Skip to content

fix: update Ollama embedding API usage #1

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
earthenvessel wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix: update Ollama embedding API usage #1

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
44 changes: 25 additions & 19 deletions app.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -98,8 +98,8 @@ def embed_documents(ollama_url: str, model: str, texts: List[str]) -> Union[List
"""
Generate embeddings for a list of text documents using the specified Ollama model.

This function sends POST requests to the Ollama server to generate embeddings
for each text document in the input list.
This function sends a POST request to the Ollama server to generate embeddings
for the input documents.

Args:
ollama_url (str): The URL of the Ollama server.
Expand All @@ -111,24 +111,28 @@ def embed_documents(ollama_url: str, model: str, texts: List[str]) -> Union[List
if successful, or None if there's an error.

Note:
If there's an error generating embeddings for any document, the function will
display an error message in the Streamlit sidebar and return None.
If there's an error generating embeddings, the function will display an error
message in the Streamlit sidebar and return None.
"""
embeddings = []
for text in texts:
try:
response = requests.post(
f'{ollama_url}/api/embeddings',
json={'model': model, 'prompt': text},
timeout=30
)
response.raise_for_status()
embedding = response.json()['embedding']
embeddings.append(embedding)
except requests.RequestException as e:
st.sidebar.error(f"Error getting embedding: {str(e)}")
return None
return embeddings
try:
response = requests.post(
f'{ollama_url}/api/embed',
json={'model': model, 'input': texts},
timeout=30
)
response.raise_for_status()

response_json = response.json()
if 'embeddings' in response_json:
return response_json['embeddings']
if 'embedding' in response_json:
return [response_json['embedding']]

st.sidebar.error("Error getting embedding: Unexpected Ollama response format.")
return None
except requests.RequestException as e:
st.sidebar.error(f"Error getting embedding: {str(e)}")
return None

def chat_with_documents(ollama_url: str, model: str, prompt: str, temperature: float):
"""
Expand Down Expand Up @@ -384,6 +388,8 @@ def hybrid_search(query: str, collection, documents: List[Dict[str, Union[str, D
"""
# Embed the query
question_embedding = embed_documents(st.session_state.ollama_url, st.session_state.embedding_model, [query])
if not question_embedding:
return []

# Perform semantic search over all documents
total_docs = len(documents)
Expand Down