minke.yu/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 12 Packages Projects Releases Wiki Activity

Add a little env var for disabling Flashinfer autotune cache (#26193)

Browse Source
This commit is contained in:
Brayden Zhong
2026-05-27 23:59:59 -07:00
committed by GitHub
parent bed20249f1
commit 97dd6aad60
3 changed files with 22 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs_new/docs/references/environment_variables.mdx
+5
View File
@@ -127,6 +127,11 @@ SGLang supports various environment variables that can be used to configure its
<td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.05)"}}>Control FlashInfer availability check</td>
<td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.02)"}}>`true`</td>
</tr>
<tr>
<td style={{padding: "9px 12px", fontWeight: 500, backgroundColor: "rgba(255,255,255,0.02)"}}>`SGLANG_FLASHINFER_AUTOTUNE_CACHE`</td>
<td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.05)"}}>Reuse persisted FlashInfer autotune results from `SGLANG_CACHE_DIR` across runs. Set to `0` to force re-autotuning on every startup; the fresh result is written to a `runs/<rank>.<timestamp>.json` sibling file (the canonical cache is left untouched).</td>
<td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.02)"}}>`true`</td>
</tr>
<tr>
<td style={{padding: "9px 12px", fontWeight: 500, backgroundColor: "rgba(255,255,255,0.02)"}}>`SGLANG_SKIP_P2P_CHECK`</td>
<td style={{padding: "9px 12px", backgroundColor: "rgba(255,255,255,0.05)"}}>Skip P2P (peer-to-peer) access check</td>