cohere.calls¶
A module for calling Cohere's chat models.
BaseCall
¶
Bases: BasePrompt
, Generic[BaseCallResponseT, BaseCallResponseChunkT, BaseToolT, MessageParamT]
, ABC
The base class abstract interface for calling LLMs.
Source code in mirascope/base/calls.py
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 |
|
call(retries=0, **kwargs)
abstractmethod
¶
A call to an LLM.
An implementation of this function must return a response that extends
BaseCallResponse
. This ensures a consistent API and convenience across e.g.
different model providers.
Source code in mirascope/base/calls.py
call_async(retries=0, **kwargs)
abstractmethod
async
¶
An asynchronous call to an LLM.
An implementation of this function must return a response that extends
BaseCallResponse
. This ensures a consistent API and convenience across e.g.
different model providers.
Source code in mirascope/base/calls.py
from_prompt(prompt_type, call_params)
classmethod
¶
Returns a call_type generated dynamically from this base call.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
prompt_type |
type[BasePromptT]
|
The prompt class to use for the call. Properties and class variables of this class will be used to create the new call class. Must be a class that can be instantiated. |
required |
call_params |
BaseCallParams
|
The call params to use for the call. |
required |
Returns:
Type | Description |
---|---|
type[BasePromptT]
|
A new call class with new call_type. |
Source code in mirascope/base/calls.py
stream(retries=0, **kwargs)
abstractmethod
¶
A call to an LLM that streams the response in chunks.
An implementation of this function must yield response chunks that extend
BaseCallResponseChunk
. This ensures a consistent API and convenience across
e.g. different model providers.
Source code in mirascope/base/calls.py
stream_async(retries=0, **kwargs)
abstractmethod
async
¶
A asynchronous call to an LLM that streams the response in chunks.
An implementation of this function must yield response chunks that extend
BaseCallResponseChunk
. This ensures a consistent API and convenience across
e.g. different model providers.
Source code in mirascope/base/calls.py
CohereCall
¶
Bases: BaseCall[CohereCallResponse, CohereCallResponseChunk, CohereTool, ChatMessage]
A base class for calling Cohere's chat models.
Example:
from mirascope.cohere import CohereCall
class BookRecommender(CohereCall):
prompt_template = "Please recommend a {genre} book"
genre: str
response = BookRecommender(genre="fantasy").call()
print(response.content)
#> There are many great books to read, it ultimately depends...
Source code in mirascope/cohere/calls.py
22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 |
|
call(retries=0, **kwargs)
¶
Makes a call to the model using this CohereCall
instance.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
**kwargs |
Any
|
Additional keyword arguments parameters to pass to the call. These
will override any existing arguments in |
{}
|
Returns:
Type | Description |
---|---|
CohereCallResponse
|
A |
Source code in mirascope/cohere/calls.py
call_async(retries=0, **kwargs)
async
¶
Makes an asynchronous call to the model using this CohereCall
.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
**kwargs |
Any
|
Additional keyword arguments parameters to pass to the call. These
will override any existing arguments in |
{}
|
Returns:
Type | Description |
---|---|
CohereCallResponse
|
An |
Source code in mirascope/cohere/calls.py
messages()
¶
Returns the template as a formatted list of messages.
Source code in mirascope/cohere/calls.py
stream(retries=0, **kwargs)
¶
Streams the response for a call using this CohereCall
.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
**kwargs |
Any
|
Additional keyword arguments parameters to pass to the call. These
will override any existing arguments in |
{}
|
Yields:
Type | Description |
---|---|
CohereCallResponseChunk
|
A |
Source code in mirascope/cohere/calls.py
stream_async(retries=0, **kwargs)
async
¶
Streams the response for an asynchronous call using this CohereCall
.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
**kwargs |
Any
|
Additional keyword arguments parameters to pass to the call. These
will override any existing arguments in |
{}
|
Yields:
Type | Description |
---|---|
AsyncGenerator[CohereCallResponseChunk, None]
|
A |
Source code in mirascope/cohere/calls.py
CohereCallParams
¶
Bases: BaseCallParams[CohereTool]
The parameters to use when calling the Cohere chat API.
Source code in mirascope/cohere/types.py
kwargs(tool_type=CohereTool, exclude=None)
¶
Returns the keyword argument call parameters.
Source code in mirascope/cohere/types.py
CohereCallResponse
¶
Bases: BaseCallResponse[NonStreamedChatResponse, CohereTool]
A convenience wrapper around the Cohere NonStreamedChatResponse
response.
When using Mirascope's convenience wrappers to interact with Cohere chat models via
CohereCall
, responses using CohereCall.call()
will return a CohereCallResponse
whereby the implemented properties allow for simpler syntax and a convenient
developer experience.
Example:
from mirascope.cohere import CohereCall
class BookRecommender(CohereCall):
prompt_template = "Please recommend a {genre} book"
genre: str
response = Bookrecommender(genre="fantasy").call()
print(response.content)
#> The Name of the Wind
print(response.message)
#> ...
print(response.choices)
#> ...
Source code in mirascope/cohere/types.py
92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 |
|
citations: Optional[list[ChatCitation]]
property
¶
Returns the citations for the 0th choice message.
content: str
property
¶
Returns the content of the chat completion for the 0th choice.
documents: Optional[list[ChatDocument]]
property
¶
Returns the documents for the 0th choice message.
finish_reasons: Optional[list[str]]
property
¶
Returns the finish reasons of the response.
id: Optional[str]
property
¶
Returns the id of the response.
input_tokens: Optional[float]
property
¶
Returns the number of input tokens.
message_param: ChatMessage
property
¶
Returns the assistant's response as a message parameter.
model: Optional[str]
property
¶
Returns the name of the response model.
Cohere does not return model, so we return None
output_tokens: Optional[float]
property
¶
Returns the number of output tokens.
search_queries: Optional[list[ChatSearchQuery]]
property
¶
Returns the search queries for the 0th choice message.
search_results: Optional[list[ChatSearchResult]]
property
¶
Returns the search results for the 0th choice message.
tool: Optional[CohereTool]
property
¶
Returns the 0th tool for the 0th choice message.
Raises:
Type | Description |
---|---|
ValidationError
|
if the tool call doesn't match the tool's schema. |
tool_calls: Optional[list[ToolCall]]
property
¶
Returns the tool calls for the 0th choice message.
tools: Optional[list[CohereTool]]
property
¶
Returns the tools for the 0th choice message.
Raises:
Type | Description |
---|---|
ValidationError
|
if a tool call doesn't match the tool's schema. |
usage: Optional[ApiMetaBilledUnits]
property
¶
Returns the usage of the response.
dump()
¶
Dumps the response to a dictionary.
tool_message_params(tools_and_outputs)
classmethod
¶
Returns the tool message parameters for tool call results.
Source code in mirascope/cohere/types.py
CohereCallResponseChunk
¶
Bases: BaseCallResponseChunk[StreamedChatResponse, CohereTool]
Convenience wrapper around chat completion streaming chunks.
When using Mirascope's convenience wrappers to interact with Cohere models via
CohereCall.stream
, responses will return an CohereCallResponseChunk
, whereby
the implemented properties allow for simpler syntax and a convenient developer
experience.
Example:
from mirascope.cohere import CohereCall
class Math(CohereCall):
prompt_template = "What is 1 + 2?"
content = ""
for chunk in Math().stream():
content += chunk.content
print(content)
#> 1
# 1 +
# 1 + 2
# 1 + 2 equals
# 1 + 2 equals
# 1 + 2 equals 3
# 1 + 2 equals 3.
Source code in mirascope/cohere/types.py
263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 |
|
citations: Optional[list[ChatCitation]]
property
¶
Returns the citations for citation-generation event type else None.
content: str
property
¶
Returns the content for the 0th choice delta.
documents: Optional[list[ChatDocument]]
property
¶
Returns the documents for search-results event type else None.
event_type: Literal['stream-start', 'search-queries-generation', 'search-results', 'text-generation', 'citation-generation', 'tool-calls-generation', 'stream-end', 'tool-calls-chunk']
property
¶
Returns the type of the chunk.
finish_reasons: Optional[list[str]]
property
¶
Returns the finish reasons of the response.
id: Optional[str]
property
¶
Returns the id of the response.
input_tokens: Optional[float]
property
¶
Returns the number of input tokens.
model: Optional[str]
property
¶
Returns the name of the response model.
Cohere does not return model, so we return None
output_tokens: Optional[float]
property
¶
Returns the number of output tokens.
response: Optional[NonStreamedChatResponse]
property
¶
Returns the full response for the stream-end event type else None.
search_queries: Optional[list[ChatSearchQuery]]
property
¶
Returns the search queries for search-query event type else None.
search_results: Optional[list[ChatSearchResult]]
property
¶
Returns the search results for search-results event type else None.
tool_calls: Optional[list[ToolCall]]
property
¶
Returns the partial tool calls for the 0th choice message.
usage: Optional[ApiMetaBilledUnits]
property
¶
Returns the usage of the response.
CohereTool
¶
Bases: BaseTool[ToolCall]
A base class for easy use of tools with the Cohere chat client.
CohereTool
internally handles the logic that allows you to use tools with simple
calls such as CohereCallResponse.tool
or CohereTool.fn
, as seen in the
examples below.
Example:
from mirascope.cohere import CohereCall, CohereCallParams
def animal_matcher(fav_food: str, fav_color: str) -> str:
"""Tells you your most likely favorite animal from personality traits.
Args:
fav_food: your favorite food.
fav_color: your favorite color.
Returns:
The animal most likely to be your favorite based on traits.
"""
return "Your favorite animal is the best one, a frog."
class AnimalMatcher(CohereCall):
prompt_template = """
Tell me my favorite animal if my favorite food is {food} and my
favorite color is {color}.
"""
food: str
color: str
call_params = CohereCallParams(tools=[animal_matcher])
response = AnimalMatcher(food="pizza", color="red").call
tool = response.tool
print(tool.fn(**tool.args))
#> Your favorite animal is the best one, a frog.
Source code in mirascope/cohere/tools.py
19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 |
|
from_base_type(base_type)
classmethod
¶
Constructs a CohereTool
type from a BaseType
type.
from_fn(fn)
classmethod
¶
from_model(model)
classmethod
¶
Constructs a CohereTool
type from a BaseModel
type.
from_tool_call(tool_call)
classmethod
¶
Extracts an instance of the tool constructed from a tool call response.
Given ToolCall
from an Cohere chat completion response, takes its function
arguments and creates an CohereTool
instance from it.
Parameters:
Name | Type | Description | Default |
---|---|---|---|
tool_call |
ToolCall
|
The |
required |
Returns:
Type | Description |
---|---|
CohereTool
|
An instance of the tool constructed from the tool call. |
Raises:
Type | Description |
---|---|
ValidationError
|
if the tool call doesn't match the tool schema. |
Source code in mirascope/cohere/tools.py
tool_schema()
classmethod
¶
Constructs a tool schema for use with the Cohere chat client.
A Mirascope CohereTool
is deconstructed into a JSON schema, and relevant keys
are renamed to match the Cohere tool schema used to make function/tool calls in
the Cohere chat API.
Returns:
Type | Description |
---|---|
Tool
|
The constructed tool schema. |
Source code in mirascope/cohere/tools.py
MessageRole
¶
Bases: _Enum
Roles that the BasePrompt
messages parser can parse from the template.
SYSTEM: A system message. USER: A user message. ASSISTANT: A message response from the assistant or chat client. MODEL: A message response from the assistant or chat client. Model is used by Google's Gemini instead of assistant, which doesn't have system messages. CHATBOT: A message response from the chat client. Chatbot is used by Cohere instead of assistant. TOOL: A message representing the output of calling a tool.
Source code in mirascope/enums.py
cohere_api_calculate_cost(usage, model='command-r')
¶
Calculate the cost of a completion using the Cohere API.
https://cohere.com/pricing
Model Input Output command-r $0.5 / 1M tokens $1.5 / 1M tokens command-r-plus $3 / 1M tokens $15 / 1M tokens
Source code in mirascope/cohere/utils.py
get_wrapped_async_client(client, self)
¶
Get a wrapped async client.
Source code in mirascope/base/ops_utils.py
get_wrapped_call(call, self, **kwargs)
¶
Wrap a call to add the llm_ops
parameter if it exists.
Source code in mirascope/base/ops_utils.py
get_wrapped_client(client, self)
¶
Get a wrapped client.
Source code in mirascope/base/ops_utils.py
retry(fn)
¶
Decorator for retrying a function.