98.8 GB

Ctrl+K

1 contributor

History: 40 commits

weilan55

index: add g3s (v7 text-only corpus)

32c9267 verified 15 days ago

g3c
g3c: TRUE final checkpoint iter 6413 (global_step 6412) — supersedes the pre-OOM 6001 snapshot; the resume wrote to .61/.62, not the original .101/.31 20 days ago
g3d
g3d: complete model/ with all 8 distcp shards (ranks 4-7 were on the second trainer node) about 1 month ago
g3e
g3e: final checkpoint + run config about 1 month ago
g3f
g3f: final checkpoint + run config about 1 month ago
g3g
g3g: README for final iter 4581; now step-aligned with g3e/g3f about 1 month ago
g3h
g3h: complete model/ with all 8 distcp shards (ranks 4-7 were on the second trainer node) about 1 month ago
g3j
g3j: final checkpoint iter 4581 — warm-start from g3e iter 4581, g3f architecture (block 3) about 1 month ago
g3k
g3k: v3h6r finetuned on v4_train, 2 epochs (2946 steps), acc 0.629 28 days ago
g3l
g3l: final checkpoint iter 7527 — v3h6 draft, v34 corpus, 2 epochs (schedule derived, not pinned) 23 days ago
g3m
g3m: final checkpoint iter 4455 — g3f at max_seq_length 40000 (drops 8000 long rows) 23 days ago
g3n
g3n: final checkpoint iter 4455 — g3m with default rope (theta 1e6) instead of yarn 22 days ago
g3o
g3o: final checkpoint iter 4455 — rope_theta 1e6->5e4 in isolation; theta measured inert (acc +0.00% vs g3n) 21 days ago
g3p
g3p: final checkpoint iter 4581 — v5 corpus at max_seq_length 65536 (two variables vs g3n) 20 days ago
g3q
g3q: final checkpoint iter 9161 — num_epochs 2->4 on v5 (single variable vs g3p) 17 days ago
g3r
g3r: final checkpoint iter 14325 — v6 text-only corpus, zero tool-call supervision 17 days ago
g3s
g3s: final checkpoint iter 18019 — v7 text-only corpus (1.15M rows, 66.5% with reasoning), single variable vs g3r 15 days ago
v3h6r
v3h6r: 4-epoch ablation off v3h4, final iter_0009161 (loss 1.491, acc 0.672) about 1 month ago
.gitattributes

9.56 kB
g3s: final checkpoint iter 18019 — v7 text-only corpus (1.15M rows, 66.5% with reasoning), single variable vs g3r 15 days ago
README.md

8.99 kB
index: add g3s (v7 text-only corpus) 15 days ago