Skip to content

Commit f1cd622

Browse files
committed
docs(quick-start): update config sections with new model types #271
Update the "Custom Config / OpenAI compatible" section to "Custom Config / OpenAI compatible (Old)" and add a new "New Config" section detailing various model types and their configurations.
1 parent 8551242 commit f1cd622

File tree

1 file changed

+40
-1
lines changed

1 file changed

+40
-1
lines changed

docs/quick-start.md

Lines changed: 40 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -23,7 +23,7 @@ So, since 2024.4, AutoDev was unavailable in the JetBrains Plugin Repository, yo
2323

2424
<img src="https://unitmesh.cc/auto-dev/autodev-2-config.png" alt="AutoDev 2 Config" width="600px"/>
2525

26-
### Custom Config / OpenAI compatible
26+
### Custom Config / OpenAI compatible (Old)
2727

2828
Tested: 零一万物([#94](https://github.com/unit-mesh/auto-dev/issues/94)), 月之暗面(Moonshot
2929
AI)、深度求索(Deepseek [#96](https://github.com/unit-mesh/auto-dev/issues/96)),ChatGLM(#90)
@@ -40,3 +40,42 @@ AI)、深度求索(Deepseek [#96](https://github.com/unit-mesh/auto-dev/issu
4040
6. Apply and OK.
4141

4242
for more, see in [Customize LLM Server](/custom/llm-server)
43+
44+
### New Config
45+
46+
modelType: `["Default", "Plan", "Act", "Completion", "Embedding", "FastApply", "Others"]`
47+
48+
- Default: the default model for all cases if not specified
49+
- Plan: for reasoning, planning, etc, like: `DeepSeek R1`, recommend to use Best Model
50+
- Act: for action, like: `DeepSeek V3`, `Qwen 72B` etc
51+
- Completion: for code completion, not support FIM yet.
52+
- Embedding: for embedding, like: `sentence-transformers/all-MiniLM-L6-v2`
53+
- FastApply: for fix patch generate issue, like: `Kortix/FastApply-1.5B-v1.0`
54+
- Others: just a placeholder, no special treatment
55+
56+
```json
57+
[
58+
{
59+
"name": "GLM4-Plus",
60+
"url": "https://open.bigmodel.cn/api/paas/v4/chat/completions",
61+
"auth": {
62+
"type": "Bearer",
63+
"token": "sk-ii"
64+
},
65+
"requestFormat": "{ \"customFields\": {\"model\": \"glm-4-plus\", \"stream\": true}}",
66+
"responseFormat": "$.choices[0].delta.content",
67+
"modelType": "FastApply"
68+
},
69+
{
70+
"name": "DeepSeek R1",
71+
"url": "https://api.deepseek.com/chat/completions",
72+
"auth": {
73+
"type": "Bearer",
74+
"token": "sk-ii"
75+
},
76+
"requestFormat": "{ \"customFields\": {\"model\": \"deepseek-reasoner\", \"stream\": true}}",
77+
"responseFormat": "$.choices[0].delta.content",
78+
"modelType": "Plan"
79+
}
80+
]
81+
```

0 commit comments

Comments
 (0)