Commit History
20 days ago
-
Upload glint_parity_eval.py with huggingface_hub
Verified signature7a7066f -
auto ckpt 160000
Verified signatured95eca8 -
card: 64M #6 flagship, board-protocol verified (glint bare-prompt: eff 77.51 / ARC 47.94)
Verified signature64a8ba9 -
revert card to pre-64M: hold 64M #5 pending board-protocol (glint bare-prompt) re-measure
Verified signaturea2ffd89 -
card: add 64M #5 flagship (eff 77.70, board-native recompute-verified)
Verified signature562ae51 -
Upload eval/bpb_bin.py with huggingface_hub
Verified signature94b0512 -
Upload eval/board_eval_fabryka.py with huggingface_hub
Verified signature5792063 -
auto ckpt 400000
Verified signature8d40ea8 -
auto ckpt 120000
Verified signature013828d -
auto ckpt 360000
Verified signature29278a2 -
auto ckpt 80000
Verified signatureb937d01 -
auto ckpt 320000
Verified signatureca4fdaf -
auto ckpt 280000
Verified signature97be15a -
auto ckpt 240000
Verified signaturec4b25ae -
auto ckpt 200000
Verified signaturebc6abf1 -
auto ckpt 120000
Verified signature4aa0680 -
v2-muon FWE/Cosmo ckpt 40k for board-native eval
Verified signature7ce52bf -
forward-hold val vals/v2_mix_val.bin
Verified signature56a8291 -
forward-hold val vals/fwe_only_val.bin
Verified signature34030ef -
arcmix-val cross-general forward-hold
Verified signature652da07 -
model-def (GPT-ref Qwen3-arch) for same-harness loader
Verified signaturee88580a -
v1-muon arcmix ckpt 160k for same-harness re-measure
Verified signature0364e96 -
v1-muon arcmix ckpt 80k for same-harness re-measure
Verified signature4f5c1d4 -
v1-muon arcmix ckpt 40k for same-harness re-measure
Verified signaturedad7c1a -
card: document ARC-MIX 9.42B corpus (32M/64M runs) + v2 FineWeb-Edu-dominant blend plan
Verified signature82b7265
21 days ago
-
card: confirmed board entry (#16 32M / #21 16M, maintainer-verified); all models + results; ARC lever = value residuals (competitive intel); 64M A/B roadmap
Verified signature179cfda -
card: Positioning CONFIRMED on board #16/#21 + Roadmap VR-ARC-lever/data-stack/64M-A/B
Verified signaturea665182 -
Upload glint_32m_muon_results.json with huggingface_hub
Verified signaturef881d8b -
Upload glint_32m_18b_results.json with huggingface_hub
Verified signature94aa4c8 -
Upload run_32m_18b/ckpt.pt with huggingface_hub
Verified signaturebe18bc4 -
Upload README.md with huggingface_hub
Verified signature7757b15 -
Upload scaling_v5.png with huggingface_hub
Verified signature0729601 -
Upload run_149m/ckpt.pt with huggingface_hub
Verified signature46e6835 -
Upload README.md with huggingface_hub
Verified signature6d4bb85 -
Upload glint_149m_results.json with huggingface_hub
Verified signaturee1fd699 -
Upload train_gpt_ref.py with huggingface_hub
Verified signatureb3d0010 -
Upload README.md with huggingface_hub
Verified signature02a1315 -
Upload README.md with huggingface_hub
Verified signatureff77f7c -
Upload glint_32m_16b_results.json with huggingface_hub
Verified signature9c8a9c7 -
Upload run_32m_16b/ckpt.pt with huggingface_hub
Verified signaturec853bd4 -
remove bogus stale-copy (v1 train had failed; real result comes post-v1)
Verified signaturea659388 -
Upload glint_32m_16b_results.json with huggingface_hub
Verified signature349d7b7 -
Upload glint_16m_arcmix_results.json with huggingface_hub
Verified signaturee053c3b -
Upload run_16m_arcmix/ckpt.pt with huggingface_hub
Verified signature15ad2a9
22 days ago
-
Upload README.md with huggingface_hub
Verified signature3ac120d -
Upload train_tokenizer.py with huggingface_hub
Verified signature2a8d44e -
Upload README.md with huggingface_hub
Verified signature6a0572b -
Upload progress_v5.png with huggingface_hub
Verified signatured5c9bc3 -
Remove abandoned KARD-synthetic pipeline (pivoted to real-data ARC-curation)
Verified signature866e9cb -
Delete charts/scaling_v5_latarnik.png with huggingface_hub
Verified signaturedd30df7