File size: 3,197 Bytes
0633d2d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 |
import subprocess
import argparse
# python run_experiments.py --dataset "Birds-Nest" --model yolov10n yolov10s yolov10m yolov10l
def run_experiment(base_command, run_mode, use_pretrained):
"""
Constructs and runs a single experiment command.
"""
command = base_command + ["--run", run_mode]
if use_pretrained:
command.append("--pretrained")
print("="*80)
print(f"Starting run: {run_mode}")
print(f"Command: {' '.join(command)}")
print("="*80)
try:
# subprocess.run is a blocking call, ensuring sequential execution.
subprocess.run(command, check=True)
print(f"\nSUCCESS: Run '{run_mode}' completed.\n")
except subprocess.CalledProcessError as e:
print(f"\nERROR: Run '{run_mode}' failed with exit code {e.returncode}.\n")
# Decide if you want to stop all subsequent runs on failure
# raise e # Uncomment to stop the entire sequence on error
def main():
"""
Parses arguments and launches a sequence of training experiments for each specified model.
"""
parser = argparse.ArgumentParser(description="Run a sequence of YOLO training experiments.")
# Define arguments that will be common to all training runs
parser.add_argument('--dataset', type=str, required=True, choices=["Birds-Nest", "Common-VALID", "Electric-Substation", "InsPLAD-det"], help='Dataset name to be used.')
parser.add_argument("--model", nargs='+', required=True, choices=["yolov8n", "yolov8s", "yolov8m", "yolov8l", "yolov10n", "yolov10s", "yolov10m", "yolov10l"], help="One or more models to use for the experiments.")
parser.add_argument("--epochs", type=int, default=1000, help="Number of epochs.")
parser.add_argument("--batch", type=int, default=16, help="Batch size.")
parser.add_argument("--plots", action="store_true", default=True, help="Generate plots for all runs.")
args = parser.parse_args()
# Define the sequence of experiments to run for each model
# Each tuple is (run_mode, use_pretrained_flag)
experiment_sequence = [
("From_Scratch", False),
("Finetuning", True),
("freeze_[P1-P3]", True),
("freeze_Backbone", True),
("freeze_[P1-23]", True)
]
# Iterate over each specified model variant
for model_name in args.model:
print(f"\n{'='*25} Starting Experiments for Model: {model_name.upper()} {'='*25}\n")
# Base command list, specific to the current model
base_command = [
"python", "main.py",
"--dataset", args.dataset,
"--epochs", str(args.epochs),
"--batch", str(args.batch),
"--model", model_name
]
if args.plots:
base_command.append("--plots")
# Execute each experiment in sequence for the current model
for run_mode, use_pretrained in experiment_sequence:
run_experiment(base_command, run_mode, use_pretrained)
print("="*80)
print("All experiments for all specified models have been completed.")
print("="*80)
if __name__ == "__main__":
main() |