Skip to content

Commit df58016

Browse files
docs: use deepseek-v31 and qwen3-30b in reasoning examples
1 parent 927f1bc commit df58016

File tree

1 file changed

+7
-14
lines changed

1 file changed

+7
-14
lines changed

website/docs/getting-started/reasoning-routing-quickstart.md

Lines changed: 7 additions & 14 deletions
Original file line numberDiff line numberDiff line change
@@ -31,7 +31,7 @@ vllm_endpoints:
3131
- name: "endpoint1"
3232
address: "127.0.0.1"
3333
port: 8000
34-
models: ["deepseek-v3", "qwen3-7b", "openai/gpt-oss-20b"]
34+
models: ["deepseek-v31", "qwen3-30b", "openai/gpt-oss-20b"]
3535
weight: 1
3636

3737
# Reasoning family configurations (how to express reasoning for a family)
@@ -54,10 +54,10 @@ default_reasoning_effort: medium # low | medium | high
5454

5555
# Map concrete model names to a reasoning family
5656
model_config:
57-
"deepseek-v3":
57+
"deepseek-v31":
5858
reasoning_family: "deepseek"
5959
preferred_endpoints: ["endpoint1"]
60-
"qwen3-7b":
60+
"qwen3-30b":
6161
reasoning_family: "qwen3"
6262
preferred_endpoints: ["endpoint1"]
6363
"openai/gpt-oss-20b":
@@ -73,19 +73,11 @@ categories:
7373
model_scores:
7474
- model: openai/gpt-oss-20b
7575
score: 1.0
76-
- model: deepseek-v3
76+
- model: deepseek-v31
7777
score: 0.8
78-
- model: qwen3-7b
78+
- model: qwen3-30b
7979
score: 0.8
8080

81-
- name: general
82-
use_reasoning: false
83-
reasoning_description: "General chit-chat doesn’t need reasoning"
84-
model_scores:
85-
- model: qwen3-7b
86-
score: 1.0
87-
- model: deepseek-v3
88-
score: 0.8
8981

9082
# A safe default when no category is confidently selected
9183
default_model: qwen3-7b
@@ -96,7 +88,8 @@ Notes
9688
- A model only gets reasoning fields if it has a model_config.<MODEL>.reasoning_family that maps to a reasoning_families entry.
9789
- DeepSeek/Qwen3 (chat_template_kwargs): the router injects chat_template_kwargs only when reasoning is enabled. When disabled, no chat_template_kwargs are added.
9890
- GPT/GPT-OSS (reasoning_effort): when reasoning is enabled, the router sets reasoning_effort based on the category (fallback to default_reasoning_effort). When reasoning is disabled, if the request already contains reasoning_effort and the model’s family type is reasoning_effort, the router preserves the original value; otherwise it is absent.
99-
- For more stable classification, you can add category descriptions in config and keep them semantically distinctive.
91+
- Category descriptions (for example, description and reasoning_description) are informational only today; they do not affect routing or classification.
92+
- Categories must be from MMLU-Pro at the moment; avoid free-form categories like "general". If you want generic categories, consider opening an issue to map them to MMLU-Pro.
10093
10194
2) Start the router
10295
Option A: Local build + Envoy

0 commit comments

Comments
 (0)