Upload RoboTransferPipeline

README.md

@@ -1,3 +1,4 @@
 ---
 license: apache-2.0
+library_name: diffusers
 ---

depth_guider_net/config.json

@@ -0,0 +1,8 @@
+{
+  "_class_name": "GuiderNet",
+  "_diffusers_version": "0.34.0",
+  "_name_or_path": "/shared_disk/users/wenkang.qin/checkpoints/robotransfer-high-resolution/depth_guider_net",
+  "in_channels": 1,
+  "mid_channels": 4,
+  "out_channels": 4
+}

depth_guider_net/diffusion_pytorch_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:b4403e73bc36dc6301f496c57e45d5591c70b21e9b82ae5a377694fd336faa9b
+size 2832

feature_extractor/preprocessor_config.json

@@ -0,0 +1,27 @@
+{
+  "crop_size": {
+    "height": 224,
+    "width": 224
+  },
+  "do_center_crop": true,
+  "do_convert_rgb": true,
+  "do_normalize": true,
+  "do_rescale": true,
+  "do_resize": true,
+  "image_mean": [
+    0.48145466,
+    0.4578275,
+    0.40821073
+  ],
+  "image_processor_type": "CLIPImageProcessor",
+  "image_std": [
+    0.26862954,
+    0.26130258,
+    0.27577711
+  ],
+  "resample": 3,
+  "rescale_factor": 0.00392156862745098,
+  "size": {
+    "shortest_edge": 224
+  }
+}

image_encoder/config.json

@@ -0,0 +1,22 @@
+{
+  "architectures": [
+    "CLIPVisionModelWithProjection"
+  ],
+  "attention_dropout": 0.0,
+  "dropout": 0.0,
+  "hidden_act": "gelu",
+  "hidden_size": 1280,
+  "image_size": 224,
+  "initializer_factor": 1.0,
+  "initializer_range": 0.02,
+  "intermediate_size": 5120,
+  "layer_norm_eps": 1e-05,
+  "model_type": "clip_vision_model",
+  "num_attention_heads": 16,
+  "num_channels": 3,
+  "num_hidden_layers": 32,
+  "patch_size": 14,
+  "projection_dim": 1024,
+  "torch_dtype": "float32",
+  "transformers_version": "4.53.1"
+}

image_encoder/model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed1e5af7b4042ca30ec29999a4a5cfcac90b7fb610fd05ace834f2dcbb763eab
+size 2528371296

model_index.json

@@ -0,0 +1,33 @@
+{
+  "_class_name": "RoboTransferPipeline",
+  "_diffusers_version": "0.34.0",
+  "_name_or_path": "/shared_disk/users/wenkang.qin/checkpoints/robotransfer-high-resolution",
+  "depth_guider_net": [
+    "robotransfer.models.guider",
+    "GuiderNet"
+  ],
+  "feature_extractor": [
+    "transformers",
+    "CLIPImageProcessor"
+  ],
+  "image_encoder": [
+    "transformers",
+    "CLIPVisionModelWithProjection"
+  ],
+  "normal_guider_net": [
+    "robotransfer.models.guider",
+    "GuiderNet"
+  ],
+  "scheduler": [
+    "diffusers",
+    "EulerDiscreteScheduler"
+  ],
+  "unet": [
+    "diffusers",
+    "UNetSpatioTemporalConditionModel"
+  ],
+  "vae": [
+    "diffusers",
+    "AutoencoderKLTemporalDecoder"
+  ]
+}

normal_guider_net/config.json

@@ -0,0 +1,8 @@
+{
+  "_class_name": "GuiderNet",
+  "_diffusers_version": "0.34.0",
+  "_name_or_path": "/shared_disk/users/wenkang.qin/checkpoints/robotransfer-high-resolution/normal_guider_net",
+  "in_channels": 3,
+  "mid_channels": 4,
+  "out_channels": 4
+}

normal_guider_net/diffusion_pytorch_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a5edc33a20bb6038cea62abc34bdc2723587c999e86917ad19bcaa3fd36d57b0
+size 3344

scheduler/scheduler_config.json

@@ -0,0 +1,24 @@
+{
+  "_class_name": "EulerDiscreteScheduler",
+  "_diffusers_version": "0.34.0",
+  "beta_end": 0.012,
+  "beta_schedule": "scaled_linear",
+  "beta_start": 0.00085,
+  "clip_sample": false,
+  "final_sigmas_type": "zero",
+  "interpolation_type": "linear",
+  "num_train_timesteps": 1000,
+  "prediction_type": "v_prediction",
+  "rescale_betas_zero_snr": false,
+  "set_alpha_to_one": false,
+  "sigma_max": 700.0,
+  "sigma_min": 0.002,
+  "skip_prk_steps": true,
+  "steps_offset": 1,
+  "timestep_spacing": "leading",
+  "timestep_type": "continuous",
+  "trained_betas": null,
+  "use_beta_sigmas": false,
+  "use_exponential_sigmas": false,
+  "use_karras_sigmas": true
+}

unet/config.json

@@ -0,0 +1,41 @@
+{
+  "_class_name": "UNetSpatioTemporalConditionModel",
+  "_diffusers_version": "0.34.0",
+  "_name_or_path": "/shared_disk/users/wenkang.qin/checkpoints/robotransfer-high-resolution/unet",
+  "addition_time_embed_dim": 256,
+  "block_out_channels": [
+    320,
+    640,
+    1280,
+    1280
+  ],
+  "caption_channels": null,
+  "cross_attention_dim": 1024,
+  "down_block_types": [
+    "CrossAttnDownBlockSpatioTemporal",
+    "CrossAttnDownBlockSpatioTemporal",
+    "CrossAttnDownBlockSpatioTemporal",
+    "DownBlockSpatioTemporal"
+  ],
+  "exvideo_cfg": null,
+  "in_channels": 16,
+  "layers_per_block": 2,
+  "num_attention_heads": [
+    5,
+    10,
+    20,
+    20
+  ],
+  "num_frames": 25,
+  "out_channels": 4,
+  "patch_size": 1,
+  "projection_class_embeddings_input_dim": 768,
+  "sample_size": 96,
+  "transformer_layers_per_block": 1,
+  "up_block_types": [
+    "UpBlockSpatioTemporal",
+    "CrossAttnUpBlockSpatioTemporal",
+    "CrossAttnUpBlockSpatioTemporal",
+    "CrossAttnUpBlockSpatioTemporal"
+  ]
+}

unet/diffusion_pytorch_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:a6e5dda5cacee67e6ab3549813cc9ccb1efefa3a781ae81e8452f0e760c1b170
+size 6098774624

vae/config.json

@@ -0,0 +1,24 @@
+{
+  "_class_name": "AutoencoderKLTemporalDecoder",
+  "_diffusers_version": "0.34.0",
+  "_name_or_path": "/shared_disk/users/wenkang.qin/checkpoints/robotransfer-high-resolution/vae",
+  "block_out_channels": [
+    128,
+    256,
+    512,
+    512
+  ],
+  "down_block_types": [
+    "DownEncoderBlock2D",
+    "DownEncoderBlock2D",
+    "DownEncoderBlock2D",
+    "DownEncoderBlock2D"
+  ],
+  "force_upcast": true,
+  "in_channels": 3,
+  "latent_channels": 4,
+  "layers_per_block": 2,
+  "out_channels": 3,
+  "sample_size": 768,
+  "scaling_factor": 0.18215
+}

vae/diffusion_pytorch_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5d92aa595a53d9da9faf594f09910ee869d5d567c8bb0362d5095673c69997d6
+size 391017740