Upload hf_config.json with huggingface_hub

hf_config.json

@@ -0,0 +1,250 @@
+{
+  "dataset_mixture": {
+    "datasets": [
+      {
+        "repo_id": "robocasa/atomic",
+        "vqa": null,
+        "root": "/fss/bot/navigate",
+        "episodes": null,
+        "image_transforms": {
+          "enable": false,
+          "max_num_transforms": 3,
+          "random_order": false,
+          "tfs": {
+            "brightness": {
+              "weight": 1,
+              "type": "ColorJitter",
+              "kwargs": {
+                "brightness": [
+                  0.8,
+                  1.2
+                ]
+              }
+            },
+            "contrast": {
+              "weight": 1,
+              "type": "ColorJitter",
+              "kwargs": {
+                "contrast": [
+                  0.8,
+                  1.2
+                ]
+              }
+            },
+            "saturation": {
+              "weight": 1,
+              "type": "ColorJitter",
+              "kwargs": {
+                "saturation": [
+                  0.5,
+                  1.5
+                ]
+              }
+            },
+            "hue": {
+              "weight": 1,
+              "type": "ColorJitter",
+              "kwargs": {
+                "hue": [
+                  -0.05,
+                  0.05
+                ]
+              }
+            },
+            "sharpness": {
+              "weight": 1,
+              "type": "SharpnessJitter",
+              "kwargs": {
+                "sharpness": [
+                  0.5,
+                  1.5
+                ]
+              }
+            }
+          }
+        },
+        "revision": null,
+        "use_imagenet_stats": true,
+        "video_backend": "pyav",
+        "stats": null,
+        "data_features_name_mapping": null,
+        "val_split_ratio": 0.05
+      }
+    ],
+    "weights": [
+      1
+    ],
+    "action_freq": 20,
+    "image_resample_strategy": "nearest",
+    "vector_resample_strategy": "nearest",
+    "val_split_ratio": 0.05
+  },
+  "policy": {
+    "type": "pi05",
+    "n_obs_steps": 1,
+    "normalization_mapping": {
+      "VISUAL": "IDENTITY",
+      "STATE": "MIN_MAX",
+      "ACTION": "MEAN_STD"
+    },
+    "input_features": {
+      "state": {
+        "type": "STATE",
+        "shape": [
+          32
+        ]
+      },
+      "camera0": {
+        "type": "VISUAL",
+        "shape": [
+          3,
+          256,
+          256
+        ]
+      },
+      "camera1": {
+        "type": "VISUAL",
+        "shape": [
+          3,
+          256,
+          256
+        ]
+      },
+      "camera2": {
+        "type": "VISUAL",
+        "shape": [
+          3,
+          256,
+          256
+        ]
+      }
+    },
+    "output_features": {
+      "actions": {
+        "type": "ACTION",
+        "shape": [
+          32
+        ]
+      }
+    },
+    "device": null,
+    "use_amp": false,
+    "pretrained_path": "TensorAuto/pzay376r-null-31000",
+    "cloud_vlm_latency_mean": 0,
+    "cloud_vlm_latency_std": 0,
+    "cloud_vlm_latency_lower": 0,
+    "cloud_vlm_latency_upper": 0,
+    "action_decoder_latency_mean": 0,
+    "action_decoder_latency_std": 0,
+    "action_decoder_latency_lower": 0,
+    "action_decoder_latency_upper": 0,
+    "chunk_size": 20,
+    "n_action_steps": 20,
+    "max_state_dim": 32,
+    "max_action_dim": 32,
+    "predict_response": false,
+    "resize_imgs_with_padding": [
+      224,
+      224
+    ],
+    "empty_cameras": 0,
+    "prompt_max_length": 256,
+    "response_max_length": 52,
+    "discrete_action_max_length": 75,
+    "proj_width": 1024,
+    "dropout": 0.1,
+    "num_steps": 10,
+    "max_delay": 0,
+    "init_strategy": "no_init",
+    "attention_implementation": "eager",
+    "freeze_vision_encoder": false,
+    "train_expert_only": false,
+    "optimizer_lr": 2.5e-05,
+    "optimizer_betas": [
+      0.9,
+      0.95
+    ],
+    "optimizer_eps": 1e-08,
+    "optimizer_weight_decay": 1e-10,
+    "scheduler_warmup_steps": 1000,
+    "scheduler_decay_steps": 30000,
+    "scheduler_decay_lr": 2.5e-06
+  },
+  "job_name": "pi05",
+  "resume": false,
+  "seed": 1000,
+  "resolution": [
+    256,
+    256
+  ],
+  "num_cams": 3,
+  "max_state_dim": 32,
+  "max_action_dim": 32,
+  "action_chunk": 20,
+  "loss_weighting": {
+    "MSE": 1,
+    "CE": 1
+  },
+  "num_workers": 16,
+  "batch_size": 10,
+  "gradient_accumulation_steps": 1,
+  "dataloader_batch_size": 10,
+  "prefetch_factor": 8,
+  "steps": 1000000,
+  "log_freq": 100,
+  "save_checkpoint": true,
+  "save_freq": 1000,
+  "use_policy_training_preset": false,
+  "optimizer": {
+    "type": "adamw",
+    "lr": 2.5e-05,
+    "weight_decay": 1e-10,
+    "grad_clip_norm": 10,
+    "betas": [
+      0.9,
+      0.95
+    ],
+    "eps": 1e-08
+  },
+  "scheduler": {
+    "type": "cosine_decay_with_warmup",
+    "num_warmup_steps": 1000,
+    "num_decay_steps": 30000,
+    "peak_lr": 2.5e-05,
+    "decay_lr": 2.5e-06
+  },
+  "wandb": {
+    "enable": true,
+    "entity": "wyautox-autox",
+    "project": "robocasa",
+    "run_id": "pzay376r",
+    "name": null,
+    "notes": "PI05 training run",
+    "tags": [],
+    "group": null,
+    "job_type": null,
+    "mode": null,
+    "allow_resume": true,
+    "disable_artifact": false
+  },
+  "debug": false,
+  "trace_nans": false,
+  "env": null,
+  "eval": {
+    "n_episodes": 16,
+    "batch_size": 16,
+    "use_async_envs": true,
+    "max_episodes_rendered": 16,
+    "grid_size": null,
+    "recording_root": null
+  },
+  "eval_freq": 0,
+  "val_freq": 1000,
+  "last_checkpoint_only": true,
+  "server": {
+    "port": 50051,
+    "max_workers": 4,
+    "max_send_message_length_mb": 100,
+    "max_receive_message_length_mb": 100
+  }
+}