Skip to content

Make Ollama context window configurable via num_ctx #29

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

Merged
rcjackson merged 1 commit into from
Jul 10, 2026
Merged

Make Ollama context window configurable via num_ctx #29

Show file tree
Hide file tree
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
3 changes: 3 additions & 0 deletions lars/nepho/config.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,6 +12,9 @@ class Config:

# Ollama Configuration
OLLAMA_BASE_URL: str = os.getenv("OLLAMA_BASE_URL", "http://localhost:11434")
# Context window size (tokens) passed to Ollama as num_ctx. Ollama's
# default is small (2048), which silently truncates long prompts/images.
OLLAMA_NUM_CTX: int = int(os.getenv("OLLAMA_NUM_CTX", "8192"))

# Default models
DEFAULT_GPT_MODEL: str = os.getenv("DEFAULT_GPT_MODEL", "gpt-4-vision-preview")
Expand Down
11 changes: 7 additions & 4 deletions lars/nepho/models/ollama_model.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -8,11 +8,12 @@
class OllamaModel(BaseModel):
"""Ollama model implementation for local models."""

def __init__(self, model_name: str = None, base_url: str = None):
def __init__(self, model_name: str = None, base_url: str = None, num_ctx: int = None):
model_name = model_name or config.DEFAULT_OLLAMA_MODEL
super().__init__(model_name)

self.base_url = base_url or config.OLLAMA_BASE_URL
self.num_ctx = num_ctx or config.OLLAMA_NUM_CTX
self.api_url = f"{self.base_url}/api/generate"
self.chat_url = f"{self.base_url}/api/chat"

Expand Down Expand Up @@ -76,7 +77,8 @@ async def chat(self, prompt: str, images: Optional[List[str]] = None) -> str:
"model": self.model_name,
"prompt": prompt,
"images": images_data,
"stream": False
"stream": False,
"options": {"num_ctx": self.num_ctx}
}

# Use generate endpoint for vision models
Expand All @@ -88,7 +90,8 @@ async def chat(self, prompt: str, images: Optional[List[str]] = None) -> str:
"messages": [
{"role": "user", "content": prompt}
],
"stream": False
"stream": False,
"options": {"num_ctx": self.num_ctx}
}
url = self.chat_url

Expand Down
Loading