Skip to content

local vision llm provider#150

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
rcholic merged 1 commit into from
Jan 16, 2026
Merged

local vision llm provider #150

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
1 change: 0 additions & 1 deletion examples/human_cursor_click_demo.py
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -38,4 +38,3 @@ def main() -> None:

if __name__ == "__main__":
main()

9 changes: 9 additions & 0 deletions pyproject.toml
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -43,6 +43,15 @@ sentience = "sentience.cli:main"
browser-use = [
"browser-use>=0.1.40",
]
vision-local = [
"pillow>=10.0.0",
"torch>=2.2.0",
"transformers>=4.46.0",
]
mlx-vlm = [
"pillow>=10.0.0",
"mlx-vlm>=0.1.0",
]
dev = [
"pytest>=7.0.0",
"pytest-asyncio>=0.21.0",
Expand Down
4 changes: 4 additions & 0 deletions sentience/__init__.py
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -52,6 +52,8 @@
LLMProvider,
LLMResponse,
LocalLLMProvider,
LocalVisionLLMProvider,
MLXVLMProvider,
OpenAIProvider,
)
from .models import ( # Agent Layer Models
Expand DownExpand Up@@ -202,6 +204,8 @@
"OpenAIProvider",
"AnthropicProvider",
"LocalLLMProvider",
"LocalVisionLLMProvider",
"MLXVLMProvider",
"SentienceAgent",
"SentienceAgentAsync",
"SentienceVisualAgent",
Expand Down
13 changes: 10 additions & 3 deletions sentience/cursor_policy.py
View file
Open in desktop
Original file line numberDiff line numberDiff line change
Expand Up@@ -44,7 +44,13 @@ def _ease_in_out(t: float) -> float:
return t * t * (3 - 2 * t)


def _bezier(p0: tuple[float, float], p1: tuple[float, float], p2: tuple[float, float], p3: tuple[float, float], t: float) -> tuple[float, float]:
def _bezier(
p0: tuple[float, float],
p1: tuple[float, float],
p2: tuple[float, float],
p3: tuple[float, float],
t: float,
) -> tuple[float, float]:
u = 1.0 - t
tt = t * t
uu = u * u
Expand DownExpand Up@@ -87,7 +93,9 @@ def build_human_cursor_path(

# Defaults based on distance (bounded)
steps = int(policy.steps if policy.steps is not None else _clamp(10 + dist / 25.0, 12, 40))
duration_ms = int(policy.duration_ms if policy.duration_ms is not None else _clamp(120 + dist * 0.9, 120, 700))
duration_ms = int(
policy.duration_ms if policy.duration_ms is not None else _clamp(120 + dist * 0.9, 120, 700)
)

# Control points: offset roughly perpendicular to travel direction
if dist < 1e-6:
Expand DownExpand Up@@ -132,4 +140,3 @@ def build_human_cursor_path(
# Keep path bounded for trace size
"path": pts[:64],
}

Loading
Loading