Add h_rdt real checkpoint for Remove_the_cap_turn_on_the_faucet_and_fill_the_bottle_with_water
Browse files
h_rdt/real/Remove_the_cap_turn_on_the_faucet_and_fill_the_bottle_with_water/README.md
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
tags:
|
| 3 |
+
- model_hub_mixin
|
| 4 |
+
- pytorch_model_hub_mixin
|
| 5 |
+
---
|
| 6 |
+
|
| 7 |
+
This model has been pushed to the Hub using the [PytorchModelHubMixin](https://huggingface.co/docs/huggingface_hub/package_reference/mixins#huggingface_hub.PyTorchModelHubMixin) integration:
|
| 8 |
+
- Code: https://huggingface.co/hongzhe2002/H-RDT/
|
| 9 |
+
- Paper: [More Information Needed]
|
| 10 |
+
- Docs: [More Information Needed]
|
h_rdt/real/Remove_the_cap_turn_on_the_faucet_and_fill_the_bottle_with_water/config.json
ADDED
|
@@ -0,0 +1,76 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"act_adaptor": "mlp3x_silu",
|
| 3 |
+
"act_adaptor_width_multiplier": 1,
|
| 4 |
+
"act_pos_emb_config": [
|
| 5 |
+
[
|
| 6 |
+
"state",
|
| 7 |
+
1
|
| 8 |
+
],
|
| 9 |
+
[
|
| 10 |
+
"action",
|
| 11 |
+
16
|
| 12 |
+
]
|
| 13 |
+
],
|
| 14 |
+
"action_dim": 36,
|
| 15 |
+
"ema": {
|
| 16 |
+
"inv_gamma": 1.0,
|
| 17 |
+
"max_value": 0.9999,
|
| 18 |
+
"min_value": 0.0,
|
| 19 |
+
"power": 0.75,
|
| 20 |
+
"update_after_step": 0
|
| 21 |
+
},
|
| 22 |
+
"hrdt": {
|
| 23 |
+
"action_decoder_hidden_multiplier": 1,
|
| 24 |
+
"depth": 16,
|
| 25 |
+
"ffn_dim_multiplier": null,
|
| 26 |
+
"hidden_size": 2176,
|
| 27 |
+
"multiple_of": 256,
|
| 28 |
+
"norm_eps": 1e-05,
|
| 29 |
+
"num_heads": 16,
|
| 30 |
+
"num_kv_heads": 8,
|
| 31 |
+
"output_size": 36,
|
| 32 |
+
"use_flash_attn": true
|
| 33 |
+
},
|
| 34 |
+
"img_adapter": "mlp2x_silu",
|
| 35 |
+
"img_pos_emb_config": [
|
| 36 |
+
[
|
| 37 |
+
"image",
|
| 38 |
+
[
|
| 39 |
+
1,
|
| 40 |
+
1,
|
| 41 |
+
-729
|
| 42 |
+
]
|
| 43 |
+
]
|
| 44 |
+
],
|
| 45 |
+
"lang_adapter": "mlp2x_silu",
|
| 46 |
+
"lang_pos_emb_config": [
|
| 47 |
+
[
|
| 48 |
+
"language",
|
| 49 |
+
-1024
|
| 50 |
+
]
|
| 51 |
+
],
|
| 52 |
+
"max_img_len": 729,
|
| 53 |
+
"max_lang_len": 1024,
|
| 54 |
+
"mode": "finetune",
|
| 55 |
+
"noise_scheduler": {
|
| 56 |
+
"num_inference_timesteps": 5,
|
| 57 |
+
"sampler_type": "uniform",
|
| 58 |
+
"timestep_max": 0.999
|
| 59 |
+
},
|
| 60 |
+
"pred_horizon": 16,
|
| 61 |
+
"pretrained_backbone_path": "./checkpoints/pretrain-0618/checkpoint-500000/checkpoints/pretrain-0618/checkpoint-500000/pytorch_model.bin",
|
| 62 |
+
"st_adaptor": "mlp3x_silu",
|
| 63 |
+
"st_adaptor_width_multiplier": 1,
|
| 64 |
+
"state_dim": 32,
|
| 65 |
+
"text": {
|
| 66 |
+
"feature_dim": 4096
|
| 67 |
+
},
|
| 68 |
+
"time_noise": {
|
| 69 |
+
"a": 5,
|
| 70 |
+
"beta_m": 100
|
| 71 |
+
},
|
| 72 |
+
"training_mode": "lang",
|
| 73 |
+
"vision": {
|
| 74 |
+
"feature_dim": 2176
|
| 75 |
+
}
|
| 76 |
+
}
|
h_rdt/real/Remove_the_cap_turn_on_the_faucet_and_fill_the_bottle_with_water/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:21869fd82df353a2e7b1f7990b31a1b86f637ce820af22d0954b7683b5af90a7
|
| 3 |
+
size 4119255599
|