %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", + "stop_token_ids": [ + 2, + 92542 + ], + "stop": [ + "", + "<|im_end|>" + ] + }, + { + "version": 1, + "context_length": 4096, + "model_name": "baichuan-2-chat", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat" + ], + "model_description": "Baichuan2-chat is a fine-tuned version of the Baichuan LLM, specializing in chatting.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "Baichuan/Baichuan2_7b_chat_pt", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 13, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "Baichuan/Baichuan2_13b_chat_pt", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{{ (messages|selectattr('role', 'equalto', 'system')|list|last).content|trim if (messages|selectattr('role', 'equalto', 'system')|list) else '' }}\n\n{% for message in messages %}\n{% if message['role'] == 'user' %}\n\n{{ message['content']|trim -}}\n{% if not loop.last %}\n\n\n{% endif %}\n{% elif message['role'] == 'assistant' %}\n\n{{ message['content']|trim -}}\n{% if not loop.last %}\n\n\n{% endif %}\n{% endif %}\n{% endfor %}\n{% if add_generation_prompt and messages[-1]['role'] != 'assistant' %}\n\n{% endif %}", + "stop_token_ids": [ + 2, + 195 + ], + "stop": [] + }, + { + "version": 1, + "context_length": 4096, + "model_name": "baichuan-2", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "generate" + ], + "model_description": "Baichuan2 is an open-source Transformer based LLM that is trained on both Chinese and English data.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "PyTorch-NPU/baichuan2_7b_base", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 13, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "Baichuan/Baichuan2_13b_base_pt", + "model_hub": "openmind_hub" + } + ] + }, + { + "version": 1, + "context_length": 32768, + "model_name": "qwen1.5-chat", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat", + "tools" + ], + "model_description": "Qwen1.5 is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "PyTorch-NPU/qwen1.5_7b_chat", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{%- macro json_to_python_type(json_spec) %}\n {%- set basic_type_map = {\n \"string\": \"str\",\n \"number\": \"float\",\n \"integer\": \"int\",\n \"boolean\": \"bool\"\n} %}\n {%- if basic_type_map[json_spec.type] is defined %}\n {{- basic_type_map[json_spec.type] }}\n {%- elif json_spec.type == \"array\" %}\n {{- \"list[\" + json_to_python_type(json_spec|items) + \"]\" }}\n {%- elif json_spec.type == \"object\" %}\n {%- if json_spec.additionalProperties is defined %}\n {{- \"dict[str, \" + json_to_python_type(json_spec.additionalProperties) + ']' }}\n {%- else %}\n {{- \"dict\" }}\n {%- endif %}\n {%- elif json_spec.type is iterable %}\n {{- \"Union[\" }}\n {%- for t in json_spec.type %}\n {{- json_to_python_type({\"type\": t}) }}\n {%- if not loop.last %}\n {{- \",\" }}\n {%- endif %}\n {%- endfor %}\n {{- \"]\" }}\n {%- else %}\n {{- \"Any\" }}\n {%- endif %}\n{%- endmacro %}\n\n{%- if tools %}\n {{- '<|im_start|>system\n' }}\n {%- if messages[0]['role'] == 'system' %}\n {{- messages[0]['content'] + '\n\n' }}\n {%- endif %}\n {{- '# Tools\n\n' }}\n {{- \"You are a function calling AI model. You are provided with function signatures within XML tags. You may call one or more functions to assist with the user query. Don't make assumptions about what values to plug into functions. Here are the available tools: \" }}\n {%- for tool in tools %}\n {%- if tool.function is defined %}\n {%- set tool = tool.function %}\n {%- endif %}\n {{- '{\"type\": \"function\", \"function\": ' }}\n {{- '{\"name\": ' + tool.name + '\", ' }}\n {{- '\"description\": \"' + tool.name + '(' }}\n {%- for param_name, param_fields in tool.parameters.properties|items %}\n {{- param_name + \": \" + json_to_python_type(param_fields) }}\n {%- if not loop.last %}\n {{- \", \" }}\n {%- endif %}\n {%- endfor %}\n {{- \")\" }}\n {%- if tool.return is defined %}\n {{- \" -> \" + json_to_python_type(tool.return) }}\n {%- endif %}\n {{- \" - \" + tool.description + \"\n\n\" }}\n {%- for param_name, param_fields in tool.parameters.properties|items %}\n {%- if loop.first %}\n {{- \" Args:\n\" }}\n {%- endif %}\n {{- \" \" + param_name + \"(\" + json_to_python_type(param_fields) + \"): \" + param_fields.description|trim }}\n {%- endfor %}\n {%- if tool.return is defined and tool.return.description is defined %}\n {{- \"\n Returns:\n \" + tool.return.description }}\n {%- endif %}\n {{- '\"' }}\n {{- ', \"parameters\": ' }}\n {%- if tool.parameters.properties | length == 0 %}\n {{- \"{}\" }}\n {%- else %}\n {{- tool.parameters|tojson }}\n {%- endif %}\n {{- \"}\" }}\n {%- if not loop.last %}\n {{- \"\n\" }}\n {%- endif %}\n {%- endfor %}\n {{- \" \" }}\n {{- 'Use the following pydantic model json schema for each tool call you will make: {\"properties\": {\"arguments\": {\"title\": \"Arguments\", \"type\": \"object\"}, \"name\": {\"title\": \"Name\", \"type\": \"string\"}}, \"required\": [\"arguments\", \"name\"], \"title\": \"FunctionCall\", \"type\": \"object\"}\n' }}\n {{- \"For each function call return a json object with function name and arguments within XML tags as follows:\n\" }}\n {{- \"\n\" }}\n {{- '{\"name\": , \"arguments\": }\n' }}\n {{- '<|im_end|>\n' }}\n{%- else %}\n {%- if messages[0]['role'] != 'system' %}\n {{- '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}\n {%- else %}\n {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}\n {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n {%- if message.role == \"user\" or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and message.tool_calls is not defined) %}\n {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}\n {%- elif message.role == \"assistant\" %}\n {{- '<|im_start|>' + message.role + '\n\n' }}\n {%- for tool_call in message.tool_calls %}\n {%- if tool_call.function is defined %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '{' }}\n {{- '\"name\": \"' }}\n {{- tool_call.name }}\n {%- if tool_call.arguments is defined %}\n {{- ', ' }}\n {{- '\"arguments\": ' }}\n {{- tool_call.arguments|tojson }}\n {%- endif %}\n {{- '\"}' }}\n {{- '\n' }}\n {%- endfor %}\n {{- '<|im_end|>\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if not message.name is defined %}\n {{- raise_exception(\"Tool response dicts require a 'name' key indicating the name of the called function!\") }}\n {%- endif %}\n {{- '<|im_start|>user\n\n' }}\n {{- '{\"name\": \"' }}\n {{- message.name }}\n {{- '\", \"content\": ' }}\n {{- message.content|tojson + '}' }}\n {{- '\n<|im_end|>\n' }}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\n' }}\n{%- endif %}", + "stop_token_ids": [ + 151643, + 151644, + 151645 + ], + "stop": [ + "<|endoftext|>", + "<|im_start|>", + "<|im_end|>" + ] + }, + { + "version": 1, + "context_length": 65536, + "model_name": "qwen1.5", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "generate" + ], + "model_description": "CodeQwen1.5 is the Code-Specific version of Qwen1.5. It is a transformer-based decoder-only language model pretrained on a large amount of data of codes.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "PyTorch-NPU/qwen1.5_7b", + "model_hub": "openmind_hub" + } + ] + }, + { + "version": 1, + "context_length": 131072, + "model_name": "glm4-chat", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat", + "tools" + ], + "model_description": "GLM4 is the open source version of the latest generation of pre-trained models in the GLM-4 series launched by Zhipu AI.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 9, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "AI-Research/glm-4-9b-chat", + "model_hub": "openmind_hub" + } + ], + "chat_template": "[gMASK]{% for item in messages %}{% if item['tools'] is defined %}<|system|>\n你是一个名为 ChatGLM 的人工智能助手。你是基于智谱AI训练的语言模型 GLM-4 模型开发的，你的任务是针对用户的问题和要求提供适当的答复和支持。\n\n# 可用工具{% set tools = item['tools'] %}{% for tool in tools %}{% if tool['type'] == 'function' %}\n\n## {{ tool['function']['name'] }}\n\n{{ tool['function'] | tojson(indent=4) }}\n在调用上述函数时，请使用 Json 格式表示调用的参数。{% elif tool['type'] == 'python' %}\n\n## python\n\n当你向 `python` 发送包含 Python 代码的消息时，该代码将会在一个有状态的 Jupyter notebook 环境中执行。\n`python` 返回代码执行的输出，或在执行 60 秒后返回超时。\n`/mnt/data` 将会持久化存储你的文件。在此会话中，`python` 无法访问互联网。不要使用 `python` 进行任何网络请求或者在线 API 调用，这些在线内容的访问将不会成功。{% elif tool['type'] == 'simple_browser' %}\n\n## simple_browser\n\n你可以使用 `simple_browser` 工具。该工具支持以下函数：\n`search(query: str, recency_days: int)`：使用搜索引擎进行查询并显示结果，可以使用 `recency_days` 参数控制搜索内容的时效性。\n`mclick(ids: list[int])`：获取一系列指定 id 的页面内容。每次调用时，须选择3-10个页面。选择多个角度的页面，同时尽可能选择可信任的信息来源。考虑到部分页面是无法加载的，你也可以多打开一些可能有用的页面而不用担心内容过多。\n`open_url(url: str)`：打开指定的 URL。\n\n使用 `【{引用 id}†{引用文本}】` 来引用内容。\n\n操作步骤：1. 使用 `search` 来获得信息列表; 2. 使用 `mclick` 来获取指定 ID 页面的内容; 3. 根据获得的内容进行回复。在回复中应当引用信息来源。\n 如果用户提供了 URL，也可以用 `open_url` 直接打开页面。\n如果初次搜索结果没有找到合适的信息，也可以再次使用 `search` 进行搜索。{% elif tool['type'] == 'cogview' %}\n\n## cogview\n\n如果用户的请求中包含了对图像的描述，你可以使用 `cogview` 来生成图像并展示给用户。你需要向 `cogview` 发送图像描述，规则：\n- 发送给 `cogview` 的消息必须使用英语。用户的中文描述必须完全翻译为英语。\n- 应当尽可能详细地描述图像生成的需求，需求描述约 100 英文单词。\n- 保持用户原始描述的意图。不要虚构内容或者没见过的人物。\n- 如无特殊说明，所在地为中国，持有中国立场并遵循中国社会主义价值观。{% endif %}{% endfor %}{% endif %}{% if item['content'] %}<|{{ item['role'] }}|>{{ item['metadata'] }}\n{{ item['content'] }}{% endif %}{% endfor %}{% if add_generation_prompt %}<|assistant|>{% endif %}", + "stop_token_ids": [ + 151329, + 151336, + 151338 + ], + "stop": [ + "<|endoftext|>", + "<|user|>", + "<|observation|>" + ] + }, + { + "version": 1, + "context_length": 1048576, + "model_name": "glm4-chat-1m", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat", + "tools" + ], + "model_description": "GLM4 is the open source version of the latest generation of pre-trained models in the GLM-4 series launched by Zhipu AI.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 9, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "AI-Research/glm-4-9b-chat-1m", + "model_hub": "openmind_hub" + } + ], + "chat_template": "[gMASK]{% for item in messages %}{% if item['tools'] is defined %}<|system|>\n你是一个名为 GLM-4 的人工智能助手。你是基于智谱AI训练的语言模型 GLM-4 模型开发的，你的任务是针对用户的问题和要求提供适当的答复和支持。\n\n# 可用工具{% set tools = item['tools'] %}{% for tool in tools %}{% if tool['type'] == 'function' %}\n\n## {{ tool['function']['name'] }}\n\n{{ tool['function'] | tojson(indent=4) }}\n在调用上述函数时，请使用 Json 格式表示调用的参数。{% elif tool['type'] == 'python' %}\n\n## python\n\n当你向 `python` 发送包含 Python 代码的消息时，该代码将会在一个有状态的 Jupyter notebook 环境中执行。\n`python` 返回代码执行的输出，或在执行 60 秒后返回超时。\n`/mnt/data` 将会持久化存储你的文件。在此会话中，`python` 无法访问互联网。不要使用 `python` 进行任何网络请求或者在线 API 调用，这些在线内容的访问将不会成功。{% elif tool['type'] == 'simple_browser' %}\n\n## simple_browser\n\n你可以使用 `simple_browser` 工具。该工具支持以下函数：\n`search(query: str, recency_days: int)`：使用搜索引擎进行查询并显示结果，可以使用 `recency_days` 参数控制搜索内容的时效性。\n`mclick(ids: list[int])`：获取一系列指定 id 的页面内容。每次调用时，须选择3-10个页面。选择多个角度的页面，同时尽可能选择可信任的信息来源。考虑到部分页面是无法加载的，你也可以多打开一些可能有用的页面而不用担心内容过多。\n`open_url(url: str)`：打开指定的 URL。\n\n使用 `【{引用 id}†{引用文本}】` 来引用内容。\n\n操作步骤：1. 使用 `search` 来获得信息列表; 2. 使用 `mclick` 来获取指定 ID 页面的内容; 3. 根据获得的内容进行回复。在回复中应当引用信息来源。\n 如果用户提供了 URL，也可以用 `open_url` 直接打开页面。\n如果初次搜索结果没有找到合适的信息，也可以再次使用 `search` 进行搜索。{% elif tool['type'] == 'cogview' %}\n\n## cogview\n\n如果用户的请求中包含了对图像的描述，你可以使用 `cogview` 来生成图像并展示给用户。你需要向 `cogview` 发送图像描述，规则：\n- 发送给 `cogview` 的消息必须使用英语。用户的中文描述必须完全翻译为英语。\n- 应当尽可能详细地描述图像生成的需求，需求描述约 100 英文单词。\n- 保持用户原始描述的意图。不要虚构内容或者没见过的人物。\n- 如无特殊说明，所在地为中国，持有中国立场并遵循中国社会主义价值观。{% endif %}{% endfor %}{% endif %}{% if item['content'] %}<|{{ item['role'] }}|>{{ item['metadata'] }}\n{{ item['content'] }}{% endif %}{% endfor %}{% if add_generation_prompt %}<|assistant|>{% endif %}", + "stop_token_ids": [ + 151329, + 151336, + 151338 + ], + "stop": [ + "<|endoftext|>", + "<|user|>", + "<|observation|>" + ] + }, + { + "version": 1, + "context_length": 8192, + "model_name": "glm-4v", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat", + "vision" + ], + "model_description": "GLM4 is the open source version of the latest generation of pre-trained models in the GLM-4 series launched by Zhipu AI.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 9, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "AI-Research/glm-4v-9b", + "model_hub": "openmind_hub" + } + ], + "chat_template": "", + "stop_token_ids": [ + 151329, + 151336, + 151338 + ], + "stop": [ + "<|endoftext|>", + "<|user|>", + "<|observation|>" + ] + }, + { + "version": 1, + "context_length": 8192, + "model_name": "llama-3-instruct", + "model_lang": [ + "en" + ], + "model_ability": [ + "chat" + ], + "model_description": "The Llama 3 instruction tuned models are optimized for dialogue use cases and outperform many of the available open source chat models on common industry benchmarks..", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 8, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "wuhaicc/Meta-Llama-3-8B-Instruct", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 70, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "wuhaicc/Meta-Llama-3-70B-Instruct", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = '<|begin_of_text|>' + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% endif %}", + "stop_token_ids": [ + 128001, + 128009 + ], + "stop": [ + "<|end_of_text|>", + "<|eot_id|>" + ] + }, + { + "version": 1, + "context_length": 131072, + "model_name": "llama-3.1", + "model_lang": [ + "en", + "de", + "fr", + "it", + "pt", + "hi", + "es", + "th" + ], + "model_ability": [ + "generate" + ], + "model_description": "Llama 3.1 is an auto-regressive language model that uses an optimized transformer architecture", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 8, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "AI-Research/Meta-Llama-3.1-8B", + "model_hub": "openmind_hub" + } + ] + }, + { + "version": 1, + "context_length": 131072, + "model_name": "llama-3.1-instruct", + "model_lang": [ + "en", + "de", + "fr", + "it", + "pt", + "hi", + "es", + "th" + ], + "model_ability": [ + "chat", + "tools" + ], + "model_description": "The Llama 3.1 instruction tuned models are optimized for dialogue use cases and outperform many of the available open source chat models on common industry benchmarks..", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 8, + "quantizations": [ + "none" + ], + "model_id": "AI-Research/Meta-Llama-3.1-8B-Instruct", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{{- bos_token }}\n{%- if custom_tools is defined %}\n {%- set tools = custom_tools %}\n{%- endif %}\n{%- if not tools_in_user_message is defined %}\n {%- set tools_in_user_message = true %}\n{%- endif %}\n{%- if not date_string is defined %}\n {%- set date_string = \"26 Jul 2024\" %}\n{%- endif %}\n{%- if not tools is defined %}\n {%- set tools = none %}\n{%- endif %}\n\n{#- This block extracts the system message, so we can slot it into the right place. #}\n{%- if messages[0]['role'] == 'system' %}\n {%- set system_message = messages[0]['content']|trim %}\n {%- set messages = messages[1:] %}\n{%- else %}\n {%- set system_message = \"\" %}\n{%- endif %}\n\n{#- System message + builtin tools #}\n{{- \"<|start_header_id|>system<|end_header_id|>\\n\\n\" }}\n{%- if builtin_tools is defined or tools is not none %}\n {{- \"Environment: ipython\\n\" }}\n{%- endif %}\n{%- if builtin_tools is defined %}\n {{- \"Tools: \" + builtin_tools | reject('equalto', 'code_interpreter') | join(\", \") + \"\\n\\n\"}}\n{%- endif %}\n{{- \"Cutting Knowledge Date: December 2023\\n\" }}\n{{- \"Today Date: \" + date_string + \"\\n\\n\" }}\n{%- if tools is not none and not tools_in_user_message %}\n {{- \"You have access to the following functions. To call a function, please respond with JSON for a function call.\" }}\n {{- 'Respond in the format {\"name\": function name, \"parameters\": dictionary of argument name and its value}.' }}\n {{- \"Do not use variables.\\n\\n\" }}\n {%- for t in tools %}\n {{- t | tojson(indent=4) }}\n {{- \"\\n\\n\" }}\n {%- endfor %}\n{%- endif %}\n{{- system_message }}\n{{- \"<|eot_id|>\" }}\n\n{#- Custom tools are passed in a user message with some extra guidance #}\n{%- if tools_in_user_message and not tools is none %}\n {#- Extract the first user message so we can plug it in here #}\n {%- if messages | length != 0 %}\n {%- set first_user_message = messages[0]['content']|trim %}\n {%- set messages = messages[1:] %}\n {%- else %}\n {{- raise_exception(\"Cannot put tools in the first user message when there's no first user message!\") }}\n{%- endif %}\n {{- '<|start_header_id|>user<|end_header_id|>\\n\\n' -}}\n {{- \"Given the following functions, please respond with a JSON for a function call \" }}\n {{- \"with its proper arguments that best answers the given prompt.\\n\\n\" }}\n {{- 'Respond in the format {\"name\": function name, \"parameters\": dictionary of argument name and its value}.' }}\n {{- \"Do not use variables.\\n\\n\" }}\n {%- for t in tools %}\n {{- t | tojson(indent=4) }}\n {{- \"\\n\\n\" }}\n {%- endfor %}\n {{- first_user_message + \"<|eot_id|>\"}}\n{%- endif %}\n\n{%- for message in messages %}\n {%- if not (message.role == 'ipython' or message.role == 'tool' or 'tool_calls' in message) %}\n {{- '<|start_header_id|>' + message['role'] + '<|end_header_id|>\\n\\n'+ message['content'] | trim + '<|eot_id|>' }}\n {%- elif 'tool_calls' in message %}\n {%- if not message.tool_calls|length == 1 %}\n {{- raise_exception(\"This model only supports single tool-calls at once!\") }}\n {%- endif %}\n {%- set tool_call = message.tool_calls[0].function %}\n {%- if builtin_tools is defined and tool_call.name in builtin_tools %}\n {{- '<|start_header_id|>assistant<|end_header_id|>\\n\\n' -}}\n {{- \"<|python_tag|>\" + tool_call.name + \".call(\" }}\n {%- for arg_name, arg_val in tool_call.arguments | items %}\n {{- arg_name + '=\"' + arg_val + '\"' }}\n {%- if not loop.last %}\n {{- \", \" }}\n {%- endif %}\n {%- endfor %}\n {{- \")\" }}\n {%- else %}\n {{- '<|start_header_id|>assistant<|end_header_id|>\\n\\n' -}}\n {{- '{\"name\": \"' + tool_call.name + '\", ' }}\n {{- '\"parameters\": ' }}\n {{- tool_call.arguments | tojson }}\n {{- \"}\" }}\n {%- endif %}\n {%- if builtin_tools is defined %}\n {#- This means we're in ipython mode #}\n {{- \"<|eom_id|>\" }}\n {%- else %}\n {{- \"<|eot_id|>\" }}\n {%- endif %}\n {%- elif message.role == \"tool\" or message.role == \"ipython\" %}\n {{- \"<|start_header_id|>ipython<|end_header_id|>\\n\\n\" }}\n {%- if message.content is mapping or message.content is iterable %}\n {{- message.content | tojson }}\n {%- else %}\n {{- message.content }}\n {%- endif %}\n {{- \"<|eot_id|>\" }}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|start_header_id|>assistant<|end_header_id|>\\n\\n' }}\n{%- endif %}\n", + "stop_token_ids": [ + 128001, + 128008, + 128009 + ], + "stop": [ + "<|end_of_text|>", + "<|eot_id|>", + "<|eom_id|>" + ] + }, + { + "version": 1, + "context_length": 32768, + "model_name": "qwen-chat", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat" + ], + "model_description": "Qwen-chat is a fine-tuned version of the Qwen LLM trained with alignment techniques, specializing in chatting.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": "1_8", + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "HangZhou_Ascend/Qwen-1_8B-Chat", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "wuhaicc/Qwen-7B-Chat", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 14, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "wuhaicc/Qwen-14B-Chat", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{% for item in messages %}{% if loop.first and item['role'] == 'system' %}{{ '<|im_start|>system\n' + item['content'] + '<|im_end|>\n' }}{% elif loop.first %}{{ '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}{% endif %}{% if item['role'] == 'user' %}{{ '<|im_start|>user\n' + item['content'] + '<|im_end|>' }}{% elif item['role'] == 'assistant' %}{{ '<|im_start|>assistant\n' + item['content'] + '<|im_end|>' }}{% endif %}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", + "stop_token_ids": [ + 151643, + 151644, + 151645 + ], + "stop": [ + "<|endoftext|>", + "<|im_start|>", + "<|im_end|>" + ] + }, + { + "version": 1, + "context_length": 32768, + "model_name": "qwen1.5-chat", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat", + "tools" + ], + "model_description": "Qwen1.5 is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": "0_5", + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "HangZhou_Ascend/Qwen1.5-0.5B-Chat", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 4, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "HangZhou_Ascend/Qwen1.5-4B-Chat", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "PyTorch-NPU/qwen1.5_7b_chat", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 14, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "State_Cloud/Qwen1.5-14B-Chat", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 32, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "State_Cloud/Qwen1.5-32b-chat", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 72, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "State_Cloud/Qwen1.5-72b-chat", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{%- macro json_to_python_type(json_spec) %}\n {%- set basic_type_map = {\n \"string\": \"str\",\n \"number\": \"float\",\n \"integer\": \"int\",\n \"boolean\": \"bool\"\n} %}\n {%- if basic_type_map[json_spec.type] is defined %}\n {{- basic_type_map[json_spec.type] }}\n {%- elif json_spec.type == \"array\" %}\n {{- \"list[\" + json_to_python_type(json_spec|items) + \"]\" }}\n {%- elif json_spec.type == \"object\" %}\n {%- if json_spec.additionalProperties is defined %}\n {{- \"dict[str, \" + json_to_python_type(json_spec.additionalProperties) + ']' }}\n {%- else %}\n {{- \"dict\" }}\n {%- endif %}\n {%- elif json_spec.type is iterable %}\n {{- \"Union[\" }}\n {%- for t in json_spec.type %}\n {{- json_to_python_type({\"type\": t}) }}\n {%- if not loop.last %}\n {{- \",\" }}\n {%- endif %}\n {%- endfor %}\n {{- \"]\" }}\n {%- else %}\n {{- \"Any\" }}\n {%- endif %}\n{%- endmacro %}\n\n{%- if tools %}\n {{- '<|im_start|>system\n' }}\n {%- if messages[0]['role'] == 'system' %}\n {{- messages[0]['content'] + '\n\n' }}\n {%- endif %}\n {{- '# Tools\n\n' }}\n {{- \"You are a function calling AI model. You are provided with function signatures within XML tags. You may call one or more functions to assist with the user query. Don't make assumptions about what values to plug into functions. Here are the available tools: \" }}\n {%- for tool in tools %}\n {%- if tool.function is defined %}\n {%- set tool = tool.function %}\n {%- endif %}\n {{- '{\"type\": \"function\", \"function\": ' }}\n {{- '{\"name\": ' + tool.name + '\", ' }}\n {{- '\"description\": \"' + tool.name + '(' }}\n {%- for param_name, param_fields in tool.parameters.properties|items %}\n {{- param_name + \": \" + json_to_python_type(param_fields) }}\n {%- if not loop.last %}\n {{- \", \" }}\n {%- endif %}\n {%- endfor %}\n {{- \")\" }}\n {%- if tool.return is defined %}\n {{- \" -> \" + json_to_python_type(tool.return) }}\n {%- endif %}\n {{- \" - \" + tool.description + \"\n\n\" }}\n {%- for param_name, param_fields in tool.parameters.properties|items %}\n {%- if loop.first %}\n {{- \" Args:\n\" }}\n {%- endif %}\n {{- \" \" + param_name + \"(\" + json_to_python_type(param_fields) + \"): \" + param_fields.description|trim }}\n {%- endfor %}\n {%- if tool.return is defined and tool.return.description is defined %}\n {{- \"\n Returns:\n \" + tool.return.description }}\n {%- endif %}\n {{- '\"' }}\n {{- ', \"parameters\": ' }}\n {%- if tool.parameters.properties | length == 0 %}\n {{- \"{}\" }}\n {%- else %}\n {{- tool.parameters|tojson }}\n {%- endif %}\n {{- \"}\" }}\n {%- if not loop.last %}\n {{- \"\n\" }}\n {%- endif %}\n {%- endfor %}\n {{- \" \" }}\n {{- 'Use the following pydantic model json schema for each tool call you will make: {\"properties\": {\"arguments\": {\"title\": \"Arguments\", \"type\": \"object\"}, \"name\": {\"title\": \"Name\", \"type\": \"string\"}}, \"required\": [\"arguments\", \"name\"], \"title\": \"FunctionCall\", \"type\": \"object\"}\n' }}\n {{- \"For each function call return a json object with function name and arguments within XML tags as follows:\n\" }}\n {{- \"\n\" }}\n {{- '{\"name\": , \"arguments\": }\n' }}\n {{- '<|im_end|>\n' }}\n{%- else %}\n {%- if messages[0]['role'] != 'system' %}\n {{- '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}\n {%- else %}\n {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}\n {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n {%- if message.role == \"user\" or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and message.tool_calls is not defined) %}\n {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}\n {%- elif message.role == \"assistant\" %}\n {{- '<|im_start|>' + message.role + '\n\n' }}\n {%- for tool_call in message.tool_calls %}\n {%- if tool_call.function is defined %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '{' }}\n {{- '\"name\": \"' }}\n {{- tool_call.name }}\n {%- if tool_call.arguments is defined %}\n {{- ', ' }}\n {{- '\"arguments\": ' }}\n {{- tool_call.arguments|tojson }}\n {%- endif %}\n {{- '\"}' }}\n {{- '\n' }}\n {%- endfor %}\n {{- '<|im_end|>\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if not message.name is defined %}\n {{- raise_exception(\"Tool response dicts require a 'name' key indicating the name of the called function!\") }}\n {%- endif %}\n {{- '<|im_start|>user\n\n' }}\n {{- '{\"name\": \"' }}\n {{- message.name }}\n {{- '\", \"content\": ' }}\n {{- message.content|tojson + '}' }}\n {{- '\n<|im_end|>\n' }}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\n' }}\n{%- endif %}", + "stop_token_ids": [ + 151643, + 151644, + 151645 + ], + "stop": [ + "<|endoftext|>", + "<|im_start|>", + "<|im_end|>" + ] + }, + { + "version": 1, + "context_length": 65536, + "model_name": "codeqwen1.5", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "generate" + ], + "model_description": "CodeQwen1.5 is the Code-Specific version of Qwen1.5. It is a transformer-based decoder-only language model pretrained on a large amount of data of codes.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "HangZhou_Ascend/CodeQwen1.5-7B", + "model_hub": "openmind_hub" + } + ] + }, + { + "version": 1, + "context_length": 65536, + "model_name": "codeqwen1.5-chat", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat" + ], + "model_description": "CodeQwen1.5 is the Code-Specific version of Qwen1.5. It is a transformer-based decoder-only language model pretrained on a large amount of data of codes.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "HangZhou_Ascend/CodeQwen1.5-7B-Chat", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", + "stop_token_ids": [ + 151643, + 151644, + 151645 + ], + "stop": [ + "<|endoftext|>", + "<|im_start|>", + "<|im_end|>" + ] + }, + { + "version": 1, + "context_length": 32768, + "model_name": "qwen2-instruct", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat", + "tools" + ], + "model_description": "Qwen2 is the new series of Qwen large language models", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": "0_5", + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "wuhaicc/Qwen2-0.5B-Instruct", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": "1_5", + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "HangZhou_Ascend/Qwen2-1.5B-Instruct", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "wuhaicc/Qwen2-7B-Instruct", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 72, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "State_Cloud/Qwen2-72B-Instruct", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{%- macro json_to_python_type(json_spec) %}\n {%- set basic_type_map = {\n \"string\": \"str\",\n \"number\": \"float\",\n \"integer\": \"int\",\n \"boolean\": \"bool\"\n} %}\n {%- if basic_type_map[json_spec.type] is defined %}\n {{- basic_type_map[json_spec.type] }}\n {%- elif json_spec.type == \"array\" %}\n {{- \"list[\" + json_to_python_type(json_spec|items) + \"]\" }}\n {%- elif json_spec.type == \"object\" %}\n {%- if json_spec.additionalProperties is defined %}\n {{- \"dict[str, \" + json_to_python_type(json_spec.additionalProperties) + ']' }}\n {%- else %}\n {{- \"dict\" }}\n {%- endif %}\n {%- elif json_spec.type is iterable %}\n {{- \"Union[\" }}\n {%- for t in json_spec.type %}\n {{- json_to_python_type({\"type\": t}) }}\n {%- if not loop.last %}\n {{- \",\" }}\n {%- endif %}\n {%- endfor %}\n {{- \"]\" }}\n {%- else %}\n {{- \"Any\" }}\n {%- endif %}\n{%- endmacro %}\n\n{%- if tools %}\n {{- '<|im_start|>system\n' }}\n {%- if messages[0]['role'] == 'system' %}\n {{- messages[0]['content'] + '\n\n' }}\n {%- endif %}\n {{- '# Tools\n\n' }}\n {{- \"You are a function calling AI model. You are provided with function signatures within XML tags. You may call one or more functions to assist with the user query. Don't make assumptions about what values to plug into functions. Here are the available tools: \" }}\n {%- for tool in tools %}\n {%- if tool.function is defined %}\n {%- set tool = tool.function %}\n {%- endif %}\n {{- '{\"type\": \"function\", \"function\": ' }}\n {{- '{\"name\": ' + tool.name + '\", ' }}\n {{- '\"description\": \"' + tool.name + '(' }}\n {%- for param_name, param_fields in tool.parameters.properties|items %}\n {{- param_name + \": \" + json_to_python_type(param_fields) }}\n {%- if not loop.last %}\n {{- \", \" }}\n {%- endif %}\n {%- endfor %}\n {{- \")\" }}\n {%- if tool.return is defined %}\n {{- \" -> \" + json_to_python_type(tool.return) }}\n {%- endif %}\n {{- \" - \" + tool.description + \"\n\n\" }}\n {%- for param_name, param_fields in tool.parameters.properties|items %}\n {%- if loop.first %}\n {{- \" Args:\n\" }}\n {%- endif %}\n {{- \" \" + param_name + \"(\" + json_to_python_type(param_fields) + \"): \" + param_fields.description|trim }}\n {%- endfor %}\n {%- if tool.return is defined and tool.return.description is defined %}\n {{- \"\n Returns:\n \" + tool.return.description }}\n {%- endif %}\n {{- '\"' }}\n {{- ', \"parameters\": ' }}\n {%- if tool.parameters.properties | length == 0 %}\n {{- \"{}\" }}\n {%- else %}\n {{- tool.parameters|tojson }}\n {%- endif %}\n {{- \"}\" }}\n {%- if not loop.last %}\n {{- \"\n\" }}\n {%- endif %}\n {%- endfor %}\n {{- \" \" }}\n {{- 'Use the following pydantic model json schema for each tool call you will make: {\"properties\": {\"arguments\": {\"title\": \"Arguments\", \"type\": \"object\"}, \"name\": {\"title\": \"Name\", \"type\": \"string\"}}, \"required\": [\"arguments\", \"name\"], \"title\": \"FunctionCall\", \"type\": \"object\"}\n' }}\n {{- \"For each function call return a json object with function name and arguments within XML tags as follows:\n\" }}\n {{- \"\n\" }}\n {{- '{\"name\": , \"arguments\": }\n' }}\n {{- '<|im_end|>\n' }}\n{%- else %}\n {%- if messages[0]['role'] != 'system' %}\n {{- '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}\n {%- else %}\n {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}\n {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n {%- if message.role == \"user\" or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and message.tool_calls is not defined) %}\n {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}\n {%- elif message.role == \"assistant\" %}\n {{- '<|im_start|>' + message.role + '\n\n' }}\n {%- for tool_call in message.tool_calls %}\n {%- if tool_call.function is defined %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '{' }}\n {{- '\"name\": \"' }}\n {{- tool_call.name }}\n {%- if tool_call.arguments is defined %}\n {{- ', ' }}\n {{- '\"arguments\": ' }}\n {{- tool_call.arguments|tojson }}\n {%- endif %}\n {{- '\"}' }}\n {{- '\n' }}\n {%- endfor %}\n {{- '<|im_end|>\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if not message.name is defined %}\n {{- raise_exception(\"Tool response dicts require a 'name' key indicating the name of the called function!\") }}\n {%- endif %}\n {{- '<|im_start|>user\n\n' }}\n {{- '{\"name\": \"' }}\n {{- message.name }}\n {{- '\", \"content\": ' }}\n {{- message.content|tojson + '}' }}\n {{- '\n<|im_end|>\n' }}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\n' }}\n{%- endif %}", + "stop_token_ids": [ + 151643, + 151644, + 151645 + ], + "stop": [ + "<|endoftext|>", + "<|im_start|>", + "<|im_end|>" + ] + }, + { + "version": 1, + "context_length": 8192, + "model_name": "mistral-v0.1", + "model_lang": [ + "en" + ], + "model_ability": [ + "generate" + ], + "model_description": "Mistral-7B is a unmoderated Transformer based LLM claiming to outperform Llama2 on all benchmarks.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "PyTorch-NPU/mistral_7b_v0.1", + "model_hub": "openmind_hub" + } + ] + }, + { + "version": 1, + "context_length": 4096, + "model_name": "Yi", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "generate" + ], + "model_description": "The Yi series models are large language models trained from scratch by developers at 01.AI.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 6, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "wuhaicc/Yi-6B", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 9, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "wuhaicc/Yi-9B", + "model_hub": "openmind_hub" + } + ] + }, + { + "version": 1, + "context_length": 262144, + "model_name": "Yi-200k", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "generate" + ], + "model_description": "The Yi series models are large language models trained from scratch by developers at 01.AI.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 6, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "wuhaicc/Yi-6B-200K", + "model_hub": "openmind_hub" + } + ] + }, + { + "version": 1, + "context_length": 4096, + "model_name": "Yi-1.5", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "generate" + ], + "model_description": "Yi-1.5 is an upgraded version of Yi. It is continuously pre-trained on Yi with a high-quality corpus of 500B tokens and fine-tuned on 3M diverse fine-tuning samples.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 6, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "HangZhou_Ascend/Yi-1.5-6B", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 9, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "HangZhou_Ascend/Yi-1.5-9B", + "model_hub": "openmind_hub" + } + ] + }, + { + "version": 1, + "context_length": 32768, + "model_name": "internlm2.5-chat", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat" + ], + "model_description": "InternLM2.5 series of the InternLM model.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": "1_8", + "quantizations": [ + "none" + ], + "model_id": "Intern/internlm2_5-1_8b-chat", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "none" + ], + "model_id": "Intern/internlm2_5-7b-chat", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 20, + "quantizations": [ + "none" + ], + "model_id": "Intern/internlm2_5-20b-chat", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{{ '' }}{% for message in messages %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", + "stop_token_ids": [ + 2, + 92542 + ], + "stop": [ + "", + "<|im_end|>" + ] + }, + { + "version": 1, + "context_length": 262144, + "model_name": "internlm2.5-chat-1m", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat" + ], + "model_description": "InternLM2.5 series of the InternLM model supports 1M long-context", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "none" + ], + "model_id": "Intern/internlm2_5-7b-chat-1m", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{{ '' }}{% for message in messages %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", + "stop_token_ids": [ + 2, + 92542 + ], + "stop": [ + "", + "<|im_end|>" + ] + }, + { + "version": 1, + "context_length": 8192, + "model_name": "gemma-it", + "model_lang": [ + "en" + ], + "model_ability": [ + "chat" + ], + "model_description": "Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 2, + "quantizations": [ + "none", + "4-bit", + "8-bit" + ], + "model_id": "SY_AICC/gemma-2b-it", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "none", + "4-bit", + "8-bit" + ], + "model_id": "SY_AICC/gemma-7b-it", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{{ '' }}{% if messages[0]['role'] == 'system' %}{{ raise_exception('System role not supported') }}{% endif %}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if (message['role'] == 'assistant') %}{% set role = 'model' %}{% else %}{% set role = message['role'] %}{% endif %}{{ '' + role + '\n' + message['content'] | trim + '\n' }}{% endfor %}{% if add_generation_prompt %}{{'model\n'}}{% endif %}", + "stop_token_ids": [ + 1, + 106, + 107 + ], + "stop": [ + "", + "", + "" + ] + }, + { + "version": 1, + "context_length": 8192, + "model_name": "gemma-2-it", + "model_lang": [ + "en" + ], + "model_ability": [ + "chat" + ], + "model_description": "Gemma is a family of lightweight, state-of-the-art open models from Google, built from the same research and technology used to create the Gemini models.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 2, + "quantizations": [ + "none", + "4-bit", + "8-bit" + ], + "model_id": "LlamaFactory/gemma-2-2b-it", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 9, + "quantizations": [ + "none", + "4-bit", + "8-bit" + ], + "model_id": "LlamaFactory/gemma-2-9b-it", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{{ '' }}{% if messages[0]['role'] == 'system' %}{{ raise_exception('System role not supported') }}{% endif %}{% for message in messages %}{% if (message['role'] == 'user') != (loop.index0 % 2 == 0) %}{{ raise_exception('Conversation roles must alternate user/assistant/user/assistant/...') }}{% endif %}{% if (message['role'] == 'assistant') %}{% set role = 'model' %}{% else %}{% set role = message['role'] %}{% endif %}{{ '' + role + '\n' + message['content'] | trim + '\n' }}{% endfor %}{% if add_generation_prompt %}{{'model\n'}}{% endif %}", + "stop_token_ids": [ + 1, + 106, + 107 + ], + "stop": [ + "", + "", + "" + ] + }, + { + "version": 1, + "context_length": 4096, + "model_name": "minicpm-2b-sft-bf16", + "model_lang": [ + "zh" + ], + "model_ability": [ + "chat" + ], + "model_description": "MiniCPM is an End-Size LLM developed by ModelBest Inc. and TsinghuaNLP, with only 2.4B parameters excluding embeddings.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 2, + "quantizations": [ + "none" + ], + "model_id": "AI-Research/MiniCPM-2B-sft-bf16", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{% for message in messages %}{% if message['role'] == 'user' %}{{'<用户>' + message['content'].strip() + ''}}{% else %}{{message['content'].strip()}}{% endif %}{% endfor %}", + "stop_token_ids": [ + 1, + 2 + ], + "stop": [ + "", + "" + ] + }, + { + "version": 1, + "context_length": 4096, + "model_name": "minicpm-2b-dpo-bf16", + "model_lang": [ + "zh" + ], + "model_ability": [ + "chat" + ], + "model_description": "MiniCPM is an End-Size LLM developed by ModelBest Inc. and TsinghuaNLP, with only 2.4B parameters excluding embeddings.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 2, + "quantizations": [ + "none" + ], + "model_id": "AI-Research/MiniCPM-2B-dpo-bf16", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{% for message in messages %}{% if message['role'] == 'user' %}{{'<用户>' + message['content'].strip() + ''}}{% else %}{{message['content'].strip()}}{% endif %}{% endfor %}", + "stop_token_ids": [ + 1, + 2 + ], + "stop": [ + "", + "" + ] + }, + { + "version": 1, + "context_length": 8192, + "model_name": "cogvlm2", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat", + "vision" + ], + "model_description": "CogVLM2 have achieved good results in many lists compared to the previous generation of CogVLM open source models. Its excellent performance can compete with some non-open source models.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 20, + "quantizations": [ + "none" + ], + "model_id": "AI-Research/cogvlm2-llama3-chinese-chat-19b", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{% if not add_generation_prompt is defined %}{% set add_generation_prompt = false %}{% endif %}{% set loop_messages = messages %}{% for message in loop_messages %}{% set content = '<|start_header_id|>' + message['role'] + '<|end_header_id|>\n\n'+ message['content'] | trim + '<|eot_id|>' %}{% if loop.index0 == 0 %}{% set content = '<|begin_of_text|>' + content %}{% endif %}{{ content }}{% endfor %}{% if add_generation_prompt %}{{ '<|start_header_id|>assistant<|end_header_id|>\n\n' }}{% else %}{{ '<|end_of_text|>' }}{% endif %}", + "stop_token_ids": [ + 128001, + 128009 + ], + "stop": [ + "<|end_of_text|>", + "<|eot_id|>" + ] + }, + { + "version": 1, + "context_length": 8192, + "model_name": "telechat", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat" + ], + "model_description": "The TeleChat is a large language model developed and trained by China Telecom Artificial Intelligence Technology Co., LTD. The 7B model base is trained with 1.5 trillion Tokens and 3 trillion Tokens and Chinese high-quality corpus.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "TeleAI/TeleChat-7B-pt", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 12, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "TeleAI/TeleChat-12B-pt", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 52, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "TeleAI/TeleChat-52B-pt", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{{ (messages|selectattr('role', 'equalto', 'system')|list|last).content|trim if (messages|selectattr('role', 'equalto', 'system')|list) else '' }}{%- for message in messages -%}{%- if message['role'] == 'user' -%}{{- '<_user>' + message['content'] +'<_bot>' -}}{%- elif message['role'] == 'assistant' -%}{{- message['content'] + '<_end>' -}}{%- endif -%}{%- endfor -%}", + "stop": [ + "<_end>", + "<_start>" + ], + "stop_token_ids": [ + 160133, + 160132 + ] + }, + { + "version":1, + "context_length":32768, + "model_name":"qwen2-vl-instruct", + "model_lang":[ + "en", + "zh" + ], + "model_ability":[ + "chat", + "vision" + ], + "model_description":"Qwen2-VL: To See the World More Clearly.Qwen2-VL is the latest version of the vision language models in the Qwen model familities.", + "model_specs":[ + { + "model_format":"pytorch", + "model_size_in_billions":2, + "quantizations":[ + "none" + ], + "model_id":"LlamaFactory/Qwen2-VL-2B-Instruct", + "model_hub": "openmind_hub" + }, + { + "model_format":"pytorch", + "model_size_in_billions":7, + "quantizations":[ + "none" + ], + "model_id":"LlamaFactory/Qwen2-VL-7B-Instruct", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{% for message in messages %}{% if loop.first and message['role'] != 'system' %}<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n{% endif %}<|im_start|>{{ message['role'] }}\n{% if message['content'] is string %}{{ message['content'] }}<|im_end|>\n{% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>\n{% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant\n{% endif %}", + "stop_token_ids": [ + 151645, + 151643 + ], + "stop": [ + "<|im_end|>", + "<|endoftext|>" + ] + }, + { + "version": 1, + "context_length": 32768, + "model_name": "minicpm3-4b", + "model_lang": [ + "zh" + ], + "model_ability": [ + "chat" + ], + "model_description": "MiniCPM3-4B is the 3rd generation of MiniCPM series. The overall performance of MiniCPM3-4B surpasses Phi-3.5-mini-Instruct and GPT-3.5-Turbo-0125, being comparable with many recent 7B~9B models.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 4, + "quantizations": [ + "none" + ], + "model_id": "AI-Research/MiniCPM3-4B", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{% for message in messages %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}", + "stop_token_ids": [ + 1, + 2 + ], + "stop": [ + "", + "" + ] + }, + { + "version": 1, + "context_length": 32768, + "model_name": "qwen2.5", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "generate" + ], + "model_description": "Qwen2.5 is the latest series of Qwen large language models. For Qwen2.5, we release a number of base language models and instruction-tuned language models ranging from 0.5 to 72 billion parameters.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": "0_5", + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "Tianjin_Ascend/qwen2.5-0.5b", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": "1_5", + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "Tianjin_Ascend/Qwen2.5-1.5B", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 3, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "Tianjin_Ascend/Qwen2.5-3B", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "AI-Research/Qwen2.5-7B", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 32, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "AI-Research/Qwen2.5-32B", + "model_hub": "openmind_hub" + } + ] + }, + { + "version": 1, + "context_length": 32768, + "model_name": "qwen2.5-instruct", + "model_lang": [ + "en", + "zh" + ], + "model_ability": [ + "chat", + "tools" + ], + "model_description": "Qwen2.5 is the latest series of Qwen large language models. For Qwen2.5, we release a number of base language models and instruction-tuned language models ranging from 0.5 to 72 billion parameters.", + "model_specs": [ + { + "model_format": "pytorch", + "model_size_in_billions": 7, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "AI-Research/Qwen2.5-7B-Instruct", + "model_hub": "openmind_hub" + }, + { + "model_format": "pytorch", + "model_size_in_billions": 32, + "quantizations": [ + "4-bit", + "8-bit", + "none" + ], + "model_id": "AI-Research/Qwen2.5-32B-Instruct", + "model_hub": "openmind_hub" + } + ], + "chat_template": "{%- if tools %}\n {{- '<|im_start|>system\\n' }}\n {%- if messages[0]['role'] == 'system' %}\n {{- messages[0]['content'] }}\n {%- else %}\n {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}\n {%- endif %}\n {{- \"\\n\\n# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within XML tags:\\n\" }}\n {%- for tool in tools %}\n {{- \"\\n\" }}\n {{- tool | tojson }}\n {%- endfor %}\n {{- \"\\n\\n\\nFor each function call, return a json object with function name and arguments within XML tags:\\n\\n{\\\"name\\\": , \\\"arguments\\\": }\\n<|im_end|>\\n\" }}\n{%- else %}\n {%- if messages[0]['role'] == 'system' %}\n {{- '<|im_start|>system\\n' + messages[0]['content'] + '<|im_end|>\\n' }}\n {%- else %}\n {{- '<|im_start|>system\\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\\n' }}\n {%- endif %}\n{%- endif %}\n{%- for message in messages %}\n {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) or (message.role == \"assistant\" and not message.tool_calls) %}\n {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n {%- elif message.role == \"assistant\" %}\n {{- '<|im_start|>' + message.role }}\n {%- if message.content %}\n {{- '\\n' + message.content }}\n {%- endif %}\n {%- for tool_call in message.tool_calls %}\n {%- if tool_call.function is defined %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '\\n\\n{\"name\": \"' }}\n {{- tool_call.name }}\n {{- '\", \"arguments\": ' }}\n {{- tool_call.arguments | tojson }}\n {{- '}\\n' }}\n {%- endfor %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != \"tool\") %}\n {{- '<|im_start|>user' }}\n {%- endif %}\n {{- '\\n\\n' }}\n {{- message.content }}\n {{- '\\n' }}\n {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n {{- '<|im_end|>\\n' }}\n {%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n{%- endif %}\n", + "stop_token_ids": [ + 151643, + 151644, + 151645 + ], + "stop": [ + "<|endoftext|>", + "<|im_start|>", + "<|im_end|>" + ] + } +] diff --git a/xinference/model/rerank/core.py b/xinference/model/rerank/core.py index de9a7d33ae..ffd1485eb5 100644 --- a/xinference/model/rerank/core.py +++ b/xinference/model/rerank/core.py @@ -347,7 +347,9 @@ def create_rerank_model_instance( devices: List[str], model_uid: str, model_name: str, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, model_path: Optional[str] = None, **kwargs, ) -> Tuple[RerankModel, RerankModelDescription]: diff --git a/xinference/model/utils.py b/xinference/model/utils.py index 52735ce089..bc641e0458 100644 --- a/xinference/model/utils.py +++ b/xinference/model/utils.py @@ -54,6 +54,13 @@ def download_from_modelscope() -> bool: return False +def download_from_openmind_hub() -> bool: + if os.environ.get(XINFERENCE_ENV_MODEL_SRC): + return os.environ.get(XINFERENCE_ENV_MODEL_SRC) == "openmind_hub" + else: + return False + + def download_from_csghub() -> bool: if os.environ.get(XINFERENCE_ENV_MODEL_SRC) == "csghub": return True diff --git a/xinference/model/video/core.py b/xinference/model/video/core.py index c7545122f0..4916031fb8 100644 --- a/xinference/model/video/core.py +++ b/xinference/model/video/core.py @@ -97,7 +97,9 @@ def generate_video_description( def match_diffusion( model_name: str, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, ) -> VideoModelFamilyV1: from ..utils import download_from_modelscope from . import BUILTIN_VIDEO_MODELS, MODELSCOPE_VIDEO_MODELS @@ -157,7 +159,9 @@ def create_video_model_instance( devices: List[str], model_uid: str, model_name: str, - download_hub: Optional[Literal["huggingface", "modelscope", "csghub"]] = None, + download_hub: Optional[ + Literal["huggingface", "modelscope", "openmind_hub", "csghub"] + ] = None, model_path: Optional[str] = None, **kwargs, ) -> Tuple[DiffUsersVideoModel, VideoModelDescription]: diff --git a/xinference/web/ui/src/scenes/launch_model/modelCard.js b/xinference/web/ui/src/scenes/launch_model/modelCard.js index bcffbace11..ea97f2a91a 100644 --- a/xinference/web/ui/src/scenes/launch_model/modelCard.js +++ b/xinference/web/ui/src/scenes/launch_model/modelCard.js @@ -1405,8 +1405,19 @@ const ModelCard = ({ label="(Optional) Download_hub" > {(csghubArr.includes(modelData.model_name) - ? ['none', 'huggingface', 'modelscope', 'csghub'] - : ['none', 'huggingface', 'modelscope'] + ? [ + 'none', + 'huggingface', + 'modelscope', + 'openmind_hub', + 'csghub', + ] + : [ + 'none', + 'huggingface', + 'modelscope', + 'openmind_hub', + ] ).map((item) => { return ( @@ -1566,13 +1577,15 @@ const ModelCard = ({ }} label="(Optional) Download_hub" > - {['none', 'huggingface', 'modelscope'].map((item) => { - return ( - - {item} - - ) - })} + {['none', 'huggingface', 'modelscope', 'openmind_hub'].map( + (item) => { + return ( + + {item} + + ) + } + )}