| { |
| "external_models": { |
| "dino": "facebook/dinov3-vitl16-pretrain-lvd1689m", |
| "rembg": "briaai/RMBG-2.0" |
| }, |
| "sparse_structure": { |
| "sampler": { |
| "args": { |
| "sigma_min": 1e-05 |
| }, |
| "params": { |
| "steps": 12, |
| "guidance_strength": 7.5, |
| "guidance_interval": [ |
| 0.6, |
| 1 |
| ], |
| "guidance_rescale": 0.7, |
| "rescale_t": 5 |
| } |
| }, |
| "model": { |
| "name": "JointDiTVGGTStage1Model", |
| "args": { |
| "dit_args": { |
| "resolution": 16, |
| "in_channels": 8, |
| "out_channels": 8, |
| "model_channels": 1536, |
| "cond_channels": 1024, |
| "num_blocks": 30, |
| "num_heads": 12, |
| "mlp_ratio": 5.3334, |
| "pe_mode": "rope", |
| "share_mod": true, |
| "initialization": "scaled", |
| "qk_rms_norm": true, |
| "qk_rms_norm_cross": true |
| }, |
| "dino_image_size": 512, |
| "vggt_img_size": 518, |
| "vggt_patch_size": 14, |
| "anchor_view": 0, |
| "fusion_vggt_layers": [ |
| 4, |
| 7, |
| 11, |
| 14, |
| 17, |
| 19, |
| 21, |
| 23 |
| ], |
| "fusion_dit_blocks": [ |
| 7, |
| 10, |
| 13, |
| 16, |
| 19, |
| 22, |
| 25, |
| 28 |
| ], |
| "cond_fusion_hidden_dim": 1024, |
| "dit_to_vggt_gate_cap": 1 |
| } |
| }, |
| "decoder": { |
| "normalization": null, |
| "checkpoint": { |
| "repo_id": "microsoft/TRELLIS-image-large", |
| "filename": "ckpts/ss_dec_conv3d_16l8_fp16", |
| "revision": "25e0d31ffbebe4b5a97464dd851910efc3002d96" |
| } |
| }, |
| "weights": { |
| "dit_base": { |
| "repo_id": "microsoft/TRELLIS.2-4B", |
| "filename": "ckpts/ss_flow_img_dit_1_3B_64_bf16.safetensors", |
| "revision": "af44b45f2e35a493886929c6d786e563ec68364d" |
| }, |
| "vggt_base": "uniforge3d/vggt/uniforge3d_stage1_structure_vggt_backbone.pt", |
| "dit_lora": { |
| "rank": 64, |
| "alpha": 128, |
| "target_modules": [ |
| "blocks.*.self_attn.to_qkv", |
| "blocks.*.self_attn.to_out", |
| "blocks.*.cross_attn.to_q", |
| "blocks.*.cross_attn.to_kv", |
| "blocks.*.cross_attn.to_out" |
| ] |
| }, |
| "vggt_lora": { |
| "rank": 32, |
| "alpha": 32, |
| "target_modules": [ |
| "aggregator.frame_blocks.*.attn.qkv", |
| "aggregator.frame_blocks.*.attn.proj", |
| "aggregator.frame_blocks.*.mlp.fc1", |
| "aggregator.frame_blocks.*.mlp.fc2", |
| "aggregator.global_blocks.*.attn.qkv", |
| "aggregator.global_blocks.*.attn.proj", |
| "aggregator.global_blocks.*.mlp.fc1", |
| "aggregator.global_blocks.*.mlp.fc2", |
| "point_head*" |
| ] |
| }, |
| "checkpoint": "uniforge3d/stage1/uniforge3d_stage1_structure_joint_model.pt" |
| } |
| }, |
| "structured_latent_512": { |
| "sampler": { |
| "name": "cfg", |
| "args": { |
| "sigma_min": 1e-05 |
| }, |
| "params": { |
| "steps": 20, |
| "guidance_strength": 3, |
| "guidance_rescale": 0, |
| "rescale_t": 1 |
| } |
| }, |
| "model": { |
| "name": "VGGTConditionedSLatFlowModel", |
| "args": { |
| "denoiser_args": { |
| "resolution": 32, |
| "in_channels": 32, |
| "out_channels": 32, |
| "model_channels": 1536, |
| "cond_channels": 1024, |
| "num_blocks": 30, |
| "num_heads": 12, |
| "mlp_ratio": 5.3334, |
| "pe_mode": "rope", |
| "share_mod": true, |
| "initialization": "scaled", |
| "qk_rms_norm": true, |
| "qk_rms_norm_cross": true, |
| "routed_layers": [ |
| 5, |
| 7, |
| 10, |
| 12, |
| 15, |
| 17, |
| 20, |
| 22, |
| 25, |
| 27 |
| ] |
| }, |
| "adapter_args": { |
| "feature_layers": [ |
| 4, |
| 11, |
| 17, |
| 23 |
| ], |
| "feature_dim": 2048, |
| "per_layer_dim": 1024, |
| "hidden_dim": 4096, |
| "output_dim": 1024 |
| } |
| } |
| }, |
| "decoder": { |
| "resolution": 512, |
| "normalization": { |
| "mean": [ |
| 0.781296, |
| 0.018091, |
| -0.495192, |
| -0.558457, |
| 1.06053, |
| 0.093252, |
| 1.518149, |
| -0.933218, |
| -0.732996, |
| 2.604095, |
| -0.118341, |
| -2.143904, |
| 0.495076, |
| -2.179512, |
| -2.130751, |
| -0.996944, |
| 0.261421, |
| -2.217463, |
| 1.260067, |
| -0.150213, |
| 3.790713, |
| 1.481266, |
| -1.046058, |
| -1.523667, |
| -0.059621, |
| 2.22078, |
| 1.621212, |
| 0.87723, |
| 0.567247, |
| -3.175944, |
| -3.186688, |
| 1.578665 |
| ], |
| "std": [ |
| 5.972266, |
| 4.706852, |
| 5.44501, |
| 5.209927, |
| 5.32022, |
| 4.547237, |
| 5.020802, |
| 5.444004, |
| 5.226681, |
| 5.683095, |
| 4.831436, |
| 5.286469, |
| 5.652043, |
| 5.367606, |
| 5.525084, |
| 4.730578, |
| 4.805265, |
| 5.124013, |
| 5.530808, |
| 5.619001, |
| 5.10393, |
| 5.41767, |
| 5.269677, |
| 5.547194, |
| 5.634698, |
| 5.235274, |
| 6.110351, |
| 5.511298, |
| 6.237273, |
| 4.879207, |
| 5.347008, |
| 5.405691 |
| ] |
| }, |
| "checkpoint": { |
| "repo_id": "microsoft/TRELLIS.2-4B", |
| "filename": "ckpts/shape_dec_next_dc_f16c32_fp16", |
| "revision": "af44b45f2e35a493886929c6d786e563ec68364d" |
| } |
| }, |
| "conditioning": { |
| "image_size": 512, |
| "grid_resolution": 32 |
| }, |
| "weights": { |
| "checkpoint": "uniforge3d/shape512/uniforge3d_stage2_shape512_denoiser.pt", |
| "adapter_checkpoint": "uniforge3d/shape512/uniforge3d_stage2_shape512_vggt_adapter.pt", |
| "lora": { |
| "r": 128, |
| "lora_alpha": 256, |
| "lora_dropout": 0, |
| "target_modules": [ |
| "to_q", |
| "to_kv", |
| "to_out", |
| "to_qkv" |
| ], |
| "bias": "none" |
| } |
| } |
| }, |
| "structured_latent_1024": { |
| "sampler": { |
| "name": "cfg", |
| "args": { |
| "sigma_min": 1e-05 |
| }, |
| "params": { |
| "steps": 20, |
| "guidance_strength": 6.5, |
| "guidance_rescale": 0.3, |
| "rescale_t": 1 |
| } |
| }, |
| "model": { |
| "name": "VGGTConditionedSLatFlowModel", |
| "args": { |
| "denoiser_args": { |
| "resolution": 64, |
| "in_channels": 32, |
| "out_channels": 32, |
| "model_channels": 1536, |
| "cond_channels": 1024, |
| "num_blocks": 30, |
| "num_heads": 12, |
| "mlp_ratio": 5.3334, |
| "pe_mode": "rope", |
| "share_mod": true, |
| "initialization": "scaled", |
| "qk_rms_norm": true, |
| "qk_rms_norm_cross": true, |
| "routed_layers": [ |
| 5, |
| 7, |
| 10, |
| 12, |
| 15, |
| 17, |
| 20, |
| 22, |
| 25, |
| 27 |
| ] |
| }, |
| "adapter_args": { |
| "feature_layers": [ |
| 4, |
| 11, |
| 17, |
| 23 |
| ], |
| "feature_dim": 2048, |
| "per_layer_dim": 1024, |
| "hidden_dim": 4096, |
| "output_dim": 1024, |
| "input_patch_grid_size": 32, |
| "upsample_scale": 2 |
| } |
| } |
| }, |
| "decoder": { |
| "resolution": 1024, |
| "normalization": { |
| "mean": [ |
| 0.781296, |
| 0.018091, |
| -0.495192, |
| -0.558457, |
| 1.06053, |
| 0.093252, |
| 1.518149, |
| -0.933218, |
| -0.732996, |
| 2.604095, |
| -0.118341, |
| -2.143904, |
| 0.495076, |
| -2.179512, |
| -2.130751, |
| -0.996944, |
| 0.261421, |
| -2.217463, |
| 1.260067, |
| -0.150213, |
| 3.790713, |
| 1.481266, |
| -1.046058, |
| -1.523667, |
| -0.059621, |
| 2.22078, |
| 1.621212, |
| 0.87723, |
| 0.567247, |
| -3.175944, |
| -3.186688, |
| 1.578665 |
| ], |
| "std": [ |
| 5.972266, |
| 4.706852, |
| 5.44501, |
| 5.209927, |
| 5.32022, |
| 4.547237, |
| 5.020802, |
| 5.444004, |
| 5.226681, |
| 5.683095, |
| 4.831436, |
| 5.286469, |
| 5.652043, |
| 5.367606, |
| 5.525084, |
| 4.730578, |
| 4.805265, |
| 5.124013, |
| 5.530808, |
| 5.619001, |
| 5.10393, |
| 5.41767, |
| 5.269677, |
| 5.547194, |
| 5.634698, |
| 5.235274, |
| 6.110351, |
| 5.511298, |
| 6.237273, |
| 4.879207, |
| 5.347008, |
| 5.405691 |
| ] |
| }, |
| "checkpoint": { |
| "repo_id": "microsoft/TRELLIS.2-4B", |
| "filename": "ckpts/shape_dec_next_dc_f16c32_fp16", |
| "revision": "af44b45f2e35a493886929c6d786e563ec68364d" |
| } |
| }, |
| "conditioning": { |
| "image_size": 1024, |
| "grid_resolution": 64 |
| }, |
| "weights": { |
| "checkpoint": "uniforge3d/shape1024/uniforge3d_stage2_shape1024_denoiser.pt", |
| "adapter_checkpoint": "uniforge3d/shape1024/uniforge3d_stage2_shape1024_vggt_adapter.pt", |
| "lora": { |
| "r": 128, |
| "lora_alpha": 256, |
| "lora_dropout": 0, |
| "target_modules": [ |
| "to_q", |
| "to_kv", |
| "to_out", |
| "to_qkv" |
| ], |
| "bias": "none" |
| } |
| } |
| }, |
| "structured_latent_handoff": { |
| "target_max_side": 0.9998, |
| "surface_voxel_resolution": 64 |
| }, |
| "mesh_postprocess": { |
| "grid_size": 1024, |
| "pre_simplify_target": 16777216, |
| "decimation_target": 1000000, |
| "remesh_band": 1, |
| "remesh_project": 0 |
| }, |
| "texture": { |
| "resolution": 1024, |
| "texture_size": 2048, |
| "decimation_target": 1000000, |
| "model": { |
| "name": "VGGTConditionedSLatFlowModel", |
| "args": { |
| "denoiser_args": { |
| "resolution": 64, |
| "in_channels": 64, |
| "out_channels": 32, |
| "model_channels": 1536, |
| "cond_channels": 1024, |
| "num_blocks": 30, |
| "num_heads": 12, |
| "mlp_ratio": 5.3334, |
| "pe_mode": "rope", |
| "share_mod": true, |
| "initialization": "scaled", |
| "qk_rms_norm": true, |
| "qk_rms_norm_cross": true, |
| "routed_layers": [ |
| 5, |
| 7, |
| 10, |
| 12, |
| 15, |
| 17, |
| 20, |
| 22, |
| 25, |
| 27 |
| ] |
| }, |
| "adapter_args": { |
| "feature_layers": [ |
| 4, |
| 11, |
| 17, |
| 23 |
| ], |
| "feature_dim": 2048, |
| "per_layer_dim": 1024, |
| "hidden_dim": 4096, |
| "output_dim": 1024, |
| "input_patch_grid_size": 32, |
| "upsample_scale": 2 |
| } |
| } |
| }, |
| "sampler": { |
| "args": { |
| "sigma_min": 1e-05 |
| }, |
| "params": { |
| "steps": 20, |
| "guidance_strength": 2.5, |
| "guidance_interval": [ |
| 0, |
| 1 |
| ], |
| "guidance_rescale": 0, |
| "rescale_t": 1.5 |
| } |
| }, |
| "shape_slat_normalization": { |
| "mean": [ |
| 0.781296, |
| 0.018091, |
| -0.495192, |
| -0.558457, |
| 1.06053, |
| 0.093252, |
| 1.518149, |
| -0.933218, |
| -0.732996, |
| 2.604095, |
| -0.118341, |
| -2.143904, |
| 0.495076, |
| -2.179512, |
| -2.130751, |
| -0.996944, |
| 0.261421, |
| -2.217463, |
| 1.260067, |
| -0.150213, |
| 3.790713, |
| 1.481266, |
| -1.046058, |
| -1.523667, |
| -0.059621, |
| 2.22078, |
| 1.621212, |
| 0.87723, |
| 0.567247, |
| -3.175944, |
| -3.186688, |
| 1.578665 |
| ], |
| "std": [ |
| 5.972266, |
| 4.706852, |
| 5.44501, |
| 5.209927, |
| 5.32022, |
| 4.547237, |
| 5.020802, |
| 5.444004, |
| 5.226681, |
| 5.683095, |
| 4.831436, |
| 5.286469, |
| 5.652043, |
| 5.367606, |
| 5.525084, |
| 4.730578, |
| 4.805265, |
| 5.124013, |
| 5.530808, |
| 5.619001, |
| 5.10393, |
| 5.41767, |
| 5.269677, |
| 5.547194, |
| 5.634698, |
| 5.235274, |
| 6.110351, |
| 5.511298, |
| 6.237273, |
| 4.879207, |
| 5.347008, |
| 5.405691 |
| ] |
| }, |
| "texture_slat_normalization": { |
| "mean": [ |
| 3.501659, |
| 2.212398, |
| 2.226094, |
| 0.251093, |
| -0.026248, |
| -0.687364, |
| 0.439898, |
| -0.928075, |
| 0.029398, |
| -0.339596, |
| -0.869527, |
| 1.038479, |
| -0.972385, |
| 0.126042, |
| -1.129303, |
| 0.455149, |
| -1.209521, |
| 2.069067, |
| 0.544735, |
| 2.569128, |
| -0.323407, |
| 2.293, |
| -1.925608, |
| -1.217717, |
| 1.213905, |
| 0.971588, |
| -0.023631, |
| 0.10675, |
| 2.021786, |
| 0.250524, |
| -0.662387, |
| -0.768862 |
| ], |
| "std": [ |
| 2.665652, |
| 2.743913, |
| 2.765121, |
| 2.595319, |
| 3.037293, |
| 2.291316, |
| 2.144656, |
| 2.911822, |
| 2.969419, |
| 2.501689, |
| 2.154811, |
| 3.163343, |
| 2.621215, |
| 2.381943, |
| 3.186697, |
| 3.021588, |
| 2.295916, |
| 3.234985, |
| 3.233086, |
| 2.26014, |
| 2.874801, |
| 2.810596, |
| 3.29272, |
| 2.674999, |
| 2.680878, |
| 2.372054, |
| 2.451546, |
| 2.353556, |
| 2.995195, |
| 2.379849, |
| 2.786195, |
| 2.77519 |
| ] |
| }, |
| "shape_decoder_checkpoint": { |
| "repo_id": "microsoft/TRELLIS.2-4B", |
| "filename": "ckpts/shape_dec_next_dc_f16c32_fp16", |
| "revision": "af44b45f2e35a493886929c6d786e563ec68364d" |
| }, |
| "texture_decoder_checkpoint": { |
| "repo_id": "microsoft/TRELLIS.2-4B", |
| "filename": "ckpts/tex_dec_next_dc_f16c32_fp16", |
| "revision": "af44b45f2e35a493886929c6d786e563ec68364d" |
| }, |
| "weights": { |
| "checkpoint": "uniforge3d/texture/uniforge3d_stage3_texture1024_denoiser.pt", |
| "adapter_checkpoint": "uniforge3d/texture/uniforge3d_stage3_texture1024_vggt_adapter.pt", |
| "lora": { |
| "r": 128, |
| "lora_alpha": 256, |
| "lora_dropout": 0, |
| "target_modules": [ |
| "to_q", |
| "to_kv", |
| "to_out", |
| "to_qkv" |
| ], |
| "bias": "none" |
| } |
| } |
| }, |
| "image_preprocess": { |
| "alpha_threshold": 0, |
| "padding_factor": 1, |
| "rembg_model": "briaai/RMBG-2.0" |
| }, |
| "format_version": 1, |
| "default_seed": 42, |
| "conditioner": { |
| "vggt_feature_extractor": { |
| "feature_layers": [ |
| 4, |
| 11, |
| 17, |
| 23 |
| ], |
| "shape_grid_resolution": 64, |
| "conditioning_layers": [ |
| 4, |
| 7, |
| 11, |
| 14, |
| 17, |
| 20, |
| 23 |
| ], |
| "dtype": "bfloat16", |
| "checkpoint_path": "uniforge3d/vggt/uniforge3d_shape_texture_vggt_conditioner.pt" |
| }, |
| "voxel_routers": { |
| "32": { |
| "confidence_keep_ratio": 0.95, |
| "component_iterations": 64, |
| "grid_resolution": 32, |
| "patch_grid_size": 32, |
| "image_size": 448 |
| }, |
| "64": { |
| "confidence_keep_ratio": 0.95, |
| "component_iterations": 64, |
| "grid_resolution": 64, |
| "patch_grid_size": 64, |
| "image_size": 448 |
| } |
| } |
| } |
| } |
|
|
