Qwen3.8-Flash-Next GSQ-RCO: 294.82 GB -> 112.75 GB

README.md

@@ -29,12 +29,17 @@ All of [unsloth/Qwen3.8-27B-GGUF](https://huggingface.co/unsloth/Qwen3.8-27B-GGU
   <img src="https://huggingface.co/spaces/neurosquad/README/resolve/main/assets/seedpack-benchmarks.png" alt="SeedPack benchmarks: the releases and other archivers" width="860">
 </p>
 
+### Qwen3.8-Flash-Next GSQ-RCO: 294.82 GB → 112.75 GB
+
+[ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF](https://huggingface.co/ISTA-DASLab/Qwen3.8-Flash-Next-GSQ-RCO-GGUF), the GGUFs most Qwen3.8-Flash-Next users run (Strata's default), has no BF16 model to grow its sizes from. SeedPack 0.4.0 codes its 4 sizes against each other instead: identical files and tensors once, and every other tensor as a chain of versions, each coded given the precision-weighted mean of the earlier ones. All four sizes fit in one archive of **112.75 GB**, 2.6× smaller than the repository. [The archive](https://huggingface.co/neurosquad/Qwen3.8-Flash-Next-GSQ-RCO-SEEDPACK) is for keeping all four sizes: one size alone is still smaller.
+
 ### Releases
 
-Each archive holds a whole Unsloth repository (`unsloth/<model>-GGUF`): the BF16 model, every quant and the rest of its files.
+Each archive holds a whole repository: every file of it, restored byte for byte.
 
 | Model | Repository → archive | Smaller |
 |---|---:|---:|
+| [Qwen3.8-Flash-Next GSQ-RCO](https://huggingface.co/neurosquad/Qwen3.8-Flash-Next-GSQ-RCO-SEEDPACK) | 294.82 GB → **112.75 GB** | 2.6× |
 | [Qwen3.8-27B](https://huggingface.co/neurosquad/Qwen3.8-27B-SEEDPACK) | 472.07 GB → **45.41 GB** | 10.4× |
 | [Qwen3.5-9B](https://huggingface.co/neurosquad/Qwen3.5-9B-SEEDPACK) | 147.81 GB → **12.40 GB** | 11.9× |
 | [Qwen3-0.6B](https://huggingface.co/neurosquad/Qwen3-0.6B-SEEDPACK) | 10.95 GB → **781.7 MB** | 14.0× |