Replies: 4 comments 8 replies
-
|
To resolve the issue where LlamaIndex does not recognize the 'gpt-4o' model in your Azure configuration, you need to ensure that the model is correctly specified and supported. Here is a solution to include 'gpt-4o' as a valid model:
from llama_index.llms.azure_openai import AzureOpenAIMultiModal
class CustomAzureOpenAIMultiModal(AzureOpenAIMultiModal):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
self.valid_models = [
"gpt-4-vision-preview",
"gpt-4-turbo-2024-04-09",
"gpt-4-turbo",
"gpt-4o",
"gpt-4o-2024-05-13",
"gpt-4" # Add this line
]
def _is_valid_engine(self, engine: Optional[str]) -> bool:
return engine in self.valid_models
# Use the custom class
mm_llm = CustomAzureOpenAIMultiModal(
model="gpt-4o",
deployment_name="gpt-4o",
azure_endpoint=azure_endpoint,
api_key=api_key,
openai_api_type="azure",
openai_api_version='2023-05-15',
temperature=0
)
response = mm_llm.complete(
prompt="Describe the image",
image_documents=image_path
)
llm = CustomAzureOpenAIMultiModal(
model="gpt-4o",
deployment_name="gpt-4o",
api_key=api_key,
azure_endpoint=azure_endpoint,
api_version=api_version,
temperature=0,
)This should resolve the |
Beta Was this translation helpful? Give feedback.
-
|
@dosu. Not working. I am getting this below error. |
Beta Was this translation helpful? Give feedback.
-
|
@dosu: BTW, my gpt-4o keys are wroking with langchain via 'AzureChatOpenAI'. Is that possible to work same way with llamaindex? |
Beta Was this translation helpful? Give feedback.
-
|
To resolve the
from llama_index.llms.azure_openai import AzureOpenAI
# Set up the AzureOpenAI instance
llm = AzureOpenAI(
model="gpt-4o", # Specify the model name
deployment_name="gpt-4o", # Specify your deployment name
azure_endpoint="https://YOUR_RESOURCE_NAME.openai.azure.com/", # Your Azure endpoint
api_key="your_api_key", # Your Azure OpenAI API key
api_version="2023-07-01-preview", # API version
temperature=0
)
# Now you can use `llm` instance in your LlamaIndex setupMake sure to replace
This setup should allow you to use the 'gpt-4o' model with LlamaIndex in a similar way to how you use it with LangChain via 'AzureChatOpenAI' [1][2][3][4][5]. |
Beta Was this translation helpful? Give feedback.
-
|
Still have the same error. Please help. It seems like 'AzureChatOpenAI' is the solution (might be). Please help. |
Beta Was this translation helpful? Give feedback.
-
|
To configure the 'gpt-4o' model in your Azure environment to work with LlamaIndex, you need to ensure that the from llama_index.core.service_context import ServiceContext
from llama_index.core.service_context_elements.llm_predictor import AzureOpenAIPredictor
from llama_index.core.response_synthesizers.factory import get_response_synthesizer
from llama_index.core.response_synthesizers.type import ResponseMode
# Configure the AzureOpenAIPredictor with your Azure OpenAI settings
azure_openai_predictor = AzureOpenAIPredictor(
api_key="your_azure_api_key",
api_base="your_azure_api_base",
deployment_name="gpt-4o"
)
# Create a ServiceContext with the AzureOpenAIPredictor
service_context = ServiceContext(
llm_predictor=azure_openai_predictor
)
# Get a response synthesizer using the configured ServiceContext
response_synthesizer = get_response_synthesizer(
service_context=service_context,
response_mode=ResponseMode.COMPACT # or any other mode you need
)
# Now you can use the response_synthesizer in your applicationMake sure to replace |
Beta Was this translation helpful? Give feedback.
-
|
@dosu Error again :-( |
Beta Was this translation helpful? Give feedback.
-
|
To resolve the
from llama_index.llms.azure_openai import AzureOpenAI
# Set up the AzureOpenAI instance
llm = AzureOpenAI(
model="gpt-4o", # Specify the model name
deployment_name="gpt-4o", # Specify your deployment name
azure_endpoint="https://YOUR_RESOURCE_NAME.openai.azure.com/", # Your Azure endpoint
api_key="your_api_key", # Your Azure OpenAI API key
api_version="2023-07-01-preview", # API version
temperature=0
)
# Now you can use `llm` instance in your LlamaIndex setupMake sure to replace
This setup should allow you to use the 'gpt-4o' model with LlamaIndex in a similar way to how you use it with LangChain via 'AzureChatOpenAI' [1][2][3]. |
Beta Was this translation helpful? Give feedback.
-
|
@dosu... you repeated the same answer. It won't resolve the error. Sorry. The above solution won't work. |
Beta Was this translation helpful? Give feedback.
-
|
@logan-markewich, this one's a bit out of my wheelhouse, could you step in and provide your guidance? |
Beta Was this translation helpful? Give feedback.
-
|
@logan-markewich thank you for your feedback and looking forward to see that. |
Beta Was this translation helpful? Give feedback.
-
|
This is issue with utilis.py (the file in the webiste has up to date with all models. However, my pc file is not updated despite I upgrade the package. I replace with below and error has been resolved. Closing the ticket. |
Beta Was this translation helpful? Give feedback.
-
We configured gpt-4o model in our environment and it is working fine for other packages. However, it is not working for llamaindex. Any thoughts?
My config:
File ~\AppData\Roaming\Python\Python311\site-packages\llama_index\core\response_synthesizers\factory.py:73, in get_response_synthesizer(llm, prompt_helper, service_context, text_qa_template, refine_template, summary_template, simple_template, response_mode, callback_manager, use_async, streaming, structured_answer_filtering, output_cls, program_factory, verbose)
67 prompt_helper = service_context.prompt_helper
68 else:
69 prompt_helper = (
70 prompt_helper
71 or Settings._prompt_helper
72 or PromptHelper.from_llm_metadata(
---> 73 llm.metadata,
74 )
75 )
...
207 f" {', '.join(ALL_AVAILABLE_MODELS.keys())}"
208 )
209 return ALL_AVAILABLE_MODELS[modelname]
ValueError: Unknown model 'gpt-4o'. Please provide a valid OpenAI model name in: gpt-4, gpt-4-32k, gpt-4-1106-preview, gpt-4-0125-preview
Beta Was this translation helpful? Give feedback.
All reactions