Reinforcement Learning
stable-baselines3
deep-reinforcement-learning
sb3-contrib
mujoco
Walker2d-v4
weight-editing
Eval Results (legacy)
Instructions to use erickvs/tqc-Walker2d-v4-hand-edited with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- stable-baselines3
How to use erickvs/tqc-Walker2d-v4-hand-edited with stable-baselines3:
from huggingface_sb3 import load_from_hub checkpoint = load_from_hub( repo_id="erickvs/tqc-Walker2d-v4-hand-edited", filename="{MODEL FILENAME}.zip", ) - Notebooks
- Google Colab
- Kaggle
Download edit_program.json from erickvs/tqc-Walker2d-v4-hand-edited: direct link, hf CLI and curl.
- Browser
- Download file 1.18 kB
-
https://huggingface.co/erickvs/tqc-Walker2d-v4-hand-edited/resolve/main/edit_program.json
- Command line
-
hf download hf://erickvs/tqc-Walker2d-v4-hand-edited/edit_program.json
-
curl -L -o edit_program.json https://huggingface.co/erickvs/tqc-Walker2d-v4-hand-edited/resolve/main/edit_program.json
1.18 kB
| { | |
| "wsp_lab_candidate": "57e344eb718fff61", | |
| "experiment": "009", | |
| "base": "sb3/tqc-Walker2d-v3", | |
| "repair": [], | |
| "program": [ | |
| { | |
| "args": { | |
| "k": 1.5, | |
| "outputs": [ | |
| "foot_joint", | |
| "foot_left_joint" | |
| ] | |
| }, | |
| "kind": "output_gain" | |
| }, | |
| { | |
| "args": { | |
| "k": 1.7, | |
| "outputs": [ | |
| "thigh_joint", | |
| "thigh_left_joint" | |
| ] | |
| }, | |
| "kind": "output_gain" | |
| }, | |
| { | |
| "args": { | |
| "delta": -0.05, | |
| "direction": "boost_z" | |
| }, | |
| "kind": "shift_input" | |
| }, | |
| { | |
| "args": { | |
| "delta": -0.025, | |
| "direction": "boost_x" | |
| }, | |
| "declared": { | |
| "note": "delta = -0.125 + 0.5*0.2 so the net mean shift stays -0.125", | |
| "settled_mean_torso_angle": 0.2, | |
| "source": "traces.query run 1068 items 0:4 steps 300:1000" | |
| }, | |
| "kind": "shift_input" | |
| }, | |
| { | |
| "args": { | |
| "by": "torso.angle", | |
| "direction": "boost_x", | |
| "gain": -0.5 | |
| }, | |
| "kind": "shift_input" | |
| } | |
| ], | |
| "changed_actor_values": { | |
| "actor.latent_pi.0.weight": 256, | |
| "actor.latent_pi.0.bias": 256, | |
| "actor.latent_pi.2.weight": 0, | |
| "actor.latent_pi.2.bias": 0, | |
| "actor.mu.weight": 1024, | |
| "actor.mu.bias": 4 | |
| }, | |
| "env": { | |
| "id": "Walker2d-v4", | |
| "kwargs": {} | |
| } | |
| } |