CoolFace
Datasetpublic

tomyimkc/sophia-training-results

Sophia Training Results All training history results from the sophia-agi repository. Contents 824 result files across all Sophia four-theme training experiments: Prosoche — interactive decision-making, behavioral batteries, focus experiments RunPod training — LoRA training eval ladders (Qwen2.5-3B, multiple seeds/epochs) Continual learning — sequential curriculum, QA judged results World model — on-device, adaptive, real-corpus experiments Coherence reframe — NLI… See the full description on the dataset page: https://huggingface.co/datasets/tomyimkc/sophia-training-results.

sourceHugging Faceapache-2.0updated 2mo agoView on Hugging Face
1likes72downloads

tomyimkc/sophia-training-results · main · files are served by the source, never re-hosted here