package instinct.llm

Get desktop application:
View/edit binary Protocol Buffers messages

Used in: AgentThought

optional Message tool_call_message = 1
should be a message container tool call requests
optional google.protobuf.Any custom = 100
reserved for user defined thought

Used in: AgentThought

string response = 1
bool is_cancelled = 2
bool is_expired = 3
bool is_failed = 4
optional google.protobuf.Any details = 6
this is for error details
optional PromptValue question = 7
original user input question
optional google.protobuf.Any custom = 100
this is for custom step data

Used in: AgentStep

repeated Message tool_messages = 1
completed tool calls
optional google.protobuf.Any custom = 100
reserved for user defined observation

Used in: AgentThought

repeated Message completed = 1
should be a group of completed tool message
optional Message tool_call_message = 2
original tool call requests
optional google.protobuf.Any custom = 100
reserved for user defined thought

repeated AgentStep previous_steps = 1
optional PromptValue input = 2
repeated FunctionTool function_tools = 3
code_interpreter and file_search should be translated to FunctionTool format

Used in: AgentState

oneof value
- AgentThought thought = 1
- AgentObservation observation = 2

Used in: AgentStep

oneof value
- AgentContinuation continuation = 1
- AgentPause pause = 2
- AgentFinish finish = 3

repeated LangaugeModelResult generations = 1
LangaugeModelOutput won't contain valid `raw_response` in BatchedLangaugeModelOutput, read content of `raw_response` below
oneof raw_response
- OllamaCompletionResponse ollama_completion = 2
- OllamaChatCompletionResponse ollama_chat = 3

string math_question = 1

double answer = 1

Used in: PromptValue

repeated Message messages = 1

Used in: assistant.v2.AssistantToolObject, AgentState, OpenAIChatCompletionRequest.ChatCompletionTool

string description = 1
* A description of what the function does, used by the model to choose when and how to call the function.
string name = 2
* The name of the function to be called. Must be a-z, A-Z, 0-9, or contain underscores and dashes, with a maximum length of 64.
optional FunctionTool.FunctionParametersSchema parameters = 3
* The parameters the functions accepts, described as a JSON Schema object. Omitting parameters defines a function with an empty parameter list.
repeated string required = 4

Used in: FunctionTool

string type = 1
map<string, FunctionParametersSchema> properties = 2
required for object type
optional FunctionParametersSchema items = 3
required for array type
string description = 4
repeated string required = 5
optional for object type
repeated string enum = 6
optional for string enum type

this is useful for toolkit related classes, tool message cannot represent invoke result lossless.

string invocation_id = 1
string return_value = 2
bool has_error = 3
string exception = 4

bool passed = 2

Used in: LangaugeModelResult

string text = 1
optional Message message = 2
bool is_chunk = 3

used in llm.AgentFinish.details

Used in: LLMCompilerTaskGraph

bool is_replan = 1
string answer = 2
string thought = 3

used in llm.AgentContinuation.custom, llm.AgentPause.custom

repeated LLMCompilerTaskGraph.LLMCompilerTask tasks = 1
string question = 2
question field is only used in joiner
optional LLMCompilerJoinerResult joiner_result = 3
field for joiner result

Used in: LLMCompilerTaskGraph

int64 index = 1
optional ToolCallObject tool_call = 2
repeated int64 dependencies = 3
string thought = 4
optional Message result = 5
optional message to save result

custom data for llm.Message.custom, assistant.v2.RunStepDetails.ToolCallDetail.custom

int64 index = 1
index in function graph
string thought = 2
repeated int64 dependencies = 3

* for both LLM and ChatModel

Used in: BatchedLangaugeModelResult

repeated Generation generations = 1
oneof raw_response
- OllamaCompletionResponse ollama_completion = 2
- OllamaChatCompletionResponse ollama_chat = 3

message Message

chat_completion_api.proto:6

Used in: AgentContinuation, AgentObservation, AgentPause, ChatPromptValue, Generation, LLMCompilerTaskGraph.LLMCompilerTask, MessageList, OpenAIChatCompletionChunk.OpenAIChatCompletionChunkChoice, OpenAIChatCompletionRequest, OpenAIChatCompletionResponse.OpenAIChatCompletionChoice

string role = 1
string content = 2
string name = 3
* An name for the participant. Provides the model information to differentiate between participants of the same role.
repeated ToolCallObject tool_calls = 4
* The tool calls generated by the model, such as function calls.
string tool_call_id = 5
* Tool call that this message is responding to.
optional google.protobuf.Any custom = 100
* custom data for different function calling agent. e.g. LLM Compiler relies on extra numeric index to build function graph.

message MessageList

llm.proto:24

repeated Message messages = 1

repeated string lines = 1

string id = 1
repeated OpenAIChatCompletionChunk.OpenAIChatCompletionChunkChoice choices = 2
int64 created = 5
string model = 6
string system_fingerprint = 7
string object = 8

Used in: OpenAIChatCompletionChunk

optional Message delta = 1
optional OpenAIChatCompletionLogProbs logprobs = 2
string finish_reason = 3
optional int32 index = 4

Used in: OpenAIChatCompletionChunk.OpenAIChatCompletionChunkChoice, OpenAIChatCompletionResponse.OpenAIChatCompletionChoice

repeated OpenAIChatCompletionLogProbs.OpenAIChatCompletionLogProbItem content = 1
* A list of message content tokens with log probability information.

Used in: OpenAIChatCompletionLogProbs

string token = 1
* The token
float logprob = 2
* The log probability of this token, if it is within the top 20 most likely tokens. Otherwise, the value -9999.0 is used to signify that the token is very unlikely.
repeated int32 bytes = 3
* A list of integers representing the UTF-8 bytes representation of the token. Useful in instances where characters are represented by multiple tokens and their byte representations must be combined to generate the correct text representation. Can be null if there is no bytes representation for the token.
repeated OpenAIChatCompletionLogProbItem top_logprobs = 4
* List of the most likely tokens and their log probability, at this token position. In rare cases, there may be fewer than the number of requested top_logprobs returned.

repeated Message messages = 1
* A list of messages comprising the conversation so far.
string model = 2
* ID of the model to use. See the model endpoint compatibility table for details on which models work with the Chat API
float frequency_penalty = 3
* Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
bool logprobs = 4
* Whether to return log probabilities of the output tokens or not. If true, returns the log probabilities of each output token returned in the content of message. This option is currently not available on the gpt-4-vision-preview model.
bool top_logprobs = 5
* An integer between 0 and 20 specifying the number of most likely tokens to return at each token position, each with an associated log probability. logprobs must be set to true if this parameter is used.
int32 max_tokens = 6
* The maximum number of tokens that can be generated in the chat completion.
int32 n = 7
* How many chat completion choices to generate for each input message. Note that you will be charged based on the number of generated tokens across all of the choices. Keep n as 1 to minimize costs.
float presence_penalty = 8
* Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
optional OpenAIChatCompletionRequest.OpenAIResponseFormat response_format = 9
* An object specifying the format that the model must output. Compatible with GPT-4 Turbo and all GPT-3.5 Turbo models newer than gpt-3.5-turbo-1106. Setting to { "type": "json_object" } enables JSON mode, which guarantees the message the model generates is valid JSON.
int32 seed = 10
* This feature is in Beta. If specified, our system will make a best effort to sample deterministically, such that repeated requests with the same seed and parameters should return the same result. Determinism is not guaranteed, and you should refer to the system_fingerprint response parameter to monitor changes in the backend.
repeated string stop = 11
* Up to 4 sequences where the API will stop generating further tokens.
optional bool stream = 12
* If set, partial message deltas will be sent, like in ChatGPT. Tokens will be sent as data-only server-sent events as they become available, with the stream terminated by a data: [DONE] message.
optional float temperature = 13
* What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
float top_p = 14
* An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.
repeated OpenAIChatCompletionRequest.ChatCompletionTool tools = 15
* A list of tools the model may call. Currently, only functions are supported as a tool. Use this to provide a list of functions the model may generate JSON inputs for. A max of 128 functions are supported.
string tool_choice = 16
* Controls which (if any) function is called by the model. none means the model will not call a function and instead generates a message. auto means the model can pick between generating a message or calling a function. Specifying a particular function via {"type": "function", "function": {"name": "my_function"}} forces the model to call that function.
oneof tool_choice { /** none means the model will not call a function and instead generates a message. auto means the model can pick between generating a message or calling a function. */ string mode = 16; /** Specifies a tool the model should use. Use to force the model to call a specific function. */ OpenAITool forced_tool = 17; }
string user = 18
* A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse. Learn more.

Used in: OpenAIChatCompletionRequest

string type = 1
always function
optional FunctionTool function = 2

Used in: OpenAIChatCompletionRequest

string type = 1

string type = 1
* The type of the tool. Currently, only function is supported.
optional OpenAITool.FunctionToolObject object = 2

Used in: OpenAITool

string name = 1
* The name of the function to call.

* https://platform.openai.com/docs/api-reference/chat/object

string id = 1
* A unique identifier for the chat completion.
repeated OpenAIChatCompletionResponse.OpenAIChatCompletionChoice choices = 2
* A list of chat completion choices. Can be more than one if n is greater than 1.
int64 created = 5
* The Unix timestamp (in seconds) of when the chat completion was created.
string model = 6
* The model used for the chat completion.
string system_fingerprint = 7
* This fingerprint represents the backend configuration that the model runs with. Can be used in conjunction with the seed request parameter to understand when backend changes have been made that might impact determinism.
string object = 8
* The object type, which is always chat.completion.
optional OpenAIUsageInResponse usage = 9
* Usage statistics for the completion request.

Used in: OpenAIChatCompletionResponse

string finish_reason = 1
* The reason the model stopped generating tokens. This will be stop if the model hit a natural stop point or a provided stop sequence, length if the maximum number of tokens specified in the request was reached, content_filter if content was omitted due to a flag from our content filters, tool_calls if the model called a tool, or function_call (deprecated) if the model called a function.
int32 index = 2
* The index of the choice in the list of choices.
optional Message message = 3
* A chat completion message generated by the model.
optional OpenAIChatCompletionLogProbs logprobs = 4
* Log probability information for the choice.

repeated string input = 1
* Input text to embed
string model = 2
* ID of the model to use. You can use the List models API to see all of your available models, or see our Model overview for descriptions of them.
string encoding_format = 3
* The format to return the embeddings in. Can be either float or base64.
uint32 dimension = 4
* The number of dimensions the resulting output embeddings should have. Only supported in text-embedding-3 and later models.
string user = 5
* A unique identifier representing your end-user, which can help OpenAI to monitor and detect abuse.

string object = 1
repeated OpenAIEmbeddingResponse.OpenAIEmbedding data = 2
string model = 3
optional OpenAIUsageInResponse usage = 4

Used in: OpenAIEmbeddingResponse

int32 index = 1
repeated float embedding = 2
string object = 3

Used in: OpenAIChatCompletionResponse, OpenAIEmbeddingResponse

optional int32 completion_tokens = 1
* Number of tokens in the generated completion.
optional int32 prompt_tokens = 2
* Number of tokens in the prompt.
optional int32 total_tokens = 3
* Total number of tokens used in the request (prompt + completion).