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

[Docs] Rename Qwen3.8-Max-DSpark to Qwen3.8-2.4T-A95B-DSpark (#34590)

Browse Source
This commit is contained in:
Xinyuan Tong
2026-08-12 15:25:48 +00:00
committed by GitHub
parent ad47dde65c
commit d21eefc94f
2 changed files with 7 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/src/snippets/configs/Qwen/qwen3.8.jsx
+5 -5
View File
@@ -215,7 +215,7 @@ export const config = {
// Gate the chip on both so the panel can't emit a command that aborts.
{ id: "dspark", label: "DSpark",
flags: ["--speculative-algorithm DSPARK",
"--speculative-draft-model-path RadixArk/Qwen3.8-Max-DSpark"],
"--speculative-draft-model-path RadixArk/Qwen3.8-2.4T-A95B-DSpark"],
disable: [
{ when: { dpAttnOn: [true] },
reason: "DSpark with DP-Attention additionally requires --enable-dp-lm-head, the built-in TP MoE (--moe-a2a-backend none) and no context parallel. Turn DP-Attention off in the Attention card above, or pick a cell that doesn't use it." },
@@ -825,7 +825,7 @@ export const config = {
"--kv-cache-dtype fp8_e4m3",
"--mamba-ssm-dtype bfloat16",
"--speculative-algorithm DSPARK",
"--speculative-draft-model-path RadixArk/Qwen3.8-Max-DSpark",
"--speculative-draft-model-path RadixArk/Qwen3.8-2.4T-A95B-DSpark",
"--mem-fraction-static 0.95",
"--chunked-prefill-size 8192",
"--max-prefill-tokens 8192",
@@ -858,7 +858,7 @@ export const config = {
"--mamba-radix-cache-strategy extra_buffer",
"--mamba-ssm-dtype bfloat16",
"--speculative-algorithm DSPARK",
"--speculative-draft-model-path RadixArk/Qwen3.8-Max-DSpark",
"--speculative-draft-model-path RadixArk/Qwen3.8-2.4T-A95B-DSpark",
"--mem-fraction-static 0.90",
"--chunked-prefill-size 8192",
"--max-prefill-tokens 8192",
@@ -884,7 +884,7 @@ export const config = {
"--tp-size 32",
"--mamba-ssm-dtype bfloat16",
"--speculative-algorithm DSPARK",
"--speculative-draft-model-path RadixArk/Qwen3.8-Max-DSpark",
"--speculative-draft-model-path RadixArk/Qwen3.8-2.4T-A95B-DSpark",
"--mem-fraction-static 0.95",
"--chunked-prefill-size 8192",
"--max-prefill-tokens 8192",
@@ -920,7 +920,7 @@ export const config = {
"--mamba-radix-cache-strategy extra_buffer",
"--mamba-ssm-dtype bfloat16",
"--speculative-algorithm DSPARK",
"--speculative-draft-model-path RadixArk/Qwen3.8-Max-DSpark",
"--speculative-draft-model-path RadixArk/Qwen3.8-2.4T-A95B-DSpark",
"--mem-fraction-static 0.80",
"--max-running-requests 128",
"--chunked-prefill-size 8192",