Puddings22
/

MIRAI-GlassBox-LLM

Text Generation

interpretability

signal-processing

causal-attribution

Model card Files Files and versions

MIRAI-GlassBox-LLM

8.67 GB

Ctrl+K

Ctrl+K

1 contributor

History: 13 commits

Puddings22's picture

Add prominent undertrained disclaimer; disable the inference widget

397e9ed verified about 2 months ago

REBUILD
REBUILD: keep only mirai5_clean (arch) and mirai5_ddp_clean (trainer); both self-contained about 2 months ago
bin
Add MIRAI code, verification suite, frozen data, paper, license about 2 months ago
cache
Add MIRAI code, verification suite, frozen data, paper, license about 2 months ago
checkpoints
Add checkpoints/m5_ft_25b4k_surgical/final.pt about 2 months ago
experiments
Add MIRAI code, verification suite, frozen data, paper, license about 2 months ago
paper
Add MIRAI code, verification suite, frozen data, paper, license about 2 months ago
tokenizer
Ship the tokenizer; correct precision (fp32 weights, bf16 compute); document the inspection panels about 2 months ago
.gitattributes

1.65 kB
Ship the tokenizer; correct precision (fp32 weights, bf16 compute); document the inspection panels about 2 months ago
CITATION.cff

810 Bytes
Point citation metadata at the real repo URL about 2 months ago
LICENSE

11.4 kB
Add MIRAI code, verification suite, frozen data, paper, license about 2 months ago
NOTICE

348 Bytes
Add MIRAI code, verification suite, frozen data, paper, license about 2 months ago
README.md

13.9 kB
Add prominent undertrained disclaimer; disable the inference widget about 2 months ago
config.json

711 Bytes
Add config.json so Hub download counting works about 2 months ago
verify_repo.py

2.97 kB
Ship the tokenizer; correct precision (fp32 weights, bf16 compute); document the inspection panels about 2 months ago