run_job_head_cs_021.job

#!/bin/bash
#SBATCH --job-name=pytorch_test             # Job name
#SBATCH --output=/work/imvia/de1450bo/repos/RNb-NeuS-fork/logs/test_output_%j.log  # Standard output
#SBATCH --error=/work/imvia/de1450bo/repos/RNb-NeuS-fork/logs/test_error_%j.log    # Standard error
#SBATCH --time=24:00:00                     # Time limit (24 hours)
#SBATCH --gres=gpu:1                        # Number of GPUs
#SBATCH --ntasks=1                          # Number of tasks
#SBATCH --cpus-per-task=4                   # Number of CPUs per task
#SBATCH --mem=16G                           # Memory allocation

# Variable to specify the case
CASE="head_cs_021"
N_ITERATIONS="300000"
BATCH_SIZE="2048" #4906
# Set the path to your configuration file
CONF_PATH="/work/imvia/de1450bo/repos/RNb-NeuS-fork/confs/wmask_rnb.conf"

# Generate the updated configuration file based on the CASE variable
cat <<EOL > $CONF_PATH
general {
    base_exp_dir = /work/imvia/de1450bo/repos/RNb-NeuS-fork/data/out/${CASE}/wmask_rnb/
    recording = [
        ./,
        ./models
    ]
}

dataset {
    data_dir = /work/imvia/de1450bo/repos/RNb-NeuS-fork/data/in/${CASE}_in/

    normal_dir = normal
    albedo_dir = albedo
    mask_dir = mask

    render_cameras_name = cameras_${CASE}.npz
    object_cameras_name = cameras_${CASE}.npz
}

train {
    learning_rate = 0.001#5e-4,
    learning_rate_alpha = 0.05,
    end_iter = ${N_ITERATIONS},#300000,
    warm_up_iter = 10,#200000,

    batch_size = ${BATCH_SIZE}#512,
    validate_resolution_level = 4,
    warm_up_end = 5000,
    anneal_end = 0,
    use_white_bkgd = False,

    save_freq = 20000, # It was 10.000 before
    val_freq = 5000,
    val_mesh_freq = 25000,
    report_freq = 500,

    igr_weight = 0.1,
    mask_weight = 0.1,
}

model {
    nerf {
        D = 8,
        d_in = 4,
        d_in_view = 3,
        W = 256,
        multires = 10,
        multires_view = 4,
        output_ch = 4,
        skips=[4],
        use_viewdirs=True
    }

    sdf_network {
        d_out = 257,
        d_in = 3,
        d_hidden = 256,
        n_layers = 8,
        skip_in = [4],
        multires = 6,
        bias = 0.5,
        scale = 1.0,
        geometric_init = True,
        weight_norm = True
    }

    variance_network {
        init_val = 0.3
    }

    rendering_network {
        d_feature = 256,
        mode = no_view_dir,
        d_in = 6,
        d_out = 3,
        d_hidden = 256,
        n_layers = 2,
        weight_norm = True,
        multires_view = 4,
        squeeze_out = True
    }

    neus_renderer {
        n_samples = 64,
        n_importance = 64,
        n_outside = 0,
        up_sample_steps = 4,    # 1 for simple coarse-to-fine sampling
        perturb = 1.0
    }
}
EOL

# Load any necessary modules
module load pytorch/2.2.2/gpu               # Load PyTorch GPU module

# Set the path to your virtual environment
VENV_PATH="/work/imvia/de1450bo/repos/RNb-NeuS-fork/.venv"

# Activate the virtual environment
source $VENV_PATH/bin/activate

# Debugging information
which python
python --version
which pip
pip list

# Path for the real-time log file
LOG_FILE="/work/imvia/de1450bo/repos/RNb-NeuS-fork/logs/${CASE}_training.log"

# Run your Python script and log output in real-time
/work/imvia/de1450bo/repos/RNb-NeuS-fork/.venv/bin/python exp_runner.py --mode train_rnb --conf $CONF_PATH --case $CASE 2>&1 | tee $LOG_FILE
# qsub -q gpu jobs/run_job_head_cs_016.job