[misc]: extend the comment sweep to every file

scripts/preprocess/launch_slurm_text.sh

@@ -1,12 +1,10 @@
 #!/bin/bash
 
-# Create output directory if it doesn't exist
 mkdir -p preprocess_output_text
 
 # Launch 8 jobs, one for each node
 # Each node processes 8 consecutive files (64 total files / 8 nodes = 8 files per node)
 for node_id in {0..7}; do
-    # Calculate the starting file number for this node
     start_file=$((node_id * 8 + 1))
     
     echo "Launching text-only node $node_id with files v2m_${start_file}.txt to v2m_$((start_file + 7)).txt"

scripts/preprocess/syn_text.slurm

@@ -29,10 +29,8 @@ NODE_ID=${2:-0}  # Node identifier (0-7)
 
 num_gpus=1
 export MODEL_BASE=Wan-AI/Wan2.1-T2V-1.3B-Diffusers
-# Start port number - we'll increment for each job
 base_port=$((29603 + NODE_ID * 100))  # Different port range per node
 
-# Create an array of CUDA device IDs
 gpu_ids=(0 1 2 3 4 5 6 7)
 
 GPU_NUM=1
@@ -43,12 +41,9 @@ echo "START_FILE: $START_FILE"
 echo "Base port for this node: $base_port"
 echo "Processing TEXT-ONLY data"
 
-# Run 8 parallel preprocessing jobs on this node
 for i in {1..8}; do
-    # Calculate port for this job
     port=$((base_port + i))
     
-    # Get GPU ID using modulo to cycle through available GPUs
     gpu=${gpu_ids[((i-1))]}
     
     # Calculate which file this GPU should process

scripts/preprocess/v1_preprocess_wan_data_i2v.sh

@@ -1,4 +1,3 @@
-# export WANDB_MODE="offline"
 GPU_NUM=1 # 2,4,8
 MODEL_PATH="Wan-AI/Wan2.1-I2V-14B-480P-Diffusers"
 MODEL_TYPE="wan"

scripts/preprocess/v1_preprocess_wan_data_t2v.sh

@@ -1,4 +1,3 @@
-# export WANDB_MODE="offline"
 GPU_NUM=1 # 2,4,8
 MODEL_PATH="Wan-AI/Wan2.1-T2V-1.3B-Diffusers"
 MODEL_TYPE="wan"