LeWAM-Simulation-Checkpoint
Collection
LeWAM v1 checkpoints for simulation tasks: OGBench scene, puzzle (3x3), cube, DexMimicGen drawer, and PushT. • 10 items • Updated
LeWAM PointMaze Large
Goal-reaching checkpoint in the LeWAM v1 format. Files are stored at the repository root:
lewam_best.pt
lewam_config.json
Download both files into $STABLEWM_HOME/checkpoints/<run_name>/, then load with the v1 code:
from lewam.eval.loaders import load_lewam
model, config = load_lewam("<run_name>")
The current v1 release does not include a task-specific evaluation config for this environment.
Dataset: LeWAM/lewam-pointmaze-large.
The trained weights are preserved exactly. Strict model loading, attention masks, encoder output, goal-conditioned action prediction, and dynamics prediction were checked against the source checkpoint. No optimizer state is included.
SHA256 of lewam_best.pt: 76e8ceb6f8bea8f31b078f51337b3f7b2bb8b7177165e374ddd8d9e00badf177.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support