The Cohere component is an AI component that allows users to connect the AI models served on the Cohere Platform.
It can carry out the following tasks:
The component definition and tasks are defined in the definition.yaml and tasks.yaml files respectively.
Setup
In order to communicate with Cohere, the following connection details need to be
provided. You may specify them directly in a pipeline recipe as key-value pairs
within the component's setup block, or you can create a Connection from
the Integration Settings
page and reference the whole setup as setup: ${connection.<my-connection-id>}.
Field
Field ID
Type
Note
API Key
api-key
string
Fill in your Cohere API key. To find your keys, visit the Cohere dashboard page.
Supported Tasks
Text Generation Chat
Cohere's text generation models (often called generative pre-trained transformers or large language models) have been trained to understand natural language, code, and images. The models provide text outputs in response to their inputs. The inputs to these models are also referred to as "prompts". Designing a prompt is essentially how you “program” a large language model model, usually by providing instructions or some examples of how to successfully complete a task.
Input
Field ID
Type
Description
Task ID (required)
task
string
TASK_TEXT_GENERATION_CHAT
Model Name (required)
model-name
string
The Cohere command model to be used. Enum values
command-r-plus
command-r
command
command-nightly
command-light
command-light-nightly
Prompt (required)
prompt
string
The prompt text.
System Message
system-message
string
The system message helps set the behavior of the assistant. For example, you can modify the personality of the assistant or provide specific instructions about how it should behave throughout the conversation. By default, the model’s behavior is using a generic message as "You are a helpful assistant.".
Documents
documents
array[string]
The documents to be used for the model, for optimal performance, the length of each document should be less than 300 words.
Prompt Images
prompt-images
array[string]
The prompt images (Note: As for 2024-06-24 Cohere models are not multimodal, so images will be ignored.).
Incorporate external chat history, specifically previous messages within the conversation. Each message should adhere to the format: : {"role": "The message role, i.e. 'USER' or 'CHATBOT'", "content": "message content"}.
Seed
seed
integer
The seed (default=42).
Temperature
temperature
number
The temperature for sampling (default=0.7).
Top K
top-k
integer
Top k for sampling (default=10).
Max New Tokens
max-new-tokens
integer
The maximum number of tokens for model to generate (default=50).
Input Objects in Text Generation Chat
Chat History
Incorporate external chat history, specifically previous messages within the conversation. Each message should adhere to the format: : {"role": "The message role, i.e. 'USER' or 'CHATBOT'", "content": "message content"}.
Token Usage on the Cohere Platform Command Models.
Output Objects in Text Generation Chat
Citations
Field
Field ID
Type
Note
End
end
integer
The end position of the citation.
Start
start
integer
The start position of the citation.
Text
text
string
The text body of the citation.
Usage
Field
Field ID
Type
Note
Input Tokens
input-tokens
number
The input tokens used by Cohere Models.
Output Tokens
output-tokens
number
The output tokens generated by Cohere Models.
Text Embeddings
An embedding is a list of floating point numbers that captures semantic information about the text that it represents.
Input
Field ID
Type
Description
Task ID (required)
task
string
TASK_TEXT_EMBEDDINGS
Embedding Type (required)
embedding-type
string
Specifies the return type of embedding, Note that 'binary'/'ubinary' options means the component will return packed unsigned binary embeddings. The length of each binary embedding is 1/8 the length of the float embeddings of the provided model. Enum values
float
int8
uint8
binary
ubinary
Input Type (required)
input-type
string
Specifies the type of input passed to the model. Enum values
Rerank models sort text inputs by semantic relevance to a specified query. They are often used to sort search results returned from an existing search solution.
Input
Field ID
Type
Description
Task ID (required)
task
string
TASK_TEXT_RERANKING
Model Name (required)
model-name
string
The Cohere rerank model to be used. Enum values
rerank-english-v3.0
rerank-multilingual-v3.0
Query (required)
query
string
The query.
Documents (required)
documents
array[string]
The documents to be used for reranking.
Top N
top-n
integer
The number of most relevant documents or indices to return. Defaults to the length of the documents (default=3).
Maximum Number of Chunks per Document
max-chunks-per-doc
integer
The maximum number of chunks to produce internally from a document (default=10).
