visible-cx
/

Gemma-4-12B-CoreAI

Text Generation

Model card Files Files and versions

Gemma-4-12B-CoreAI

70.2 GB

Ctrl+K

Ctrl+K

2 contributors

History: 28 commits

HiramKHackenbacker's picture

HiramKHackenbacker

Publish verified Gemma 4 12B raw-Metal 65k pack

a6f63a1 verified about 1 month ago

dense2in-int8
Add dense2in-int8/gemma4_12b_qat_decode_int8lin_pf64 2 months ago
dense2in-pf64
Add dense2in-pf64/gemma4_12b_qat_decode_int4linsym_pf64 2 months ago
dense2in
dense2in: add the missing tokenizer/generation_config.json (fix by addition; byte-identical to the file the two dense2in-pf64 folders already ship) 2 months ago
raw-metal
Publish verified Gemma 4 12B raw-Metal 65k pack about 1 month ago
tiered-v3
12B v3 manifest: register the verify role about 2 months ago
tiered
tiered: ladder README (staged status, digests, per-rung requirements) about 2 months ago
v4
v4: tied-table 12B — one bound vocab file, verify at the rungs that load, accepted_count, activations about 2 months ago
v5
12B v5: remove doubly-nested sidecar path about 1 month ago
.gitattributes

4.71 kB
gemma4_12b_int4_rawmetal raw-Metal int4 pack on the native QAT grid — tokenizer, generation config and stop ids inside; manifest carries sizes and sha256s about 1 month ago
README.md

28.1 kB
model card: v5 ring-extent bundle (single-pass speculative verify) about 1 month ago