Fix non-ASCII characters in README

README.md

@@ -8,9 +8,9 @@ tags:
 
 [![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.20279247.svg)](https://doi.org/10.5281/zenodo.20279247)
 
-# Qwen 3 4B Instruct — Lossless Compressed
+# Qwen 3 4B Instruct -- Lossless Compressed
 
-> **7.49 GB → 4.95 GB (34% smaller). Bit-identical weights. Drop-in replacement.**
+> **7.49 GB - 4.95 GB (34% smaller). Bit-identical weights. Drop-in replacement.**
 
 ## Use it in 2 lines
 
@@ -53,7 +53,7 @@ model = BigSmallStreamingModel.from_pretrained(
 )
 ```
 
-Uses up to ~12× less VRAM than standard loading by streaming layers on demand.
+Uses up to ~12- less VRAM than standard loading by streaming layers on demand.
 
 ## Stream straight from the Hub (no disk)
 
@@ -91,8 +91,8 @@ See [github.com/wpferrell/Bigsmall](https://github.com/wpferrell/Bigsmall) for t
 
 ## License
 
-- **Model weights:** apache-2.0 — same as [Qwen/Qwen3-4B-Instruct-2507](https://huggingface.co/Qwen/Qwen3-4B-Instruct-2507).
-- **BigSmall format:** [Elastic License 2.0](https://github.com/wpferrell/Bigsmall/blob/main/LICENSE) — free for personal, research, and commercial use.
+- **Model weights:** apache-2.0 -- same as [Qwen/Qwen3-4B-Instruct-2507](https://huggingface.co/Qwen/Qwen3-4B-Instruct-2507).
+- **BigSmall format:** [Elastic License 2.0](https://github.com/wpferrell/Bigsmall/blob/main/LICENSE) -- free for personal, research, and commercial use.
 - **Commercial SaaS licensing:** wpferrell@gmail.com
 
 ## Citation