remove sf for now

This commit is contained in:
H1yori233
2026-03-24 15:25:05 -07:00
parent 49f08b5482
commit 4c3e5624f9
3 changed files with 0 additions and 1878 deletions
@@ -1,163 +0,0 @@
#!/bin/bash
#SBATCH --job-name=mg-test
#SBATCH --partition=main
#SBATCH --nodes=4
#SBATCH --ntasks=4
#SBATCH --ntasks-per-node=1
#SBATCH --gres=gpu:8
#SBATCH --cpus-per-task=128
#SBATCH --mem=1440G
#SBATCH --output=sf_train_output/sf_%j.out
#SBATCH --error=sf_train_output/sf_%j.err
#SBATCH --exclusive
set -e -x
# Environment Setup
source ~/conda/miniconda/bin/activate
conda activate /mnt/weka/home/hao.zhang/conda/miniconda/envs/alexfv
export PYTHONPATH="/mnt/weka/home/hao.zhang/kaiqin/FastVideo_clean:$PYTHONPATH"
# Basic Info
export WANDB_API_KEY="7ff8b6e8356924f7a6dd51a0342dd1a422ea9352"
export WANDB_MODE="online"
export NCCL_P2P_DISABLE=1
export MASTER_PORT=29500
export NODE_RANK=$SLURM_PROCID
nodes=( $(scontrol show hostnames $SLURM_JOB_NODELIST) )
export MASTER_ADDR=${nodes[0]}
export TOKENIZERS_PARALLELISM=false
echo "MASTER_ADDR: $MASTER_ADDR"
echo "NODE_RANK: $NODE_RANK"
RUN_NAME=$(date +"%m%d_%H%M")
echo "RUN_NAME: $RUN_NAME"
# Model paths for Self-Forcing DMD distillation:
GENERATOR_MODEL_PATH="../mg_models/Solaris-Causal-30K-8K-18K"
REAL_SCORE_MODEL_PATH="../mg_models/Solaris-30K-8K-18K" # Teacher mode
FAKE_SCORE_MODEL_PATH="../mg_models/Solaris-30K-8K-18K" # Critic model
# GENERATOR_MODEL_PATH="../mg_models/Solaris-6Ksteps-Causal-0228"
# REAL_SCORE_MODEL_PATH="../mg_models/Solaris-10Ksteps-0301" # Teacher mode
# FAKE_SCORE_MODEL_PATH="../mg_models/Solaris-10Ksteps-0301" # Critic model
DATA_DIR="../solaris/datasets/train/preprocessed_0228_81"
VALIDATION_DATASET_FILE="examples/distill/MatrixGame2.0/validation.json"
# Training arguments
training_args=(
--tracker_project_name "matrixgame_sf"
--output_dir "checkpoints/matrixgame_sf_${RUN_NAME}"
--wandb_run_name "${RUN_NAME}_test"
--max_train_steps 1200
--train_batch_size 1
--train_sp_batch_size 1
--gradient_accumulation_steps 1
--num_latent_t 21
--num_height 352
--num_width 640
# --enable_gradient_checkpointing_type "full"
--log_visualization
--simulate_generator_forward
--num_frames 81
--num_frame_per_block 3 # Frame generation block size for self-forcing
# --enable_gradient_masking
# --gradient_mask_last_n_frames 21
--use-diagonal-denoising
)
# Parallel arguments
parallel_args=(
--num_gpus 32
--sp_size 1
--tp_size 1
--hsdp_replicate_dim 1
--hsdp_shard_dim 32
)
model_args=(
--model_path $GENERATOR_MODEL_PATH # TODO: check if you can remove this in this script
--pretrained_model_name_or_path $GENERATOR_MODEL_PATH
--real_score_model_path $REAL_SCORE_MODEL_PATH
--fake_score_model_path $FAKE_SCORE_MODEL_PATH
)
dataset_args=(
--data_path "$DATA_DIR"
--dataloader_num_workers 4
)
# Validation arguments
validation_args=(
--log_validation
--log_visualization
--visualization-steps 200
--validation_dataset_file "$VALIDATION_DATASET_FILE"
--validation_steps 200
--validation_sampling_steps "4"
--validation_guidance_scale "6.0"
)
# Optimizer arguments
optimizer_args=(
--learning_rate 3e-6
--mixed_precision "bf16"
--weight_only_checkpointing_steps 400
--training_state_checkpointing_steps 400
--weight_decay 0
--betas "0.9,0.95"
--max_grad_norm 1.0
)
# Miscellaneous arguments
miscellaneous_args=(
--inference_mode False
--checkpoints_total_limit 3
--training_cfg_rate 0.0
--dit_precision "fp32"
--flow_shift 5
--seed 1000
--use_ema True
--ema_decay 0.99
--ema_start_step 200
)
dmd_args=(
--dmd_denoising_steps '1000,100'
--diagonal-warmup-mid-steps '750,500'
--min_timestep_ratio 0.02
--max_timestep_ratio 0.98
--dfake_gen_update_ratio 5
--real_score_guidance_scale 3.0
--fake_score_learning_rate 3e-7
--fake_score_betas "0.9,0.95"
--warp_denoising_step
)
self_forcing_args=(
--independent_first_frame False # Whether to treat first frame independently
--same_step_across_blocks True # Whether to use same denoising step across all blocks
--last_step_only False # Whether to only use the last denoising step
--context_noise 0 # Amount of noise to add during context caching (0 = no noise)
)
mkdir -p sf_train_output
srun torchrun \
--nnodes $SLURM_JOB_NUM_NODES \
--nproc_per_node 8 \
--node_rank $SLURM_PROCID \
--rdzv_backend=c10d \
--rdzv_endpoint="$MASTER_ADDR:$MASTER_PORT" \
fastvideo/training/matrixgame_self_forcing_distillation_pipeline.py \
"${parallel_args[@]}" \
"${model_args[@]}" \
"${dataset_args[@]}" \
"${training_args[@]}" \
"${optimizer_args[@]}" \
"${validation_args[@]}" \
"${miscellaneous_args[@]}" \
"${dmd_args[@]}" \
"${self_forcing_args[@]}"
@@ -1,324 +0,0 @@
{
"data": [
{
"caption": "000009_Bravo_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000009_Bravo_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000009_Bravo_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000009_Bravo_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000009_Bravo_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000009_Bravo_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000009_Bravo_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000009_Bravo_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000010_Alpha_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Alpha_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Alpha_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Alpha_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000010_Alpha_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Alpha_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Alpha_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Alpha_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000010_Bravo_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Bravo_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Bravo_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Bravo_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000010_Bravo_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Bravo_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Bravo_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000010_Bravo_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000011_Alpha_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Alpha_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Alpha_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Alpha_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000011_Alpha_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Alpha_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Alpha_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Alpha_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000011_Bravo_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Bravo_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Bravo_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Bravo_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000011_Bravo_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Bravo_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Bravo_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000011_Bravo_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000012_Alpha_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000012_Alpha_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000012_Alpha_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000012_Alpha_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000012_Alpha_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000012_Alpha_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000012_Alpha_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000012_Alpha_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000012_Bravo_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000012_Bravo_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000012_Bravo_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/translationEval/test/000012_Bravo_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000008_Bravo_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000008_Bravo_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000008_Bravo_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000008_Bravo_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000008_Bravo_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000008_Bravo_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000008_Bravo_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000008_Bravo_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000009_Alpha_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Alpha_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Alpha_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Alpha_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000009_Alpha_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Alpha_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Alpha_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Alpha_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000009_Bravo_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Bravo_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Bravo_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Bravo_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000009_Bravo_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Bravo_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Bravo_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000009_Bravo_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000010_Alpha_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Alpha_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Alpha_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Alpha_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000010_Alpha_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Alpha_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Alpha_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Alpha_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000010_Bravo_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Bravo_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Bravo_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Bravo_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000010_Bravo_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Bravo_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Bravo_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Bravo_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000010_Bravo_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Bravo_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Bravo_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000010_Bravo_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000011_Alpha_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Alpha_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Alpha_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Alpha_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000011_Alpha_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Alpha_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Alpha_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Alpha_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000011_Bravo_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Bravo_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Bravo_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Bravo_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000011_Bravo_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Bravo_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Bravo_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/oneLooksAwayEval/test/000011_Bravo_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000015_Alpha_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Alpha_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Alpha_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Alpha_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000015_Alpha_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Alpha_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Alpha_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Alpha_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000015_Bravo_instance_000",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Bravo_instance_000.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Bravo_instance_000_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Bravo_instance_000_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
},
{
"caption": "000015_Bravo_instance_001",
"image_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Bravo_instance_001.png",
"ref_video": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Bravo_instance_001_camera.mp4",
"action_path": "/mnt/weka/home/hao.zhang/kaiqin/solaris/datasets/eval/structureEval/test/000015_Bravo_instance_001_camera.npy",
"num_inference_steps": 4,
"height": 352,
"width": 640,
"num_frames": 81
}
]
}
File diff suppressed because it is too large Load Diff