phoenix/litellm-mirror
0
0
Fork 1
mirror of https://github.com/BerriAI/litellm.git synced 2025-04-25 18:54:30 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

fix supports vision

Browse source
This commit is contained in:
Ishaan Jaff 2024-07-11 12:59:42 -07:00
parent 6e86e4291a
commit 341f88d191
3 changed files with 10 additions and 5 deletions
Download patch file Download diff file Expand all files Collapse all files

3
litellm/proxy/proxy_config.yaml
View file

@ -4,6 +4,9 @@ model_list:
model: openai/fake
api_key: fake-key
api_base: https://exampleopenaiendpoint-production.up.railway.app/
- model_name: gemini-flash
litellm_params:
model: gemini/gemini-1.5-flash
general_settings:
master_key: sk-1234

1
litellm/types/utils.py
View file

@ -73,6 +73,7 @@ class ModelInfo(TypedDict, total=False):
supported_openai_params: Required[Optional[List[str]]]
supports_system_messages: Optional[bool]
supports_response_schema: Optional[bool]
supports_vision: Optional[bool]
class GenericStreamingChunk(TypedDict):

11
litellm/utils.py
View file

@ -4829,6 +4829,7 @@ def get_model_info(model: str, custom_llm_provider: Optional[str] = None) -> Mod
supports_response_schema=_model_info.get(
"supports_response_schema", None
),
supports_vision=_model_info.get("supports_vision", None),
)
except Exception:
raise Exception(
@ -8126,7 +8127,7 @@ class CustomStreamWrapper:
if chunk.startswith(self.complete_response):
# Remove last_sent_chunk only if it appears at the start of the new chunk
chunk = chunk[len(self.complete_response):]
chunk = chunk[len(self.complete_response) :]
self.complete_response += chunk
return chunk
@ -10124,7 +10125,7 @@ def mock_completion_streaming_obj(
model_response, mock_response, model, n: Optional[int] = None
):
for i in range(0, len(mock_response), 3):
completion_obj = Delta(role="assistant", content=mock_response[i: i + 3])
completion_obj = Delta(role="assistant", content=mock_response[i : i + 3])
if n is None:
model_response.choices[0].delta = completion_obj
else:
@ -10133,7 +10134,7 @@ def mock_completion_streaming_obj(
_streaming_choice = litellm.utils.StreamingChoices(
index=j,
delta=litellm.utils.Delta(
role="assistant", content=mock_response[i: i + 3]
role="assistant", content=mock_response[i : i + 3]
),
)
_all_choices.append(_streaming_choice)
@ -10145,7 +10146,7 @@ async def async_mock_completion_streaming_obj(
model_response, mock_response, model, n: Optional[int] = None
):
for i in range(0, len(mock_response), 3):
completion_obj = Delta(role="assistant", content=mock_response[i: i + 3])
completion_obj = Delta(role="assistant", content=mock_response[i : i + 3])
if n is None:
model_response.choices[0].delta = completion_obj
else:
@ -10154,7 +10155,7 @@ async def async_mock_completion_streaming_obj(
_streaming_choice = litellm.utils.StreamingChoices(
index=j,
delta=litellm.utils.Delta(
role="assistant", content=mock_response[i: i + 3]
role="assistant", content=mock_response[i : i + 3]
),
)
_all_choices.append(_streaming_choice)