diff --git a/010000/pretrained_model/config.json b/010000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/010000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/010000/pretrained_model/model.safetensors b/010000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5b669222c73ac3e09b09be7a18f2aed758956386 --- /dev/null +++ b/010000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f2be3e767de5b69047dce6d3deae1412c9ff25e70e70b3f446ff8111656dab66 +size 206701072 diff --git a/010000/pretrained_model/train_config.json b/010000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/010000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/010000/training_state/optimizer_param_groups.json b/010000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/010000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/010000/training_state/optimizer_state.safetensors b/010000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b80551460e5c72e3d8de7872260344fa4585ce8b --- /dev/null +++ b/010000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f42439e1e031c7d4eb5d61bc7c2e56398cbd4e812c88696bdda4e48419ba344 +size 412817652 diff --git a/010000/training_state/rng_state.safetensors b/010000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..caa42d1236d5442a6cb3cea8dc99793051255f53 --- /dev/null +++ b/010000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e14bfda74fe31cd5ca9fe1a24d4a9f796d60c3643a5c9a6eac3a394864b30774 +size 15708 diff --git a/010000/training_state/training_step.json b/010000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..7cb7c0986e9e7461ca851ce71e95d235ae3d2732 --- /dev/null +++ b/010000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 10000 +} \ No newline at end of file diff --git a/020000/pretrained_model/config.json b/020000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/020000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/020000/pretrained_model/model.safetensors b/020000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..11efe2bd79df80a5c875537d9bccaf83b91b2a68 --- /dev/null +++ b/020000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cae60d023ad60a5a2c37f0236cc153bb8ac809f44f5f7875c0dd477db8f836d5 +size 206701072 diff --git a/020000/pretrained_model/train_config.json b/020000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/020000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/020000/training_state/optimizer_param_groups.json b/020000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/020000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/020000/training_state/optimizer_state.safetensors b/020000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f0f763726a8b3788d5754e2724ffe553c5990fc7 --- /dev/null +++ b/020000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:75abf768e22eda7215d0854d15469adea7e18cdd92f1bff2d41252b75e52e8c0 +size 412817652 diff --git a/020000/training_state/rng_state.safetensors b/020000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ecd69eecaba6424cdd46444d3427524c12809cfc --- /dev/null +++ b/020000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5a4c53e0f13fdc6cf0bb428ea264386c2626b99e9d1623772dfb839b9ccc08da +size 15708 diff --git a/020000/training_state/training_step.json b/020000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..dc9bb47026c5d5237ca6fc5dbff6020dd122ea05 --- /dev/null +++ b/020000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 20000 +} \ No newline at end of file diff --git a/030000/pretrained_model/config.json b/030000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/030000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/030000/pretrained_model/model.safetensors b/030000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f4b338014cd17f6ff8d1877f230c78f27dddfdff --- /dev/null +++ b/030000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e9a4c8e8b952493fe458f66feb7ea83b50e023e65d6bab490dcdd6ad3cd52892 +size 206701072 diff --git a/030000/pretrained_model/train_config.json b/030000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/030000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/030000/training_state/optimizer_param_groups.json b/030000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/030000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/030000/training_state/optimizer_state.safetensors b/030000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7ddb11f1b9ced31c7dbed41e876d213fd70c34e8 --- /dev/null +++ b/030000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0f3086e55f64b6510db203a99f6daee07aa25addd290bd3bed7605a7a6314bba +size 412817652 diff --git a/030000/training_state/rng_state.safetensors b/030000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..97fd9f6ac162b023c0b28768b6e9db9f42ea246e --- /dev/null +++ b/030000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c07cc7d521fb88bbb5bb2918d3642a6b9fe78421d0e7e87cb77afccb16379e14 +size 15708 diff --git a/030000/training_state/training_step.json b/030000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..f4945f660f45b332883dccfccf18d8b8815d916a --- /dev/null +++ b/030000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 30000 +} \ No newline at end of file diff --git a/040000/pretrained_model/config.json b/040000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/040000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/040000/pretrained_model/model.safetensors b/040000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b565896b227829efcd1bbf261794d5e4a2d623f4 --- /dev/null +++ b/040000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ecb58b03e1dbb5d3484d3938c4a3b505f89425af2fc465c078d6d655ce916fd9 +size 206701072 diff --git a/040000/pretrained_model/train_config.json b/040000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/040000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/040000/training_state/optimizer_param_groups.json b/040000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/040000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/040000/training_state/optimizer_state.safetensors b/040000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..90fc255d8f3fba92b5d3060c9751ac7fa5a685f6 --- /dev/null +++ b/040000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0863c6cf05eedb48bd8ada2be66f09c7212136bccf81a7f7d9b75e4a6387fce1 +size 412817652 diff --git a/040000/training_state/rng_state.safetensors b/040000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..336c1abb949db48821f4e86bcf1ee3e26394001f --- /dev/null +++ b/040000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:68dc1564cc35a25dfef9cf6d6742e89464db920303352b420c2bd66c447df987 +size 15708 diff --git a/040000/training_state/training_step.json b/040000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..fe40d8ff4312c7e5a8fba9bcc932a43a1384ba77 --- /dev/null +++ b/040000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 40000 +} \ No newline at end of file diff --git a/050000/pretrained_model/config.json b/050000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/050000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/050000/pretrained_model/model.safetensors b/050000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..35ff02264f9a2a577cb43cb5aff94a8c7420df69 --- /dev/null +++ b/050000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:59bd7d5ba5ad1d9dd9c4bbe86f7f3ab84dadec79d2af59da24d50b2aaece9938 +size 206701072 diff --git a/050000/pretrained_model/train_config.json b/050000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/050000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/050000/training_state/optimizer_param_groups.json b/050000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/050000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/050000/training_state/optimizer_state.safetensors b/050000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..adee09b3a680dcc891dd9f3d7a57544777f35a19 --- /dev/null +++ b/050000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f58198b870b1c361d642581b5d71abf997e9fe1772706a8d9f0a6a107dd55e6d +size 412817652 diff --git a/050000/training_state/rng_state.safetensors b/050000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d4c2947ec435965cf350c6878709cf5d798a0639 --- /dev/null +++ b/050000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d6b8d103ec9c3b608a1229cfdcf33a2c72faad536ec21307e8cb9a38e6fc2de +size 15708 diff --git a/050000/training_state/training_step.json b/050000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..b3c03563f17d6d98552117ccf00705b693aa3fbd --- /dev/null +++ b/050000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 50000 +} \ No newline at end of file diff --git a/060000/pretrained_model/config.json b/060000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/060000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/060000/pretrained_model/model.safetensors b/060000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..89265f22bae194591b7a90a143a0b91f99eb5cf1 --- /dev/null +++ b/060000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:901e7a6abb623c20bcc3d6a8f4daef54b1d52bfc6be96370f83b1c699c28e2d1 +size 206701072 diff --git a/060000/pretrained_model/train_config.json b/060000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/060000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/060000/training_state/optimizer_param_groups.json b/060000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/060000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/060000/training_state/optimizer_state.safetensors b/060000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..199cfb4b0b5f1efa9be347c08b72889f26d026d1 --- /dev/null +++ b/060000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a7aba8014c093ecdba82faa26271138f6ef2e22158bc137b23a4dd4e7a7b8a58 +size 412817652 diff --git a/060000/training_state/rng_state.safetensors b/060000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4a1d6b393b114ad724c401b57c84e552609ed2a2 --- /dev/null +++ b/060000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a36a0670fbdaee8d67354dd29ed6bebd93bc982ddaec6003e8ff64fafa982c2b +size 15708 diff --git a/060000/training_state/training_step.json b/060000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..d360b6037f1c8c2c85d38e951160eb02ace507dd --- /dev/null +++ b/060000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 60000 +} \ No newline at end of file diff --git a/070000/pretrained_model/config.json b/070000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/070000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/070000/pretrained_model/model.safetensors b/070000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ced294297f7e03db1b640e9bace6a82c9ee9130f --- /dev/null +++ b/070000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2c424eea3a58898e86a216eef356c2d022e33fde47282a83bf4698d9426aeaf2 +size 206701072 diff --git a/070000/pretrained_model/train_config.json b/070000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/070000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/070000/training_state/optimizer_param_groups.json b/070000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/070000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/070000/training_state/optimizer_state.safetensors b/070000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9444cedf39b54120850665a266550b204f8639ed --- /dev/null +++ b/070000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:27d20978c847e4f54049bd07c9bf2520ee60923a3160faa77fdd06f42c942aa4 +size 412817652 diff --git a/070000/training_state/rng_state.safetensors b/070000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0ed1c5513cf8227b02b031745838f050488291ac --- /dev/null +++ b/070000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0d793d4dfc16f6a056c217510ec744a09c8354ba028a623ef0538893dcfd7c1a +size 15708 diff --git a/070000/training_state/training_step.json b/070000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..1ec0c1eb0f525e34fa3a57743f750b2b34b5b77b --- /dev/null +++ b/070000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 70000 +} \ No newline at end of file diff --git a/080000/pretrained_model/config.json b/080000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/080000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/080000/pretrained_model/model.safetensors b/080000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8cf529ec7528061e0ad5bb25f9c0283fbd831db0 --- /dev/null +++ b/080000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6f381734f45143d1050d35303a1e618ed98d17a98e3d6ad3e3ca9131615227f6 +size 206701072 diff --git a/080000/pretrained_model/train_config.json b/080000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/080000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/080000/training_state/optimizer_param_groups.json b/080000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/080000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/080000/training_state/optimizer_state.safetensors b/080000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..80530bba557929c9397577e22acd50cfb50947ea --- /dev/null +++ b/080000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:976007b0fdd7d3249c7dbca6354a6949b2d03043468645e49e0071513342f3fa +size 412817652 diff --git a/080000/training_state/rng_state.safetensors b/080000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fffd3fc91aefff15963329a516fea122beca85e7 --- /dev/null +++ b/080000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2d3a1d475a4d5d04dd49ff4d415639a16fc3b9e8c51ef9076241d254dbd35aee +size 15708 diff --git a/080000/training_state/training_step.json b/080000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..384eace4ecc2a6cba352aa7cf27f04405b7319c3 --- /dev/null +++ b/080000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 80000 +} \ No newline at end of file diff --git a/090000/pretrained_model/config.json b/090000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/090000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/090000/pretrained_model/model.safetensors b/090000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..eb0c67ab3fb3034ba1f3b8562c1fad2108a0e03a --- /dev/null +++ b/090000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b86b01ba79c48679364428ee5f9443bee58bd1cb4aada48f6858f7ab1e1d23df +size 206701072 diff --git a/090000/pretrained_model/train_config.json b/090000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/090000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/090000/training_state/optimizer_param_groups.json b/090000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/090000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/090000/training_state/optimizer_state.safetensors b/090000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7e2f9ff38bca84ec93ef3f61851cb7d761645602 --- /dev/null +++ b/090000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8ab4ebd4f36e887a21005a8ce46b77b2bb41621359bb1d569c1f4ee51f12b6b3 +size 412817652 diff --git a/090000/training_state/rng_state.safetensors b/090000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..15f6de233f23f0411d2663aca2d7baed3d71cff3 --- /dev/null +++ b/090000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2f83f6dcd9292222c3bda2cc69e0cebd663f59486953355ff4bf4e4bf2bb961b +size 15708 diff --git a/090000/training_state/training_step.json b/090000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..cec965b83b7f817ac99559792f84cbb69a721bfd --- /dev/null +++ b/090000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 90000 +} \ No newline at end of file diff --git a/100000/pretrained_model/config.json b/100000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/100000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/100000/pretrained_model/model.safetensors b/100000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..05d9a20cd8a634713728f23697412b96109ac84e --- /dev/null +++ b/100000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c04bd5396d164a452f6451309f1a3545bf6822e5869f0e13bc21ffe33b258f75 +size 206701072 diff --git a/100000/pretrained_model/train_config.json b/100000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/100000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/100000/training_state/optimizer_param_groups.json b/100000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/100000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/100000/training_state/optimizer_state.safetensors b/100000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..d71b5cf916a6eae0c9413a5cd23069c9adca2d69 --- /dev/null +++ b/100000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a42baf99ee35f3c7910b1f6c57d33505b8c48d4c3bb0eecd1dfa0fdb7630f9fb +size 412817652 diff --git a/100000/training_state/rng_state.safetensors b/100000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..71374d2bdcbba9b79366a028ce8d5efb7ea0d6ba --- /dev/null +++ b/100000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:129aa86b9cda8a64d25d39e203ec26e6843dd30064db3b93b350fe53c859ae81 +size 15708 diff --git a/100000/training_state/training_step.json b/100000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..8eb73c13d28bca88058c08796abbf931c3f9b012 --- /dev/null +++ b/100000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 100000 +} \ No newline at end of file diff --git a/110000/pretrained_model/config.json b/110000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/110000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/110000/pretrained_model/model.safetensors b/110000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fc78d918553be991bc262767899c341602b04456 --- /dev/null +++ b/110000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7d6c3ce9085784607b092d5ededcd29649309adbc9c9f5726cf6ed766df8b668 +size 206701072 diff --git a/110000/pretrained_model/train_config.json b/110000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/110000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/110000/training_state/optimizer_param_groups.json b/110000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/110000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/110000/training_state/optimizer_state.safetensors b/110000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7c788cb2880e1fcc64f03c6321b136615be21140 --- /dev/null +++ b/110000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:407a7e20d605b85b5aadd39a943112805fe1dbf15b2223dc088bc779310f2ea4 +size 412817652 diff --git a/110000/training_state/rng_state.safetensors b/110000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..806b6390d0d2605deb3febc94d4c868dd19f2b26 --- /dev/null +++ b/110000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5a32b8d5ad2149469090884f28c4d4f16d5b534040a89bdead6666a39ae06560 +size 15708 diff --git a/110000/training_state/training_step.json b/110000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..064896ea461ac4e49e4b688e1f5b6dc2ff4abd53 --- /dev/null +++ b/110000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 110000 +} \ No newline at end of file diff --git a/120000/pretrained_model/config.json b/120000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/120000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/120000/pretrained_model/model.safetensors b/120000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6cbfff3d642ea3f5d1027640319fd9a38007db81 --- /dev/null +++ b/120000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c165b8d3dc432bdff30ec01da1defa9f54e51b0f1f0d5a8200ccbcf98f52eb82 +size 206701072 diff --git a/120000/pretrained_model/train_config.json b/120000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/120000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/120000/training_state/optimizer_param_groups.json b/120000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/120000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/120000/training_state/optimizer_state.safetensors b/120000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3132fee9c18c7ee89c20d611a3a8bbc7dc834580 --- /dev/null +++ b/120000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:91b1b471f50a8e7070beab8baf19121ab1a6b4b6a1ce4270a6979e9d98c02fed +size 412817652 diff --git a/120000/training_state/rng_state.safetensors b/120000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..02b52eaff7df33ffde35954d81eeb66ded3bf273 --- /dev/null +++ b/120000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:13b5210facc6a8ea84cd72cb480e6336b7ed284d357848c24ab09977dc5ca993 +size 15708 diff --git a/120000/training_state/training_step.json b/120000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..6bbb8719463d819c865c104597046a0165eea278 --- /dev/null +++ b/120000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 120000 +} \ No newline at end of file diff --git a/130000/pretrained_model/config.json b/130000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/130000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/130000/pretrained_model/model.safetensors b/130000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..17e8a229534d42f75580df2053cd48358ed2c117 --- /dev/null +++ b/130000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:b35d88398438354772046eeef0e17d7d8a2bd55388e3d100aed6e6f3295b4404 +size 206701072 diff --git a/130000/pretrained_model/train_config.json b/130000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/130000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/130000/training_state/optimizer_param_groups.json b/130000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/130000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/130000/training_state/optimizer_state.safetensors b/130000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..43017b8c4db80ccb3c520f0cf67011ea2322d8ea --- /dev/null +++ b/130000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4cc6fb4960a93c885d208ecdee6b938a6ce8b433944136bdffcea2fabb87daf4 +size 412817652 diff --git a/130000/training_state/rng_state.safetensors b/130000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ce50d605cdbcac2cd7284aaa46a1cb700f232c60 --- /dev/null +++ b/130000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cfa612989a77067410f9c000b5ae5480813ce0e9eeba7a055253eee6e4dfdd8f +size 15708 diff --git a/130000/training_state/training_step.json b/130000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..cbd692b6bcb0d10cdb491cd7c943e192d8df0441 --- /dev/null +++ b/130000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 130000 +} \ No newline at end of file diff --git a/140000/pretrained_model/config.json b/140000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/140000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/140000/pretrained_model/model.safetensors b/140000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..25f7efae443fa4716c16dcb0d77c09af1ac2ee28 --- /dev/null +++ b/140000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2ad1a83313f8ede3507afddbea1403f1d631ec87c810854c944c2291ff02c87a +size 206701072 diff --git a/140000/pretrained_model/train_config.json b/140000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/140000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/140000/training_state/optimizer_param_groups.json b/140000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/140000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/140000/training_state/optimizer_state.safetensors b/140000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ef7a0271199926dab2b13fc00f1e64380405e26d --- /dev/null +++ b/140000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:20516269e676c96493c47b4ac5bff11f7d600aff2e7731e2a39216cd1b8401c1 +size 412817652 diff --git a/140000/training_state/rng_state.safetensors b/140000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6d4826f462974e06db3ac36bc05b42909432f5a7 --- /dev/null +++ b/140000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a36c2c4742c780019eceef224271e3ecc504deee6452f78abc5f92fa8b6edc4b +size 15708 diff --git a/140000/training_state/training_step.json b/140000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..f3695dfae1578a72a581cd61c827ae9be8dcaa64 --- /dev/null +++ b/140000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 140000 +} \ No newline at end of file diff --git a/150000/pretrained_model/config.json b/150000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/150000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/150000/pretrained_model/model.safetensors b/150000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..53e6f1bba12c853a8dd697dc4d33707e012be509 --- /dev/null +++ b/150000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:da96e88e11ca7e5c913ce083e252f980a05cc13bcea008007f557b8b60ac2ede +size 206701072 diff --git a/150000/pretrained_model/train_config.json b/150000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/150000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/150000/training_state/optimizer_param_groups.json b/150000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/150000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/150000/training_state/optimizer_state.safetensors b/150000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4a73e73841b2b6be1d4e8f1f20b4dbd48ed5c6f9 --- /dev/null +++ b/150000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:66dfb80cc51a4b237a5e2aff59c359c63e9df61a127d6f315f38a3142f2f3644 +size 412817652 diff --git a/150000/training_state/rng_state.safetensors b/150000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2aa8e34b8e16f0e5fbcdff521c224ba84a85e120 --- /dev/null +++ b/150000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:08ae4067429b5f87e7fd78bd5f60188fd2f4fd55771c63f65110e864c8a5b4c0 +size 15708 diff --git a/150000/training_state/training_step.json b/150000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..941897fdfccbe971211529c488ac32a17d6dc0a0 --- /dev/null +++ b/150000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 150000 +} \ No newline at end of file diff --git a/160000/pretrained_model/config.json b/160000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/160000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/160000/pretrained_model/model.safetensors b/160000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..778b6b9e1f3443aebea22ece9fd78e3b2d94a5a9 --- /dev/null +++ b/160000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3ed6bfe3ec5f374dabb8cce986215acf99c333d62a8e4f99d28e57423f250080 +size 206701072 diff --git a/160000/pretrained_model/train_config.json b/160000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/160000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/160000/training_state/optimizer_param_groups.json b/160000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/160000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/160000/training_state/optimizer_state.safetensors b/160000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4a3db4904a50d3006e24cf4887c2c42a09180391 --- /dev/null +++ b/160000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9fc596a1f35d6602a205998112930a79fd153d73404f9d94bf1e1ba096ac8c37 +size 412817652 diff --git a/160000/training_state/rng_state.safetensors b/160000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..3cea86a3ac9c208eeef5f12d30d2ef71d0cd823b --- /dev/null +++ b/160000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:188b0345f36e21dd51b48d253b72edf53ce8ab50ac2d2347948f44d2224a0534 +size 15708 diff --git a/160000/training_state/training_step.json b/160000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..18f112ab3b2ba641c0049e21daf4fb129368a0a0 --- /dev/null +++ b/160000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 160000 +} \ No newline at end of file diff --git a/170000/pretrained_model/config.json b/170000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/170000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/170000/pretrained_model/model.safetensors b/170000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ce683559d0250d67ff51458fc1d1ab4d50909076 --- /dev/null +++ b/170000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6f3596785fcc6953b45dba469121659b56d8083e2513f0771cd525b2026db9c4 +size 206701072 diff --git a/170000/pretrained_model/train_config.json b/170000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/170000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/170000/training_state/optimizer_param_groups.json b/170000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/170000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/170000/training_state/optimizer_state.safetensors b/170000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f23e8073bceff2e11d7bc2e7b9f6cde77bdd9243 --- /dev/null +++ b/170000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9fc5b3d2765c52b8c9b95c5e7663d6b796134b5e8164885b3a6ef9d007e3c28b +size 412817652 diff --git a/170000/training_state/rng_state.safetensors b/170000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5ffec1700a10b72716dd85252b9782d1f0fb8813 --- /dev/null +++ b/170000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:681f85cc63d0a39630ca3aaceb94d9b4871ed5512d728e9737b9d9ab414eb82c +size 15708 diff --git a/170000/training_state/training_step.json b/170000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..6a4324fbfa4744b587756639bb720731190f7a3e --- /dev/null +++ b/170000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 170000 +} \ No newline at end of file diff --git a/180000/pretrained_model/config.json b/180000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/180000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/180000/pretrained_model/model.safetensors b/180000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..58b987c124318f4e48215e3647cafa81b15b30d5 --- /dev/null +++ b/180000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:61dc591bb2dafb83cb3daf0d53debb62d88c89630deca6eec77178fb91f91c6e +size 206701072 diff --git a/180000/pretrained_model/train_config.json b/180000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/180000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/180000/training_state/optimizer_param_groups.json b/180000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/180000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/180000/training_state/optimizer_state.safetensors b/180000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c704d6f75dea7825eb65ff3e8a9c191c00d15895 --- /dev/null +++ b/180000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:23ea0e91e7c1170afafd57b9dfed0fa8f3b6dbd2a693e65eb28bcc7eadd5bbf4 +size 412817652 diff --git a/180000/training_state/rng_state.safetensors b/180000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e7881008e195dd5f051cfabd8d2b5144334a30f9 --- /dev/null +++ b/180000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f5d858c45995f111319338f5f30deed632e906674767a57f3859258df501e663 +size 15708 diff --git a/180000/training_state/training_step.json b/180000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..3da013f3092359f6c4ebb97dfc69e8300dbb32e4 --- /dev/null +++ b/180000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 180000 +} \ No newline at end of file diff --git a/190000/pretrained_model/config.json b/190000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/190000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/190000/pretrained_model/model.safetensors b/190000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5ff5081cebb11cd276d81bd9edbca86e9e3a517a --- /dev/null +++ b/190000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:461bbf8770cd705118d2cabb0dd5dd678d0698c2890b427b1308c3eb72dc7b0c +size 206701072 diff --git a/190000/pretrained_model/train_config.json b/190000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/190000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/190000/training_state/optimizer_param_groups.json b/190000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/190000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/190000/training_state/optimizer_state.safetensors b/190000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2983022e55a91d461678833cac614880e5faad0e --- /dev/null +++ b/190000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ea9deb535863fdd168889a1ffba8204127194e81ce5645f2e248a1ab6af79b97 +size 412817652 diff --git a/190000/training_state/rng_state.safetensors b/190000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..171fa9cded9125f4b620756739bbe569ef6fc1e8 --- /dev/null +++ b/190000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5e075f80044ee709dbd208bb2534651b2127f4e7a821264b581fcabc2cbedbb5 +size 15708 diff --git a/190000/training_state/training_step.json b/190000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..2103fe39ddacbae2425d4373d5ed361ec66f3ee7 --- /dev/null +++ b/190000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 190000 +} \ No newline at end of file diff --git a/200000/pretrained_model/config.json b/200000/pretrained_model/config.json new file mode 100644 index 0000000000000000000000000000000000000000..682d39b157fa2fd54c6b0ce77fef0db20d521f46 --- /dev/null +++ b/200000/pretrained_model/config.json @@ -0,0 +1,69 @@ +{ + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 +} \ No newline at end of file diff --git a/200000/pretrained_model/model.safetensors b/200000/pretrained_model/model.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..991e404c593ae6252bab304b1ea378db5dec49e0 --- /dev/null +++ b/200000/pretrained_model/model.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:491501d3374962cf9a593858e20b50ee97206c8813b92af52ad97ee5a1bb1015 +size 206701072 diff --git a/200000/pretrained_model/train_config.json b/200000/pretrained_model/train_config.json new file mode 100644 index 0000000000000000000000000000000000000000..1fa22e2f6422bf07c2aa7f7904a19d4437ad38a3 --- /dev/null +++ b/200000/pretrained_model/train_config.json @@ -0,0 +1,175 @@ +{ + "dataset": { + "repo_id": "dleon23/record-so101_2", + "root": null, + "episodes": null, + "image_transforms": { + "enable": false, + "max_num_transforms": 3, + "random_order": false, + "tfs": { + "brightness": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "brightness": [ + 0.8, + 1.2 + ] + } + }, + "contrast": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "contrast": [ + 0.8, + 1.2 + ] + } + }, + "saturation": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "saturation": [ + 0.5, + 1.5 + ] + } + }, + "hue": { + "weight": 1.0, + "type": "ColorJitter", + "kwargs": { + "hue": [ + -0.05, + 0.05 + ] + } + }, + "sharpness": { + "weight": 1.0, + "type": "SharpnessJitter", + "kwargs": { + "sharpness": [ + 0.5, + 1.5 + ] + } + } + } + }, + "revision": null, + "use_imagenet_stats": true, + "video_backend": "torchcodec" + }, + "env": null, + "policy": { + "type": "act", + "n_obs_steps": 1, + "normalization_mapping": { + "VISUAL": "MEAN_STD", + "STATE": "MEAN_STD", + "ACTION": "MEAN_STD" + }, + "input_features": { + "observation.state": { + "type": "STATE", + "shape": [ + 6 + ] + }, + "observation.images.front": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + }, + "observation.images.overhead": { + "type": "VISUAL", + "shape": [ + 3, + 480, + 640 + ] + } + }, + "output_features": { + "action": { + "type": "ACTION", + "shape": [ + 6 + ] + } + }, + "device": "cuda", + "use_amp": false, + "push_to_hub": true, + "repo_id": "dleon23/act-so101_2", + "private": null, + "tags": null, + "license": null, + "chunk_size": 100, + "n_action_steps": 100, + "vision_backbone": "resnet18", + "pretrained_backbone_weights": "ResNet18_Weights.IMAGENET1K_V1", + "replace_final_stride_with_dilation": false, + "pre_norm": false, + "dim_model": 512, + "n_heads": 8, + "dim_feedforward": 3200, + "feedforward_activation": "relu", + "n_encoder_layers": 4, + "n_decoder_layers": 1, + "use_vae": true, + "latent_dim": 32, + "n_vae_encoder_layers": 4, + "temporal_ensemble_coeff": null, + "dropout": 0.1, + "kl_weight": 10.0, + "optimizer_lr": 1e-05, + "optimizer_weight_decay": 0.0001, + "optimizer_lr_backbone": 1e-05 + }, + "output_dir": "outputs/train/so101_act", + "job_name": "so101_act_training", + "resume": false, + "seed": 1000, + "num_workers": 4, + "batch_size": 8, + "steps": 200000, + "eval_freq": 10000, + "log_freq": 200, + "save_checkpoint": true, + "save_freq": 10000, + "use_policy_training_preset": true, + "optimizer": { + "type": "adamw", + "lr": 1e-05, + "weight_decay": 0.0001, + "grad_clip_norm": 10.0, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08 + }, + "scheduler": null, + "eval": { + "n_episodes": 50, + "batch_size": 50, + "use_async_envs": false + }, + "wandb": { + "enable": false, + "disable_artifact": false, + "project": "lerobot", + "entity": null, + "notes": null, + "run_id": null, + "mode": null + } +} \ No newline at end of file diff --git a/200000/training_state/optimizer_param_groups.json b/200000/training_state/optimizer_param_groups.json new file mode 100644 index 0000000000000000000000000000000000000000..33506df097b87c64b2aa15765ab89706780c0dc0 --- /dev/null +++ b/200000/training_state/optimizer_param_groups.json @@ -0,0 +1,191 @@ +[ + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 0, + 1, + 2, + 3, + 4, + 5, + 6, + 7, + 8, + 9, + 10, + 11, + 12, + 13, + 14, + 15, + 16, + 17, + 18, + 19, + 20, + 21, + 22, + 23, + 24, + 25, + 26, + 27, + 28, + 29, + 30, + 31, + 32, + 33, + 34, + 35, + 36, + 37, + 38, + 39, + 40, + 41, + 42, + 43, + 44, + 45, + 46, + 47, + 48, + 49, + 50, + 51, + 52, + 53, + 54, + 55, + 56, + 57, + 58, + 59, + 60, + 61, + 62, + 63, + 64, + 65, + 66, + 67, + 68, + 69, + 70, + 71, + 72, + 73, + 74, + 75, + 76, + 77, + 78, + 79, + 80, + 81, + 82, + 83, + 84, + 85, + 86, + 87, + 88, + 89, + 90, + 91, + 92, + 93, + 94, + 95, + 96, + 97, + 98, + 99, + 100, + 101, + 102, + 103, + 104, + 105, + 106, + 107, + 108, + 109, + 110, + 111, + 112, + 113, + 114, + 115, + 116, + 117, + 118, + 119, + 120, + 121, + 122, + 123, + 124, + 125, + 126, + 127, + 128, + 129, + 130, + 131, + 132 + ] + }, + { + "lr": 1e-05, + "betas": [ + 0.9, + 0.999 + ], + "eps": 1e-08, + "weight_decay": 0.0001, + "amsgrad": false, + "maximize": false, + "foreach": null, + "capturable": false, + "differentiable": false, + "fused": null, + "decoupled_weight_decay": true, + "params": [ + 133, + 134, + 135, + 136, + 137, + 138, + 139, + 140, + 141, + 142, + 143, + 144, + 145, + 146, + 147, + 148, + 149, + 150, + 151, + 152 + ] + } +] \ No newline at end of file diff --git a/200000/training_state/optimizer_state.safetensors b/200000/training_state/optimizer_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..2301d63c69ed1d0d2214949450cbb0ea2948bf6e --- /dev/null +++ b/200000/training_state/optimizer_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c8edb8ee64f01be59e48c0f7fce1bb57278c349ec5de315dcf696b6ad1d17c9f +size 412817652 diff --git a/200000/training_state/rng_state.safetensors b/200000/training_state/rng_state.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..41f8ec4ab241503c90227660db0d384f659ba86b --- /dev/null +++ b/200000/training_state/rng_state.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:51fba9ba76b6b2d1a0340591e81f02e693d4833f6b55a55db3a1f5315369f5da +size 15708 diff --git a/200000/training_state/training_step.json b/200000/training_state/training_step.json new file mode 100644 index 0000000000000000000000000000000000000000..bcef8d8b05253341579ace0987c46cee6885c961 --- /dev/null +++ b/200000/training_state/training_step.json @@ -0,0 +1,3 @@ +{ + "step": 200000 +} \ No newline at end of file