Unified checkpoint collection for the RoboSynChallenge submission (code: https://github.com/OuyangZhenqiu/RoboSynChallenge-submission). One subdirectory per task; run bash eval.sh from the code repo (checkpoints default to checkpoints/, i.e. these subdirectory names). - All ten tasks are self-trained. sampleloading switched to the AWR advantage-weighted checkpoint on 2026-10-03 (13/90 = 14.4% merged over three 30-ep runs: seed1000 7/30 + 2/30, seed2000 4/30; official ckpt same caliber 3.3%). Single-run spread on this task is large (sim epsilon flips borderline episodes); the merged value is the unbiased estimate. - Self-trained ten-task average 59.1% (nine-task average 60.7%).
Independent publisher
Ouyang Zhenqiu
oyzq941353193
Models in Library1
Datasets in Library0
Models on Hugging Face10
Followers—
Compare the published evidence
This profile connects the 1 model and dataset record shown below to their upstream cards and sources. The library counts describe SAVRN's coverage, rather than this publisher's entire output.
Compare model specifications, source dates, and hardware estimates on the individual records. Publisher identity verification is separate from a model's quality, deployment readiness, or suitability for your work.