Added pretrained model

config.json

@@ -0,0 +1,220 @@
+{
+  "_name_or_path": "/workspace/multiformer_m0",
+  "activation_dropout": 0.0,
+  "activation_function": "relu",
+  "architectures": [
+    "Multiformer"
+  ],
+  "attention_dropout": 0.0,
+  "auxiliary_loss": false,
+  "backbone": "pvt_v2",
+  "backbone_config": {
+    "_out_features": [
+      "stage1",
+      "stage2",
+      "stage3",
+      "stage4"
+    ],
+    "_out_indices": [
+      0,
+      1,
+      2,
+      3
+    ],
+    "do_reduce_labels": true,
+    "id2label": {
+      "0": "unlabeled",
+      "1": "building",
+      "2": "fence",
+      "3": "other",
+      "4": "pedestrian",
+      "5": "pole",
+      "6": "road line",
+      "7": "road",
+      "8": "sidewalk",
+      "9": "vegetation",
+      "10": "vehicle",
+      "11": "wall",
+      "12": "traffic sign",
+      "13": "sky",
+      "14": "ground",
+      "15": "bridge",
+      "16": "rail track",
+      "17": "guard rail",
+      "18": "traffic light",
+      "19": "static",
+      "20": "dynamic",
+      "21": "water",
+      "22": "terrain"
+    },
+    "label2id": {
+      "bridge": 15,
+      "building": 1,
+      "dynamic": 20,
+      "fence": 2,
+      "ground": 14,
+      "guard rail": 17,
+      "other": 3,
+      "pedestrian": 4,
+      "pole": 5,
+      "rail track": 16,
+      "road": 7,
+      "road line": 6,
+      "sidewalk": 8,
+      "sky": 13,
+      "static": 19,
+      "terrain": 22,
+      "traffic light": 18,
+      "traffic sign": 12,
+      "unlabeled": 0,
+      "vegetation": 9,
+      "vehicle": 10,
+      "wall": 11,
+      "water": 21
+    },
+    "mlp_ratios": [
+      4,
+      4,
+      4,
+      4
+    ],
+    "model_type": "pvt_v2",
+    "output_hidden_states": true
+  },
+  "bbox_cost": 5,
+  "bbox_loss_coefficient": 5,
+  "class_cost": 1,
+  "d_model": 256,
+  "decoder_attention_heads": 8,
+  "decoder_ffn_dim": 256,
+  "decoder_layers": 3,
+  "decoder_n_points": 4,
+  "depth_decoder_dim": 64,
+  "det2d_box_keep_prob": 0.35,
+  "det2d_extra_feature_levels": 1,
+  "det2d_fuse_depth": false,
+  "det2d_fuse_semantic": false,
+  "det2d_input_feature_levels": [
+    1,
+    2,
+    3
+  ],
+  "det2d_input_proj_groups": 32,
+  "det2d_input_proj_kernels": [
+    1,
+    1,
+    1
+  ],
+  "det2d_input_proj_pads": [
+    0,
+    0,
+    0
+  ],
+  "det2d_input_proj_strides": [
+    1,
+    1,
+    1
+  ],
+  "det2d_use_pos_embed": true,
+  "det3d_num_heading_bins": 12,
+  "det3d_predict_class": false,
+  "det3d_type_mean_sizes": {
+    "bicycle": [
+      1.3252513842854072,
+      0.5087623590366082,
+      1.6023777324693749
+    ],
+    "bus": [
+      2.012248454530893,
+      1.8628641001034811,
+      4.327287051409026
+    ],
+    "car": [
+      1.6351819382478878,
+      1.8992701184837708,
+      4.471483058450479
+    ],
+    "motorcycle": [
+      1.2625147692222876,
+      0.6884931937235099,
+      2.030941778814124
+    ],
+    "pedestrian": [
+      1.773017496887137,
+      0.38524578907445944,
+      0.38524578907445944
+    ],
+    "truck": [
+      2.774033790320722,
+      2.507875293310309,
+      6.471412330091904
+    ]
+  },
+  "dice_loss_coefficient": 1,
+  "dilation": false,
+  "disable_custom_kernels": false,
+  "dropout": 0.1,
+  "encoder_attention_heads": 8,
+  "encoder_ffn_dim": 256,
+  "encoder_layerdrop": 0.0,
+  "encoder_layers": 3,
+  "encoder_n_points": 4,
+  "eos_coefficient": 0.1,
+  "focal_alpha": 0.25,
+  "frozen_batch_norm": false,
+  "giou_cost": 2,
+  "giou_loss_coefficient": 2,
+  "head_in_index": -1,
+  "id2label": {
+    "0": "pedestrian",
+    "1": "car",
+    "2": "truck",
+    "3": "bus",
+    "4": "motorcycle",
+    "5": "bicycle"
+  },
+  "init_std": 0.02,
+  "init_xavier_std": 1.0,
+  "is_encoder_decoder": true,
+  "label2id": {
+    "bicycle": 5,
+    "bus": 3,
+    "car": 1,
+    "motorcycle": 4,
+    "pedestrian": 0,
+    "truck": 2
+  },
+  "mask_loss_coefficient": 1,
+  "max_depth": 10,
+  "max_position_embeddings": 1024,
+  "model_type": "multiformer",
+  "num_channels": 3,
+  "num_feature_levels": 4,
+  "num_queries": 300,
+  "omit_heads": [
+    "det_3d"
+  ],
+  "position_embedding_type": "sine",
+  "semantic_classifier_dropout": 0.1,
+  "semantic_decoder_dim": 256,
+  "semantic_loss_ignore_index": 255,
+  "silog_lambda": 0.25,
+  "tasks": [
+    "semseg",
+    "depth",
+    "det_2d"
+  ],
+  "torch_dtype": "float32",
+  "train_backbone": true,
+  "train_tasks": [
+    "semseg",
+    "depth",
+    "det_2d"
+  ],
+  "transformers_version": "4.36.0.dev0",
+  "two_stage": false,
+  "two_stage_num_proposals": 300,
+  "use_pretrained_backbone": true,
+  "use_timm_backbone": false,
+  "with_box_refine": false
+}

model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5f0978f2e67d0faa0be82ca1038b219d84d03a3606272dcfa48d2f9f9a6ede75
+size 32212096