update readme for epoch 17

README.md

@@ -6,10 +6,10 @@ base_model:
 
 **Puzzle Box XL**
 
-A latent diffusion model (LDM) geared toward illustration, style composability, and sample variety. Addresses a few deficiencies with the SDXL base model.
+A latent diffusion model (LDM) geared toward illustration, style composability, and sample variety. Addresses a few deficiencies with the SDXL base model; feels more like an SD 1.x with better resolution and much better prompt adherence.
 
 * Architecture: SD XL (base model is v1.0)
-* Training procedure: U-Net fully unfrozen, all-parameter continued pretraining at LR between 3e-8 and 3e-7 for 16,950,000 steps (at epoch 16, batch size 4).
+* Training procedure: U-Net fully unfrozen, all-parameter continued pretraining at LR between 3e-8 and 3e-7 for 18,000,000 steps (at epoch 17, batch size 4).
 
 Trained on the Puzzle Box dataset, a large collection of permissively licensed images from the public Internet (or generated by previous Puzzle Box models). Each image 
 has from 3 to 17 different captions which are used interchangably during training. There are 9.3 million images and 62 million captions in the dataset.
@@ -27,7 +27,8 @@ megapixels for epoch 15+. CFG scales between 2 and 7 can work well with Puzzle B
 
 **Captioning:** About 1.4 million of the captions in the dataset are human-written. The remainder come from a variety of ML models, either vision transformers or 
 classifers. Models used in captioning the Puzzle Box dataset include: Qwen 2 VL 72b, BLIP 2 OPT-6.5B COCO, Llava 1.5, MiniCPM 2.6, bakllava, Moondream, DeepSeek Janus 7b,
-Mistral Pixtral 12b, CapPa, Gemma 3 27b, JoyCaption, and wd-eva02-large-tagger-v3. Only open-weights models were used.
+Mistral Pixtral 12b, CapPa, Gemma 3 27b, JoyCaption, CLIP Interrogator 0.6.0, and wd-eva02-large-tagger-v3. Deepseek v3 is used to create detailed consensus captions from 
+the others. Only open-weights models were used.
 
 In addition to human/machine-generated main caption, there are a large number of additional human-provided tags referring to style ("pointillism", "caricature", "Winsor McKay"), 
 genre ("pop art", "advertising", "pixel art"), source ("wikiart", "library of congress"), or image content ("fluid expression", "pin-up", "squash and stretch").
@@ -46,12 +47,27 @@ This allows later checkpoints to generate 1+ megapixel images without tiling or
 
 Model checkpoints currently available:
 
+- from epoch 17, **18000k** training steps, 06 July 2025
 - from epoch 16, **16950k** training steps, 05 May 2025
 - from epoch 15, **15800k** training steps, 08 March 2025
 - from epoch 14, **14290k** training steps, 02 December 2024
 - from epoch 13, **11930k** training steps, 15 August 2024
 - from epoch 12, **10570k** training steps, 21 June 2024
 
+*Which checkpoint's best?* Later checkpoints have better aesthetics and better prompt adherence at higher resolution and lower CFG scale, but are also more 'opinionated'; longer conditioning may be necessary to get the generation as you like it. In particular, the latest checkpoints are trained on the most consensus captions, which are highly accurate but also quite long. Earlier checkpoints may give larger sample variety on short conditioning, which (at lower resolution) may make them useful drafting models: searching for good noise seeds, etc. Earlier checkpoints may also be better for merging with other LDMs based on SD XL.
+
 This model has been trained carefully on top of the SDXL base, with a widely diverse training set at low learning rate. Accordingly, it should *merge* well with most other 
 LDMs built off SDXL base. (Merging LDMs built off the same base is a form of transfer learning; you can add Puzzle Box concepts to other SDXL models this way. Spherical 
 interpolation is best.) 
+
+The U-Net self-attention layers are the layers most modified by the continued pretrain; comparing those layers to SD XL 1.0, the correlation is:
+| Epoch |    Date    | R-squared |
+| ----- | ---------- | --------- |
+|  17   | 2025-07-06 |  97.705%  |
+|  16   | 2025-05-05 |  97.917%  |
+|  15   | 2025-03-08 |  98.312%  |
+|  14   | 2024-12-02 |  98.573%  |
+|  13   | 2024-08-05 |  98.876%  |
+|  12   | 2024-06-21 |  99.167%  |
+
+(For reference, Pony-family models, which are also based on SD XL 1.0 but are trained at much higher LR, paving the model, are around 40%, and Playground-derived models, which are trained on SD XL architecture from static, are below 25%.)