Research artefact of SN56 text tournament replay; validator-style round-trip. The other files are the trainer's output directory, uploaded unchanged. The repo is downloaded again and scored on the task's held-out rows to check that the Hub copy scores exactly as the local copy did.
Independent publisher
Zaydens
zaydens
LLM post-training — preference learning (DPO), reinforcement learning (GRPO), and instruction tuning. Interested in adaptive hyperparameter strategies, reward shaping, and model architecture analysis.
Models in Library2
Datasets in Library0
Models on Hugging Face45
Followers—
Compare the published evidence
This profile connects the 2 model and dataset records 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.
Models
Research artefact of SN56 text tournament replay; validator-style round-trip. The other files are the trainer's output directory, uploaded unchanged. The repo is downloaded again and scored on the task's held-out rows to check that the Hub copy scores exactly as the local copy did.