Add files using upload-large-folder tool

light_scannet/scene0630_03.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8d64e10f01f8951b9d25f2158385a0436d84ab88fdf13106b192b9f63d15a46b
+size 169390352

results/sceneverse_scannet_exp1_b48_Pretrain_all_scannet_training_run1/b48_contrastive_best_caption/ckpt/ckpt_100.pth/custom_checkpoint_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:405ffe7614188abeeb34cf689548a654410844e0be3e49d003451baa8036b975
+size 1040

results/sceneverse_scannet_exp1_b48_Pretrain_all_scannet_training_run1/b48_contrastive_best_caption/ckpt/ckpt_100.pth/model_1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60d95b10b6e140a9626a7058d5038528f2ff80148dc4569b881db56052046509
+size 40

results/sceneverse_scannet_exp1_b48_Pretrain_all_scannet_training_run1/b48_contrastive_best_caption/ckpt/ckpt_100.pth/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4877f9848632672910825ef000880ece4fb5d3b10f690be6369790296461b063
+size 15124

results/sceneverse_scannet_exp1_b48_Pretrain_all_scannet_training_run1/b48_contrastive_best_caption/ckpt/ckpt_100.pth/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:59ab72209a69cb206ee8d1eb30cd64613735aff7450e766f00c4103bbd00a487
+size 1064

results/sceneverse_scannet_exp1_b48_Pretrain_all_scannet_training_run1/b48_contrastive_best_caption/config.yaml

@@ -0,0 +1,338 @@
+name: sceneverse_scannet_exp1
+rng_seed: 42
+num_gpu: 4
+mode: train
+note: scannet_training_run1
+naming_keywords:
+- dataloader.batchsize
+- task
+- note
+- time
+base_dir: results
+exp_dir: !!python/object/apply:pathlib.PosixPath
+- results
+- sceneverse_scannet_exp1_b48_Pretrain_all_scannet_training_run1
+- 2025-06-24-19:53:19.624883
+save_frequency: 10
+resume: false
+debug:
+  flag: false
+  debug_size: 20
+  hard_debug: false
+logger:
+  name: wandb
+  entity: yem
+  project: 3DReason
+  run_id: okngyypm
+data:
+  note: all
+  train:
+  - ScanNetSpatialRefer
+  val:
+  - ScanNetSpatialRefer
+  test:
+  - ScanNetSpatialRefer
+  args:
+    max_obj_len: 80
+    max_seq_len: 50
+    num_points: 1024
+    pc_type: pred
+    sem_type: '607'
+    filter_lang: false
+    txt_mask_ratio: 0.15
+    pc_mask_ratio: 0.1
+    rot_aug: true
+    mask_strategy: random
+    use_scene_cap: true
+    max_scene_cap_len: 300
+  ScanNetSpatialRefer:
+    train:
+      sources:
+      - view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        - nr3d
+        sr3d_plus_aug: true
+      sgrefer:
+        anno_type:
+        - rel2_gpt
+        - rel2_template
+        - relm_gpt
+        - relm_template
+        - star_gpt
+        - star_template
+      sgcaption:
+        anno_type:
+        - gpt
+        - template
+    val:
+      sources:
+      - view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+    test:
+      sources:
+      - view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+  RScanSpatialRefer:
+    train:
+      sources:
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+    test:
+      sources:
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+  MultiScanSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+  ARKitSceneSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+  HMSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+  use_voxel: false
+  scan_family_base: existing_datasets/ScanNet
+  rscan_base: existing_datasets/3RScan
+  arkitscene_base: existing_datasets/ARKitScenes
+  multiscan_base: existing_datasets/MultiScan
+  hm_base: existing_datasets/HM3D
+  procthor_base: existing_datasets/ProcThor
+  s3d_base: existing_datasets/Structured3D'
+data_aug:
+  aug_list:
+  - scene_aug
+  scene_aug:
+    translation:
+      enabled: false
+      value:
+      - 1.0
+      - 1.0
+      - 1.0
+      p: 1.0
+    scaling:
+      enabled: false
+      p: 1.0
+      value:
+      - 0.9
+      - 1.1
+    flip:
+      enabled: false
+      p: 0.5
+    rotation:
+      enabled: true
+      p: 1.0
+      axis_align: true
+      value:
+      - 0.0
+      - 0.0
+      - 1.0
+      shuffle: true
+    color_jitter: false
+    order_shuffle: false
+  obj_aug:
+    translation:
+      enabled: false
+      value:
+      - 0.1
+      - 0.1
+      - 0.1
+      p: 1.0
+    rotation:
+      enabled: false
+      p: 1.0
+      axis_align: false
+      value:
+      - 0.0
+      - 0.0
+      - 0.1
+      shuffle: true
+    random_jitter:
+      enabled: false
+      value: 0.01
+      accord_to_size: false
+      p: 1.0
+    pts_shuffle: true
+task: Pretrain
+data_wrapper:
+  train: SceneDatasetWrapper
+  val: SceneDatasetWrapper
+  test: SceneDatasetWrapper
+trainer: OpenVocabTrainer
+ckpt_path: ''
+pretrain_ckpt_path: ''
+dataloader:
+  batchsize: 12
+  num_workers: 4
+  balance_dataset: false
+  filter_empty_annotations: false
+solver:
+  gradient_accumulation_steps: 1
+  epochs_per_save: 20
+  epochs_per_eval: 1
+  lr: 0.0005
+  grad_norm: 5.0
+  epochs: 150
+  optim:
+    name: AdamW
+    args:
+      betas:
+      - 0.9
+      - 0.98
+  sched:
+    name: warmup_cosine
+    args:
+      warmup_steps: 500
+      minimum_ratio: 0.1
+eval:
+  train:
+    name: PretrainEval
+  val:
+    name: ScanReferEval
+  save: false
+model:
+  name: OpenVocab
+  language:
+    name: SigLIPLanguageEncoder
+    args:
+      weights: google/siglip-base-patch16-224
+    lr: 1.0e-05
+  vision:
+    name: PointOpenVocabEncoder
+    args:
+      backbone: pointnet++
+      hidden_size: 768
+      freeze: true
+      path: pretrained_weights/pointnetpp-open-bert
+      num_attention_heads: 12
+      spatial_dim: 5
+      num_layers: 4
+      dim_loc: 6
+      dim_feedforward: 2048
+      attn_type: spatial
+      pairwise_rel_type: center
+      use_matmul_label: false
+      lang_type: bert
+      lang_path: pretrained_weights/607_text_embeddings
+    lr: 0.0001
+  grounding:
+    name: UnifiedSpatialCrossEncoderV2
+    args:
+      hidden_size: 768
+      num_attention_heads: 12
+      num_layers: 4
+      dim_feedforward: 2048
+      dim_loc: 6
+    lr: 0.0001
+  inter: before
+  heads:
+    head_list:
+    - pretrain_head
+    pretrain_head:
+      name: OVPretrainHead
+      args:
+        hidden_size: 768
+        vocab_size: 30522
+  loss_type: ListLoss
+  loss_list:
+  - SceneViewPM_loss
+  vis_loss_list:
+  - SceneViewPM_loss

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.100232/config.yaml

@@ -0,0 +1,318 @@
+name: sceneverse_scannet_exp1
+rng_seed: 42
+num_gpu: 8
+mode: train
+note: scannet_training_run1
+naming_keywords:
+- dataloader.batchsize
+- task
+- note
+- time
+base_dir: results
+exp_dir: !!python/object/apply:pathlib.PosixPath
+- results
+- sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1
+- 2025-07-07-09:21:27.100232
+save_frequency: 10
+resume: false
+is_pretrain: true
+debug:
+  flag: false
+  debug_size: 20
+  hard_debug: false
+logger:
+  name: wandb
+  entity: yem
+  project: 3DReason
+  run_id: b3xyp9hh
+data:
+  note: all
+  train:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  val:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  test:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  args:
+    max_obj_len: 80
+    max_seq_len: 50
+    num_points: 1024
+    pc_type: pred
+    sem_type: '607'
+    filter_lang: false
+    txt_mask_ratio: 0.15
+    pc_mask_ratio: 0.1
+    rot_aug: true
+    mask_strategy: random
+    use_scene_cap: true
+    max_scene_cap_len: 300
+  ScanNetSpatialRefer:
+    train:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        - nr3d
+        sr3d_plus_aug: true
+      sgrefer:
+        anno_type:
+        - rel2_gpt
+        - rel2_template
+        - relm_gpt
+        - relm_template
+        - star_gpt
+        - star_template
+      sgcaption:
+        anno_type:
+        - gpt
+        - template
+    val:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+    test:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+  RScanSpatialRefer:
+    train:
+      sources:
+      - 3rscan_view_cap
+    val:
+      sources:
+      - 3rscan_view_cap
+    test:
+      sources:
+      - 3rscan_view_cap
+  MultiScanSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+  ARKitSceneSpatialRefer:
+    train:
+      sources:
+      - arkitscenes_view_cap
+    val:
+      sources:
+      - arkitscenes_view_cap
+    test:
+      sources:
+      - arkitscenes_view_cap
+  HMSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+  use_voxel: false
+  scan_family_base: existing_datasets/ScanNet
+  rscan_base: existing_datasets/3RScan
+  arkitscene_base: existing_datasets/Arkitscenes
+  multiscan_base: existing_datasets/MultiScan
+  hm_base: existing_datasets/HM3D
+  procthor_base: existing_datasets/ProcThor
+  s3d_base: existing_datasets/Structured3D'
+data_aug:
+  aug_list:
+  - scene_aug
+  scene_aug:
+    translation:
+      enabled: false
+      value:
+      - 1.0
+      - 1.0
+      - 1.0
+      p: 1.0
+    scaling:
+      enabled: false
+      p: 1.0
+      value:
+      - 0.9
+      - 1.1
+    flip:
+      enabled: false
+      p: 0.5
+    rotation:
+      enabled: true
+      p: 1.0
+      axis_align: true
+      value:
+      - 0.0
+      - 0.0
+      - 1.0
+      shuffle: true
+    color_jitter: false
+    order_shuffle: false
+  obj_aug:
+    translation:
+      enabled: false
+      value:
+      - 0.1
+      - 0.1
+      - 0.1
+      p: 1.0
+    rotation:
+      enabled: false
+      p: 1.0
+      axis_align: false
+      value:
+      - 0.0
+      - 0.0
+      - 0.1
+      shuffle: true
+    random_jitter:
+      enabled: false
+      value: 0.01
+      accord_to_size: false
+      p: 1.0
+    pts_shuffle: true
+task: Pretrain
+data_wrapper:
+  train: SceneDatasetWrapper
+  val: SceneDatasetWrapper
+  test: SceneDatasetWrapper
+trainer: OpenVocabTrainer
+ckpt_path: ''
+pretrain_ckpt_path: ''
+dataloader:
+  batchsize: 8
+  num_workers: 2
+  balance_dataset: false
+  filter_empty_annotations: false
+solver:
+  gradient_accumulation_steps: 1
+  epochs_per_save: 20
+  epochs_per_eval: 1
+  lr: 0.0005
+  grad_norm: 5.0
+  epochs: 300
+  optim:
+    name: AdamW
+    args:
+      betas:
+      - 0.9
+      - 0.98
+  sched:
+    name: warmup_cosine
+    args:
+      warmup_steps: 500
+      minimum_ratio: 0.1
+eval:
+  train:
+    name: PretrainEval
+  val:
+    name: ScanReferEval
+  save: false
+model:
+  name: OpenVocab
+  language:
+    name: SigLIPLanguageEncoder
+    args:
+      weights: google/siglip-base-patch16-224
+    lr: 1.0e-05
+  vision:
+    name: qihoo360/fg-clip-base
+    args:
+      backbone: pointnet++
+      hidden_size: 768
+      freeze: true
+      path: pretrained_weights/pointnetpp-open-bert
+      num_attention_heads: 12
+      spatial_dim: 5
+      num_layers: 4
+      dim_loc: 6
+      dim_feedforward: 2048
+      attn_type: spatial
+      pairwise_rel_type: center
+      use_matmul_label: false
+      lang_type: bert
+      lang_path: pretrained_weights/607_text_embeddings
+    lr: 0.0001
+  grounding:
+    name: UnifiedSpatialCrossEncoderV2
+    args:
+      hidden_size: 768
+      num_attention_heads: 12
+      num_layers: 4
+      dim_feedforward: 2048
+      dim_loc: 6
+    lr: 0.0001
+  inter: before
+  heads:
+    head_list:
+    - pretrain_head
+    pretrain_head:
+      name: OVPretrainHead
+      args:
+        hidden_size: 768
+        vocab_size: 30522
+  loss_type: ListLoss
+  loss_list:
+  - SceneViewPM_loss
+  vis_loss_list:
+  - SceneViewPM_loss

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.107655/config.yaml

@@ -0,0 +1,318 @@
+name: sceneverse_scannet_exp1
+rng_seed: 42
+num_gpu: 8
+mode: train
+note: scannet_training_run1
+naming_keywords:
+- dataloader.batchsize
+- task
+- note
+- time
+base_dir: results
+exp_dir: !!python/object/apply:pathlib.PosixPath
+- results
+- sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1
+- 2025-07-07-09:21:27.107655
+save_frequency: 10
+resume: false
+is_pretrain: true
+debug:
+  flag: false
+  debug_size: 20
+  hard_debug: false
+logger:
+  name: wandb
+  entity: yem
+  project: 3DReason
+  run_id: avoo7irx
+data:
+  note: all
+  train:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  val:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  test:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  args:
+    max_obj_len: 80
+    max_seq_len: 50
+    num_points: 1024
+    pc_type: pred
+    sem_type: '607'
+    filter_lang: false
+    txt_mask_ratio: 0.15
+    pc_mask_ratio: 0.1
+    rot_aug: true
+    mask_strategy: random
+    use_scene_cap: true
+    max_scene_cap_len: 300
+  ScanNetSpatialRefer:
+    train:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        - nr3d
+        sr3d_plus_aug: true
+      sgrefer:
+        anno_type:
+        - rel2_gpt
+        - rel2_template
+        - relm_gpt
+        - relm_template
+        - star_gpt
+        - star_template
+      sgcaption:
+        anno_type:
+        - gpt
+        - template
+    val:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+    test:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+  RScanSpatialRefer:
+    train:
+      sources:
+      - 3rscan_view_cap
+    val:
+      sources:
+      - 3rscan_view_cap
+    test:
+      sources:
+      - 3rscan_view_cap
+  MultiScanSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+  ARKitSceneSpatialRefer:
+    train:
+      sources:
+      - arkitscenes_view_cap
+    val:
+      sources:
+      - arkitscenes_view_cap
+    test:
+      sources:
+      - arkitscenes_view_cap
+  HMSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+  use_voxel: false
+  scan_family_base: existing_datasets/ScanNet
+  rscan_base: existing_datasets/3RScan
+  arkitscene_base: existing_datasets/Arkitscenes
+  multiscan_base: existing_datasets/MultiScan
+  hm_base: existing_datasets/HM3D
+  procthor_base: existing_datasets/ProcThor
+  s3d_base: existing_datasets/Structured3D'
+data_aug:
+  aug_list:
+  - scene_aug
+  scene_aug:
+    translation:
+      enabled: false
+      value:
+      - 1.0
+      - 1.0
+      - 1.0
+      p: 1.0
+    scaling:
+      enabled: false
+      p: 1.0
+      value:
+      - 0.9
+      - 1.1
+    flip:
+      enabled: false
+      p: 0.5
+    rotation:
+      enabled: true
+      p: 1.0
+      axis_align: true
+      value:
+      - 0.0
+      - 0.0
+      - 1.0
+      shuffle: true
+    color_jitter: false
+    order_shuffle: false
+  obj_aug:
+    translation:
+      enabled: false
+      value:
+      - 0.1
+      - 0.1
+      - 0.1
+      p: 1.0
+    rotation:
+      enabled: false
+      p: 1.0
+      axis_align: false
+      value:
+      - 0.0
+      - 0.0
+      - 0.1
+      shuffle: true
+    random_jitter:
+      enabled: false
+      value: 0.01
+      accord_to_size: false
+      p: 1.0
+    pts_shuffle: true
+task: Pretrain
+data_wrapper:
+  train: SceneDatasetWrapper
+  val: SceneDatasetWrapper
+  test: SceneDatasetWrapper
+trainer: OpenVocabTrainer
+ckpt_path: ''
+pretrain_ckpt_path: ''
+dataloader:
+  batchsize: 8
+  num_workers: 2
+  balance_dataset: false
+  filter_empty_annotations: false
+solver:
+  gradient_accumulation_steps: 1
+  epochs_per_save: 20
+  epochs_per_eval: 1
+  lr: 0.0005
+  grad_norm: 5.0
+  epochs: 300
+  optim:
+    name: AdamW
+    args:
+      betas:
+      - 0.9
+      - 0.98
+  sched:
+    name: warmup_cosine
+    args:
+      warmup_steps: 500
+      minimum_ratio: 0.1
+eval:
+  train:
+    name: PretrainEval
+  val:
+    name: ScanReferEval
+  save: false
+model:
+  name: OpenVocab
+  language:
+    name: SigLIPLanguageEncoder
+    args:
+      weights: google/siglip-base-patch16-224
+    lr: 1.0e-05
+  vision:
+    name: qihoo360/fg-clip-base
+    args:
+      backbone: pointnet++
+      hidden_size: 768
+      freeze: true
+      path: pretrained_weights/pointnetpp-open-bert
+      num_attention_heads: 12
+      spatial_dim: 5
+      num_layers: 4
+      dim_loc: 6
+      dim_feedforward: 2048
+      attn_type: spatial
+      pairwise_rel_type: center
+      use_matmul_label: false
+      lang_type: bert
+      lang_path: pretrained_weights/607_text_embeddings
+    lr: 0.0001
+  grounding:
+    name: UnifiedSpatialCrossEncoderV2
+    args:
+      hidden_size: 768
+      num_attention_heads: 12
+      num_layers: 4
+      dim_feedforward: 2048
+      dim_loc: 6
+    lr: 0.0001
+  inter: before
+  heads:
+    head_list:
+    - pretrain_head
+    pretrain_head:
+      name: OVPretrainHead
+      args:
+        hidden_size: 768
+        vocab_size: 30522
+  loss_type: ListLoss
+  loss_list:
+  - SceneViewPM_loss
+  vis_loss_list:
+  - SceneViewPM_loss

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.120351/config.yaml

@@ -0,0 +1,318 @@
+name: sceneverse_scannet_exp1
+rng_seed: 42
+num_gpu: 8
+mode: train
+note: scannet_training_run1
+naming_keywords:
+- dataloader.batchsize
+- task
+- note
+- time
+base_dir: results
+exp_dir: !!python/object/apply:pathlib.PosixPath
+- results
+- sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1
+- 2025-07-07-09:21:27.120351
+save_frequency: 10
+resume: false
+is_pretrain: true
+debug:
+  flag: false
+  debug_size: 20
+  hard_debug: false
+logger:
+  name: wandb
+  entity: yem
+  project: 3DReason
+  run_id: f2blnuci
+data:
+  note: all
+  train:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  val:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  test:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  args:
+    max_obj_len: 80
+    max_seq_len: 50
+    num_points: 1024
+    pc_type: pred
+    sem_type: '607'
+    filter_lang: false
+    txt_mask_ratio: 0.15
+    pc_mask_ratio: 0.1
+    rot_aug: true
+    mask_strategy: random
+    use_scene_cap: true
+    max_scene_cap_len: 300
+  ScanNetSpatialRefer:
+    train:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        - nr3d
+        sr3d_plus_aug: true
+      sgrefer:
+        anno_type:
+        - rel2_gpt
+        - rel2_template
+        - relm_gpt
+        - relm_template
+        - star_gpt
+        - star_template
+      sgcaption:
+        anno_type:
+        - gpt
+        - template
+    val:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+    test:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+  RScanSpatialRefer:
+    train:
+      sources:
+      - 3rscan_view_cap
+    val:
+      sources:
+      - 3rscan_view_cap
+    test:
+      sources:
+      - 3rscan_view_cap
+  MultiScanSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+  ARKitSceneSpatialRefer:
+    train:
+      sources:
+      - arkitscenes_view_cap
+    val:
+      sources:
+      - arkitscenes_view_cap
+    test:
+      sources:
+      - arkitscenes_view_cap
+  HMSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+  use_voxel: false
+  scan_family_base: existing_datasets/ScanNet
+  rscan_base: existing_datasets/3RScan
+  arkitscene_base: existing_datasets/Arkitscenes
+  multiscan_base: existing_datasets/MultiScan
+  hm_base: existing_datasets/HM3D
+  procthor_base: existing_datasets/ProcThor
+  s3d_base: existing_datasets/Structured3D'
+data_aug:
+  aug_list:
+  - scene_aug
+  scene_aug:
+    translation:
+      enabled: false
+      value:
+      - 1.0
+      - 1.0
+      - 1.0
+      p: 1.0
+    scaling:
+      enabled: false
+      p: 1.0
+      value:
+      - 0.9
+      - 1.1
+    flip:
+      enabled: false
+      p: 0.5
+    rotation:
+      enabled: true
+      p: 1.0
+      axis_align: true
+      value:
+      - 0.0
+      - 0.0
+      - 1.0
+      shuffle: true
+    color_jitter: false
+    order_shuffle: false
+  obj_aug:
+    translation:
+      enabled: false
+      value:
+      - 0.1
+      - 0.1
+      - 0.1
+      p: 1.0
+    rotation:
+      enabled: false
+      p: 1.0
+      axis_align: false
+      value:
+      - 0.0
+      - 0.0
+      - 0.1
+      shuffle: true
+    random_jitter:
+      enabled: false
+      value: 0.01
+      accord_to_size: false
+      p: 1.0
+    pts_shuffle: true
+task: Pretrain
+data_wrapper:
+  train: SceneDatasetWrapper
+  val: SceneDatasetWrapper
+  test: SceneDatasetWrapper
+trainer: OpenVocabTrainer
+ckpt_path: ''
+pretrain_ckpt_path: ''
+dataloader:
+  batchsize: 8
+  num_workers: 2
+  balance_dataset: false
+  filter_empty_annotations: false
+solver:
+  gradient_accumulation_steps: 1
+  epochs_per_save: 20
+  epochs_per_eval: 1
+  lr: 0.0005
+  grad_norm: 5.0
+  epochs: 300
+  optim:
+    name: AdamW
+    args:
+      betas:
+      - 0.9
+      - 0.98
+  sched:
+    name: warmup_cosine
+    args:
+      warmup_steps: 500
+      minimum_ratio: 0.1
+eval:
+  train:
+    name: PretrainEval
+  val:
+    name: ScanReferEval
+  save: false
+model:
+  name: OpenVocab
+  language:
+    name: SigLIPLanguageEncoder
+    args:
+      weights: google/siglip-base-patch16-224
+    lr: 1.0e-05
+  vision:
+    name: qihoo360/fg-clip-base
+    args:
+      backbone: pointnet++
+      hidden_size: 768
+      freeze: true
+      path: pretrained_weights/pointnetpp-open-bert
+      num_attention_heads: 12
+      spatial_dim: 5
+      num_layers: 4
+      dim_loc: 6
+      dim_feedforward: 2048
+      attn_type: spatial
+      pairwise_rel_type: center
+      use_matmul_label: false
+      lang_type: bert
+      lang_path: pretrained_weights/607_text_embeddings
+    lr: 0.0001
+  grounding:
+    name: UnifiedSpatialCrossEncoderV2
+    args:
+      hidden_size: 768
+      num_attention_heads: 12
+      num_layers: 4
+      dim_feedforward: 2048
+      dim_loc: 6
+    lr: 0.0001
+  inter: before
+  heads:
+    head_list:
+    - pretrain_head
+    pretrain_head:
+      name: OVPretrainHead
+      args:
+        hidden_size: 768
+        vocab_size: 30522
+  loss_type: ListLoss
+  loss_list:
+  - SceneViewPM_loss
+  vis_loss_list:
+  - SceneViewPM_loss

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.121545/config.yaml

@@ -0,0 +1,318 @@
+name: sceneverse_scannet_exp1
+rng_seed: 42
+num_gpu: 8
+mode: train
+note: scannet_training_run1
+naming_keywords:
+- dataloader.batchsize
+- task
+- note
+- time
+base_dir: results
+exp_dir: !!python/object/apply:pathlib.PosixPath
+- results
+- sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1
+- 2025-07-07-09:21:27.121545
+save_frequency: 10
+resume: false
+is_pretrain: true
+debug:
+  flag: false
+  debug_size: 20
+  hard_debug: false
+logger:
+  name: wandb
+  entity: yem
+  project: 3DReason
+  run_id: i7fiqjmc
+data:
+  note: all
+  train:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  val:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  test:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  args:
+    max_obj_len: 80
+    max_seq_len: 50
+    num_points: 1024
+    pc_type: pred
+    sem_type: '607'
+    filter_lang: false
+    txt_mask_ratio: 0.15
+    pc_mask_ratio: 0.1
+    rot_aug: true
+    mask_strategy: random
+    use_scene_cap: true
+    max_scene_cap_len: 300
+  ScanNetSpatialRefer:
+    train:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        - nr3d
+        sr3d_plus_aug: true
+      sgrefer:
+        anno_type:
+        - rel2_gpt
+        - rel2_template
+        - relm_gpt
+        - relm_template
+        - star_gpt
+        - star_template
+      sgcaption:
+        anno_type:
+        - gpt
+        - template
+    val:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+    test:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+  RScanSpatialRefer:
+    train:
+      sources:
+      - 3rscan_view_cap
+    val:
+      sources:
+      - 3rscan_view_cap
+    test:
+      sources:
+      - 3rscan_view_cap
+  MultiScanSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+  ARKitSceneSpatialRefer:
+    train:
+      sources:
+      - arkitscenes_view_cap
+    val:
+      sources:
+      - arkitscenes_view_cap
+    test:
+      sources:
+      - arkitscenes_view_cap
+  HMSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+  use_voxel: false
+  scan_family_base: existing_datasets/ScanNet
+  rscan_base: existing_datasets/3RScan
+  arkitscene_base: existing_datasets/Arkitscenes
+  multiscan_base: existing_datasets/MultiScan
+  hm_base: existing_datasets/HM3D
+  procthor_base: existing_datasets/ProcThor
+  s3d_base: existing_datasets/Structured3D'
+data_aug:
+  aug_list:
+  - scene_aug
+  scene_aug:
+    translation:
+      enabled: false
+      value:
+      - 1.0
+      - 1.0
+      - 1.0
+      p: 1.0
+    scaling:
+      enabled: false
+      p: 1.0
+      value:
+      - 0.9
+      - 1.1
+    flip:
+      enabled: false
+      p: 0.5
+    rotation:
+      enabled: true
+      p: 1.0
+      axis_align: true
+      value:
+      - 0.0
+      - 0.0
+      - 1.0
+      shuffle: true
+    color_jitter: false
+    order_shuffle: false
+  obj_aug:
+    translation:
+      enabled: false
+      value:
+      - 0.1
+      - 0.1
+      - 0.1
+      p: 1.0
+    rotation:
+      enabled: false
+      p: 1.0
+      axis_align: false
+      value:
+      - 0.0
+      - 0.0
+      - 0.1
+      shuffle: true
+    random_jitter:
+      enabled: false
+      value: 0.01
+      accord_to_size: false
+      p: 1.0
+    pts_shuffle: true
+task: Pretrain
+data_wrapper:
+  train: SceneDatasetWrapper
+  val: SceneDatasetWrapper
+  test: SceneDatasetWrapper
+trainer: OpenVocabTrainer
+ckpt_path: ''
+pretrain_ckpt_path: ''
+dataloader:
+  batchsize: 8
+  num_workers: 2
+  balance_dataset: false
+  filter_empty_annotations: false
+solver:
+  gradient_accumulation_steps: 1
+  epochs_per_save: 20
+  epochs_per_eval: 1
+  lr: 0.0005
+  grad_norm: 5.0
+  epochs: 300
+  optim:
+    name: AdamW
+    args:
+      betas:
+      - 0.9
+      - 0.98
+  sched:
+    name: warmup_cosine
+    args:
+      warmup_steps: 500
+      minimum_ratio: 0.1
+eval:
+  train:
+    name: PretrainEval
+  val:
+    name: ScanReferEval
+  save: false
+model:
+  name: OpenVocab
+  language:
+    name: SigLIPLanguageEncoder
+    args:
+      weights: google/siglip-base-patch16-224
+    lr: 1.0e-05
+  vision:
+    name: qihoo360/fg-clip-base
+    args:
+      backbone: pointnet++
+      hidden_size: 768
+      freeze: true
+      path: pretrained_weights/pointnetpp-open-bert
+      num_attention_heads: 12
+      spatial_dim: 5
+      num_layers: 4
+      dim_loc: 6
+      dim_feedforward: 2048
+      attn_type: spatial
+      pairwise_rel_type: center
+      use_matmul_label: false
+      lang_type: bert
+      lang_path: pretrained_weights/607_text_embeddings
+    lr: 0.0001
+  grounding:
+    name: UnifiedSpatialCrossEncoderV2
+    args:
+      hidden_size: 768
+      num_attention_heads: 12
+      num_layers: 4
+      dim_feedforward: 2048
+      dim_loc: 6
+    lr: 0.0001
+  inter: before
+  heads:
+    head_list:
+    - pretrain_head
+    pretrain_head:
+      name: OVPretrainHead
+      args:
+        hidden_size: 768
+        vocab_size: 30522
+  loss_type: ListLoss
+  loss_list:
+  - SceneViewPM_loss
+  vis_loss_list:
+  - SceneViewPM_loss

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.122672/config.yaml

@@ -0,0 +1,318 @@
+name: sceneverse_scannet_exp1
+rng_seed: 42
+num_gpu: 8
+mode: train
+note: scannet_training_run1
+naming_keywords:
+- dataloader.batchsize
+- task
+- note
+- time
+base_dir: results
+exp_dir: !!python/object/apply:pathlib.PosixPath
+- results
+- sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1
+- 2025-07-07-09:21:27.122672
+save_frequency: 10
+resume: false
+is_pretrain: true
+debug:
+  flag: false
+  debug_size: 20
+  hard_debug: false
+logger:
+  name: wandb
+  entity: yem
+  project: 3DReason
+  run_id: i9ljysdk
+data:
+  note: all
+  train:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  val:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  test:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  args:
+    max_obj_len: 80
+    max_seq_len: 50
+    num_points: 1024
+    pc_type: pred
+    sem_type: '607'
+    filter_lang: false
+    txt_mask_ratio: 0.15
+    pc_mask_ratio: 0.1
+    rot_aug: true
+    mask_strategy: random
+    use_scene_cap: true
+    max_scene_cap_len: 300
+  ScanNetSpatialRefer:
+    train:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        - nr3d
+        sr3d_plus_aug: true
+      sgrefer:
+        anno_type:
+        - rel2_gpt
+        - rel2_template
+        - relm_gpt
+        - relm_template
+        - star_gpt
+        - star_template
+      sgcaption:
+        anno_type:
+        - gpt
+        - template
+    val:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+    test:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+  RScanSpatialRefer:
+    train:
+      sources:
+      - 3rscan_view_cap
+    val:
+      sources:
+      - 3rscan_view_cap
+    test:
+      sources:
+      - 3rscan_view_cap
+  MultiScanSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+  ARKitSceneSpatialRefer:
+    train:
+      sources:
+      - arkitscenes_view_cap
+    val:
+      sources:
+      - arkitscenes_view_cap
+    test:
+      sources:
+      - arkitscenes_view_cap
+  HMSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+  use_voxel: false
+  scan_family_base: existing_datasets/ScanNet
+  rscan_base: existing_datasets/3RScan
+  arkitscene_base: existing_datasets/Arkitscenes
+  multiscan_base: existing_datasets/MultiScan
+  hm_base: existing_datasets/HM3D
+  procthor_base: existing_datasets/ProcThor
+  s3d_base: existing_datasets/Structured3D'
+data_aug:
+  aug_list:
+  - scene_aug
+  scene_aug:
+    translation:
+      enabled: false
+      value:
+      - 1.0
+      - 1.0
+      - 1.0
+      p: 1.0
+    scaling:
+      enabled: false
+      p: 1.0
+      value:
+      - 0.9
+      - 1.1
+    flip:
+      enabled: false
+      p: 0.5
+    rotation:
+      enabled: true
+      p: 1.0
+      axis_align: true
+      value:
+      - 0.0
+      - 0.0
+      - 1.0
+      shuffle: true
+    color_jitter: false
+    order_shuffle: false
+  obj_aug:
+    translation:
+      enabled: false
+      value:
+      - 0.1
+      - 0.1
+      - 0.1
+      p: 1.0
+    rotation:
+      enabled: false
+      p: 1.0
+      axis_align: false
+      value:
+      - 0.0
+      - 0.0
+      - 0.1
+      shuffle: true
+    random_jitter:
+      enabled: false
+      value: 0.01
+      accord_to_size: false
+      p: 1.0
+    pts_shuffle: true
+task: Pretrain
+data_wrapper:
+  train: SceneDatasetWrapper
+  val: SceneDatasetWrapper
+  test: SceneDatasetWrapper
+trainer: OpenVocabTrainer
+ckpt_path: ''
+pretrain_ckpt_path: ''
+dataloader:
+  batchsize: 8
+  num_workers: 2
+  balance_dataset: false
+  filter_empty_annotations: false
+solver:
+  gradient_accumulation_steps: 1
+  epochs_per_save: 20
+  epochs_per_eval: 1
+  lr: 0.0005
+  grad_norm: 5.0
+  epochs: 300
+  optim:
+    name: AdamW
+    args:
+      betas:
+      - 0.9
+      - 0.98
+  sched:
+    name: warmup_cosine
+    args:
+      warmup_steps: 500
+      minimum_ratio: 0.1
+eval:
+  train:
+    name: PretrainEval
+  val:
+    name: ScanReferEval
+  save: false
+model:
+  name: OpenVocab
+  language:
+    name: SigLIPLanguageEncoder
+    args:
+      weights: google/siglip-base-patch16-224
+    lr: 1.0e-05
+  vision:
+    name: qihoo360/fg-clip-base
+    args:
+      backbone: pointnet++
+      hidden_size: 768
+      freeze: true
+      path: pretrained_weights/pointnetpp-open-bert
+      num_attention_heads: 12
+      spatial_dim: 5
+      num_layers: 4
+      dim_loc: 6
+      dim_feedforward: 2048
+      attn_type: spatial
+      pairwise_rel_type: center
+      use_matmul_label: false
+      lang_type: bert
+      lang_path: pretrained_weights/607_text_embeddings
+    lr: 0.0001
+  grounding:
+    name: UnifiedSpatialCrossEncoderV2
+    args:
+      hidden_size: 768
+      num_attention_heads: 12
+      num_layers: 4
+      dim_feedforward: 2048
+      dim_loc: 6
+    lr: 0.0001
+  inter: before
+  heads:
+    head_list:
+    - pretrain_head
+    pretrain_head:
+      name: OVPretrainHead
+      args:
+        hidden_size: 768
+        vocab_size: 30522
+  loss_type: ListLoss
+  loss_list:
+  - SceneViewPM_loss
+  vis_loss_list:
+  - SceneViewPM_loss

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.123870/config.yaml

@@ -0,0 +1,318 @@
+name: sceneverse_scannet_exp1
+rng_seed: 42
+num_gpu: 8
+mode: train
+note: scannet_training_run1
+naming_keywords:
+- dataloader.batchsize
+- task
+- note
+- time
+base_dir: results
+exp_dir: !!python/object/apply:pathlib.PosixPath
+- results
+- sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1
+- 2025-07-07-09:21:27.123870
+save_frequency: 10
+resume: false
+is_pretrain: true
+debug:
+  flag: false
+  debug_size: 20
+  hard_debug: false
+logger:
+  name: wandb
+  entity: yem
+  project: 3DReason
+  run_id: 10nrbapp
+data:
+  note: all
+  train:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  val:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  test:
+  - ScanNetSpatialRefer
+  - RScanSpatialRefer
+  - ARKitSceneSpatialRefer
+  args:
+    max_obj_len: 80
+    max_seq_len: 50
+    num_points: 1024
+    pc_type: pred
+    sem_type: '607'
+    filter_lang: false
+    txt_mask_ratio: 0.15
+    pc_mask_ratio: 0.1
+    rot_aug: true
+    mask_strategy: random
+    use_scene_cap: true
+    max_scene_cap_len: 300
+  ScanNetSpatialRefer:
+    train:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        - nr3d
+        sr3d_plus_aug: true
+      sgrefer:
+        anno_type:
+        - rel2_gpt
+        - rel2_template
+        - relm_gpt
+        - relm_template
+        - star_gpt
+        - star_template
+      sgcaption:
+        anno_type:
+        - gpt
+        - template
+    val:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+    test:
+      sources:
+      - scannet_view_cap
+      referit3d:
+        anno_type:
+        - sr3d
+        sr3d_plus_aug: false
+      sgrefer:
+        anno_type:
+        - template
+      sgcaption:
+        anno_type:
+        - gpt
+  RScanSpatialRefer:
+    train:
+      sources:
+      - 3rscan_view_cap
+    val:
+      sources:
+      - 3rscan_view_cap
+    test:
+      sources:
+      - 3rscan_view_cap
+  MultiScanSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+      - star_gpt
+  ARKitSceneSpatialRefer:
+    train:
+      sources:
+      - arkitscenes_view_cap
+    val:
+      sources:
+      - arkitscenes_view_cap
+    test:
+      sources:
+      - arkitscenes_view_cap
+  HMSpatialRefer:
+    train:
+      sources:
+      - anno
+      - rel2_template
+      - rel2_gpt
+      - relm_template
+      - relm_gpt
+      - star_template
+      - star_gpt
+    val:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+    test:
+      sources:
+      - anno
+      - rel2_template
+      - relm_gpt
+      - relm_template
+      - star_template
+  use_voxel: false
+  scan_family_base: existing_datasets/ScanNet
+  rscan_base: existing_datasets/3RScan
+  arkitscene_base: existing_datasets/Arkitscenes
+  multiscan_base: existing_datasets/MultiScan
+  hm_base: existing_datasets/HM3D
+  procthor_base: existing_datasets/ProcThor
+  s3d_base: existing_datasets/Structured3D'
+data_aug:
+  aug_list:
+  - scene_aug
+  scene_aug:
+    translation:
+      enabled: false
+      value:
+      - 1.0
+      - 1.0
+      - 1.0
+      p: 1.0
+    scaling:
+      enabled: false
+      p: 1.0
+      value:
+      - 0.9
+      - 1.1
+    flip:
+      enabled: false
+      p: 0.5
+    rotation:
+      enabled: true
+      p: 1.0
+      axis_align: true
+      value:
+      - 0.0
+      - 0.0
+      - 1.0
+      shuffle: true
+    color_jitter: false
+    order_shuffle: false
+  obj_aug:
+    translation:
+      enabled: false
+      value:
+      - 0.1
+      - 0.1
+      - 0.1
+      p: 1.0
+    rotation:
+      enabled: false
+      p: 1.0
+      axis_align: false
+      value:
+      - 0.0
+      - 0.0
+      - 0.1
+      shuffle: true
+    random_jitter:
+      enabled: false
+      value: 0.01
+      accord_to_size: false
+      p: 1.0
+    pts_shuffle: true
+task: Pretrain
+data_wrapper:
+  train: SceneDatasetWrapper
+  val: SceneDatasetWrapper
+  test: SceneDatasetWrapper
+trainer: OpenVocabTrainer
+ckpt_path: ''
+pretrain_ckpt_path: ''
+dataloader:
+  batchsize: 8
+  num_workers: 2
+  balance_dataset: false
+  filter_empty_annotations: false
+solver:
+  gradient_accumulation_steps: 1
+  epochs_per_save: 20
+  epochs_per_eval: 1
+  lr: 0.0005
+  grad_norm: 5.0
+  epochs: 300
+  optim:
+    name: AdamW
+    args:
+      betas:
+      - 0.9
+      - 0.98
+  sched:
+    name: warmup_cosine
+    args:
+      warmup_steps: 500
+      minimum_ratio: 0.1
+eval:
+  train:
+    name: PretrainEval
+  val:
+    name: ScanReferEval
+  save: false
+model:
+  name: OpenVocab
+  language:
+    name: SigLIPLanguageEncoder
+    args:
+      weights: google/siglip-base-patch16-224
+    lr: 1.0e-05
+  vision:
+    name: qihoo360/fg-clip-base
+    args:
+      backbone: pointnet++
+      hidden_size: 768
+      freeze: true
+      path: pretrained_weights/pointnetpp-open-bert
+      num_attention_heads: 12
+      spatial_dim: 5
+      num_layers: 4
+      dim_loc: 6
+      dim_feedforward: 2048
+      attn_type: spatial
+      pairwise_rel_type: center
+      use_matmul_label: false
+      lang_type: bert
+      lang_path: pretrained_weights/607_text_embeddings
+    lr: 0.0001
+  grounding:
+    name: UnifiedSpatialCrossEncoderV2
+    args:
+      hidden_size: 768
+      num_attention_heads: 12
+      num_layers: 4
+      dim_feedforward: 2048
+      dim_loc: 6
+    lr: 0.0001
+  inter: before
+  heads:
+    head_list:
+    - pretrain_head
+    pretrain_head:
+      name: OVPretrainHead
+      args:
+        hidden_size: 768
+        vocab_size: 30522
+  loss_type: ListLoss
+  loss_list:
+  - SceneViewPM_loss
+  vis_loss_list:
+  - SceneViewPM_loss

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_100.pth/custom_checkpoint_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:94b16c562129ade67bb3d956f54d743d3a4a32f176d47863dd68a69c4a47c996
+size 1040

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_100.pth/model_1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60d95b10b6e140a9626a7058d5038528f2ff80148dc4569b881db56052046509
+size 40

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_100.pth/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b9ce0a2b33b768415ca93e3014b12286fdbc02c6dea9dbdaea78a45a3ae5b57d
+size 37850810

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_100.pth/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d5dddf0ff0ce57d6d7fa8aab284f25702b555bdc27d86f8634743dc3d7e249fb
+size 16036

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_100.pth/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5363664b0a4643747fa98b0705c048352a70f23b3c69d7c0c837b36b8eea6844
+size 1128

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_120.pth/custom_checkpoint_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be526b0bf55235b17f0641b9d44a5deb393581fa43f3c12f6a34834ae24bd65c
+size 1040

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_120.pth/model_1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60d95b10b6e140a9626a7058d5038528f2ff80148dc4569b881db56052046509
+size 40

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_120.pth/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:93134fd770997a682aa1ed20decf41577fdf71dfe63886688f845f12b63ebd66
+size 37850810

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_120.pth/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6ce9be254e32ccea0919e31ade5f7e3e267622eae1db8cd83b67cc9bf78b45d8
+size 16036

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_120.pth/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bce866ee7f3ede5cd9046e6ffc64712ab0e00ac1788028ddc885594d2c66084b
+size 1128

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_140.pth/custom_checkpoint_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:781c4909c4bef3c68ae07a29fd292beb096a2221a5d34b3aa03e7e140a7abd3a
+size 1040

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_140.pth/model_1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60d95b10b6e140a9626a7058d5038528f2ff80148dc4569b881db56052046509
+size 40

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_140.pth/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:37f652781eb740b0daeb86bb4e1fc2f3dc509e1caab15d1b10cbfff2b5ab4f71
+size 37850810

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_140.pth/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:dcd2443a478db6711aec3026dfabed83d067d8811e1762ba0347c9d4bbdc9849
+size 16036

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_140.pth/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:53d24d926677f17f54060fb6dcc7d9127e6cb62d0c899578a5417f3a72367095
+size 1128

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_160.pth/custom_checkpoint_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1198419e20802c22a8265d2e11959e7f6fc70de0d29c04741ebdf071d8512c2f
+size 1040

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_160.pth/model_1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60d95b10b6e140a9626a7058d5038528f2ff80148dc4569b881db56052046509
+size 40

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_160.pth/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:017a37a67614de87894f85ee52ac966f71f664be078e0680c850d22c3608b2bd
+size 37850810

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_160.pth/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:17b05b83af1072a93af3900827f92dc96a09290e7348e36702f22e26cd3db48e
+size 16036

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_160.pth/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:97363ad41c48026710bf636c473c3ff27913066f310f5dcf72d602cde2d4b5d7
+size 1128

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_180.pth/custom_checkpoint_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26aeaf1aa5c51b2b0e4532fd6a9fe73590e112023b1d690efe9b845e5aa07d69
+size 1040

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_180.pth/model_1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60d95b10b6e140a9626a7058d5038528f2ff80148dc4569b881db56052046509
+size 40

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_180.pth/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:940ed504fbd6d61154e7b2bad04a28e514bb2b772f5763e8c09e7cbd73205d34
+size 37850810

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_180.pth/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b7d4f123defe2cc3064bf33b0e58d6d22f335b5003af6896740dd9e68758abee
+size 16036

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_180.pth/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c15ae73185be3ecfdc86b4cfde3ff26559a5bb88cd81e70a02599eeeba1f592a
+size 1128

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_20.pth/custom_checkpoint_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ba4e17e3cebc85e84d267396621008c2749c147eb02bbbd98e4c3f2f536b4b4f
+size 1040

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_20.pth/model_1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60d95b10b6e140a9626a7058d5038528f2ff80148dc4569b881db56052046509
+size 40

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_20.pth/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:533fd6d3fa88c1edf14d24ce3462700c0359df3bc9c07ca63686def16e6e24ee
+size 37850810

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_20.pth/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a93800ec3c063f6ed822ec48941b3127c0035c98da8a0b724b66b470745b126d
+size 16036

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_20.pth/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:9b1eedbd67dc651c5a7db4a31ea9b215123db7c71cb8dd23fb3d019e74bdb0c5
+size 1128

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_200.pth/custom_checkpoint_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8026c61642fafd241718b631086628e202c45be79629f78d88d8340345db0346
+size 1040

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_200.pth/model_1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60d95b10b6e140a9626a7058d5038528f2ff80148dc4569b881db56052046509
+size 40

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_200.pth/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b095c5a061a185341c645152c86e188a122fd359b885b2076a94654987da1f8e
+size 37850810

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_200.pth/random_states_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bd98345159d5c7c414679b85d0865677ba407066d2499eaef9899d80a82a95a6
+size 16036

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_200.pth/scheduler.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:598e91f2dc6a64f9878e5fa0ed60db98005fbcd398d0e8baed26743f06bde4e0
+size 1128

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_220.pth/custom_checkpoint_0.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4849df77223a463c470dbe07d6fbc547c6dcb442bfc791d0894d6d3050ab4196
+size 1040

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_220.pth/model_1.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:60d95b10b6e140a9626a7058d5038528f2ff80148dc4569b881db56052046509
+size 40

results/sceneverse_scannet_exp1_b64_Pretrain_all_scannet_training_run1/2025-07-07-09:21:27.125816/ckpt/ckpt_220.pth/optimizer.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:3ed8360d1b5031dc84788b6babe6e71eca294c4bc1da6520b141eba8e8dd06d6
+size 37850810