ZUO66/imagewam-r1lite-place-multi-cups

ImageWAM: R1Lite place multi cups

Final 20-epoch / step 15,280 inference checkpoint for the R1Lite right arm. This is a full fine-tune, not a LoRA adapter. It contains visual/action MoT weights and the trained proprio encoder, not optimizer or resumable state. No robot rollout success rate is claimed by this release.

Training and inference code and real-robot guide.

Files

• step_015280.pt: final visual/action/proprio inference weights.
• config.yaml: resolved training architecture/protocol with portable paths.
• dataset_stats.json: training normalization statistics; use these unchanged.
• manifest.json: byte counts and SHA256 checksums.

Protocol

110 episodes, 61,068 frames at 15 FPS, fixed dataset revision 935cd0d7f397573616435d245f558ab160da27c5. Head RGB and right-wrist RGB are independently resized to 224 x 224 and joined horizontally into 224 x 448. Actions are 7D absolute right-arm targets: six joints followed by right gripper. Current 7D proprio conditions the model. Past 16 actions predict future 16; q01/q99 normalization, boundary padding retained.

I2I source is (current latent + Gaussian noise) / sqrt(2). A2A source is normalized action history plus 0.5 * Gaussian noise on all seven dimensions. Visual and action endpoint-consistency objectives are enabled. Action-loss gradients into the visual conditioning branch are blocked. Qwen context length is 128; the current proprio adds one condition token. Training used 8 x A100 80 GB, microbatch 10, global batch 80, no accumulation, bf16, ZeRO-1, cosine LR 1e-4 to 1e-6, 764 warmup steps and 20 epochs.

Download and inference

hf download ZUO66/imagewam-r1lite-place-multi-cups --local-dir checkpoints/r1lite_cups

Use scripts/r1lite/infer.py in HumanWAM with the matching model config, external FLUX.2 base weights/autoencoder and Qwen3-4B. The release does not redistribute these frozen dependencies. The full checkpoint overwrites the visual/action initialization; the action initializer is not needed for inference. Prompt: A video recorded from a robot's point of view executing the following instruction: place multi cups. Use one visual step and one action step with fresh noise for each policy call. Use executed command history rather than predicted-but-unexecuted actions.

Limitations and attribution

This checkpoint is task-, camera- and robot-specific. Validate offline first; do not send unverified predictions directly to a physical robot. The visual/action initialization derives from FLUX.2-klein-base-4B (Apache-2.0); ImageWAM implementation retains its upstream MIT license. Dataset attribution and terms remain with at237299966/r1lite_place_multi_cups.