020768b261
README
2026-10-25 01:32:32 +02:00
3ead3824c9
README
2026-10-25 01:26:46 +02:00
fada5550d3
oprava
2026-10-25 01:18:39 +02:00
ac0a0e386f
oprava
2026-10-25 00:59:11 +02:00
67061cf373
Add Aya 300k learning-rate study and benchmark results
...
Document six completed LR runs as preparation for full Aya training on Perun. Record selection of LR 0.00010 by lowest final validation loss, selected-model lm-eval and partial SK-bench results, reproducible provenance, and dataset-access/context-truncation limitations. Preserve existing 30k and Titan material.
2026-10-25 00:33:24 +02:00
29f94a93d9
readme_image
2026-10-17 22:28:48 +02:00
8113c49e4f
readme_images
2026-10-17 22:26:19 +02:00
2302243974
30k_experiment_hruska
2026-10-17 22:23:38 +02:00
94d0fce9f4
README
2026-10-04 23:59:25 +02:00
6c9ad5bd3d
README
2026-10-04 23:56:42 +02:00
a45cf5fdee
README
2026-10-04 23:51:17 +02:00
581ae22dbb
lmeval
2026-10-04 23:50:18 +02:00
d4b00793b3
README
2026-09-30 22:25:01 +02:00
cacb9a6375
change
2026-09-30 21:33:57 +02:00
796b29f91b
README
2026-09-30 21:25:31 +02:00
c92fec1e6b
README
2026-09-30 21:05:56 +02:00
e18a0f0853
README
2026-09-30 21:03:13 +02:00
4f47e276ff
update
2026-08-28 02:02:57 +02:00
0dd438adca
update_README
2026-08-28 01:50:38 +02:00
0bfde352ab
update_README
2026-08-28 01:49:13 +02:00
20bcf01be7
update
2026-08-20 22:59:57 +02:00
6962418bb1
update
2026-08-19 02:34:46 +02:00
b927cd6dad
update
2026-08-19 01:31:43 +02:00
a9b8a5a6fc
update
2026-08-19 01:29:49 +02:00
50d0581424
update
2026-08-19 01:28:10 +02:00
cdd3b3fc30
Oprava
2026-06-11 22:16:27 +02:00
0ef6cf229c
first commit
2026-06-11 21:21:18 +02:00
