14.4 GB

Ctrl+K

1 contributor

History: 102 commits

Maggio33

Upload eval_push.py with huggingface_hub

5c776db verified 11 days ago

bpe16m_10B
Upload bpe16m_10B/ckpt.pt with huggingface_hub 15 days ago
bpe16m_3.2B
Upload bpe16m_3.2B/ckpt.pt with huggingface_hub 15 days ago
bpe16m_6B
Upload bpe16m_6B/ckpt.pt with huggingface_hub 15 days ago
bpe32m_baseline
Upload bpe32m_baseline/ckpt.pt with huggingface_hub 14 days ago
decontam_verify
decontam-verify kept sample for Hart 11 days ago
eval
Upload eval/bpb_bin.py with huggingface_hub 12 days ago
forkB_arcmix_edu
auto ckpt ckpt_340k.pt 11 days ago
run_149m
Upload run_149m/ckpt.pt with huggingface_hub 13 days ago
run_16m_arcmix
Upload run_16m_arcmix/ckpt.pt with huggingface_hub 14 days ago
run_16m_expanded
Upload run_16m_expanded/ckpt.pt with huggingface_hub 14 days ago
run_32m_16b
Upload run_32m_16b/ckpt.pt with huggingface_hub 13 days ago
run_32m_18b
Upload run_32m_18b/ckpt.pt with huggingface_hub 13 days ago
v1_128m
auto ckpt 120000 12 days ago
v1_muon
remove mislabeled ckpt (was step-320000, autopush-race); canonical=ckpt_320k 11 days ago
v2_muon
auto ckpt 160000 12 days ago
v2b
auto ckpt 80000 12 days ago
vals
forward-hold val vals/v2_mix_val.bin 12 days ago
.gitattributes

1.77 kB
decontam-verify kept sample for Hart 11 days ago
README.md

16.1 kB
card: honest wiki correction (byte_ppl 2.372 / BPB 1.246 / eff ~75.9); maintainer re-benchmark PR#78; ARC-Easy 47.94 exact-match 12 days ago
blend_edu.py

2.95 kB
Upload blend_edu.py with huggingface_hub 11 days ago
board_overlay_v5.png

154 kB
xet

Upload board_overlay_v5.png with huggingface_hub 14 days ago
build_decontam_index.py

4.03 kB
Upload build_decontam_index.py with huggingface_hub 11 days ago
build_v2_blend.py

9.37 kB
Upload build_v2_blend.py with huggingface_hub 11 days ago
ckpt_autopush.py

3.09 kB
Upload ckpt_autopush.py with huggingface_hub 11 days ago
dump_kept_sample.py

2.17 kB
Upload dump_kept_sample.py with huggingface_hub 11 days ago
eval_push.py

1.74 kB
Upload eval_push.py with huggingface_hub 11 days ago
fabryka_push.py

4.83 kB
Upload fabryka_push.py with huggingface_hub 11 days ago
fork_launch.sh

2.34 kB
Upload fork_launch.sh with huggingface_hub 11 days ago
glint_149m_results.json

113 Bytes
Upload glint_149m_results.json with huggingface_hub 13 days ago
glint_16m_arcmix_results.json

114 Bytes
Upload glint_16m_arcmix_results.json with huggingface_hub 14 days ago
glint_16m_expanded_results.json

113 Bytes
Upload glint_16m_expanded_results.json with huggingface_hub 14 days ago
glint_32m_16b_results.json

113 Bytes
Upload glint_32m_16b_results.json with huggingface_hub 13 days ago
glint_32m_18b_results.json

113 Bytes
Upload glint_32m_18b_results.json with huggingface_hub 13 days ago
glint_32m_muon_results.json

114 Bytes
Upload glint_32m_muon_results.json with huggingface_hub 13 days ago
glint_parity_eval.py

13.8 kB
glint_parity_eval: arch-aware loader (Qwen3 cfg from ckpt) - fix 64M maintainer load 12 days ago
make_board_overlay.py

2.61 kB
Upload make_board_overlay.py with huggingface_hub 14 days ago
make_progress_chart.py

3.55 kB
Upload make_progress_chart.py with huggingface_hub 14 days ago
pod_prep.sh

1.84 kB
Upload pod_prep.sh with huggingface_hub 11 days ago
progress_v5.png

136 kB
xet

Upload progress_v5.png with huggingface_hub 14 days ago
scaling_v5.png

97.8 kB
Upload scaling_v5.png with huggingface_hub 13 days ago
tokenizer.json

830 kB
Upload tokenizer.json with huggingface_hub 15 days ago
train_gpt_ref.py

24 kB
model-def (GPT-ref Qwen3-arch) for same-harness loader 12 days ago
train_tokenizer.py

2.88 kB
Upload train_tokenizer.py with huggingface_hub 14 days ago