diff --git a/module/ai/svg.go b/ai-provider/local/ollama.go similarity index 94% rename from module/ai/svg.go rename to ai-provider/local/ollama.go index 92b50ec8..31660541 100644 --- a/module/ai/svg.go +++ b/ai-provider/local/ollama.go @@ -1,7 +1,9 @@ -package ai +package ai_provider_local var ( - ollamaSvg = ` + OllamaBase = "http://apipark-ollama:11434" + OllamaConfig = "{\n \"mirostat\": 0,\n \"mirostat_eta\": 0.1,\n \"mirostat_tau\": 5.0,\n \"num_ctx\": 4096,\n \"repeat_last_n\":64,\n \"repeat_penalty\": 1.1,\n \"temperature\": 0.7,\n \"seed\": 42,\n \"num_predict\": 42,\n \"top_k\": 40,\n \"top_p\": 0.9,\n \"min_p\": 0.5\n}\n" + OllamaSvg = ` Union[LLMResult, Generator]: - cred_with_endpoint = self._update_endpoint_url(credentials=credentials) - return super()._invoke(model, cred_with_endpoint, prompt_messages, model_parameters, tools, stop, stream, user) - - def validate_credentials(self, model: str, credentials: dict) -> None: - cred_with_endpoint = self._update_endpoint_url(credentials=credentials) - self._add_custom_parameters(credentials, model) - return super().validate_credentials(model, cred_with_endpoint) - - @classmethod - def _add_custom_parameters(cls, credentials: dict, model: str) -> None: - credentials["mode"] = "chat" - - def _generate( - self, - model: str, - credentials: dict, - prompt_messages: list[PromptMessage], - model_parameters: dict, - tools: Optional[list[PromptMessageTool]] = None, - stop: Optional[list[str]] = None, - stream: bool = True, - user: Optional[str] = None, - ) -> Union[LLMResult, Generator]: - cred_with_endpoint = self._update_endpoint_url(credentials=credentials) - return super()._generate( - model, cred_with_endpoint, prompt_messages, model_parameters, tools, stop, stream, user - ) - - def get_customizable_model_schema(self, model: str, credentials: dict) -> AIModelEntity: - cred_with_endpoint = self._update_endpoint_url(credentials=credentials) - - return super().get_customizable_model_schema(model, cred_with_endpoint) - - def get_num_tokens( - self, - model: str, - credentials: dict, - prompt_messages: list[PromptMessage], - tools: Optional[list[PromptMessageTool]] = None, - ) -> int: - cred_with_endpoint = self._update_endpoint_url(credentials=credentials) - - return super().get_num_tokens(model, cred_with_endpoint, prompt_messages, tools) diff --git a/ai-provider/model-runtime/model-providers/novita/llm/mistral-nemo.yaml b/ai-provider/model-runtime/model-providers/novita/llm/mistral-nemo.yaml new file mode 100644 index 00000000..6f116738 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/novita/llm/mistral-nemo.yaml @@ -0,0 +1,41 @@ +model: mistralai/mistral-nemo +label: + zh_Hans: Mistral Nemo + en_US: Mistral Nemo +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 131072 +parameter_rules: + - name: temperature + use_template: temperature + min: 0 + max: 2 + default: 1 + - name: top_p + use_template: top_p + min: 0 + max: 1 + default: 1 + - name: max_tokens + use_template: max_tokens + min: 1 + max: 2048 + default: 512 + - name: frequency_penalty + use_template: frequency_penalty + min: -2 + max: 2 + default: 0 + - name: presence_penalty + use_template: presence_penalty + min: -2 + max: 2 + default: 0 +pricing: + input: '0.0017' + output: '0.0017' + unit: '0.0001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/novita/llm/openchat-7b.yaml b/ai-provider/model-runtime/model-providers/novita/llm/openchat-7b.yaml new file mode 100644 index 00000000..b21ea301 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/novita/llm/openchat-7b.yaml @@ -0,0 +1,41 @@ +model: openchat/openchat-7b +label: + zh_Hans: OpenChat 7B + en_US: OpenChat 7B +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 4096 +parameter_rules: + - name: temperature + use_template: temperature + min: 0 + max: 2 + default: 1 + - name: top_p + use_template: top_p + min: 0 + max: 1 + default: 1 + - name: max_tokens + use_template: max_tokens + min: 1 + max: 2048 + default: 512 + - name: frequency_penalty + use_template: frequency_penalty + min: -2 + max: 2 + default: 0 + - name: presence_penalty + use_template: presence_penalty + min: -2 + max: 2 + default: 0 +pricing: + input: '0.0006' + output: '0.0006' + unit: '0.0001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/novita/llm/qwen-2-72b-instruct.yaml b/ai-provider/model-runtime/model-providers/novita/llm/qwen-2-72b-instruct.yaml new file mode 100644 index 00000000..069f9096 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/novita/llm/qwen-2-72b-instruct.yaml @@ -0,0 +1,41 @@ +model: qwen/qwen-2-72b-instruct +label: + zh_Hans: Qwen2 72B Instruct + en_US: Qwen2 72B Instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + min: 0 + max: 2 + default: 1 + - name: top_p + use_template: top_p + min: 0 + max: 1 + default: 1 + - name: max_tokens + use_template: max_tokens + min: 1 + max: 2048 + default: 512 + - name: frequency_penalty + use_template: frequency_penalty + min: -2 + max: 2 + default: 0 + - name: presence_penalty + use_template: presence_penalty + min: -2 + max: 2 + default: 0 +pricing: + input: '0.0034' + output: '0.0039' + unit: '0.0001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/novita/llm/qwen-2-7b-instruct.yaml b/ai-provider/model-runtime/model-providers/novita/llm/qwen-2-7b-instruct.yaml new file mode 100644 index 00000000..afc627f1 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/novita/llm/qwen-2-7b-instruct.yaml @@ -0,0 +1,41 @@ +model: qwen/qwen-2-7b-instruct +label: + zh_Hans: Qwen 2 7B Instruct + en_US: Qwen 2 7B Instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + min: 0 + max: 2 + default: 1 + - name: top_p + use_template: top_p + min: 0 + max: 1 + default: 1 + - name: max_tokens + use_template: max_tokens + min: 1 + max: 2048 + default: 512 + - name: frequency_penalty + use_template: frequency_penalty + min: -2 + max: 2 + default: 0 + - name: presence_penalty + use_template: presence_penalty + min: -2 + max: 2 + default: 0 +pricing: + input: '0.00054' + output: '0.00054' + unit: '0.0001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/novita/llm/qwen-2-vl-72b-instruct.yaml b/ai-provider/model-runtime/model-providers/novita/llm/qwen-2-vl-72b-instruct.yaml new file mode 100644 index 00000000..06bdf0c8 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/novita/llm/qwen-2-vl-72b-instruct.yaml @@ -0,0 +1,41 @@ +model: qwen/qwen-2-vl-72b-instruct +label: + zh_Hans: Qwen 2 VL 72B Instruct + en_US: Qwen 2 VL 72B Instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + min: 0 + max: 2 + default: 1 + - name: top_p + use_template: top_p + min: 0 + max: 1 + default: 1 + - name: max_tokens + use_template: max_tokens + min: 1 + max: 2048 + default: 512 + - name: frequency_penalty + use_template: frequency_penalty + min: -2 + max: 2 + default: 0 + - name: presence_penalty + use_template: presence_penalty + min: -2 + max: 2 + default: 0 +pricing: + input: '0.0045' + output: '0.0045' + unit: '0.0001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/novita/llm/qwen-2.5-72b-instruct.yaml b/ai-provider/model-runtime/model-providers/novita/llm/qwen-2.5-72b-instruct.yaml new file mode 100644 index 00000000..97f5af35 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/novita/llm/qwen-2.5-72b-instruct.yaml @@ -0,0 +1,41 @@ +model: qwen/qwen-2.5-72b-instruct +label: + zh_Hans: Qwen 2.5 72B Instruct + en_US: Qwen 2.5 72B Instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32000 +parameter_rules: + - name: temperature + use_template: temperature + min: 0 + max: 2 + default: 1 + - name: top_p + use_template: top_p + min: 0 + max: 1 + default: 1 + - name: max_tokens + use_template: max_tokens + min: 1 + max: 2048 + default: 512 + - name: frequency_penalty + use_template: frequency_penalty + min: -2 + max: 2 + default: 0 + - name: presence_penalty + use_template: presence_penalty + min: -2 + max: 2 + default: 0 +pricing: + input: '0.0038' + output: '0.004' + unit: '0.0001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/nvidia/llm/deepseek-r1.yaml b/ai-provider/model-runtime/model-providers/nvidia/llm/deepseek-r1.yaml new file mode 100644 index 00000000..159941ec --- /dev/null +++ b/ai-provider/model-runtime/model-providers/nvidia/llm/deepseek-r1.yaml @@ -0,0 +1,35 @@ +model: deepseek-ai/deepseek-r1 +label: + en_US: deepseek-ai/deepseek-r1 +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 128000 +parameter_rules: + - name: temperature + use_template: temperature + min: 0 + max: 1 + default: 0.5 + - name: top_p + use_template: top_p + min: 0 + max: 1 + default: 1 + - name: max_tokens + use_template: max_tokens + min: 1 + max: 1024 + default: 1024 + - name: frequency_penalty + use_template: frequency_penalty + min: -2 + max: 2 + default: 0 + - name: presence_penalty + use_template: presence_penalty + min: -2 + max: 2 + default: 0 diff --git a/ai-provider/model-runtime/model-providers/openAI/llm/gpt-4o-2024-11-20.yaml b/ai-provider/model-runtime/model-providers/openAI/llm/gpt-4o-2024-11-20.yaml new file mode 100644 index 00000000..2c7c1c6e --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openAI/llm/gpt-4o-2024-11-20.yaml @@ -0,0 +1,47 @@ +model: gpt-4o-2024-11-20 +label: + zh_Hans: gpt-4o-2024-11-20 + en_US: gpt-4o-2024-11-20 +model_type: llm +features: + - multi-tool-call + - agent-thought + - stream-tool-call + - vision +model_properties: + mode: chat + context_size: 128000 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: presence_penalty + use_template: presence_penalty + - name: frequency_penalty + use_template: frequency_penalty + - name: max_tokens + use_template: max_tokens + default: 16384 + min: 1 + max: 16384 + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object + - json_schema + - name: json_schema + use_template: json_schema +pricing: + input: '2.50' + output: '10.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openAI/llm/gpt-4o-audio-preview.yaml b/ai-provider/model-runtime/model-providers/openAI/llm/gpt-4o-audio-preview.yaml new file mode 100644 index 00000000..e707acc5 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openAI/llm/gpt-4o-audio-preview.yaml @@ -0,0 +1,44 @@ +model: gpt-4o-audio-preview +label: + zh_Hans: gpt-4o-audio-preview + en_US: gpt-4o-audio-preview +model_type: llm +features: + - multi-tool-call + - agent-thought + - stream-tool-call + - audio +model_properties: + mode: chat + context_size: 128000 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: presence_penalty + use_template: presence_penalty + - name: frequency_penalty + use_template: frequency_penalty + - name: max_tokens + use_template: max_tokens + default: 16384 + min: 1 + max: 16384 + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '5.00' + output: '15.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openAI/llm/o1-2024-12-17.yaml b/ai-provider/model-runtime/model-providers/openAI/llm/o1-2024-12-17.yaml new file mode 100644 index 00000000..643258a2 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openAI/llm/o1-2024-12-17.yaml @@ -0,0 +1,48 @@ +model: o1-2024-12-17 +label: + en_US: o1-2024-12-17 +model_type: llm +features: + - multi-tool-call + - agent-thought + - stream-tool-call + - vision +model_properties: + mode: chat + context_size: 200000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + default: 50000 + min: 1 + max: 50000 + - name: reasoning_effort + label: + zh_Hans: 推理工作 + en_US: reasoning_effort + type: string + help: + zh_Hans: 限制推理模型的推理工作 + en_US: constrains effort on reasoning for reasoning models + required: false + options: + - low + - medium + - high + - name: response_format + label: + zh_Hans: 回复格式 + en_US: response_format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '15.00' + output: '60.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openAI/llm/o1.yaml b/ai-provider/model-runtime/model-providers/openAI/llm/o1.yaml new file mode 100644 index 00000000..53d5f719 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openAI/llm/o1.yaml @@ -0,0 +1,49 @@ +model: o1 +label: + zh_Hans: o1 + en_US: o1 +model_type: llm +features: + - multi-tool-call + - agent-thought + - stream-tool-call + - vision +model_properties: + mode: chat + context_size: 200000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + default: 50000 + min: 1 + max: 50000 + - name: reasoning_effort + label: + zh_Hans: 推理工作 + en_US: reasoning_effort + type: string + help: + zh_Hans: 限制推理模型的推理工作 + en_US: constrains effort on reasoning for reasoning models + required: false + options: + - low + - medium + - high + - name: response_format + label: + zh_Hans: 回复格式 + en_US: response_format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '15.00' + output: '60.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openAI/llm/o3-mini-2025-01-31.yaml b/ai-provider/model-runtime/model-providers/openAI/llm/o3-mini-2025-01-31.yaml new file mode 100644 index 00000000..3f717bf0 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openAI/llm/o3-mini-2025-01-31.yaml @@ -0,0 +1,46 @@ +model: o3-mini-2025-01-31 +label: + zh_Hans: o3-mini-2025-01-31 + en_US: o3-mini-2025-01-31 +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 200000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + default: 100000 + min: 1 + max: 100000 + - name: reasoning_effort + label: + zh_Hans: 推理工作 + en_US: reasoning_effort + type: string + help: + zh_Hans: 限制推理模型的推理工作 + en_US: constrains effort on reasoning for reasoning models + required: false + options: + - low + - medium + - high + - name: response_format + label: + zh_Hans: 回复格式 + en_US: response_format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '1.10' + output: '4.40' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openAI/llm/o3-mini.yaml b/ai-provider/model-runtime/model-providers/openAI/llm/o3-mini.yaml new file mode 100644 index 00000000..755fc005 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openAI/llm/o3-mini.yaml @@ -0,0 +1,46 @@ +model: o3-mini +label: + zh_Hans: o3-mini + en_US: o3-mini +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 200000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + default: 100000 + min: 1 + max: 100000 + - name: reasoning_effort + label: + zh_Hans: 推理工作 + en_US: reasoning_effort + type: string + help: + zh_Hans: 限制推理模型的推理工作 + en_US: constrains effort on reasoning for reasoning models + required: false + options: + - low + - medium + - high + - name: response_format + label: + zh_Hans: 回复格式 + en_US: response_format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '1.10' + output: '4.40' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openrouter/llm/claude-3-5-haiku.yaml b/ai-provider/model-runtime/model-providers/openrouter/llm/claude-3-5-haiku.yaml new file mode 100644 index 00000000..de45093a --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openrouter/llm/claude-3-5-haiku.yaml @@ -0,0 +1,38 @@ +model: anthropic/claude-3-5-haiku +label: + en_US: claude-3-5-haiku +model_type: llm +features: + - agent-thought + - tool-call + - stream-tool-call +model_properties: + mode: chat + context_size: 200000 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: max_tokens + use_template: max_tokens + required: true + default: 8192 + min: 1 + max: 8192 + - name: response_format + use_template: response_format +pricing: + input: "1" + output: "5" + unit: "0.000001" + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openrouter/llm/deepseek-r1.yaml b/ai-provider/model-runtime/model-providers/openrouter/llm/deepseek-r1.yaml new file mode 100644 index 00000000..488b7dd4 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openrouter/llm/deepseek-r1.yaml @@ -0,0 +1,59 @@ +model: deepseek/deepseek-r1 +label: + en_US: deepseek-r1 +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 163840 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 1 + min: 0.0 + max: 2.0 + help: + zh_Hans: 控制生成结果的多样性和随机性。数值越小，越严谨；数值越大，越发散。 + en_US: Control the diversity and randomness of generated results. The smaller the value, the more rigorous it is; the larger the value, the more divergent it is. + - name: max_tokens + use_template: max_tokens + type: int + default: 4096 + min: 1 + max: 4096 + help: + zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: top_p + use_template: top_p + type: float + default: 1 + min: 0.01 + max: 1.00 + help: + zh_Hans: 控制生成结果的随机性。数值越小，随机性越弱；数值越大，随机性越强。一般而言，top_p 和 temperature 两个参数选择一个进行调整即可。 + en_US: Control the randomness of generated results. The smaller the value, the weaker the randomness; the larger the value, the stronger the randomness. Generally speaking, you can adjust one of the two parameters top_p and temperature. + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: frequency_penalty + use_template: frequency_penalty + default: 0 + min: -2.0 + max: 2.0 + help: + zh_Hans: 介于 -2.0 和 2.0 之间的数字。如果该值为正，那么新 token 会根据其在已有文本中的出现频率受到相应的惩罚，降低模型重复相同内容的可能性。 + en_US: A number between -2.0 and 2.0. If the value is positive, new tokens are penalized based on their frequency of occurrence in existing text, reducing the likelihood that the model will repeat the same content. +pricing: + input: "3" + output: "8" + unit: "0.000001" + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-11b-vision-instruct.yaml b/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-11b-vision-instruct.yaml new file mode 100644 index 00000000..6ad2c26c --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-11b-vision-instruct.yaml @@ -0,0 +1,46 @@ +model: meta-llama/llama-3.2-11b-vision-instruct +label: + zh_Hans: llama-3.2-11b-vision-instruct + en_US: llama-3.2-11b-vision-instruct +model_type: llm +features: + - agent-thought + - vision +model_properties: + mode: chat + context_size: 131072 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + - name: max_tokens + use_template: max_tokens + - name: context_length_exceeded_behavior + default: None + label: + zh_Hans: 上下文长度超出行为 + en_US: Context Length Exceeded Behavior + help: + zh_Hans: 上下文长度超出行为 + en_US: Context Length Exceeded Behavior + type: string + options: + - None + - truncate + - error + - name: response_format + use_template: response_format +pricing: + input: '0.055' + output: '0.055' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-1b-instruct.yaml b/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-1b-instruct.yaml new file mode 100644 index 00000000..657ef168 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-1b-instruct.yaml @@ -0,0 +1,45 @@ +model: meta-llama/llama-3.2-1b-instruct +label: + zh_Hans: llama-3.2-1b-instruct + en_US: llama-3.2-1b-instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 131072 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + - name: max_tokens + use_template: max_tokens + - name: context_length_exceeded_behavior + default: None + label: + zh_Hans: 上下文长度超出行为 + en_US: Context Length Exceeded Behavior + help: + zh_Hans: 上下文长度超出行为 + en_US: Context Length Exceeded Behavior + type: string + options: + - None + - truncate + - error + - name: response_format + use_template: response_format +pricing: + input: '0.01' + output: '0.02' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-3b-instruct.yaml b/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-3b-instruct.yaml new file mode 100644 index 00000000..7f6e24e5 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-3b-instruct.yaml @@ -0,0 +1,45 @@ +model: meta-llama/llama-3.2-3b-instruct +label: + zh_Hans: llama-3.2-3b-instruct + en_US: llama-3.2-3b-instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 131072 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + - name: max_tokens + use_template: max_tokens + - name: context_length_exceeded_behavior + default: None + label: + zh_Hans: 上下文长度超出行为 + en_US: Context Length Exceeded Behavior + help: + zh_Hans: 上下文长度超出行为 + en_US: Context Length Exceeded Behavior + type: string + options: + - None + - truncate + - error + - name: response_format + use_template: response_format +pricing: + input: '0.03' + output: '0.05' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-90b-vision-instruct.yaml b/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-90b-vision-instruct.yaml new file mode 100644 index 00000000..c264db0f --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openrouter/llm/llama-3.2-90b-vision-instruct.yaml @@ -0,0 +1,46 @@ +model: meta-llama/llama-3.2-90b-vision-instruct +label: + zh_Hans: llama-3.2-90b-vision-instruct + en_US: llama-3.2-90b-vision-instruct +model_type: llm +features: + - agent-thought + - vision +model_properties: + mode: chat + context_size: 131072 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + - name: max_tokens + use_template: max_tokens + - name: context_length_exceeded_behavior + default: None + label: + zh_Hans: 上下文长度超出行为 + en_US: Context Length Exceeded Behavior + help: + zh_Hans: 上下文长度超出行为 + en_US: Context Length Exceeded Behavior + type: string + options: + - None + - truncate + - error + - name: response_format + use_template: response_format +pricing: + input: '0.35' + output: '0.4' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openrouter/llm/o3-mini-2025-01-31.yaml b/ai-provider/model-runtime/model-providers/openrouter/llm/o3-mini-2025-01-31.yaml new file mode 100644 index 00000000..0cb38c11 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openrouter/llm/o3-mini-2025-01-31.yaml @@ -0,0 +1,49 @@ +model: openai/o3-mini-2025-01-31 +label: + en_US: o3-mini-2025-01-31 +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 200000 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: presence_penalty + use_template: presence_penalty + - name: frequency_penalty + use_template: frequency_penalty + - name: max_tokens + use_template: max_tokens + default: 512 + min: 1 + max: 100000 + - name: response_format + label: + zh_Hans: 回复格式 + en_US: response_format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: "1.10" + output: "4.40" + unit: "0.000001" + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openrouter/llm/o3-mini.yaml b/ai-provider/model-runtime/model-providers/openrouter/llm/o3-mini.yaml new file mode 100644 index 00000000..6fe38bd3 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openrouter/llm/o3-mini.yaml @@ -0,0 +1,49 @@ +model: openai/o3-mini +label: + en_US: o3-mini +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 200000 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: presence_penalty + use_template: presence_penalty + - name: frequency_penalty + use_template: frequency_penalty + - name: max_tokens + use_template: max_tokens + default: 512 + min: 1 + max: 100000 + - name: response_format + label: + zh_Hans: 回复格式 + en_US: response_format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: "1.10" + output: "4.40" + unit: "0.000001" + currency: USD diff --git a/ai-provider/model-runtime/model-providers/openrouter/llm/qwen2.5-72b-instruct.yaml b/ai-provider/model-runtime/model-providers/openrouter/llm/qwen2.5-72b-instruct.yaml new file mode 100644 index 00000000..5392b111 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/openrouter/llm/qwen2.5-72b-instruct.yaml @@ -0,0 +1,39 @@ +model: qwen/qwen-2.5-72b-instruct +label: + en_US: qwen-2.5-72b-instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 131072 +parameter_rules: + - name: temperature + use_template: temperature + - name: max_tokens + use_template: max_tokens + type: int + default: 512 + min: 1 + max: 8192 + help: + zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: frequency_penalty + use_template: frequency_penalty +pricing: + input: "0.35" + output: "0.4" + unit: "0.000001" + currency: USD diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/Internvl2-26b.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/Internvl2-26b.yaml new file mode 100644 index 00000000..f7b03e12 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/Internvl2-26b.yaml @@ -0,0 +1,84 @@ +model: OpenGVLab/InternVL2-26B +label: + en_US: OpenGVLab/InternVL2-26B +model_type: llm +features: + - vision +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 2000 + min: 1 + max: 2000 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '21' + output: '21' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/Internvl2-8b.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/Internvl2-8b.yaml new file mode 100644 index 00000000..1e858bb4 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/Internvl2-8b.yaml @@ -0,0 +1,84 @@ +model: Pro/OpenGVLab/InternVL2-8B +label: + en_US: Pro/OpenGVLab/InternVL2-8B +model_type: llm +features: + - vision +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 2000 + min: 1 + max: 2000 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '21' + output: '21' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-llama-70B.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-llama-70B.yaml new file mode 100644 index 00000000..59e0b4d6 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-llama-70B.yaml @@ -0,0 +1,21 @@ +model: deepseek-ai/DeepSeek-R1-Distill-Llama-70B +label: + zh_Hans: deepseek-ai/DeepSeek-R1-Distill-Llama-70B + en_US: deepseek-ai/DeepSeek-R1-Distill-Llama-70B +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + min: 1 + max: 8192 + default: 4096 +pricing: + input: "0.00" + output: "4.3" + unit: "0.000001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-llama-8B.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-llama-8B.yaml new file mode 100644 index 00000000..f3256aa5 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-llama-8B.yaml @@ -0,0 +1,21 @@ +model: deepseek-ai/DeepSeek-R1-Distill-Llama-8B +label: + zh_Hans: deepseek-ai/DeepSeek-R1-Distill-Llama-8B + en_US: deepseek-ai/DeepSeek-R1-Distill-Llama-8B +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + min: 1 + max: 8192 + default: 4096 +pricing: + input: "0.00" + output: "0.00" + unit: "0.000001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-1.5B.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-1.5B.yaml new file mode 100644 index 00000000..72972786 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-1.5B.yaml @@ -0,0 +1,21 @@ +model: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B +label: + zh_Hans: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B + en_US: deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + min: 1 + max: 8192 + default: 4096 +pricing: + input: "0.00" + output: "1.26" + unit: "0.000001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-14B.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-14B.yaml new file mode 100644 index 00000000..24b5c89e --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-14B.yaml @@ -0,0 +1,21 @@ +model: deepseek-ai/DeepSeek-R1-Distill-Qwen-14B +label: + zh_Hans: deepseek-ai/DeepSeek-R1-Distill-Qwen-14B + en_US: deepseek-ai/DeepSeek-R1-Distill-Qwen-14B +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + min: 1 + max: 8192 + default: 4096 +pricing: + input: "0.00" + output: "0.70" + unit: "0.000001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-32B.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-32B.yaml new file mode 100644 index 00000000..2a8cce1f --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-32B.yaml @@ -0,0 +1,21 @@ +model: deepseek-ai/DeepSeek-R1-Distill-Qwen-32B +label: + zh_Hans: deepseek-ai/DeepSeek-R1-Distill-Qwen-32B + en_US: deepseek-ai/DeepSeek-R1-Distill-Qwen-32B +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + min: 1 + max: 8192 + default: 4096 +pricing: + input: "0.00" + output: "1.26" + unit: "0.000001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-7B.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-7B.yaml new file mode 100644 index 00000000..cde1c14a --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1-distill-qwen-7B.yaml @@ -0,0 +1,21 @@ +model: deepseek-ai/DeepSeek-R1-Distill-Qwen-7B +label: + zh_Hans: deepseek-ai/DeepSeek-R1-Distill-Qwen-7B + en_US: deepseek-ai/DeepSeek-R1-Distill-Qwen-7B +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + min: 1 + max: 8192 + default: 4096 +pricing: + input: "0.00" + output: "0.00" + unit: "0.000001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1.yaml new file mode 100644 index 00000000..44c6a9d5 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-r1.yaml @@ -0,0 +1,21 @@ +model: deepseek-ai/DeepSeek-R1 +label: + zh_Hans: deepseek-ai/DeepSeek-R1 + en_US: deepseek-ai/DeepSeek-R1 +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 64000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + min: 1 + max: 8192 + default: 4096 +pricing: + input: "4" + output: "16" + unit: "0.000001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-v3.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-v3.yaml new file mode 100644 index 00000000..ed1a5f00 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/deepseek-v3.yaml @@ -0,0 +1,53 @@ +model: deepseek-ai/DeepSeek-V3 +label: + en_US: deepseek-ai/DeepSeek-V3 +model_type: llm +features: + - agent-thought + - tool-call + - stream-tool-call +model_properties: + mode: chat + context_size: 64000 +parameter_rules: + - name: temperature + use_template: temperature + - name: max_tokens + use_template: max_tokens + type: int + default: 512 + min: 1 + max: 4096 + help: + zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: frequency_penalty + use_template: frequency_penalty + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: "1" + output: "2" + unit: "0.000001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/hunyuan-a52b-instruct.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/hunyuan-a52b-instruct.yaml new file mode 100644 index 00000000..51d6c024 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/hunyuan-a52b-instruct.yaml @@ -0,0 +1,85 @@ +model: Tencent/Hunyuan-A52B-Instruct +label: + en_US: Tencent/Hunyuan-A52B-Instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 2000 + min: 1 + max: 2000 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '21' + output: '21' + unit: '0.000001' + currency: RMB +deprecated: true diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/internlm2_5-20b-chat.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/internlm2_5-20b-chat.yaml new file mode 100644 index 00000000..a5ae3674 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/internlm2_5-20b-chat.yaml @@ -0,0 +1,51 @@ +model: internlm/internlm2_5-20b-chat +label: + en_US: internlm/internlm2_5-20b-chat +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + - name: max_tokens + use_template: max_tokens + type: int + default: 512 + min: 1 + max: 4096 + help: + zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: frequency_penalty + use_template: frequency_penalty + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '1' + output: '1' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/internvl2-llama3-76b.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/internvl2-llama3-76b.yaml new file mode 100644 index 00000000..b5443df1 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/internvl2-llama3-76b.yaml @@ -0,0 +1,85 @@ +model: OpenGVLab/InternVL2-Llama3-76B +label: + en_US: OpenGVLab/InternVL2-Llama3-76B +model_type: llm +features: + - vision +model_properties: + mode: chat + context_size: 8192 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 2000 + min: 1 + max: 2000 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '21' + output: '21' + unit: '0.000001' + currency: RMB +deprecated: true diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/janus-pro-7B.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/janus-pro-7B.yaml new file mode 100644 index 00000000..dabbd745 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/janus-pro-7B.yaml @@ -0,0 +1,22 @@ +model: deepseek-ai/Janus-Pro-7B +label: + zh_Hans: deepseek-ai/Janus-Pro-7B + en_US: deepseek-ai/Janus-Pro-7B +model_type: llm +features: + - agent-thought + - vision +model_properties: + mode: chat + context_size: 32000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + min: 1 + max: 8192 + default: 4096 +pricing: + input: "0.00" + output: "0.00" + unit: "0.000001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/meta-llama-3.3-70b-instruct.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/meta-llama-3.3-70b-instruct.yaml new file mode 100644 index 00000000..9373a8f4 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/meta-llama-3.3-70b-instruct.yaml @@ -0,0 +1,53 @@ +model: meta-llama/Llama-3.3-70B-Instruct +label: + en_US: meta-llama/Llama-3.3-70B-Instruct +model_type: llm +features: + - agent-thought + - tool-call + - stream-tool-call +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + - name: max_tokens + use_template: max_tokens + type: int + default: 512 + min: 1 + max: 4096 + help: + zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: frequency_penalty + use_template: frequency_penalty + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '4.13' + output: '4.13' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen-qvq-72B-preview.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen-qvq-72B-preview.yaml new file mode 100644 index 00000000..dada6bb8 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen-qvq-72B-preview.yaml @@ -0,0 +1,54 @@ +model: Qwen/QVQ-72B-Preview +label: + en_US: Qwen/QVQ-72B-Preview +model_type: llm +features: + - agent-thought + - tool-call + - stream-tool-call + - vision +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + - name: max_tokens + use_template: max_tokens + type: int + default: 8192 + min: 1 + max: 16384 + help: + zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: frequency_penalty + use_template: frequency_penalty + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '9.90' + output: '9.90' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen-qwq-32B-preview.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen-qwq-32B-preview.yaml new file mode 100644 index 00000000..e73c5d20 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen-qwq-32B-preview.yaml @@ -0,0 +1,53 @@ +model: Qwen/QwQ-32B-Preview +label: + en_US: Qwen/QwQ-32B-Preview +model_type: llm +features: + - agent-thought + - tool-call + - stream-tool-call +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + - name: max_tokens + use_template: max_tokens + type: int + default: 4096 + min: 1 + max: 8192 + help: + zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: frequency_penalty + use_template: frequency_penalty + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '1.26' + output: '1.26' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2-vl-72b-instruct.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2-vl-72b-instruct.yaml new file mode 100644 index 00000000..f5180b41 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2-vl-72b-instruct.yaml @@ -0,0 +1,84 @@ +model: Qwen/Qwen2-VL-72B-Instruct +label: + en_US: Qwen/Qwen2-VL-72B-Instruct +model_type: llm +features: + - vision +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 2000 + min: 1 + max: 2000 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '4.13' + output: '4.13' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2-vl-7b-Instruct.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2-vl-7b-Instruct.yaml new file mode 100644 index 00000000..0ffbaee3 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2-vl-7b-Instruct.yaml @@ -0,0 +1,84 @@ +model: Pro/Qwen/Qwen2-VL-7B-Instruct +label: + en_US: Pro/Qwen/Qwen2-VL-7B-Instruct +model_type: llm +features: + - vision +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 2000 + min: 1 + max: 2000 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '0.35' + output: '0.35' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-72b-instruct-128k.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-72b-instruct-128k.yaml new file mode 100644 index 00000000..79f94da3 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-72b-instruct-128k.yaml @@ -0,0 +1,51 @@ +model: Qwen/Qwen2.5-72B-Instruct-128K +label: + en_US: Qwen/Qwen2.5-72B-Instruct-128K +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 131072 +parameter_rules: + - name: temperature + use_template: temperature + - name: max_tokens + use_template: max_tokens + type: int + default: 512 + min: 1 + max: 4096 + help: + zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: frequency_penalty + use_template: frequency_penalty + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '4.13' + output: '4.13' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-72b-instruct-vendorA.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-72b-instruct-vendorA.yaml new file mode 100644 index 00000000..fdbe38ff --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-72b-instruct-vendorA.yaml @@ -0,0 +1,51 @@ +model: Vendor-A/Qwen/Qwen2.5-72B-Instruct +label: + en_US: Vendor-A/Qwen/Qwen2.5-72B-Instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + - name: max_tokens + use_template: max_tokens + type: int + default: 512 + min: 1 + max: 4096 + help: + zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: frequency_penalty + use_template: frequency_penalty + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '1.00' + output: '1.00' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-coder-32b-instruct.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-coder-32b-instruct.yaml new file mode 100644 index 00000000..de2224a6 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-coder-32b-instruct.yaml @@ -0,0 +1,84 @@ +model: Qwen/Qwen2.5-Coder-32B-Instruct +label: + en_US: Qwen/Qwen2.5-Coder-32B-Instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32768 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 8192 + min: 1 + max: 8192 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '1.26' + output: '1.26' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-coder-7b-instruct.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-coder-7b-instruct.yaml new file mode 100644 index 00000000..c31a338c --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-coder-7b-instruct.yaml @@ -0,0 +1,84 @@ +model: Qwen/Qwen2.5-Coder-7B-Instruct +label: + en_US: Qwen/Qwen2.5-Coder-7B-Instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 131072 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 8192 + min: 1 + max: 8192 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '0' + output: '0' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-math-72b-instruct.yaml b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-math-72b-instruct.yaml new file mode 100644 index 00000000..40c9ab48 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/siliconflow/llm/qwen2.5-math-72b-instruct.yaml @@ -0,0 +1,85 @@ +model: Qwen/Qwen2.5-Math-72B-Instruct +label: + en_US: Qwen/Qwen2.5-Math-72B-Instruct +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 4096 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 2000 + min: 1 + max: 2000 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '4.13' + output: '4.13' + unit: '0.000001' + currency: RMB +deprecated: true diff --git a/ai-provider/model-runtime/model-providers/spark/llm/spark-1.5.yaml b/ai-provider/model-runtime/model-providers/spark/llm/spark-1.5.yaml new file mode 100644 index 00000000..fcd65c24 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/spark/llm/spark-1.5.yaml @@ -0,0 +1,34 @@ +model: spark-1.5 +deprecated: true +label: + en_US: Spark V1.5 +model_type: llm +model_properties: + mode: chat +parameter_rules: + - name: temperature + use_template: temperature + default: 0.5 + help: + zh_Hans: 核采样阈值。用于决定结果随机性，取值越高随机性越强即相同的问题得到的不同答案的可能性越高。 + en_US: Kernel sampling threshold. Used to determine the randomness of the results. The higher the value, the stronger the randomness, that is, the higher the possibility of getting different answers to the same question. + - name: max_tokens + use_template: max_tokens + default: 512 + min: 1 + max: 4096 + help: + zh_Hans: 模型回答的tokens的最大长度。 + en_US: 模型回答的tokens的最大长度。 + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + default: 4 + min: 1 + max: 6 + help: + zh_Hans: 从 k 个候选中随机选择一个（非等概率）。 + en_US: Randomly select one from k candidates (non-equal probability). + required: false diff --git a/ai-provider/model-runtime/model-providers/spark/llm/spark-2.yaml b/ai-provider/model-runtime/model-providers/spark/llm/spark-2.yaml new file mode 100644 index 00000000..2db6805a --- /dev/null +++ b/ai-provider/model-runtime/model-providers/spark/llm/spark-2.yaml @@ -0,0 +1,34 @@ +model: spark-2 +deprecated: true +label: + en_US: Spark V2.0 +model_type: llm +model_properties: + mode: chat +parameter_rules: + - name: temperature + use_template: temperature + default: 0.5 + help: + zh_Hans: 核采样阈值。用于决定结果随机性，取值越高随机性越强即相同的问题得到的不同答案的可能性越高。 + en_US: Kernel sampling threshold. Used to determine the randomness of the results. The higher the value, the stronger the randomness, that is, the higher the possibility of getting different answers to the same question. + - name: max_tokens + use_template: max_tokens + default: 2048 + min: 1 + max: 8192 + help: + zh_Hans: 模型回答的tokens的最大长度。 + en_US: 模型回答的tokens的最大长度。 + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + default: 4 + min: 1 + max: 6 + help: + zh_Hans: 从 k 个候选中随机选择一个（非等概率）。 + en_US: Randomly select one from k candidates (non-equal probability). + required: false diff --git a/ai-provider/model-runtime/model-providers/spark/llm/spark-3.5.yaml b/ai-provider/model-runtime/model-providers/spark/llm/spark-3.5.yaml new file mode 100644 index 00000000..86617a53 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/spark/llm/spark-3.5.yaml @@ -0,0 +1,34 @@ +model: spark-3.5 +deprecated: true +label: + en_US: Spark V3.5 +model_type: llm +model_properties: + mode: chat +parameter_rules: + - name: temperature + use_template: temperature + default: 0.5 + help: + zh_Hans: 核采样阈值。用于决定结果随机性，取值越高随机性越强即相同的问题得到的不同答案的可能性越高。 + en_US: Kernel sampling threshold. Used to determine the randomness of the results. The higher the value, the stronger the randomness, that is, the higher the possibility of getting different answers to the same question. + - name: max_tokens + use_template: max_tokens + default: 2048 + min: 1 + max: 8192 + help: + zh_Hans: 模型回答的tokens的最大长度。 + en_US: 模型回答的tokens的最大长度。 + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + default: 4 + min: 1 + max: 6 + help: + zh_Hans: 从 k 个候选中随机选择一个（非等概率）。 + en_US: Randomly select one from k candidates (non-equal probability). + required: false diff --git a/ai-provider/model-runtime/model-providers/spark/llm/spark-3.yaml b/ai-provider/model-runtime/model-providers/spark/llm/spark-3.yaml new file mode 100644 index 00000000..9f296c68 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/spark/llm/spark-3.yaml @@ -0,0 +1,34 @@ +model: spark-3 +deprecated: true +label: + en_US: Spark V3.0 +model_type: llm +model_properties: + mode: chat +parameter_rules: + - name: temperature + use_template: temperature + default: 0.5 + help: + zh_Hans: 核采样阈值。用于决定结果随机性，取值越高随机性越强即相同的问题得到的不同答案的可能性越高。 + en_US: Kernel sampling threshold. Used to determine the randomness of the results. The higher the value, the stronger the randomness, that is, the higher the possibility of getting different answers to the same question. + - name: max_tokens + use_template: max_tokens + default: 2048 + min: 1 + max: 8192 + help: + zh_Hans: 模型回答的tokens的最大长度。 + en_US: 模型回答的tokens的最大长度。 + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + default: 4 + min: 1 + max: 6 + help: + zh_Hans: 从 k 个候选中随机选择一个（非等概率）。 + en_US: Randomly select one from k candidates (non-equal probability). + required: false diff --git a/ai-provider/model-runtime/model-providers/spark/llm/spark-4.yaml b/ai-provider/model-runtime/model-providers/spark/llm/spark-4.yaml new file mode 100644 index 00000000..4b5529e8 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/spark/llm/spark-4.yaml @@ -0,0 +1,34 @@ +model: spark-4 +deprecated: true +label: + en_US: Spark V4.0 +model_type: llm +model_properties: + mode: chat +parameter_rules: + - name: temperature + use_template: temperature + default: 0.5 + help: + zh_Hans: 核采样阈值。用于决定结果随机性，取值越高随机性越强即相同的问题得到的不同答案的可能性越高。 + en_US: Kernel sampling threshold. Used to determine the randomness of the results. The higher the value, the stronger the randomness, that is, the higher the possibility of getting different answers to the same question. + - name: max_tokens + use_template: max_tokens + default: 4096 + min: 1 + max: 8192 + help: + zh_Hans: 模型回答的tokens的最大长度。 + en_US: 模型回答的tokens的最大长度。 + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + default: 4 + min: 1 + max: 6 + help: + zh_Hans: 从 k 个候选中随机选择一个（非等概率）。 + en_US: Randomly select one from k candidates (non-equal probability). + required: false diff --git a/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-r1-distill-qwen-14B.yaml b/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-r1-distill-qwen-14B.yaml new file mode 100644 index 00000000..2bce8805 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-r1-distill-qwen-14B.yaml @@ -0,0 +1,21 @@ +model: deepseek-r1-distill-qwen-14b +label: + zh_Hans: DeepSeek-R1-Distill-Qwen-14B + en_US: DeepSeek-R1-Distill-Qwen-14B +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + min: 1 + max: 8192 + default: 4096 +pricing: + input: "0.001" + output: "0.003" + unit: "0.001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-r1-distill-qwen-32B.yaml b/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-r1-distill-qwen-32B.yaml new file mode 100644 index 00000000..dfc155ff --- /dev/null +++ b/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-r1-distill-qwen-32B.yaml @@ -0,0 +1,21 @@ +model: deepseek-r1-distill-qwen-32b +label: + zh_Hans: DeepSeek-R1-Distill-Qwen-32B + en_US: DeepSeek-R1-Distill-Qwen-32B +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 32000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + min: 1 + max: 8192 + default: 4096 +pricing: + input: "0.002" + output: "0.006" + unit: "0.001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-r1.yaml b/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-r1.yaml new file mode 100644 index 00000000..742e7f00 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-r1.yaml @@ -0,0 +1,21 @@ +model: deepseek-r1 +label: + zh_Hans: DeepSeek-R1 + en_US: DeepSeek-R1 +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 64000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + min: 1 + max: 8192 + default: 4096 +pricing: + input: "0.004" + output: "0.016" + unit: '0.001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-v3.yaml b/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-v3.yaml new file mode 100644 index 00000000..23f38d60 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/tongyi/llm/deepseek-v3.yaml @@ -0,0 +1,52 @@ +model: deepseek-v3 +label: + zh_Hans: DeepSeek-V3 + en_US: DeepSeek-V3 +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 64000 +parameter_rules: + - name: temperature + use_template: temperature + - name: max_tokens + use_template: max_tokens + type: int + default: 512 + min: 1 + max: 4096 + help: + zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: frequency_penalty + use_template: frequency_penalty + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: "0.002" + output: "0.008" + unit: "0.001" + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/tongyi/llm/farui-plus.yaml b/ai-provider/model-runtime/model-providers/tongyi/llm/farui-plus.yaml new file mode 100644 index 00000000..34a57d1f --- /dev/null +++ b/ai-provider/model-runtime/model-providers/tongyi/llm/farui-plus.yaml @@ -0,0 +1,77 @@ +# for more details, please refer to https://help.aliyun.com/zh/model-studio/getting-started/models +model: farui-plus +label: + en_US: farui-plus +model_type: llm +features: + - multi-tool-call + - agent-thought + - stream-tool-call +model_properties: + mode: chat + context_size: 12288 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 2000 + min: 1 + max: 2000 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + use_template: response_format +pricing: + input: '0.02' + output: '0.02' + unit: '0.001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/tongyi/llm/qwen2.5-14b-instruct-1m.yaml b/ai-provider/model-runtime/model-providers/tongyi/llm/qwen2.5-14b-instruct-1m.yaml new file mode 100644 index 00000000..c3d72ec4 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/tongyi/llm/qwen2.5-14b-instruct-1m.yaml @@ -0,0 +1,75 @@ +# for more details, please refer to https://help.aliyun.com/zh/model-studio/getting-started/models +model: qwen2.5-14b-instruct-1m +label: + en_US: qwen2.5-14b-instruct-1m +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 1000000 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 8192 + min: 1 + max: 8192 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + use_template: response_format +pricing: + input: '0.001' + output: '0.003' + unit: '0.001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/tongyi/llm/qwen2.5-7b-instruct-1m.yaml b/ai-provider/model-runtime/model-providers/tongyi/llm/qwen2.5-7b-instruct-1m.yaml new file mode 100644 index 00000000..44968e54 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/tongyi/llm/qwen2.5-7b-instruct-1m.yaml @@ -0,0 +1,75 @@ +# for more details, please refer to https://help.aliyun.com/zh/model-studio/getting-started/models +model: qwen2.5-7b-instruct-1m +label: + en_US: qwen2.5-7b-instruct-1m +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 1000000 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 用于控制随机性和多样性的程度。具体来说，temperature值控制了生成文本时对每个候选词的概率分布进行平滑的程度。较高的temperature值会降低概率分布的峰值，使得更多的低概率词被选择，生成结果更加多样化；而较低的temperature值则会增强概率分布的峰值，使得高概率词更容易被选择，生成结果更加确定。 + en_US: Used to control the degree of randomness and diversity. Specifically, the temperature value controls the degree to which the probability distribution of each candidate word is smoothed when generating text. A higher temperature value will reduce the peak value of the probability distribution, allowing more low-probability words to be selected, and the generated results will be more diverse; while a lower temperature value will enhance the peak value of the probability distribution, making it easier for high-probability words to be selected. , the generated results are more certain. + - name: max_tokens + use_template: max_tokens + type: int + default: 8192 + min: 1 + max: 8192 + help: + zh_Hans: 用于指定模型在生成内容时token的最大数量，它定义了生成的上限，但不保证每次都会生成到这个数量。 + en_US: It is used to specify the maximum number of tokens when the model generates content. It defines the upper limit of generation, but does not guarantee that this number will be generated every time. + - name: top_p + use_template: top_p + type: float + default: 0.8 + min: 0.1 + max: 0.9 + help: + zh_Hans: 生成过程中核采样方法概率阈值，例如，取值为0.8时，仅保留概率加起来大于等于0.8的最可能token的最小集合作为候选集。取值范围为（0,1.0)，取值越大，生成的随机性越高；取值越低，生成的确定性越高。 + en_US: The probability threshold of the kernel sampling method during the generation process. For example, when the value is 0.8, only the smallest set of the most likely tokens with a sum of probabilities greater than or equal to 0.8 is retained as the candidate set. The value range is (0,1.0). The larger the value, the higher the randomness generated; the lower the value, the higher the certainty generated. + - name: top_k + type: int + min: 0 + max: 99 + label: + zh_Hans: 取样数量 + en_US: Top k + help: + zh_Hans: 生成时，采样候选集的大小。例如，取值为50时，仅将单次生成中得分最高的50个token组成随机采样的候选集。取值越大，生成的随机性越高；取值越小，生成的确定性越高。 + en_US: The size of the sample candidate set when generated. For example, when the value is 50, only the 50 highest-scoring tokens in a single generation form a randomly sampled candidate set. The larger the value, the higher the randomness generated; the smaller the value, the higher the certainty generated. + - name: seed + required: false + type: int + default: 1234 + label: + zh_Hans: 随机种子 + en_US: Random seed + help: + zh_Hans: 生成时使用的随机数种子，用户控制模型生成内容的随机性。支持无符号64位整数，默认值为 1234。在使用seed时，模型将尽可能生成相同或相似的结果，但目前不保证每次生成的结果完全相同。 + en_US: The random number seed used when generating, the user controls the randomness of the content generated by the model. Supports unsigned 64-bit integers, default value is 1234. When using seed, the model will try its best to generate the same or similar results, but there is currently no guarantee that the results will be exactly the same every time. + - name: repetition_penalty + required: false + type: float + default: 1.1 + label: + zh_Hans: 重复惩罚 + en_US: Repetition penalty + help: + zh_Hans: 用于控制模型生成时的重复度。提高repetition_penalty时可以降低模型生成的重复度。1.0表示不做惩罚。 + en_US: Used to control the repeatability when generating models. Increasing repetition_penalty can reduce the duplication of model generation. 1.0 means no punishment. + - name: response_format + use_template: response_format +pricing: + input: '0.0005' + output: '0.001' + unit: '0.001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/vertex_ai/llm/anthropic.claude-3.5-sonnet-v2.yaml b/ai-provider/model-runtime/model-providers/vertex_ai/llm/anthropic.claude-3.5-sonnet-v2.yaml new file mode 100644 index 00000000..37b9f30c --- /dev/null +++ b/ai-provider/model-runtime/model-providers/vertex_ai/llm/anthropic.claude-3.5-sonnet-v2.yaml @@ -0,0 +1,55 @@ +model: claude-3-5-sonnet-v2@20241022 +label: + en_US: Claude 3.5 Sonnet v2 +model_type: llm +features: + - agent-thought + - vision +model_properties: + mode: chat + context_size: 200000 +parameter_rules: + - name: max_tokens + use_template: max_tokens + required: true + type: int + default: 8192 + min: 1 + max: 8192 + help: + zh_Hans: 停止前生成的最大令牌数。请注意，Anthropic Claude 模型可能会在达到 max_tokens 的值之前停止生成令牌。不同的 Anthropic Claude 模型对此参数具有不同的最大值。 + en_US: The maximum number of tokens to generate before stopping. Note that Anthropic Claude models might stop generating tokens before reaching the value of max_tokens. Different Anthropic Claude models have different maximum values for this parameter. + - name: temperature + use_template: temperature + required: false + type: float + default: 1 + min: 0.0 + max: 1.0 + help: + zh_Hans: 生成内容的随机性。 + en_US: The amount of randomness injected into the response. + - name: top_p + required: false + type: float + default: 0.999 + min: 0.000 + max: 1.000 + help: + zh_Hans: 在核采样中，Anthropic Claude 按概率递减顺序计算每个后续标记的所有选项的累积分布，并在达到 top_p 指定的特定概率时将其切断。您应该更改温度或top_p，但不能同时更改两者。 + en_US: In nucleus sampling, Anthropic Claude computes the cumulative distribution over all the options for each subsequent token in decreasing probability order and cuts it off once it reaches a particular probability specified by top_p. You should alter either temperature or top_p, but not both. + - name: top_k + required: false + type: int + default: 0 + min: 0 + # tip docs from aws has error, max value is 500 + max: 500 + help: + zh_Hans: 对于每个后续标记，仅从前 K 个选项中进行采样。使用 top_k 删除长尾低概率响应。 + en_US: Only sample from the top K options for each subsequent token. Use top_k to remove long tail low probability responses. +pricing: + input: '0.003' + output: '0.015' + unit: '0.001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-001.yaml b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-001.yaml new file mode 100644 index 00000000..bef7ca5e --- /dev/null +++ b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-001.yaml @@ -0,0 +1,41 @@ +model: gemini-2.0-flash-001 +label: + en_US: Gemini 2.0 Flash 001 +model_type: llm +features: + - agent-thought + - vision + - tool-call + - stream-tool-call + - document + - video + - audio +model_properties: + mode: chat + context_size: 1048576 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: max_output_tokens + use_template: max_tokens + default: 8192 + min: 1 + max: 8192 + - name: json_schema + use_template: json_schema +pricing: + input: '0.00' + output: '0.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-exp.yaml b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-exp.yaml new file mode 100644 index 00000000..bcd59623 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-exp.yaml @@ -0,0 +1,39 @@ +model: gemini-2.0-flash-exp +label: + en_US: Gemini 2.0 Flash Exp +model_type: llm +features: + - agent-thought + - vision + - tool-call + - stream-tool-call + - document +model_properties: + mode: chat + context_size: 1048576 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: max_output_tokens + use_template: max_tokens + default: 8192 + min: 1 + max: 8192 + - name: json_schema + use_template: json_schema +pricing: + input: '0.00' + output: '0.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-lite-preview-02-05.yaml b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-lite-preview-02-05.yaml new file mode 100644 index 00000000..9c0a1e06 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-lite-preview-02-05.yaml @@ -0,0 +1,41 @@ +model: gemini-2.0-flash-lite-preview-02-05 +label: + en_US: Gemini 2.0 Flash Lite Preview 0205 +model_type: llm +features: + - agent-thought + - vision + - tool-call + - stream-tool-call + - document + - video + - audio +model_properties: + mode: chat + context_size: 1048576 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: max_output_tokens + use_template: max_tokens + default: 8192 + min: 1 + max: 8192 + - name: json_schema + use_template: json_schema +pricing: + input: '0.00' + output: '0.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-thinking-exp-01-21.yaml b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-thinking-exp-01-21.yaml new file mode 100644 index 00000000..6e2fc767 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-thinking-exp-01-21.yaml @@ -0,0 +1,39 @@ +model: gemini-2.0-flash-thinking-exp-01-21 +label: + en_US: Gemini 2.0 Flash Thinking Exp 0121 +model_type: llm +features: + - agent-thought + - vision + - document + - video + - audio +model_properties: + mode: chat + context_size: 32767 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: max_output_tokens + use_template: max_tokens + default: 8192 + min: 1 + max: 8192 + - name: json_schema + use_template: json_schema +pricing: + input: '0.00' + output: '0.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-thinking-exp-1219.yaml b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-thinking-exp-1219.yaml new file mode 100644 index 00000000..dfcf8fd0 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-flash-thinking-exp-1219.yaml @@ -0,0 +1,39 @@ +model: gemini-2.0-flash-thinking-exp-1219 +label: + en_US: Gemini 2.0 Flash Thinking Exp 1219 +model_type: llm +features: + - agent-thought + - vision + - document + - video + - audio +model_properties: + mode: chat + context_size: 32767 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: max_output_tokens + use_template: max_tokens + default: 8192 + min: 1 + max: 8192 + - name: json_schema + use_template: json_schema +pricing: + input: '0.00' + output: '0.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-pro-exp-02-05.yaml b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-pro-exp-02-05.yaml new file mode 100644 index 00000000..96926a17 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-2.0-pro-exp-02-05.yaml @@ -0,0 +1,37 @@ +model: gemini-2.0-pro-exp-02-05 +label: + en_US: Gemini 2.0 Pro Exp 0205 +model_type: llm +features: + - agent-thought + - document +model_properties: + mode: chat + context_size: 2000000 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + en_US: Top k + type: int + help: + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: presence_penalty + use_template: presence_penalty + - name: frequency_penalty + use_template: frequency_penalty + - name: max_output_tokens + use_template: max_tokens + required: true + default: 8192 + min: 1 + max: 8192 +pricing: + input: '0.00' + output: '0.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-exp-1114.yaml b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-exp-1114.yaml new file mode 100644 index 00000000..bd49b476 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-exp-1114.yaml @@ -0,0 +1,41 @@ +model: gemini-exp-1114 +label: + en_US: Gemini exp 1114 +model_type: llm +features: + - agent-thought + - vision + - tool-call + - stream-tool-call + - document + - video + - audio +model_properties: + mode: chat + context_size: 32767 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: max_output_tokens + use_template: max_tokens + default: 8192 + min: 1 + max: 8192 + - name: json_schema + use_template: json_schema +pricing: + input: '0.00' + output: '0.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-exp-1121.yaml b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-exp-1121.yaml new file mode 100644 index 00000000..8e3f218d --- /dev/null +++ b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-exp-1121.yaml @@ -0,0 +1,41 @@ +model: gemini-exp-1121 +label: + en_US: Gemini exp 1121 +model_type: llm +features: + - agent-thought + - vision + - tool-call + - stream-tool-call + - document + - video + - audio +model_properties: + mode: chat + context_size: 32767 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: max_output_tokens + use_template: max_tokens + default: 8192 + min: 1 + max: 8192 + - name: json_schema + use_template: json_schema +pricing: + input: '0.00' + output: '0.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-exp-1206.yaml b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-exp-1206.yaml new file mode 100644 index 00000000..7a7c361c --- /dev/null +++ b/ai-provider/model-runtime/model-providers/vertex_ai/llm/gemini-exp-1206.yaml @@ -0,0 +1,41 @@ +model: gemini-exp-1206 +label: + en_US: Gemini exp 1206 +model_type: llm +features: + - agent-thought + - vision + - tool-call + - stream-tool-call + - document + - video + - audio +model_properties: + mode: chat + context_size: 2097152 +parameter_rules: + - name: temperature + use_template: temperature + - name: top_p + use_template: top_p + - name: top_k + label: + zh_Hans: 取样数量 + en_US: Top k + type: int + help: + zh_Hans: 仅从每个后续标记的前 K 个选项中采样。 + en_US: Only sample from the top K options for each subsequent token. + required: false + - name: max_output_tokens + use_template: max_tokens + default: 8192 + min: 1 + max: 8192 + - name: json_schema + use_template: json_schema +pricing: + input: '0.00' + output: '0.00' + unit: '0.000001' + currency: USD diff --git a/ai-provider/model-runtime/model-providers/wenxin/llm/ernie-4.0-turbo-128k.yaml b/ai-provider/model-runtime/model-providers/wenxin/llm/ernie-4.0-turbo-128k.yaml new file mode 100644 index 00000000..f8d56406 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/wenxin/llm/ernie-4.0-turbo-128k.yaml @@ -0,0 +1,40 @@ +model: ernie-4.0-turbo-128k +label: + en_US: Ernie-4.0-turbo-128K +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 131072 +parameter_rules: + - name: temperature + use_template: temperature + min: 0.1 + max: 1.0 + default: 0.8 + - name: top_p + use_template: top_p + - name: max_tokens + use_template: max_tokens + default: 1024 + min: 2 + max: 4096 + - name: presence_penalty + use_template: presence_penalty + default: 1.0 + min: 1.0 + max: 2.0 + - name: frequency_penalty + use_template: frequency_penalty + - name: response_format + use_template: response_format + - name: disable_search + label: + zh_Hans: 禁用搜索 + en_US: Disable Search + type: boolean + help: + zh_Hans: 禁用模型自行进行外部搜索。 + en_US: Disable the model to perform external search. + required: false diff --git a/ai-provider/model-runtime/model-providers/wenxin/llm/ernie-lite-pro-128k.yaml b/ai-provider/model-runtime/model-providers/wenxin/llm/ernie-lite-pro-128k.yaml new file mode 100644 index 00000000..4f5832c8 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/wenxin/llm/ernie-lite-pro-128k.yaml @@ -0,0 +1,42 @@ +model: ernie-lite-pro-128k +label: + en_US: Ernie-Lite-Pro-128K +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 128000 +parameter_rules: + - name: temperature + use_template: temperature + min: 0.1 + max: 1.0 + default: 0.8 + - name: top_p + use_template: top_p + - name: min_output_tokens + label: + en_US: "Min Output Tokens" + zh_Hans: "最小输出Token数" + use_template: max_tokens + min: 2 + max: 2048 + help: + zh_Hans: 指定模型最小输出token数 + en_US: Specifies the lower limit on the length of generated results. + - name: max_output_tokens + label: + en_US: "Max Output Tokens" + zh_Hans: "最大输出Token数" + use_template: max_tokens + min: 2 + max: 2048 + default: 2048 + help: + zh_Hans: 指定模型最大输出token数 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: presence_penalty + use_template: presence_penalty + - name: frequency_penalty + use_template: frequency_penalty diff --git a/ai-provider/model-runtime/model-providers/yi/llm/yi-lightning.yaml b/ai-provider/model-runtime/model-providers/yi/llm/yi-lightning.yaml new file mode 100644 index 00000000..fccf1b3a --- /dev/null +++ b/ai-provider/model-runtime/model-providers/yi/llm/yi-lightning.yaml @@ -0,0 +1,43 @@ +model: yi-lightning +label: + zh_Hans: yi-lightning + en_US: yi-lightning +model_type: llm +features: + - agent-thought +model_properties: + mode: chat + context_size: 16384 +parameter_rules: + - name: temperature + use_template: temperature + type: float + default: 0.3 + min: 0.0 + max: 2.0 + help: + zh_Hans: 控制生成结果的多样性和随机性。数值越小，越严谨；数值越大，越发散。 + en_US: Control the diversity and randomness of generated results. The smaller the value, the more rigorous it is; the larger the value, the more divergent it is. + - name: max_tokens + use_template: max_tokens + type: int + default: 1024 + min: 1 + max: 4000 + help: + zh_Hans: 指定生成结果长度的上限。如果生成结果截断，可以调大该参数。 + en_US: Specifies the upper limit on the length of generated results. If the generated results are truncated, you can increase this parameter. + - name: top_p + use_template: top_p + type: float + default: 0.9 + min: 0.01 + max: 1.00 + help: + zh_Hans: 控制生成结果的随机性。数值越小，随机性越弱；数值越大，随机性越强。一般而言，top_p 和 temperature 两个参数选择一个进行调整即可。 + en_US: Control the randomness of generated results. The smaller the value, the weaker the randomness; the larger the value, the stronger the randomness. Generally speaking, you can adjust one of the two parameters top_p and temperature. +pricing: + input: '0.99' + output: '0.99' + unit: '0.000001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/zhipuai/llm/glm-4-air-0111.yaml b/ai-provider/model-runtime/model-providers/zhipuai/llm/glm-4-air-0111.yaml new file mode 100644 index 00000000..8d301fc6 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/zhipuai/llm/glm-4-air-0111.yaml @@ -0,0 +1,66 @@ +model: glm-4-air-0111 +label: + en_US: glm-4-air-0111 +model_type: llm +features: + - multi-tool-call + - agent-thought + - stream-tool-call +model_properties: + mode: chat + context_size: 131072 +parameter_rules: + - name: temperature + use_template: temperature + default: 0.95 + min: 0.0 + max: 1.0 + help: + zh_Hans: 采样温度，控制输出的随机性，必须为正数取值范围是：(0.0,1.0]，不能等于 0,默认值为 0.95 值越大，会使输出更随机，更具创造性；值越小，输出会更加稳定或确定建议您根据应用场景调整 top_p 或 temperature 参数，但不要同时调整两个参数。 + en_US: Sampling temperature, controls the randomness of the output, must be a positive number. The value range is (0.0,1.0], which cannot be equal to 0. The default value is 0.95. The larger the value, the more random and creative the output will be; the smaller the value, The output will be more stable or certain. It is recommended that you adjust the top_p or temperature parameters according to the application scenario, but do not adjust both parameters at the same time. + - name: top_p + use_template: top_p + default: 0.7 + help: + zh_Hans: 用温度取样的另一种方法，称为核取样取值范围是：(0.0, 1.0) 开区间，不能等于 0 或 1，默认值为 0.7 模型考虑具有 top_p 概率质量tokens的结果例如：0.1 意味着模型解码器只考虑从前 10% 的概率的候选集中取 tokens 建议您根据应用场景调整 top_p 或 temperature 参数，但不要同时调整两个参数。 + en_US: Another method of temperature sampling is called kernel sampling. The value range is (0.0, 1.0) open interval, which cannot be equal to 0 or 1. The default value is 0.7. The model considers the results with top_p probability mass tokens. For example 0.1 means The model decoder only considers tokens from the candidate set with the top 10% probability. It is recommended that you adjust the top_p or temperature parameters according to the application scenario, but do not adjust both parameters at the same time. + - name: do_sample + label: + zh_Hans: 采样策略 + en_US: Sampling strategy + type: boolean + help: + zh_Hans: do_sample 为 true 时启用采样策略，do_sample 为 false 时采样策略 temperature、top_p 将不生效。默认值为 true。 + en_US: When `do_sample` is set to true, the sampling strategy is enabled. When `do_sample` is set to false, the sampling strategies such as `temperature` and `top_p` will not take effect. The default value is true. + default: true + - name: max_tokens + use_template: max_tokens + default: 1024 + min: 1 + max: 4095 + - name: web_search + type: boolean + label: + zh_Hans: 联网搜索 + en_US: Web Search + default: false + help: + zh_Hans: 模型内置了互联网搜索服务，该参数控制模型在生成文本时是否参考使用互联网搜索结果。启用互联网搜索，模型会将搜索结果作为文本生成过程中的参考信息，但模型会基于其内部逻辑“自行判断”是否使用互联网搜索结果。 + en_US: The model has a built-in Internet search service. This parameter controls whether the model refers to Internet search results when generating text. When Internet search is enabled, the model will use the search results as reference information in the text generation process, but the model will "judge" whether to use Internet search results based on its internal logic. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '0.0005' + output: '0.0005' + unit: '0.001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/zhipuai/llm/glm-4-flashx.yaml b/ai-provider/model-runtime/model-providers/zhipuai/llm/glm-4-flashx.yaml new file mode 100644 index 00000000..abd95a3a --- /dev/null +++ b/ai-provider/model-runtime/model-providers/zhipuai/llm/glm-4-flashx.yaml @@ -0,0 +1,66 @@ +model: glm-4-flashx +label: + en_US: glm-4-flashx +model_type: llm +features: + - multi-tool-call + - agent-thought + - stream-tool-call +model_properties: + mode: chat + context_size: 131072 +parameter_rules: + - name: temperature + use_template: temperature + default: 0.95 + min: 0.0 + max: 1.0 + help: + zh_Hans: 采样温度，控制输出的随机性，必须为正数取值范围是：(0.0,1.0]，不能等于 0,默认值为 0.95 值越大，会使输出更随机，更具创造性；值越小，输出会更加稳定或确定建议您根据应用场景调整 top_p 或 temperature 参数，但不要同时调整两个参数。 + en_US: Sampling temperature, controls the randomness of the output, must be a positive number. The value range is (0.0,1.0], which cannot be equal to 0. The default value is 0.95. The larger the value, the more random and creative the output will be; the smaller the value, The output will be more stable or certain. It is recommended that you adjust the top_p or temperature parameters according to the application scenario, but do not adjust both parameters at the same time. + - name: top_p + use_template: top_p + default: 0.7 + help: + zh_Hans: 用温度取样的另一种方法，称为核取样取值范围是：(0.0, 1.0) 开区间，不能等于 0 或 1，默认值为 0.7 模型考虑具有 top_p 概率质量tokens的结果例如：0.1 意味着模型解码器只考虑从前 10% 的概率的候选集中取 tokens 建议您根据应用场景调整 top_p 或 temperature 参数，但不要同时调整两个参数。 + en_US: Another method of temperature sampling is called kernel sampling. The value range is (0.0, 1.0) open interval, which cannot be equal to 0 or 1. The default value is 0.7. The model considers the results with top_p probability mass tokens. For example 0.1 means The model decoder only considers tokens from the candidate set with the top 10% probability. It is recommended that you adjust the top_p or temperature parameters according to the application scenario, but do not adjust both parameters at the same time. + - name: do_sample + label: + zh_Hans: 采样策略 + en_US: Sampling strategy + type: boolean + help: + zh_Hans: do_sample 为 true 时启用采样策略，do_sample 为 false 时采样策略 temperature、top_p 将不生效。默认值为 true。 + en_US: When `do_sample` is set to true, the sampling strategy is enabled. When `do_sample` is set to false, the sampling strategies such as `temperature` and `top_p` will not take effect. The default value is true. + default: true + - name: max_tokens + use_template: max_tokens + default: 1024 + min: 1 + max: 4095 + - name: web_search + type: boolean + label: + zh_Hans: 联网搜索 + en_US: Web Search + default: false + help: + zh_Hans: 模型内置了互联网搜索服务，该参数控制模型在生成文本时是否参考使用互联网搜索结果。启用互联网搜索，模型会将搜索结果作为文本生成过程中的参考信息，但模型会基于其内部逻辑“自行判断”是否使用互联网搜索结果。 + en_US: The model has a built-in Internet search service. This parameter controls whether the model refers to Internet search results when generating text. When Internet search is enabled, the model will use the search results as reference information in the text generation process, but the model will "judge" whether to use Internet search results based on its internal logic. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '0' + output: '0' + unit: '0.001' + currency: RMB diff --git a/ai-provider/model-runtime/model-providers/zhipuai/llm/glm_4v_flash.yaml b/ai-provider/model-runtime/model-providers/zhipuai/llm/glm_4v_flash.yaml new file mode 100644 index 00000000..9e3a7f20 --- /dev/null +++ b/ai-provider/model-runtime/model-providers/zhipuai/llm/glm_4v_flash.yaml @@ -0,0 +1,64 @@ +model: glm-4v-flash +label: + en_US: glm-4v-flash +model_type: llm +model_properties: + mode: chat + context_size: 2048 +features: + - vision +parameter_rules: + - name: temperature + use_template: temperature + default: 0.95 + min: 0.0 + max: 1.0 + help: + zh_Hans: 采样温度，控制输出的随机性，必须为正数取值范围是：(0.0,1.0]，不能等于 0,默认值为 0.95 值越大，会使输出更随机，更具创造性；值越小，输出会更加稳定或确定建议您根据应用场景调整 top_p 或 temperature 参数，但不要同时调整两个参数。 + en_US: Sampling temperature, controls the randomness of the output, must be a positive number. The value range is (0.0,1.0], which cannot be equal to 0. The default value is 0.95. The larger the value, the more random and creative the output will be; the smaller the value, The output will be more stable or certain. It is recommended that you adjust the top_p or temperature parameters according to the application scenario, but do not adjust both parameters at the same time. + - name: top_p + use_template: top_p + default: 0.6 + help: + zh_Hans: 用温度取样的另一种方法，称为核取样取值范围是：(0.0, 1.0) 开区间，不能等于 0 或 1，默认值为 0.7 模型考虑具有 top_p 概率质量tokens的结果例如：0.1 意味着模型解码器只考虑从前 10% 的概率的候选集中取 tokens 建议您根据应用场景调整 top_p 或 temperature 参数，但不要同时调整两个参数。 + en_US: Another method of temperature sampling is called kernel sampling. The value range is (0.0, 1.0) open interval, which cannot be equal to 0 or 1. The default value is 0.7. The model considers the results with top_p probability mass tokens. For example 0.1 means The model decoder only considers tokens from the candidate set with the top 10% probability. It is recommended that you adjust the top_p or temperature parameters according to the application scenario, but do not adjust both parameters at the same time. + - name: do_sample + label: + zh_Hans: 采样策略 + en_US: Sampling strategy + type: boolean + help: + zh_Hans: do_sample 为 true 时启用采样策略，do_sample 为 false 时采样策略 temperature、top_p 将不生效。默认值为 true。 + en_US: When `do_sample` is set to true, the sampling strategy is enabled. When `do_sample` is set to false, the sampling strategies such as `temperature` and `top_p` will not take effect. The default value is true. + default: true + - name: max_tokens + use_template: max_tokens + default: 1024 + min: 1 + max: 1024 + - name: web_search + type: boolean + label: + zh_Hans: 联网搜索 + en_US: Web Search + default: false + help: + zh_Hans: 模型内置了互联网搜索服务，该参数控制模型在生成文本时是否参考使用互联网搜索结果。启用互联网搜索，模型会将搜索结果作为文本生成过程中的参考信息，但模型会基于其内部逻辑“自行判断”是否使用互联网搜索结果。 + en_US: The model has a built-in Internet search service. This parameter controls whether the model refers to Internet search results when generating text. When Internet search is enabled, the model will use the search results as reference information in the text generation process, but the model will "judge" whether to use Internet search results based on its internal logic. + - name: response_format + label: + zh_Hans: 回复格式 + en_US: Response Format + type: string + help: + zh_Hans: 指定模型必须输出的格式 + en_US: specifying the format that the model must output + required: false + options: + - text + - json_object +pricing: + input: '0.00' + output: '0.00' + unit: '0.000001' + currency: RMB diff --git a/controller/ai-local/iml.go b/controller/ai-local/iml.go index 11ec8cf0..4ba66ca7 100644 --- a/controller/ai-local/iml.go +++ b/controller/ai-local/iml.go @@ -37,10 +37,6 @@ import ( "github.com/google/uuid" ) -var ( - ollamaConfig = "{\n \"mirostat\": 0,\n \"mirostat_eta\": 0.1,\n \"mirostat_tau\": 5.0,\n \"num_ctx\": 4096,\n \"repeat_last_n\":64,\n \"repeat_penalty\": 1.1,\n \"temperature\": 0.7,\n \"seed\": 42,\n \"num_predict\": 42,\n \"top_k\": 40,\n \"top_p\": 0.9,\n \"min_p\": 0.5\n}\n" -) - var _ ILocalModelController = &imlLocalModelController{} type imlLocalModelController struct { @@ -231,7 +227,7 @@ func (i *imlLocalModelController) initAILocalService(ctx context.Context, model } aiModel := &ai_api_dto.AiModel{ Id: model, - Config: ollamaConfig, + Config: ai_provider_local.OllamaConfig, Provider: providerId, Type: "local", } diff --git a/controller/service/iml.go b/controller/service/iml.go index 2f2f54fa..c21ecb88 100644 --- a/controller/service/iml.go +++ b/controller/service/iml.go @@ -8,6 +8,8 @@ import ( "strings" "time" + ai_provider_local "github.com/APIParkLab/APIPark/ai-provider/local" + subscribe_dto "github.com/APIParkLab/APIPark/module/subscribe/dto" "github.com/APIParkLab/APIPark/module/subscribe" @@ -55,9 +57,9 @@ import ( "github.com/google/uuid" ) -var ( - ollamaConfig = "{\n \"mirostat\": 0,\n \"mirostat_eta\": 0.1,\n \"mirostat_tau\": 5.0,\n \"num_ctx\": 4096,\n \"repeat_last_n\":64,\n \"repeat_penalty\": 1.1,\n \"temperature\": 0.7,\n \"seed\": 42,\n \"num_predict\": 42,\n \"top_k\": 40,\n \"top_p\": 0.9,\n \"min_p\": 0.5\n}\n" -) +//var ( +// ollamaConfig = "{\n \"mirostat\": 0,\n \"mirostat_eta\": 0.1,\n \"mirostat_tau\": 5.0,\n \"num_ctx\": 4096,\n \"repeat_last_n\":64,\n \"repeat_penalty\": 1.1,\n \"temperature\": 0.7,\n \"seed\": 42,\n \"num_predict\": 42,\n \"top_k\": 40,\n \"top_p\": 0.9,\n \"min_p\": 0.5\n}\n" +//) var ( _ IServiceController = (*imlServiceController)(nil) @@ -343,7 +345,7 @@ func (i *imlServiceController) createAIService(ctx *gin.Context, teamID string, return nil, fmt.Errorf("no local model") } modelId = list[0].Id - modelCfg = ollamaConfig + modelCfg = ai_provider_local.OllamaConfig } else { pv, err := i.providerModule.Provider(ctx, *input.Provider) if err != nil { diff --git a/module/ai-balance/iml.go b/module/ai-balance/iml.go index 6f37d240..4b93c857 100644 --- a/module/ai-balance/iml.go +++ b/module/ai-balance/iml.go @@ -6,6 +6,8 @@ import ( "fmt" "sort" + ai_provider_local "github.com/APIParkLab/APIPark/ai-provider/local" + model_runtime "github.com/APIParkLab/APIPark/ai-provider/model-runtime" "gorm.io/gorm" @@ -93,18 +95,18 @@ func (i *imlBalanceModule) Create(ctx context.Context, input *ai_balance_dto.Cre } -var ( - ollamaConfig = "{\n \"mirostat\": 0,\n \"mirostat_eta\": 0.1,\n \"mirostat_tau\": 5.0,\n \"num_ctx\": 4096,\n \"repeat_last_n\":64,\n \"repeat_penalty\": 1.1,\n \"temperature\": 0.7,\n \"seed\": 42,\n \"num_predict\": 42,\n \"top_k\": 40,\n \"top_p\": 0.9,\n \"min_p\": 0.5\n}\n" - ollamaBase = "http://apipark-ollama:11434" -) +//var ( +// ollamaConfig = "{\n \"mirostat\": 0,\n \"mirostat_eta\": 0.1,\n \"mirostat_tau\": 5.0,\n \"num_ctx\": 4096,\n \"repeat_last_n\":64,\n \"repeat_penalty\": 1.1,\n \"temperature\": 0.7,\n \"seed\": 42,\n \"num_predict\": 42,\n \"top_k\": 40,\n \"top_p\": 0.9,\n \"min_p\": 0.5\n}\n" +// ollamaBase = "http://apipark-ollama:11434" +//) func newRelease(item *ai_balance.Balance) *gateway.DynamicRelease { cfg := make(map[string]interface{}) cfg["provider"] = item.Id cfg["model"] = item.Model - cfg["model_config"] = ollamaConfig - cfg["base"] = ollamaBase + cfg["model_config"] = ai_provider_local.OllamaConfig + cfg["base"] = ai_provider_local.OllamaBase return &gateway.DynamicRelease{ BasicItem: &gateway.BasicItem{ ID: item.Id, diff --git a/module/ai-local/dto/output.go b/module/ai-local/dto/output.go index 74cc946a..e0397e3e 100644 --- a/module/ai-local/dto/output.go +++ b/module/ai-local/dto/output.go @@ -62,6 +62,7 @@ type SimpleItem struct { Id string `json:"id"` Name string `json:"name"` DefaultConfig string `json:"default_config"` + Logo string `json:"logo"` } type LocalModelItem struct { diff --git a/module/ai-local/iml.go b/module/ai-local/iml.go index 55ddb380..ebc08b5d 100644 --- a/module/ai-local/iml.go +++ b/module/ai-local/iml.go @@ -54,8 +54,8 @@ type imlLocalModel struct { } var ( - ollamaConfig = "{\n \"mirostat\": 0,\n \"mirostat_eta\": 0.1,\n \"mirostat_tau\": 5.0,\n \"num_ctx\": 4096,\n \"repeat_last_n\":64,\n \"repeat_penalty\": 1.1,\n \"temperature\": 0.7,\n \"seed\": 42,\n \"num_predict\": 42,\n \"top_k\": 40,\n \"top_p\": 0.9,\n \"min_p\": 0.5\n}\n" - ollamaBase = "http://apipark-ollama:11434" + // ollamaConfig = "{\n \"mirostat\": 0,\n \"mirostat_eta\": 0.1,\n \"mirostat_tau\": 5.0,\n \"num_ctx\": 4096,\n \"repeat_last_n\":64,\n \"repeat_penalty\": 1.1,\n \"temperature\": 0.7,\n \"seed\": 42,\n \"num_predict\": 42,\n \"top_k\": 40,\n \"top_p\": 0.9,\n \"min_p\": 0.5\n}\n" + ollamaBase = "http://apipark-ollama:11434" ) func init() { @@ -79,7 +79,8 @@ func (i *imlLocalModel) SimpleList(ctx context.Context) ([]*ai_local_dto.SimpleI return &ai_local_dto.SimpleItem{ Id: s.Id, Name: s.Name, - DefaultConfig: ollamaConfig, + DefaultConfig: ai_provider_local.OllamaConfig, + Logo: ai_provider_local.OllamaSvg, } }, func(l *ai_local.LocalModel) bool { if l.State != ai_local_dto.LocalModelStateNormal.Int() && l.State != ai_local_dto.LocalModelStateDisable.Int() { @@ -234,7 +235,7 @@ func (i *imlLocalModel) pullHook() func(msg ai_provider_local.PullMessage) error cfg := make(map[string]interface{}) cfg["provider"] = "ollama" cfg["model"] = msg.Model - cfg["model_config"] = ollamaConfig + cfg["model_config"] = ai_provider_local.OllamaConfig cfg["priority"] = 0 cfg["base"] = ollamaBase @@ -488,7 +489,7 @@ func (i *imlLocalModel) getLocalModels(ctx context.Context) ([]*gateway.DynamicR cfg := make(map[string]interface{}) cfg["provider"] = "ollama" cfg["model"] = l.Id - cfg["model_config"] = ollamaConfig + cfg["model_config"] = ai_provider_local.OllamaSvg cfg["base"] = ollamaBase releases = append(releases, &gateway.DynamicRelease{ BasicItem: &gateway.BasicItem{ diff --git a/module/ai/iml.go b/module/ai/iml.go index ed31d2ee..54145790 100644 --- a/module/ai/iml.go +++ b/module/ai/iml.go @@ -8,6 +8,8 @@ import ( "sort" "time" + ai_provider_local "github.com/APIParkLab/APIPark/ai-provider/local" + "github.com/eolinker/go-common/register" "github.com/eolinker/go-common/server" @@ -243,7 +245,7 @@ func (i *imlProviderModule) SimpleConfiguredProviders(ctx context.Context, all b items = append(items, &ai_dto.SimpleProviderItem{ Id: "ollama", Name: "Ollama", - Logo: ollamaSvg, + Logo: ai_provider_local.OllamaSvg, Configured: true, DefaultConfig: "", Status: ai_dto.ProviderEnabled,