{
  "bs_dim": 6,
  "feature_dim": 512,
  "period": 30,
  "max_seq_len": 5000,
  "batch_size": 128,
  "gpu0_bs": 32,
  "valid_batch_size": 4,
  "experiment_name": "pose2posi_hubertbase_tf",
  "data_root": "/mnt/ssd/PianoPose-new",
  "preload": true,
  "tiny": false,
  "adjust": true,
  "is_random": true,
  "return_beta": false,
  "up_list": [
    "1467634",
    "66685747"
  ],
  "continue_train": false,
  "pretrained_model_path": "pretrained_model/EmoTalk.pth",
  "pretrained_mode": "wo_bs_map",
  "wav2vec_path": "./checkpoints/hubert-base-ls960",
  "fine_map": 0,
  "encoder_type": "transformer",
  "num_layer": 2,
  "pe": false,
  "latest_layer": "tanh",
  "loss_mode": "naive_l1",
  "weight_rec": 1.0,
  "weight_vel": 1.0,
  "iterations": 100000,
  "train_sec": 8,
  "fps": 30,
  "lr": 2e-05,
  "check_val_every_n_iteration": 1000,
  "limit_val_batches": 0.6,
  "save_every_n_iteration": 1000,
  "save_top_k": 5,
  "logdir": "logs"
}