31.8 GB

Ctrl+K

1 contributor

History: 135 commits

Maggio33

auto ckpt ckpt_320k.pt

ffabb00 verified 8 days ago

bpe16m_10B
Upload bpe16m_10B/ckpt.pt with huggingface_hub 11 days ago
bpe16m_3.2B
Upload bpe16m_3.2B/ckpt.pt with huggingface_hub 11 days ago
bpe16m_6B
Upload bpe16m_6B/ckpt.pt with huggingface_hub 11 days ago
bpe32m_baseline
Upload bpe32m_baseline/ckpt.pt with huggingface_hub 11 days ago
ctrl_arcmix_resume
Model card update (BLiMP recomputed after harness fix; data-attribution experiments) + READMEs for 4 experiment folders 8 days ago
decontam_verify
decontam-verify kept sample for Hart 8 days ago
eval
Upload eval/bpb_bin.py with huggingface_hub 9 days ago
forkB_arcmix_edu
Model card update (BLiMP recomputed after harness fix; data-attribution experiments) + READMEs for 4 experiment folders 8 days ago
r3A_arcmix_edu_clean
Model card update (BLiMP recomputed after harness fix; data-attribution experiments) + READMEs for 4 experiment folders 8 days ago
r3B_arcmix_qa2x
Model card update (BLiMP recomputed after harness fix; data-attribution experiments) + READMEs for 4 experiment folders 8 days ago
r4B_arcmix_qa2x_seed1338
README for round-4 arms (r4B seed control, r4K anchor): why they exist, results 340-400k 8 days ago
r4K_anchor_arcmix_qa1
README for round-4 arms (r4B seed control, r4K anchor): why they exist, results 340-400k 8 days ago
run_149m
Upload run_149m/ckpt.pt with huggingface_hub 10 days ago
run_16m_arcmix
Upload run_16m_arcmix/ckpt.pt with huggingface_hub 10 days ago
run_16m_expanded
Upload run_16m_expanded/ckpt.pt with huggingface_hub 11 days ago
run_32m_16b
Upload run_32m_16b/ckpt.pt with huggingface_hub 10 days ago
run_32m_18b
Upload run_32m_18b/ckpt.pt with huggingface_hub 10 days ago
v1_128m
auto ckpt ckpt_320k.pt 8 days ago
v1_muon
remove mislabeled ckpt (was step-320000, autopush-race); canonical=ckpt_320k 8 days ago
v2_muon
auto ckpt 160000 9 days ago
v2b
auto ckpt 80000 8 days ago
vals
forward-hold val vals/v2_mix_val.bin 9 days ago
.gitattributes

1.77 kB
decontam-verify kept sample for Hart 8 days ago
README.md

14.1 kB
Model card update (BLiMP recomputed after harness fix; data-attribution experiments) + READMEs for 4 experiment folders 8 days ago
blend_edu.py

2.95 kB
Upload blend_edu.py with huggingface_hub 8 days ago
blend_rb4.py

6.22 kB
RB4 runda 3: blend_rb4/extract/launch_r3 + build_v2_blend fix (eot z tokenizera, brak klucza synth = stop); skan egress GREEN 8 days ago
board_overlay_v5.png

154 kB
xet

Upload board_overlay_v5.png with huggingface_hub 11 days ago
build_decontam_index.py

4.03 kB
Upload build_decontam_index.py with huggingface_hub 8 days ago
build_v2_blend.py

9.87 kB
RB4 runda 3: blend_rb4/extract/launch_r3 + build_v2_blend fix (eot z tokenizera, brak klucza synth = stop); skan egress GREEN 8 days ago
ckpt_autopush.py

3.09 kB
Upload ckpt_autopush.py with huggingface_hub 8 days ago
dump_kept_sample.py

2.17 kB
Upload dump_kept_sample.py with huggingface_hub 8 days ago
eval_push.py

1.95 kB
Upload eval_push.py with huggingface_hub 8 days ago
extract_edu_from_blend.py

2.59 kB
RB4 runda 3: blend_rb4/extract/launch_r3 + build_v2_blend fix (eot z tokenizera, brak klucza synth = stop); skan egress GREEN 8 days ago
fabryka_push.py

4.83 kB
Upload fabryka_push.py with huggingface_hub 8 days ago
fork_launch.sh

2.34 kB
Upload fork_launch.sh with huggingface_hub 8 days ago
glint_149m_results.json

113 Bytes
Upload glint_149m_results.json with huggingface_hub 10 days ago
glint_16m_arcmix_results.json

114 Bytes
Upload glint_16m_arcmix_results.json with huggingface_hub 10 days ago
glint_16m_expanded_results.json

113 Bytes
Upload glint_16m_expanded_results.json with huggingface_hub 11 days ago
glint_32m_16b_results.json

113 Bytes
Upload glint_32m_16b_results.json with huggingface_hub 10 days ago
glint_32m_18b_results.json

113 Bytes
Upload glint_32m_18b_results.json with huggingface_hub 10 days ago
glint_32m_muon_results.json

114 Bytes
Upload glint_32m_muon_results.json with huggingface_hub 9 days ago
glint_parity_eval.py

14 kB
fix: loader dopasowywał fenomeny po podciągu (6 fenomenów ×2, 73000 zamiast 67000); dokładne dopasowanie + asercja 67000 8 days ago
launch_r3.sh

2.08 kB
RB4 runda 3: blend_rb4/extract/launch_r3 + build_v2_blend fix (eot z tokenizera, brak klucza synth = stop); skan egress GREEN 8 days ago
make_board_overlay.py

2.61 kB
Upload make_board_overlay.py with huggingface_hub 11 days ago
make_progress_chart.py

3.55 kB
Upload make_progress_chart.py with huggingface_hub 11 days ago
pod_prep.sh

1.84 kB
Upload pod_prep.sh with huggingface_hub 8 days ago
progress_v5.png

136 kB
xet

Upload progress_v5.png with huggingface_hub 11 days ago
scaling_v5.png

97.8 kB
Upload scaling_v5.png with huggingface_hub 10 days ago
tokenizer.json

830 kB
Upload tokenizer.json with huggingface_hub 11 days ago
train_gpt_ref.py

24 kB
model-def (GPT-ref Qwen3-arch) for same-harness loader 9 days ago