Skip to content

Commit

Permalink
Merge pull request #228 from openai/erinkav/update
Browse files Browse the repository at this point in the history 
Add temperature, top_p and response format to Assistants API
  • Loading branch information
@erinkav-openai
erinkav-openai authored Apr 16, 2024
2 parents 3ebe2fc + a20659d commit af97d0a
Showing 1 changed file with 138 additions and 44 deletions.
182 changes: 138 additions & 44 deletions openapi.yaml
View file
Original file line number Diff line number Diff line change
Expand Up @@ -2888,7 +2888,10 @@ paths:
"instructions": "You are a helpful assistant designed to make me better at coding!",
"tools": [],
"file_ids": [],
"metadata": {}
"metadata": {},
"top_p": 1.0,
"temperature": 1.0,
"response_format": "auto"
},
{
"id": "asst_abc456",
Expand All @@ -2900,7 +2903,10 @@ paths:
"instructions": "You are a helpful assistant designed to make me better at coding!",
"tools": [],
"file_ids": [],
"metadata": {}
"metadata": {},
"top_p": 1.0,
"temperature": 1.0,
"response_format": "auto"
},
{
"id": "asst_abc789",
Expand All @@ -2912,7 +2918,10 @@ paths:
"instructions": null,
"tools": [],
"file_ids": [],
"metadata": {}
"metadata": {},
"top_p": 1.0,
"temperature": 1.0,
"response_format": "auto"
}
],
"first_id": "asst_abc123",
Expand Down Expand Up @@ -3001,7 +3010,10 @@ paths:
}
],
"file_ids": [],
"metadata": {}
"metadata": {},
"top_p": 1.0,
"temperature": 1.0,
"response_format": "auto"
}
- title: Files
request:
Expand Down Expand Up @@ -3064,7 +3076,10 @@ paths:
"file_ids": [
"file-abc123"
],
"metadata": {}
"metadata": {},
"top_p": 1.0,
"temperature": 1.0,
"response_format": "auto"
}
/assistants/{assistant_id}:
Expand Down Expand Up @@ -3238,7 +3253,10 @@ paths:
"file-abc123",
"file-abc456"
],
"metadata": {}
"metadata": {},
"top_p": 1.0,
"temperature": 1.0,
"response_format": "auto"
}
delete:
operationId: deleteAssistant
Expand Down Expand Up @@ -4501,7 +4519,8 @@ paths:
"completion_tokens": 456,
"total_tokens": 579
},
"temperature": 1,
"temperature": 1.0,
"top_p": 1.0,
"max_prompt_tokens": 1000,
"max_completion_tokens": 1000,
"truncation_strategy": {
Expand Down Expand Up @@ -4542,7 +4561,8 @@ paths:
"completion_tokens": 456,
"total_tokens": 579
},
"temperature": 1,
"temperature": 1.0,
"top_p": 1.0,
"max_prompt_tokens": 1000,
"max_completion_tokens": 1000,
"truncation_strategy": {
Expand Down Expand Up @@ -4651,7 +4671,8 @@ paths:
],
"metadata": {},
"usage": null,
"temperature": 1,
"temperature": 1.0,
"top_p": 1.0,
"max_prompt_tokens": 1000,
"max_completion_tokens": 1000,
"truncation_strategy": {
Expand Down Expand Up @@ -4996,7 +5017,8 @@ paths:
"completion_tokens": 456,
"total_tokens": 579
},
"temperature": 1,
"temperature": 1.0,
"top_p": 1.0,
"max_prompt_tokens": 1000,
"max_completion_tokens": 1000,
"truncation_strategy": {
Expand Down Expand Up @@ -5119,7 +5141,8 @@ paths:
"completion_tokens": 456,
"total_tokens": 579
},
"temperature": 1,
"temperature": 1.0,
"top_p": 1.0,
"max_prompt_tokens": 1000,
"max_completion_tokens": 1000,
"truncation_strategy": {
Expand Down Expand Up @@ -5266,7 +5289,8 @@ paths:
"file_ids": [],
"metadata": {},
"usage": null,
"temperature": 1,
"temperature": 1.0,
"top_p": 1.0,
"max_prompt_tokens": 1000,
"max_completion_tokens": 1000,
"truncation_strategy": {
Expand Down Expand Up @@ -5474,7 +5498,8 @@ paths:
"file_ids": [],
"metadata": {},
"usage": null,
"temperature": 1
"temperature": 1.0,
"top_p": 1.0,
}
/threads/{thread_id}/runs/{run_id}/steps:
Expand Down Expand Up @@ -7775,7 +7800,7 @@ components:
description: |
The ID of an uploaded file that contains training data.
See [upload file](/docs/api-reference/files/upload) for how to upload a file.
See [upload file](/docs/api-reference/files/create) for how to upload a file.
Your dataset must be formatted as a JSONL file. Additionally, you must upload your file with the purpose `fine-tune`.
Expand Down Expand Up @@ -8770,6 +8795,33 @@ components:
- total_tokens
nullable: true

AssistantsApiResponseFormatOption:
description: |
Specifies the format that the model must output. Compatible with [GPT-4 Turbo](/docs/models/gpt-4-and-gpt-4-turbo) and all GPT-3.5 Turbo models since `gpt-3.5-turbo-1106`.
Setting to `{ "type": "json_object" }` enables JSON mode, which guarantees the message the model generates is valid JSON.
**Important:** when using JSON mode, you **must** also instruct the model to produce JSON yourself via a system or user message. Without this, the model may generate an unending stream of whitespace until the generation reaches the token limit, resulting in a long-running and seemingly "stuck" request. Also note that the message content may be partially cut off if `finish_reason="length"`, which indicates the generation exceeded `max_tokens` or the conversation exceeded the max context length.
oneOf:
- type: string
description: >
`auto` is the default value
enum: [none, auto]
- $ref: "#/components/schemas/AssistantsApiResponseFormat"
x-oaiExpandable: true

AssistantsApiResponseFormat:
type: object
description: |
An object describing the expected output of the model. If `json_object` only `function` type `tools` are allowed to be passed to the Run. If `text` the model can return text or any value needed.
properties:
type:
type: string
enum: ["text", "json_object"]
example: "json_object"
default: "text"
description: Must be one of `text` or `json_object`.

AssistantObject:
type: object
title: Assistant
Expand Down Expand Up @@ -8918,6 +8970,29 @@ components:
type: object
x-oaiTypeLabel: map
nullable: true
temperature:
description: &run_temperature_description |
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.
type: number
minimum: 0
maximum: 2
default: 1
example: 1
nullable: true
top_p:
type: number
minimum: 0
maximum: 1
default: 1
example: 1
nullable: true
description: &run_top_p_description |
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.

We generally recommend altering this or temperature but not both.
response_format:
$ref: "#/components/schemas/AssistantsApiResponseFormatOption"
nullable: true
required:
- model

Expand Down Expand Up @@ -8968,6 +9043,29 @@ components:
type: object
x-oaiTypeLabel: map
nullable: true
temperature:
description: &run_temperature_description |
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.
type: number
minimum: 0
maximum: 2
default: 1
example: 1
nullable: true
top_p:
type: number
minimum: 0
maximum: 1
default: 1
example: 1
nullable: true
description: &run_top_p_description |
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.

We generally recommend altering this or temperature but not both.
response_format:
$ref: "#/components/schemas/AssistantsApiResponseFormatOption"
nullable: true

DeleteAssistantResponse:
type: object
Expand Down Expand Up @@ -9101,33 +9199,6 @@ components:
required:
- type

AssistantsApiResponseFormatOption:
description: |
Specifies the format that the model must output. Compatible with [GPT-4 Turbo](/docs/models/gpt-4-and-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.
**Important:** when using JSON mode, you **must** also instruct the model to produce JSON yourself via a system or user message. Without this, the model may generate an unending stream of whitespace until the generation reaches the token limit, resulting in a long-running and seemingly "stuck" request. Also note that the message content may be partially cut off if `finish_reason="length"`, which indicates the generation exceeded `max_tokens` or the conversation exceeded the max context length.
oneOf:
- type: string
description: >
`auto` is the default value
enum: [none, auto]
- $ref: "#/components/schemas/AssistantsApiResponseFormat"
x-oaiExpandable: true

AssistantsApiResponseFormat:
type: object
description: |
An object describing the expected output of the model. If `json_object` only `function` type `tools` are allowed to be passed to the Run. If `text` the model can return text or any value needed.
properties:
type:
type: string
enum: ["text", "json_object"]
example: "json_object"
default: "text"
description: Must be one of `text` or `json_object`.

RunObject:
type: object
title: A run on a thread
Expand Down Expand Up @@ -9264,6 +9335,10 @@ components:
description: The sampling temperature used for this run. If not set, defaults to 1.
type: number
nullable: true
top_p:
description: The nucleus sampling value used for this run. If not set, defaults to 1.
type: number
nullable: true
max_prompt_tokens:
type: integer
nullable: true
Expand Down Expand Up @@ -9339,7 +9414,8 @@ components:
"completion_tokens": 456,
"total_tokens": 579
},
"temperature": 1,
"temperature": 1.0,
"top_p": 1.0,
"max_prompt_tokens": 1000,
"max_completion_tokens": 1000,
"truncation_strategy": {
Expand Down Expand Up @@ -9422,8 +9498,18 @@ components:
default: 1
example: 1
nullable: true
description: &run_temperature_description |
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.
description: *run_temperature_description
top_p:
type: number
minimum: 0
maximum: 1
default: 1
example: 1
nullable: true
description: &run_top_p_description |
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.

We generally recommend altering this or temperature but not both.
stream:
type: boolean
nullable: true
Expand Down Expand Up @@ -9606,6 +9692,14 @@ components:
example: 1
nullable: true
description: *run_temperature_description
top_p:
type: number
minimum: 0
maximum: 1
default: 1
example: 1
nullable: true
description: *run_top_p_description
stream:
type: boolean
nullable: true
Expand Down

0 comments on commit af97d0a

Please sign in to comment.