Skip to content
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

Fix SAM2 constant reinstall #11

Open
wants to merge 2 commits into
base: main
Choose a base branch
from
Open

Fix SAM2 constant reinstall #11

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
414c0dd
Fix SAM2 reinstalling itself on every run
lukaboljevic Feb 1, 2025
487a34d
Run code quality commands
lukaboljevic Feb 1, 2025
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
21 changes: 17 additions & 4 deletions autodistill_grounded_sam_2/grounded_sam_2.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,7 +13,11 @@
from autodistill.helpers import load_image
from autodistill_florence_2 import Florence2

from autodistill_grounded_sam_2.helpers import load_SAM, load_grounding_dino, combine_detections
from autodistill_grounded_sam_2.helpers import (
combine_detections,
load_grounding_dino,
load_SAM,
)

HOME = os.path.expanduser("~")
DEVICE = torch.device("cuda" if torch.cuda.is_available() else "cpu")
Expand All @@ -22,16 +26,25 @@

SUPPORTED_GROUNDING_MODELS = ["Florence 2", "Grounding DINO"]


@dataclass
class GroundedSAM2(DetectionBaseModel):
ontology: CaptionOntology
box_threshold: float
text_threshold: float

def __init__(self, ontology: CaptionOntology, model = "Florence 2", grounding_dino_box_threshold = 0.35, grounding_dino_text_threshold = 0.25):
def __init__(
self,
ontology: CaptionOntology,
model="Florence 2",
grounding_dino_box_threshold=0.35,
grounding_dino_text_threshold=0.25,
):
if model not in SUPPORTED_GROUNDING_MODELS:
raise ValueError(f"Grounding model {model} is not supported. Supported models are {SUPPORTED_GROUNDING_MODELS}")

raise ValueError(
f"Grounding model {model} is not supported. Supported models are {SUPPORTED_GROUNDING_MODELS}"
)

self.ontology = ontology
if model == "Florence 2":
self.florence_2_predictor = Florence2(ontology=ontology)
Expand Down
21 changes: 11 additions & 10 deletions autodistill_grounded_sam_2/helpers.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,17 +2,18 @@
import subprocess
import sys
import urllib.request
from groundingdino.util.inference import Model

import torch
import numpy as np
import supervision as sv
import torch
from groundingdino.util.inference import Model

DEVICE = torch.device("cuda" if torch.cuda.is_available() else "cpu")

if not torch.cuda.is_available():
print("WARNING: CUDA not available. GroundingDINO will run very slowly.")


def load_grounding_dino():
AUTODISTILL_CACHE_DIR = os.path.expanduser("~/.cache/autodistill")

Expand Down Expand Up @@ -56,21 +57,25 @@ def load_grounding_dino():

return grounding_dino_model


def load_SAM():
cur_dir = os.getcwd()

AUTODISTILL_CACHE_DIR = os.path.expanduser("~/.cache/autodistill")
SAM_CACHE_DIR = os.path.join(AUTODISTILL_CACHE_DIR, "segment_anything_2")
SAM_CHECKPOINT_PATH = os.path.join(SAM_CACHE_DIR, "sam2_hiera_base_plus.pth")

SAM_REPOSITORY_NAME = "segment-anything-2"
SAM_REPOSITORY_DIR = os.path.join(SAM_CACHE_DIR, SAM_REPOSITORY_NAME)

url = "https://dl.fbaipublicfiles.com/segment_anything_2/072824/sam2_hiera_base_plus.pt"

# Create the destination directory if it doesn't exist
os.makedirs(os.path.dirname(SAM_CHECKPOINT_PATH), exist_ok=True)

os.chdir(SAM_CACHE_DIR)

if not os.path.isdir("~/.cache/autodistill/segment_anything_2/segment-anything-2"):
if not os.path.isdir(SAM_REPOSITORY_DIR):
subprocess.run(
[
"git",
Expand All @@ -79,11 +84,11 @@ def load_SAM():
]
)

os.chdir("segment-anything-2")
os.chdir(SAM_REPOSITORY_NAME)

subprocess.run(["pip", "install", "-e", "."])

sys.path.append("~/.cache/autodistill/segment_anything_2/segment-anything-2")
sys.path.append(SAM_REPOSITORY_DIR)

# Download the file if it doesn't exist
if not os.path.isfile(SAM_CHECKPOINT_PATH):
Expand All @@ -92,17 +97,14 @@ def load_SAM():
from sam2.build_sam import build_sam2
from sam2.sam2_image_predictor import SAM2ImagePredictor

checkpoint = "~/.cache/autodistill/segment_anything_2/sam2_hiera_base_plus.pth"
checkpoint = os.path.expanduser(checkpoint)
model_cfg = "sam2_hiera_b+.yaml"
predictor = SAM2ImagePredictor(build_sam2(model_cfg, checkpoint))
predictor = SAM2ImagePredictor(build_sam2(model_cfg, SAM_CHECKPOINT_PATH))

os.chdir(cur_dir)

return predictor



def combine_detections(detections_list, overwrite_class_ids):
if len(detections_list) == 0:
return sv.Detections.empty()
Expand Down Expand Up @@ -156,4 +158,3 @@ def combine_detections(detections_list, overwrite_class_ids):
class_id=class_id,
tracker_id=tracker_id,
)