mirror of
https://github.com/BerriAI/litellm.git
synced 2025-04-25 18:54:30 +00:00
refactor: add black formatting
This commit is contained in:
Krrish Dholakia
parent
b87d630b0a
commit
4905929de3
156 changed files with 19723 additions and 10869 deletions
|
|
@ -8,17 +8,22 @@ import litellm
|
|||
import sys, httpx
|
||||
from .prompt_templates.factory import prompt_factory, custom_prompt
|
||||
|
||||
|
||||
class GeminiError(Exception):
|
||||
def __init__(self, status_code, message):
|
||||
self.status_code = status_code
|
||||
self.message = message
|
||||
self.request = httpx.Request(method="POST", url="https://developers.generativeai.google/api/python/google/generativeai/chat")
|
||||
self.request = httpx.Request(
|
||||
method="POST",
|
||||
url="https://developers.generativeai.google/api/python/google/generativeai/chat",
|
||||
)
|
||||
self.response = httpx.Response(status_code=status_code, request=self.request)
|
||||
super().__init__(
|
||||
self.message
|
||||
) # Call the base class constructor with the parameters it needs
|
||||
|
||||
class GeminiConfig():
|
||||
|
||||
class GeminiConfig:
|
||||
"""
|
||||
Reference: https://ai.google.dev/api/python/google/generativeai/GenerationConfig
|
||||
|
||||
|
|
@ -37,33 +42,44 @@ class GeminiConfig():
|
|||
- `top_k` (int): Optional. The maximum number of tokens to consider when sampling.
|
||||
"""
|
||||
|
||||
candidate_count: Optional[int]=None
|
||||
stop_sequences: Optional[list]=None
|
||||
max_output_tokens: Optional[int]=None
|
||||
temperature: Optional[float]=None
|
||||
top_p: Optional[float]=None
|
||||
top_k: Optional[int]=None
|
||||
candidate_count: Optional[int] = None
|
||||
stop_sequences: Optional[list] = None
|
||||
max_output_tokens: Optional[int] = None
|
||||
temperature: Optional[float] = None
|
||||
top_p: Optional[float] = None
|
||||
top_k: Optional[int] = None
|
||||
|
||||
def __init__(self,
|
||||
candidate_count: Optional[int]=None,
|
||||
stop_sequences: Optional[list]=None,
|
||||
max_output_tokens: Optional[int]=None,
|
||||
temperature: Optional[float]=None,
|
||||
top_p: Optional[float]=None,
|
||||
top_k: Optional[int]=None) -> None:
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
candidate_count: Optional[int] = None,
|
||||
stop_sequences: Optional[list] = None,
|
||||
max_output_tokens: Optional[int] = None,
|
||||
temperature: Optional[float] = None,
|
||||
top_p: Optional[float] = None,
|
||||
top_k: Optional[int] = None,
|
||||
) -> None:
|
||||
locals_ = locals()
|
||||
for key, value in locals_.items():
|
||||
if key != 'self' and value is not None:
|
||||
if key != "self" and value is not None:
|
||||
setattr(self.__class__, key, value)
|
||||
|
||||
|
||||
@classmethod
|
||||
def get_config(cls):
|
||||
return {k: v for k, v in cls.__dict__.items()
|
||||
if not k.startswith('__')
|
||||
and not isinstance(v, (types.FunctionType, types.BuiltinFunctionType, classmethod, staticmethod))
|
||||
and v is not None}
|
||||
|
||||
return {
|
||||
k: v
|
||||
for k, v in cls.__dict__.items()
|
||||
if not k.startswith("__")
|
||||
and not isinstance(
|
||||
v,
|
||||
(
|
||||
types.FunctionType,
|
||||
types.BuiltinFunctionType,
|
||||
classmethod,
|
||||
staticmethod,
|
||||
),
|
||||
)
|
||||
and v is not None
|
||||
}
|
||||
|
||||
|
||||
def completion(
|
||||
|
|
@ -83,42 +99,50 @@ def completion(
|
|||
try:
|
||||
import google.generativeai as genai
|
||||
except:
|
||||
raise Exception("Importing google.generativeai failed, please run 'pip install -q google-generativeai")
|
||||
raise Exception(
|
||||
"Importing google.generativeai failed, please run 'pip install -q google-generativeai"
|
||||
)
|
||||
genai.configure(api_key=api_key)
|
||||
|
||||
|
||||
if model in custom_prompt_dict:
|
||||
# check if the model has a registered custom prompt
|
||||
model_prompt_details = custom_prompt_dict[model]
|
||||
prompt = custom_prompt(
|
||||
role_dict=model_prompt_details["roles"],
|
||||
initial_prompt_value=model_prompt_details["initial_prompt_value"],
|
||||
final_prompt_value=model_prompt_details["final_prompt_value"],
|
||||
messages=messages
|
||||
)
|
||||
# check if the model has a registered custom prompt
|
||||
model_prompt_details = custom_prompt_dict[model]
|
||||
prompt = custom_prompt(
|
||||
role_dict=model_prompt_details["roles"],
|
||||
initial_prompt_value=model_prompt_details["initial_prompt_value"],
|
||||
final_prompt_value=model_prompt_details["final_prompt_value"],
|
||||
messages=messages,
|
||||
)
|
||||
else:
|
||||
prompt = prompt_factory(model=model, messages=messages, custom_llm_provider="gemini")
|
||||
prompt = prompt_factory(
|
||||
model=model, messages=messages, custom_llm_provider="gemini"
|
||||
)
|
||||
|
||||
|
||||
## Load Config
|
||||
inference_params = copy.deepcopy(optional_params)
|
||||
inference_params.pop("stream", None) # palm does not support streaming, so we handle this by fake streaming in main.py
|
||||
config = litellm.GeminiConfig.get_config()
|
||||
for k, v in config.items():
|
||||
if k not in inference_params: # completion(top_k=3) > gemini_config(top_k=3) <- allows for dynamic variables to be passed in
|
||||
inference_params.pop(
|
||||
"stream", None
|
||||
) # palm does not support streaming, so we handle this by fake streaming in main.py
|
||||
config = litellm.GeminiConfig.get_config()
|
||||
for k, v in config.items():
|
||||
if (
|
||||
k not in inference_params
|
||||
): # completion(top_k=3) > gemini_config(top_k=3) <- allows for dynamic variables to be passed in
|
||||
inference_params[k] = v
|
||||
|
||||
|
||||
## LOGGING
|
||||
logging_obj.pre_call(
|
||||
input=prompt,
|
||||
api_key="",
|
||||
additional_args={"complete_input_dict": {"inference_params": inference_params}},
|
||||
)
|
||||
input=prompt,
|
||||
api_key="",
|
||||
additional_args={"complete_input_dict": {"inference_params": inference_params}},
|
||||
)
|
||||
## COMPLETION CALL
|
||||
try:
|
||||
_model = genai.GenerativeModel(f'models/{model}')
|
||||
response = _model.generate_content(contents=prompt, generation_config=genai.types.GenerationConfig(**inference_params))
|
||||
try:
|
||||
_model = genai.GenerativeModel(f"models/{model}")
|
||||
response = _model.generate_content(
|
||||
contents=prompt,
|
||||
generation_config=genai.types.GenerationConfig(**inference_params),
|
||||
)
|
||||
except Exception as e:
|
||||
raise GeminiError(
|
||||
message=str(e),
|
||||
|
|
@ -127,11 +151,11 @@ def completion(
|
|||
|
||||
## LOGGING
|
||||
logging_obj.post_call(
|
||||
input=prompt,
|
||||
api_key="",
|
||||
original_response=response,
|
||||
additional_args={"complete_input_dict": {}},
|
||||
)
|
||||
input=prompt,
|
||||
api_key="",
|
||||
original_response=response,
|
||||
additional_args={"complete_input_dict": {}},
|
||||
)
|
||||
print_verbose(f"raw model_response: {response}")
|
||||
## RESPONSE OBJECT
|
||||
completion_response = response
|
||||
|
|
@ -142,31 +166,34 @@ def completion(
|
|||
message_obj = Message(content=item.content.parts[0].text)
|
||||
else:
|
||||
message_obj = Message(content=None)
|
||||
choice_obj = Choices(index=idx+1, message=message_obj)
|
||||
choice_obj = Choices(index=idx + 1, message=message_obj)
|
||||
choices_list.append(choice_obj)
|
||||
model_response["choices"] = choices_list
|
||||
except Exception as e:
|
||||
traceback.print_exc()
|
||||
raise GeminiError(message=traceback.format_exc(), status_code=response.status_code)
|
||||
|
||||
try:
|
||||
raise GeminiError(
|
||||
message=traceback.format_exc(), status_code=response.status_code
|
||||
)
|
||||
|
||||
try:
|
||||
completion_response = model_response["choices"][0]["message"].get("content")
|
||||
except:
|
||||
raise GeminiError(status_code=400, message=f"No response received. Original response - {response}")
|
||||
raise GeminiError(
|
||||
status_code=400,
|
||||
message=f"No response received. Original response - {response}",
|
||||
)
|
||||
|
||||
## CALCULATING USAGE
|
||||
prompt_str = ""
|
||||
prompt_str = ""
|
||||
for m in messages:
|
||||
if isinstance(m["content"], str):
|
||||
prompt_str += m["content"]
|
||||
elif isinstance(m["content"], list):
|
||||
for content in m["content"]:
|
||||
if content["type"] == "text":
|
||||
if content["type"] == "text":
|
||||
prompt_str += content["text"]
|
||||
|
||||
prompt_tokens = len(
|
||||
encoding.encode(prompt_str)
|
||||
)
|
||||
prompt_tokens = len(encoding.encode(prompt_str))
|
||||
completion_tokens = len(
|
||||
encoding.encode(model_response["choices"][0]["message"].get("content", ""))
|
||||
)
|
||||
|
|
@ -174,13 +201,14 @@ def completion(
|
|||
model_response["created"] = int(time.time())
|
||||
model_response["model"] = "gemini/" + model
|
||||
usage = Usage(
|
||||
prompt_tokens=prompt_tokens,
|
||||
completion_tokens=completion_tokens,
|
||||
total_tokens=prompt_tokens + completion_tokens
|
||||
)
|
||||
prompt_tokens=prompt_tokens,
|
||||
completion_tokens=completion_tokens,
|
||||
total_tokens=prompt_tokens + completion_tokens,
|
||||
)
|
||||
model_response.usage = usage
|
||||
return model_response
|
||||
|
||||
|
||||
def embedding():
|
||||
# logic for parsing in - calling - parsing out model embedding calls
|
||||
pass
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue