{ "data_root": "/data/oss_bucket_0/yq/dataset/imagenet256_sdvae_ft_ema_latent_kaggle_scaled", "num_classes": 1000, "latent_size": 32, "in_channels": 4, "preset": "custom", "input_size": 32, "patch_size": 2, "hidden_size": 1152, "depth": 28, "num_heads": 16, "mlp_ratio": 4.0, "epochs": 100, "batch_size": 64, "num_workers": 16, "lr": 0.0004, "weight_decay": 0.0001, "grad_clip": 1.0, "grad_accum_steps": 1, "seed": 42, "mask_ratio": 0.25, "student_layer": 8, "teacher_layer": 20, "rep_weight": 0.8, "relation_weight": 1.0, "rep_token_mode": "all", "rep_token_ratio": 1.0, "rep_every": 1, "ema_decay": 0.9999, "teacher_t_mode": "min", "target_mode": "clean_minus_noise", "save_dir": "/data/oss_bucket_0/yq/exp/self-flow/relation_self_flow", "save_every": 5, "val_every": 1, "use_bf16": true, "zero_stage": 1, "resume": "/data/oss_bucket_0/yq/exp/self-flow/relation_self_flow/deepspeed_ckpt", "resume_tag": "", "loss_csv": "", "loss_jsonl": "", "local_rank": 0, "profile_every": 0 }