Sunflower Inference

curl --request POST \
  --url https://api.sunbird.ai/tasks/sunflower_inference \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "messages": [
    {
      "role": "<string>",
      "content": "<string>"
    }
  ],
  "model_type": "qwen",
  "temperature": 0.3,
  "stream": false,
  "system_message": "<string>"
}
'

{
  "content": "<string>",
  "model_type": "<string>",
  "usage": {
    "completion_tokens": 123,
    "prompt_tokens": 123,
    "total_tokens": 123
  },
  "processing_time": 123,
  "inference_time": 0,
  "message_count": 0
}

Sunflower Chat

Sunflower Chat

Professional Sunflower inference endpoint for multilingual chat completions. This endpoint provides access to Sunbird AI's Sunflower model, specialized in: - Multilingual conversations in Ugandan languages (Luganda, Acholi, Ateso, etc.) - Cross-lingual translations and explanations - Cultural context understanding - Educational content in local languages Features: - Automatic retry with exponential backoff - Context-aware responses - Usage tracking and monitoring - Support for custom system messages - Message history management Args: request: The FastAPI request object (required for rate limiting). chat_request: The chat request containing messages and parameters. background_tasks: FastAPI background tasks for async operations. current_user: The authenticated user (enforces authentication). service: The inference service instance. Returns: SunflowerChatResponse containing the AI's response and usage stats. Raises: BadRequestError: For validation errors. ValidationError: For invalid message format. ServiceUnavailableError: If the model is loading or request times out. ExternalServiceError: For empty model response or unexpected errors. Example: Request body with message history: { "messages": [ {"role": "system", "content": "You are Sunflower..."}, {"role": "user", "content": "Translate 'How are you?' to Luganda."}, {"role": "assistant", "content": "In Luganda, 'How are you?' is 'Oli otya?'."}, {"role": "user", "content": "How do I say 'Good morning' in Acholi?"} ], "model_type": "qwen", "temperature": 0.3 } Response: { "content": "In Acholi, 'Good morning' is 'Icwiny atir'.", "model_type": "qwen", "usage": {"completion_tokens": 15, "prompt_tokens": 50, "total_tokens": 65}, "processing_time": 2.5, "inference_time": 2.0, "message_count": 5 }

POST

/

tasks

/

sunflower_inference

Sunflower Inference

curl --request POST \
  --url https://api.sunbird.ai/tasks/sunflower_inference \
  --header 'Authorization: Bearer <token>' \
  --header 'Content-Type: application/json' \
  --data '
{
  "messages": [
    {
      "role": "<string>",
      "content": "<string>"
    }
  ],
  "model_type": "qwen",
  "temperature": 0.3,
  "stream": false,
  "system_message": "<string>"
}
'

{
  "content": "<string>",
  "model_type": "<string>",
  "usage": {
    "completion_tokens": 123,
    "prompt_tokens": 123,
    "total_tokens": 123
  },
  "processing_time": 123,
  "inference_time": 0,
  "message_count": 0
}

Authorizations

Authorization

string

header

required

The access token received from the authorization server in the OAuth 2.0 flow.

Body

application/json

Request model for Sunflower chat inference.

Attributes: messages: List of conversation messages. model_type: The model to use ('qwen'). temperature: Sampling temperature (0.0 to 2.0). stream: Whether to stream the response. system_message: Optional custom system message.

messages

SunflowerChatMessage · object[]

required

List of conversation messages

Show child attributes

model_type

string

default:qwen

Model type: 'qwen'

temperature

number

default:0.3

Sampling temperature

Required range: 0 <= x <= 2

stream

boolean

default:false

Whether to stream the response

system_message

string | null

Custom system message

Response

Successful Response

Response model from Sunflower chat inference.

Attributes: content: The AI's response text. model_type: The model used for inference. usage: Token usage statistics. processing_time: Total processing time in seconds. inference_time: Model inference time in seconds. message_count: Number of messages processed.

content

string

required

The AI's response

model_type

string

required

Model used for inference

usage

SunflowerUsageStats · object

required

Token usage statistics

Show child attributes

processing_time

number

required

Total processing time in seconds

inference_time

number

default:0

Model inference time in seconds

message_count

integer

default:0

Number of messages processed

Language ID Sunflower Simple