{
    "models": {
        "encoder": {
            "name": "SparseUnetVaeEncoder",
            "args": {
                "in_channels": 20,
                "model_channels": [64, 128, 256, 512, 1024],
                "latent_channels": 32,
                "num_blocks": [0, 4, 8, 16, 4],
                "block_type": [
                    "SparseConvNeXtBlock3d",
                    "SparseConvNeXtBlock3d",
                    "SparseConvNeXtBlock3d",
                    "SparseConvNeXtBlock3d",
                    "SparseConvNeXtBlock3d"
                ],
                "down_block_type": [
                    "SparseResBlockS2C3d",
                    "SparseResBlockS2C3d",
                    "SparseResBlockS2C3d",
                    "SparseResBlockS2C3d"
                ],
                "block_args": [{}, {}, {}, {}, {}],
                "use_fp16": true
            }
        },
        "decoder": {
            "name": "SparseUnetVaeDecoder",
            "args": {
                "out_channels": 20,
                "model_channels": [1024, 512, 256, 128, 64],
                "latent_channels": 32,
                "num_blocks": [4, 16, 8, 4, 0],
                "block_type": [
                    "SparseConvNeXtBlock3d",
                    "SparseConvNeXtBlock3d",
                    "SparseConvNeXtBlock3d",
                    "SparseConvNeXtBlock3d",
                    "SparseConvNeXtBlock3d"
                ],
                "up_block_type": [
                    "SparseResBlockC2S3d",
                    "SparseResBlockC2S3d",
                    "SparseResBlockC2S3d",
                    "SparseResBlockC2S3d"
                ],
                "block_args": [{}, {}, {}, {}, {}],
                "use_fp16": true,
                "pred_subdiv": false
            }
        }
    },
    "dataset": {
        "name": "SegVaeDataset",
        "args": {
            "resolution": 512,
            "num_classes": 20,
            "max_tokens": 30000,
            "min_tokens": 1
        }
    },
    "trainer": {
        "name": "SegVaeTrainer",
        "args": {
            "max_steps": 50000,
            "batch_size_per_gpu": 8,
            "batch_split": 1,
            "optimizer": {
                "name": "AdamW",
                "args": {
                    "lr": 1e-5,
                    "weight_decay": 0.0
                }
            },
            "ema_rate": [0.9999],
            "fp16_mode": "inflat_all",
            "fp16_scale_growth": 0.001,
            "grad_clip": {
                "name": "AdaptiveGradClipper",
                "args": {
                    "max_norm": 1.0,
                    "clip_percentile": 95
                }
            },
            "finetune_ckpt": {
                "encoder": "/cos_zw1/share_304064442/hunyuan/easonyzhu/model/TRELLIS.2-4B/ckpts/tex_enc_next_dc_f16c32_fp16.safetensors",
                "decoder": "/cos_zw1/share_304064442/hunyuan/easonyzhu/model/TRELLIS.2-4B/ckpts/tex_dec_next_dc_f16c32_fp16.safetensors"
            },
            "lambda_kl": 1e-5,
            "i_print": 1,
            "i_log": 10,
            "i_sample": 2000,
            "i_save": 500
        }
    }
}