GITBOOK-20: No subject

Author: XiaotongJiang

sglang-cookbook/gpt-oss/usage-guide.md

@@ -4,13 +4,13 @@
 
 {% stepper %}
 {% step %}
-### Install SGLang
+#### Install SGLang
 
-Following [the instruction](https://app.gitbook.com/o/TvLfyTxdRQeudJH7e5QW/s/FFtIWT8LEMaYiYzz0p8P/~/changes/11/sglang-cookbook/installation/nvidia-h-series-a-series-and-rtx-gpus)
+Following [the instruction](../installation/nvidia-h-series-a-series-and-rtx-gpus.md)
 {% endstep %}
 
 {% step %}
-### Serve the model
+#### Serve the model
 
 {% code overflow="wrap" %}
 ```bash
@@ -28,7 +28,7 @@ python3 -m sglang.launch_server --model-path openai/gpt-oss-120b --mem-fraction-
 {% endstep %}
 
 {% step %}
-### Benchmark
+#### Benchmark
 
 SGLang version (0.5.1)
 
@@ -50,13 +50,13 @@ SGLang version (0.5.1)
 
 {% stepper %}
 {% step %}
-### Install SGLang
+#### Install SGLang
 
-Following [the instruction](https://app.gitbook.com/o/TvLfyTxdRQeudJH7e5QW/s/FFtIWT8LEMaYiYzz0p8P/~/changes/11/sglang-cookbook/installation/nvidia-h-series-a-series-and-rtx-gpus)
+Following [the instruction](../installation/nvidia-h-series-a-series-and-rtx-gpus.md)
 {% endstep %}
 
 {% step %}
-### Serve the model
+#### Serve the model
 
 {% code overflow="wrap" %}
 ```bash
@@ -67,23 +67,23 @@ python3 -m sglang.launch_server --model-path openai/gpt-oss-120b --tp 2
 {% endstep %}
 
 {% step %}
-### Benchmark
+#### Benchmark
 
-<table><thead><tr><th width="209.78515625">BS/Input/Output Length</th><th width="109.6328125">TTFT(s)</th><th width="101.75390625">ITL(ms)</th><th>Input Throughput</th><th>Output Throughput</th></tr></thead><tbody><tr><td colspan="5" style="text-align: center;">Benchmark results will be added here</td></tr></tbody></table>
+<table><thead><tr><th width="209.78515625">BS/Input/Output Length</th><th width="109.6328125">TTFT(s)</th><th width="101.75390625">ITL(ms)</th><th>Input Throughput</th><th>Output Throughput</th></tr></thead><tbody><tr><td>Benchmark results will be added here</td><td></td><td></td><td></td><td></td></tr></tbody></table>
 {% endstep %}
 {% endstepper %}
 
 ### <mark style="background-color:green;">Serving with 1 x B200</mark>
 
 {% stepper %}
 {% step %}
-### Install SGLang
+#### Install SGLang
 
 Following [the instruction](../installation/nvidia-blackwell-gpus.md)
 {% endstep %}
 
 {% step %}
-### Serve the model
+#### Serve the model
 
 {% code overflow="wrap" %}
 ```bash
@@ -101,6 +101,12 @@ python3 -m sglang.launch_server --model-path openai/gpt-oss-120b
 {% endstep %}
 
 {% step %}
+#### Benchmark
+
+<table><thead><tr><th width="209.78515625">BS/Input/Output Length</th><th width="109.6328125">TTFT(s)</th><th width="101.75390625">ITL(ms)</th><th>Input Throughput</th><th>Output Throughput</th></tr></thead><tbody><tr><td>Benchmark results will be added here</td><td></td><td></td><td></td><td></td></tr></tbody></table>
+{% endstep %}
+{% endstepper %}
+
 ### With Speculative Decoding
 
 {% code overflow="wrap" %}
@@ -121,14 +127,6 @@ python3 -m sglang.launch_server --model openai/gpt-oss-120b --speculative-algo E
 python3 -m sglang.launch_server --model openai/gpt-oss-120b --speculative-algo EAGLE3 --speculative-draft lmsys/EAGLE3-gpt-oss-120b-bf16 --speculative-num-steps 5 --speculative-eagle-topk 4 --speculative-num-draft-tokens 8 --attention-backend triton --tp 4
 ```
 {% endcode %}
-{% endstep %}
-
-{% step %}
-### Benchmark
-
-<table><thead><tr><th width="209.78515625">BS/Input/Output Length</th><th width="109.6328125">TTFT(s)</th><th width="101.75390625">ITL(ms)</th><th>Input Throughput</th><th>Output Throughput</th></tr></thead><tbody><tr><td colspan="5" style="text-align: center;">Benchmark results will be added here</td></tr></tbody></table>
-{% endstep %}
-{% endstepper %}
 
 ### Responses API & Built-in Tools