Update README.md

README.md

@@ -20,15 +20,10 @@ pipeline_tag: text-to-image
 # Nitro-E · WikiArt Themes & Feelings (EMA-aligned)
 
 A LoRA post-training of **[amd/Nitro-E](https://huggingface.co/amd/Nitro-E)** (304M E-MMDiT,
-flow-matching, MIT) that renders subjects in fine-art styles conditioned on **art movement +
-genre + emotion/theme**. This repo holds the **EMA-aligned, merged, serveable** checkpoint
+flow-matching, MIT) from AMD Brain that renders subjects in fine-art styles conditioned on art movement, genre, emotion and theme.
+This repo holds the EMA-aligned, merged, serveable checkpoint
 (`nitro_cc_ema_merged.safetensors`) plus the LoRA adapter it was merged from.
 
-The **EMA** (exponential moving average over the LoRA weights, rate 0.9999) is the key fix:
-a rank-16 LoRA on ~29k images overfits within a couple thousand steps and the raw weights
-collapse to noise past the knee; the EMA-averaged weights stay coherent. This is the
-checkpoint served in the live demo.
-
 - **Base:** `amd/Nitro-E` (Nitro-E-512px), text encoder `meta-llama/Llama-3.2-1B` (gated), VAE `mit-han-lab/dc-ae-f32c32-sana-1.0`.
 - **Resolution:** 512x512. **License:** MIT (base) — see data notes below.
 - **Tracking:** [https://wandb.ai/imaging-ai/more-art-than-science](https://wandb.ai/imaging-ai/more-art-than-science)
@@ -38,8 +33,7 @@ checkpoint served in the live demo.
 - `adapter_config.json` + `adapter_model.safetensors` — the PEFT LoRA adapter (apply on top of `amd/Nitro-E` if you prefer not to use the merged weights).
 
 ## Training data
-[`emotiongoes/wikiart-themes-feelings`](https://huggingface.co/datasets/emotiongoes/wikiart-themes-feelings)
-— 32,061 WikiArt works (metadata + image URLs). ~29,000 images were downloadable and
+— A private dataset of 32,061 WikiArt works (metadata + image URLs). ~29,000 images were downloadable and
 used. The `consolidated-caption` field carries the emotional/style signal used as the prompt.
 
 ## Recipe
@@ -60,4 +54,4 @@ Example prompt: `"a tranquil river landscape, in the style of Impressionism, evo
 - Inherits Nitro-E 304M quality ceiling; can overfit WikiArt style cues.
 - Requires gated `meta-llama/Llama-3.2-1B` access to run.
 - **Data rights:** trained on WikiArt images but does not redistribute them; the dataset is
-  metadata-only. Many works are public domain but not all — use outputs accordingly.
+  metadata-only. Many works are public domain but not all so use outputs accordingly.