Spaces:
Runtime error
Runtime error
File size: 8,226 Bytes
ed4d993 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 |
from typing import Any, AsyncIterator, Dict, Iterator, List, Optional
from langchain_core._api.deprecation import deprecated
from langchain_core.callbacks import (
AsyncCallbackManagerForLLMRun,
CallbackManagerForLLMRun,
)
from langchain_core.language_models.chat_models import (
BaseChatModel,
agenerate_from_stream,
generate_from_stream,
)
from langchain_core.messages import (
AIMessage,
AIMessageChunk,
BaseMessage,
ChatMessage,
HumanMessage,
SystemMessage,
)
from langchain_core.outputs import ChatGeneration, ChatGenerationChunk, ChatResult
from langchain_community.llms.cohere import BaseCohere
def get_role(message: BaseMessage) -> str:
"""Get the role of the message.
Args:
message: The message.
Returns:
The role of the message.
Raises:
ValueError: If the message is of an unknown type.
"""
if isinstance(message, ChatMessage) or isinstance(message, HumanMessage):
return "User"
elif isinstance(message, AIMessage):
return "Chatbot"
elif isinstance(message, SystemMessage):
return "System"
else:
raise ValueError(f"Got unknown type {message}")
def get_cohere_chat_request(
messages: List[BaseMessage],
*,
connectors: Optional[List[Dict[str, str]]] = None,
**kwargs: Any,
) -> Dict[str, Any]:
"""Get the request for the Cohere chat API.
Args:
messages: The messages.
connectors: The connectors.
**kwargs: The keyword arguments.
Returns:
The request for the Cohere chat API.
"""
documents = (
None
if "source_documents" not in kwargs
else [
{
"snippet": doc.page_content,
"id": doc.metadata.get("id") or f"doc-{str(i)}",
}
for i, doc in enumerate(kwargs["source_documents"])
]
)
kwargs.pop("source_documents", None)
maybe_connectors = connectors if documents is None else None
# by enabling automatic prompt truncation, the probability of request failure is
# reduced with minimal impact on response quality
prompt_truncation = (
"AUTO" if documents is not None or connectors is not None else None
)
req = {
"message": messages[-1].content,
"chat_history": [
{"role": get_role(x), "message": x.content} for x in messages[:-1]
],
"documents": documents,
"connectors": maybe_connectors,
"prompt_truncation": prompt_truncation,
**kwargs,
}
return {k: v for k, v in req.items() if v is not None}
@deprecated(
since="0.0.30", removal="0.3.0", alternative_import="langchain_cohere.ChatCohere"
)
class ChatCohere(BaseChatModel, BaseCohere):
"""`Cohere` chat large language models.
To use, you should have the ``cohere`` python package installed, and the
environment variable ``COHERE_API_KEY`` set with your API key, or pass
it as a named parameter to the constructor.
Example:
.. code-block:: python
from langchain_community.chat_models import ChatCohere
from langchain_core.messages import HumanMessage
chat = ChatCohere(model="command", max_tokens=256, temperature=0.75)
messages = [HumanMessage(content="knock knock")]
chat.invoke(messages)
"""
class Config:
"""Configuration for this pydantic object."""
allow_population_by_field_name = True
arbitrary_types_allowed = True
@property
def _llm_type(self) -> str:
"""Return type of chat model."""
return "cohere-chat"
@property
def _default_params(self) -> Dict[str, Any]:
"""Get the default parameters for calling Cohere API."""
return {
"temperature": self.temperature,
}
@property
def _identifying_params(self) -> Dict[str, Any]:
"""Get the identifying parameters."""
return {**{"model": self.model}, **self._default_params}
def _stream(
self,
messages: List[BaseMessage],
stop: Optional[List[str]] = None,
run_manager: Optional[CallbackManagerForLLMRun] = None,
**kwargs: Any,
) -> Iterator[ChatGenerationChunk]:
request = get_cohere_chat_request(messages, **self._default_params, **kwargs)
if hasattr(self.client, "chat_stream"): # detect and support sdk v5
stream = self.client.chat_stream(**request)
else:
stream = self.client.chat(**request, stream=True)
for data in stream:
if data.event_type == "text-generation":
delta = data.text
chunk = ChatGenerationChunk(message=AIMessageChunk(content=delta))
if run_manager:
run_manager.on_llm_new_token(delta, chunk=chunk)
yield chunk
async def _astream(
self,
messages: List[BaseMessage],
stop: Optional[List[str]] = None,
run_manager: Optional[AsyncCallbackManagerForLLMRun] = None,
**kwargs: Any,
) -> AsyncIterator[ChatGenerationChunk]:
request = get_cohere_chat_request(messages, **self._default_params, **kwargs)
if hasattr(self.async_client, "chat_stream"): # detect and support sdk v5
stream = await self.async_client.chat_stream(**request)
else:
stream = await self.async_client.chat(**request, stream=True)
async for data in stream:
if data.event_type == "text-generation":
delta = data.text
chunk = ChatGenerationChunk(message=AIMessageChunk(content=delta))
if run_manager:
await run_manager.on_llm_new_token(delta, chunk=chunk)
yield chunk
def _get_generation_info(self, response: Any) -> Dict[str, Any]:
"""Get the generation info from cohere API response."""
return {
"documents": response.documents,
"citations": response.citations,
"search_results": response.search_results,
"search_queries": response.search_queries,
"token_count": response.token_count,
}
def _generate(
self,
messages: List[BaseMessage],
stop: Optional[List[str]] = None,
run_manager: Optional[CallbackManagerForLLMRun] = None,
**kwargs: Any,
) -> ChatResult:
if self.streaming:
stream_iter = self._stream(
messages, stop=stop, run_manager=run_manager, **kwargs
)
return generate_from_stream(stream_iter)
request = get_cohere_chat_request(messages, **self._default_params, **kwargs)
response = self.client.chat(**request)
message = AIMessage(content=response.text)
generation_info = None
if hasattr(response, "documents"):
generation_info = self._get_generation_info(response)
return ChatResult(
generations=[
ChatGeneration(message=message, generation_info=generation_info)
]
)
async def _agenerate(
self,
messages: List[BaseMessage],
stop: Optional[List[str]] = None,
run_manager: Optional[AsyncCallbackManagerForLLMRun] = None,
**kwargs: Any,
) -> ChatResult:
if self.streaming:
stream_iter = self._astream(
messages, stop=stop, run_manager=run_manager, **kwargs
)
return await agenerate_from_stream(stream_iter)
request = get_cohere_chat_request(messages, **self._default_params, **kwargs)
response = self.client.chat(**request)
message = AIMessage(content=response.text)
generation_info = None
if hasattr(response, "documents"):
generation_info = self._get_generation_info(response)
return ChatResult(
generations=[
ChatGeneration(message=message, generation_info=generation_info)
]
)
def get_num_tokens(self, text: str) -> int:
"""Calculate number of tokens."""
return len(self.client.tokenize(text=text).tokens)
|