Name the proper-noun column for what it measures

Name the proper-noun column for what it measures

Name the proper-noun column for what it measures

Name the proper-noun column for what it measures

Name the proper-noun column for what it measures

v3.1 matrices, three-seed benchmark and eleven-language demo (part 2)

Projection, seed and GPU model all act on the same free space

The plain matrices hold on a tightly written prompt

The plain matrices are not blind to what is stated once

v3: calibrated against the stock 32B, image tokens for the 8B, demo folder

Point people arriving from older tutorials at the current files

Point people arriving from older tutorials at the current files

The -mlp residuals in fp16: half the download, and half the VRAM with node 0.1.4

README for 0.1.3: named people, the -mlp compatibility trap, corpus sources

Restore the .pt files alongside the safetensors until the reader has propagated

Correct the README: both formats are published during the transition

Promote the condition_proj-weighted matrices to the root, move the unweighted ones to obsolete/

Promote the condition_proj-weighted matrices to the root, move the unweighted ones to obsolete/

Add the condition_proj-weighted matrices, calibrated against what the DiT actually reads

ClipProj v0.1.1: attention-sink vector added to every matrix, plus 8B and int8_convrot

ClipProj v0.1.0: projection matrices and controls