Release new model, fix bugs and warnings (#52)

* Release temporal Maskgit of LiDAR generation
* Fix video streaming generation due to the code change of temporal VAE
* Fix security warning of transformers

Author: wzhgba

.gitattributes

@@ -91,8 +91,8 @@ You can download our pre-trained tokenzier and generation model in the following
 | | nuscene, waymo, argoverse, kitti360 | [Config](configs/lidar/lidar_vqvae_nwak.json) | [checkpoint](http://103.237.29.236:10030/lidar_vqvae_nwak_80k.pth), [blank code](http://103.237.29.236:10030/lidar_vqvae_nwak_80k_blank_code.pkl) |
 | MaskGIT | nuscene | [Config](configs/lidar/lidar_maskgit_layout_ns.json) | [ckpt_with_vqvae_nwa](http://103.237.29.236:10030/lidar_maskgit_nusc_150k.pth) <br> [ckpt_with_vqvae_nwak](http://103.237.29.236:10030/lidar_maskgit_vq80k_layout_ns_120k.pth) |
 | | kitti360 | [Config](configs/lidar/lidar_maskgit_vq80k_layout_kt.json) | [checkpoint](http://103.237.29.236:10030/lidar_maskgit_vq80k_layout_kt_120k.pth)|
-| Temporal MaskGIT | nuscene | [Config](configs/lidar/lidar_maskgit_temporal_vq80k_layout_ns.json) | checkpoint(TODO) |
-| | kitti360 | [Config](configs/lidar/lidar_maskgit_temporal_vq80k_layout_kt.json) | checkpoint(TODO)|
+| Temporal MaskGIT | nuscene | [Config](configs/lidar/lidar_maskgit_temporal_vq80k_layout_ns.json) | [checkpoint](http://103.237.29.236:10030/lidar_maskgit_temporal_vq80k_layout_kt_150k.pth) |
+| | kitti360 | [Config](configs/lidar/lidar_maskgit_temporal_vq80k_layout_kt.json) | [checkpoint](http://103.237.29.236:10030/lidar_maskgit_temporal_vq80k_layout_ns_150k.pth)|
 ## Examples
 
 ### T2I, T2V generation with CTSD pipeline

README.md

@@ -622,6 +622,13 @@
         "persistent_workers": true
     },
     "informations": {
-        "todo": "todo"
+        "voxel_iou": 0.11702190,
+        "chamfer_distance": 3.34721875,
+        "chamfer_distance_30": 1.88547862,
+        "chamfer_distance_40": 2.57636571,
+        "mmd": 0.00411365,
+        "jsd": 0.31335828,
+        "total_batch_sizes": 24,
+        "steps": 150000
     }
 }

configs/lidar/lidar_maskgit_temporal_vq80k_layout_kt.json

@@ -786,6 +786,11 @@
         "persistent_workers": true
     },
     "informations": {
-        "todo": "todo"
+        "voxel_iou": 0.12602140,
+        "chamfer_distance": 3.48719001,
+        "chamfer_distance_30": 1.59294820,
+        "chamfer_distance_40": 2.51212931,
+        "total_batch_sizes": 24,
+        "steps": 150000
     }
 }

configs/lidar/lidar_maskgit_temporal_vq80k_layout_ns.json

@@ -77,28 +77,28 @@
             "bias_init": -5.0
         },
         "bi_directional_Transformer": {
-                "_class_name": "dwm.models.maskgit_model.BidirectionalTransformerWithAdapter",
-                "use_extra_embedding": true,
-                "n_e": 2048,
-                "e_dim": 1024,
-                "hidden_dim": 1280,
-                "img_size": [
-                    80,
-                    80
-                ],
-                "num_heads": 40,
-                "use_checkpoint": true,
-                "depth": 24,
-                "condition_adapter": {
-                    "_class_name": "dwm.models.adapters.ImageAdapter",
-                    "in_channels": 6,
-                    "channels": [1280, 1280, 1280, 1280, 1280, 1280, 1280, 1280],
-                    "downscale_factor": 8,
-                    "is_downblocks": [false, false, false, false, false, false, false, false],
-                    "use_zero_convs": true,
-                    "gradient_checkpointing": true
-                }
-            },
+            "_class_name": "dwm.models.maskgit_model.BidirectionalTransformerWithAdapter",
+            "use_extra_embedding": true,
+            "n_e": 2048,
+            "e_dim": 1024,
+            "hidden_dim": 1280,
+            "img_size": [
+                80,
+                80
+            ],
+            "num_heads": 40,
+            "use_checkpoint": true,
+            "depth": 24,
+            "condition_adapter": {
+                "_class_name": "dwm.models.adapters.ImageAdapter",
+                "in_channels": 6,
+                "channels": [1280, 1280, 1280, 1280, 1280, 1280, 1280, 1280],
+                "downscale_factor": 8,
+                "is_downblocks": [false, false, false, false, false, false, false, false],
+                "use_zero_convs": true,
+                "gradient_checkpointing": true
+            }
+        },
         "vq_point_cloud_ckpt_path": "/cache/aoss-v2.st-sh-01.sensecoreapi-oss.cn/users/wuzehuan/workspaces/worldmodels/models/lidar_vqvae_nwa_60k.pth",
         "vq_blank_code_path": "/cache/aoss-v2.st-sh-01.sensecoreapi-oss.cn/users/wuzehuan/workspaces/worldmodels/models/lidar_vqvae_nwa_60k_blank_code.pkl",
         "model_ckpt_path": "/mnt/storage/user/liuyichen/tasks/lidar_maskgit_nusc_full-align-line/checkpoints/150k.pth",