22.8 GB

Ctrl+K

2 contributors

card: single-source = Latarnik draft verbatim (base-model framing, EOS disclosure, converged/plateau, 4-decimal tail, 25-row, dataset-linked) - end churn

0687131 verified 1 day ago

.gitattributes

1.52 kB
initial commit 4 days ago
README.md

9.9 kB
card: single-source = Latarnik draft verbatim (base-model framing, EOS disclosure, converged/plateau, 4-decimal tail, 25-row, dataset-linked) - end churn 1 day ago
ckpt_10000.safetensors

992 MB
xet

Add ckpt_10000 snapshot (val 2.878, BPB 0.961, ~20%); tighten forecast 2.5-2.63 4 days ago
ckpt_12000.safetensors

992 MB
xet

Add ckpt_12000 snapshot (val 2.846, BPB 0.947, ~25%); eos_rate 0.2 (model zaczyna emitowac EOS) 4 days ago
ckpt_14000.safetensors

992 MB
xet

Add ckpt_14000 snapshot (val 2.849, BPB 0.938, ~29%); BPB monotone 0.978->0.938, val/loss jumps = single-batch noise (rozstrzygniete) 4 days ago
ckpt_16000.safetensors

992 MB
xet

Add ckpt_16000 snapshot (val 2.789, BPB 0.932, ~33%); BPB monotone 0.978->0.932 4 days ago
ckpt_18000.safetensors

992 MB
xet

Add ckpt_18000 safetensors (step 18000, 36.9% train) 3 days ago
ckpt_2000.safetensors

992 MB
xet

GoLLeM v4 250M PL - research trajectory snapshots (ckpt 2000/6000/8000) 4 days ago
ckpt_20000.safetensors

992 MB
xet

DR: add ckpt_20000 safetensors (Monter) 3 days ago
ckpt_22000.safetensors

992 MB
xet

DR: add ckpt_22000 safetensors (Monter) 3 days ago
ckpt_24000.safetensors

992 MB
xet

DR: add ckpt_24000 safetensors (Monter) 3 days ago
ckpt_26000.safetensors

992 MB
xet

DR: add ckpt_26000 safetensors (Monter) 3 days ago
ckpt_28000.safetensors

992 MB
xet

DR: add ckpt_28000 safetensors (Monter) 3 days ago
ckpt_30000.safetensors

992 MB
xet

DR: add ckpt_30000 safetensors (Monter) 3 days ago
ckpt_32000.safetensors

992 MB
xet

DR: add ckpt_32000 safetensors (Monter) 3 days ago
ckpt_34000.safetensors

992 MB
xet

DR: add ckpt_34000 safetensors (Monter) 2 days ago
ckpt_36000.safetensors

992 MB
xet

DR#8: ckpt_36000 safetensors (992MB, 208 keys) 2 days ago
ckpt_38000.safetensors

992 MB
xet

DR#9: ckpt_38000 safetensors (992MB, 208 keys) 2 days ago
ckpt_40000.safetensors

992 MB
xet

DR#10: ckpt_40000 safetensors 2 days ago
ckpt_42000.safetensors

992 MB
xet

DR#11: ckpt_42000 safetensors 2 days ago
ckpt_46000.safetensors

992 MB
xet

GoLLeM v4: final ckpt 46000 safetensors (endpoint 0.8700 measured @48828, recompute-verified) 1 day ago
ckpt_48000.safetensors

992 MB
xet

GoLLeM v4: final ckpt 48000 safetensors (endpoint 0.8700 measured @48828, recompute-verified) 1 day ago
ckpt_48828.safetensors

992 MB
xet

GoLLeM v4: final ckpt 48828 safetensors (endpoint 0.8700 measured @48828, recompute-verified) 1 day ago
ckpt_6000.safetensors

992 MB
xet

GoLLeM v4 250M PL - research trajectory snapshots (ckpt 2000/6000/8000) 4 days ago
ckpt_8000.safetensors

992 MB
xet

GoLLeM v4 250M PL - research trajectory snapshots (ckpt 2000/6000/8000) 4 days ago
config.json

3.4 kB
GoLLeM v4: config snapshots -> 48828 (final; 44k rotated-gap honest) 1 day ago
generate.py

1.15 kB
GoLLeM v4 250M PL - research trajectory snapshots (ckpt 2000/6000/8000) 4 days ago
modeling_gollem.py

4.22 kB
GoLLeM v4 250M PL - research trajectory snapshots (ckpt 2000/6000/8000) 4 days ago
tokenizer.json

2.51 MB
GoLLeM v4 250M PL - research trajectory snapshots (ckpt 2000/6000/8000) 4 days ago