Skip to content

[None][doc] Add Qwen3.8 deployment guide and configs #17594

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Wanli-Jiang merged 1 commit into from
Aug 13, 2026
Merged

[None][doc] Add Qwen3.8 deployment guide and configs #17594

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
52 changes: 52 additions & 0 deletions docs/source/_static/config_db.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -36,6 +36,54 @@
"model_url": "https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Thinking",
"scenario": "Max Throughput"
},
{
"command": "trtllm-serve Qwen/Qwen3.8-2.4T-A95B-FP8 --config ${TRTLLM_DIR}/examples/configs/curated/qwen3.8-low-latency.yaml",
"config_filename": "qwen3.8-low-latency.yaml",
"config_github_url": "https://github.com/NVIDIA/TensorRT-LLM/blob/main/examples/configs/curated/qwen3.8-low-latency.yaml",
"config_path": "examples/configs/curated/qwen3.8-low-latency.yaml",
"config_raw_url": "https://raw.githubusercontent.com/NVIDIA/TensorRT-LLM/main/examples/configs/curated/qwen3.8-low-latency.yaml",
"gpu_compatibility": "B200, B300, GB200, GB300",
"model": "Qwen/Qwen3.8-2.4T-A95B-FP8",
"model_display_name": "Qwen3.8-2.4T-A95B (FP8)",
"model_url": "https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B-FP8",
"scenario": "Min Latency (No MTP)"
},
{
"command": "trtllm-serve Qwen/Qwen3.8-2.4T-A95B-FP8 --config ${TRTLLM_DIR}/examples/configs/curated/qwen3.8-low-latency-mtp3.yaml",
"config_filename": "qwen3.8-low-latency-mtp3.yaml",
"config_github_url": "https://github.com/NVIDIA/TensorRT-LLM/blob/main/examples/configs/curated/qwen3.8-low-latency-mtp3.yaml",
"config_path": "examples/configs/curated/qwen3.8-low-latency-mtp3.yaml",
"config_raw_url": "https://raw.githubusercontent.com/NVIDIA/TensorRT-LLM/main/examples/configs/curated/qwen3.8-low-latency-mtp3.yaml",
"gpu_compatibility": "B200, B300, GB200, GB300",
"model": "Qwen/Qwen3.8-2.4T-A95B-FP8",
"model_display_name": "Qwen3.8-2.4T-A95B (FP8)",
"model_url": "https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B-FP8",
"scenario": "Min Latency (MTP3)"
},
{
"command": "cd \"${TRTLLM_DIR}\" && trtllm-serve Qwen/Qwen3.8-2.4T-A95B-FP8 --config ${TRTLLM_DIR}/examples/configs/curated/qwen3.8-high-throughput.yaml",
"config_filename": "qwen3.8-high-throughput.yaml",
"config_github_url": "https://github.com/NVIDIA/TensorRT-LLM/blob/main/examples/configs/curated/qwen3.8-high-throughput.yaml",
"config_path": "examples/configs/curated/qwen3.8-high-throughput.yaml",
"config_raw_url": "https://raw.githubusercontent.com/NVIDIA/TensorRT-LLM/main/examples/configs/curated/qwen3.8-high-throughput.yaml",
"gpu_compatibility": "B200, B300, GB200, GB300",
"model": "Qwen/Qwen3.8-2.4T-A95B-FP8",
"model_display_name": "Qwen3.8-2.4T-A95B (FP8)",
"model_url": "https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B-FP8",
"scenario": "Max Throughput (Static EPLB, No MTP)"
},
{
"command": "cd \"${TRTLLM_DIR}\" && trtllm-serve Qwen/Qwen3.8-2.4T-A95B-FP8 --config ${TRTLLM_DIR}/examples/configs/curated/qwen3.8-high-throughput-mtp3.yaml",
"config_filename": "qwen3.8-high-throughput-mtp3.yaml",
"config_github_url": "https://github.com/NVIDIA/TensorRT-LLM/blob/main/examples/configs/curated/qwen3.8-high-throughput-mtp3.yaml",
"config_path": "examples/configs/curated/qwen3.8-high-throughput-mtp3.yaml",
"config_raw_url": "https://raw.githubusercontent.com/NVIDIA/TensorRT-LLM/main/examples/configs/curated/qwen3.8-high-throughput-mtp3.yaml",
"gpu_compatibility": "B200, B300, GB200, GB300",
"model": "Qwen/Qwen3.8-2.4T-A95B-FP8",
"model_display_name": "Qwen3.8-2.4T-A95B (FP8)",
"model_url": "https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B-FP8",
"scenario": "Max Throughput (Static EPLB, MTP3)"
},
{
"command": "trtllm-serve nvidia/Qwen3.5-397B-A17B-NVFP4 --config ${TRTLLM_DIR}/examples/configs/curated/qwen3.5.yaml",
"config_filename": "qwen3.5.yaml",
Expand Down Expand Up @@ -4226,6 +4274,10 @@
"display_name": "Qwen3-Next-80B",
"url": "https://huggingface.co/Qwen/Qwen3-Next-80B-A3B-Thinking"
},
"Qwen/Qwen3.8-2.4T-A95B-FP8": {
"display_name": "Qwen3.8-2.4T-A95B (FP8)",
"url": "https://huggingface.co/Qwen/Qwen3.8-2.4T-A95B-FP8"
},
"deepseek-ai/DeepSeek-R1-0528": {
"display_name": "DeepSeek-R1",
"url": "https://huggingface.co/deepseek-ai/DeepSeek-R1-0528"
Expand Down
14 changes: 9 additions & 5 deletions docs/source/_static/config_selector.js
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -594,11 +594,15 @@
const model = entry._hfModel || entry.model || "";
const configPath = entry.config_path || "";
if (!model || !configPath) return entry.command || "";
return [
`export TRTLLM_DIR=/app/tensorrt_llm`,
`trtllm-serve ${model} \\`,
` --config \${TRTLLM_DIR}/${configPath}`,
].join("\n");
const lines = [`export TRTLLM_DIR=/app/tensorrt_llm`];
// Configs that reference another repo file through a relative path (for example
// moe_config.load_balancer) only resolve it when served from the TensorRT LLM
// directory. generate_config_table.py marks those entries with a cd prefix.
if ((entry.command || "").startsWith("cd ")) {
lines.push(`cd "\${TRTLLM_DIR}"`);
}
lines.push(`trtllm-serve ${model} \\`, ` --config \${TRTLLM_DIR}/${configPath}`);
return lines.join("\n");
}

function curatedEntriesForModel(curatedEntries, model) {
Expand Down
257 changes: 0 additions & 257 deletions docs/source/deployment-guide/deployment-guide-for-qwen3.5-on-trtllm.md
View file
Open in desktop

This file was deleted.

Loading
Loading