Adds config.json with a full parameter breakdown (total, per-layer,
text-encoder/VAE = 0) and embeds the same info directly in the
model.safetensors header metadata, so total parameter count can be
verified by reading either file without running any code.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Add safetensors weights and standalone INFERENCE.py
Costi+1: Claude Sonnet 5
b451f49
Exports the same 3 weight matrices already encoded in model.png to
model.safetensors via convert_to_safetensors.py, and adds INFERENCE.py
as a standalone entry point that loads from the safetensors file
instead of the PNG. Verified byte-identical output between main.py
and INFERENCE.py for the same prompt. README documents both paths.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Add real eval results and reproduction guide
Costi+1: Claude Sonnet 5
d9a3537
Runs an actual COCO-captions FID/CLIP Score eval against the model's
real generated outputs rather than filling the leaderboard template
with placeholder numbers, and documents the model's actual capability
(toy MLP, not general text-to-image) alongside the measurements.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
