Delete pytorch_model.bin with huggingface_hub

Delete final_logic_giant/pytorch_model.bin with huggingface_hub

Delete final_logic_giant/config.json with huggingface_hub

Delete imprint/sft_final.pt with huggingface_hub

Delete imprint/sft_epoch3.pt with huggingface_hub

Delete imprint/sft_epoch2.pt with huggingface_hub

Delete imprint/sft_epoch1.pt with huggingface_hub

Delete base/pearl_meta.json with huggingface_hub

Delete base/probe_baseline.json with huggingface_hub

Delete base/weights_step1499.pt with huggingface_hub

rename base/weights_step1500.pt → base/Anvaya-Rabbit-2.3B-base.pt

docs: reframe as PoC, add step-1500 internal evals, honest lm-eval disclaimer

Upload folder using huggingface_hub

Upload eval/rabbit_eval_base.txt with huggingface_hub

Upload eval/rabbit_eval_base.txt with huggingface_hub

Upload folder using huggingface_hub

model card: CC-BY-NC-SA 4.0, lineage, dataset curriculum, base/imprint structure

Upload eval/rabbit_eval_report.txt with huggingface_hub

Upload folder using huggingface_hub

Upload final_logic_giant/config.json with huggingface_hub

Upload final_logic_giant/pytorch_model.bin with huggingface_hub

Upload EVAL_REPORT_20260502.txt with huggingface_hub

Golden Checkpoint: Step 1397 (Logic Giant Saturation)

initial commit