Card: red-teaming (CSAM, nudity) and training-speed statements

README.md

@@ -29,7 +29,11 @@ SD 2.1 (0.50) and PixArt-α (0.48). It is small enough to run in under two secon
 **Why we built it.** [LogoLabs](https://logolabs.org) builds AI logo generation. Agate is our attempt at a better
 *small* model for icon generation. We want one that places things where you ask, keeps colours bound to the right
 objects, runs on a consumer GPU, and can be trained from scratch on a research budget. This preview is the
-general-purpose base model; it is not yet specialised for icons. More of our work is on
+general-purpose base model; it is not yet specialised for icons.
+
+**It trains ultra-fast.** The full training run took 144.7 GPU-hours on GH200s, about 18 hours on a single
+8-GPU node. That makes retraining from scratch, on your own data or for a new domain, a one-day job rather than
+a research programme. More of our work is on
 [our Hugging Face page](https://huggingface.co/Logolabs): the Inkvec vectoriser and the LogoBrief-10K dataset.
 
 It is a research preview: it renders at **256 × 256 only**, cannot spell reliably, and ignores negation. The card
@@ -264,10 +268,19 @@ carries the concepts, and the jump to 512 tokens stops long captions from being
 **Why FLUX-Reason-6M.** We chose the training set because it is curated and contains no adult images. Agate never
 saw explicit material during training, so it has no learned way to produce it.
 
-**Internal safety tests.** We deliberately prompted Agate for nudity and sexual content. We could not get it to
-generate genitals in any attempt. The worst output was a blurry nude female figure with no anatomical detail
-(no areolae). That is a finding from our own red-teaming, not a guarantee. Agate ships without a safety classifier,
-so deployments facing the public should add one.
+**Red-teaming.** Before release we tested Agate adversarially for sexual content and for child sexual abuse material
+(CSAM).
+- **CSAM:** Agate did not produce CSAM or any sexualised depiction of a minor.
+- **Adult nudity:** we could not get Agate to generate genitals in any attempt. The worst output was a blurry nude
+  female figure with no anatomical detail (no areolae).
+
+We consider explicit or abusive output practically infeasible for this model, because it was never trained on
+such material, and misuse unlikely. That is our assessment, not a guarantee. Agate ships without a safety
+classifier, so public-facing deployments should add one.
+
+Using Agate, or any model derived from it, to depict minors sexually, to create non-consensual intimate imagery,
+or to depict real people deceptively is prohibited and, in most jurisdictions, illegal. Report any such output to
+LogoLabs through [logolabs.org](https://logolabs.org).
 
 ## Compute and energy