h4x3rotab commited on
Commit
8dc3a42
·
verified ·
1 Parent(s): 3cd8b80

Drop internal index.json fix note from model card

Browse files
Files changed (1) hide show
  1. README.md +0 -3
README.md CHANGED
@@ -34,9 +34,6 @@ Two things make it useful:
34
 
35
  **Tested with SGLang only** ([v0.5.13.post1](https://github.com/sgl-project/sglang)). The 4-bit layout and speculative-decoding path are SGLang-specific and untried on vLLM or other engines.
36
 
37
- > [!IMPORTANT]
38
- > **2026-06-19 fix:** if you downloaded this repo before this date, re-download `model.safetensors.index.json` — it's the only changed file, and it enables EAGLE/MTP speculative decoding (previously a no-op). The weights are unchanged.
39
-
40
  ## Benchmarks
41
 
42
  Measured on 8×H200 with SGLang at GLM-5.2's recommended sampling (temperature 1.0, top_p 0.95) with reasoning enabled. The reference column shows the published GLM-5.2 results.
 
34
 
35
  **Tested with SGLang only** ([v0.5.13.post1](https://github.com/sgl-project/sglang)). The 4-bit layout and speculative-decoding path are SGLang-specific and untried on vLLM or other engines.
36
 
 
 
 
37
  ## Benchmarks
38
 
39
  Measured on 8×H200 with SGLang at GLM-5.2's recommended sampling (temperature 1.0, top_p 0.95) with reasoning enabled. The reference column shows the published GLM-5.2 results.