Views
No views yet
docker/benchmark/). It is published so the on-device benchmark job can load
real trained weights: half precision saturates on the outlier activation
channels that only appear in trained weights, so an FP16 engine checked against
a randomly initialised policy passes a check it would fail in deployment.smolvla_recap (10-step weights run with snapflow_enabled)sancov/smolvla_recap_libero_spatialobservation.images.image, observation.images.image2 (3x256x256), observation.state (8)action (7), chunk size 20snapflow_enabled: true in config.json; it is
not a SnapFlow-distilled model. For that, use
sancov/snapflow_ctpf08-test.
Kept because earlier benchmark runs used this configuration.