File size: 4,821 Bytes
1d8c46a | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 | """LLM / embedding provider configuration for students without local models.
Resolution order: CLI args > environment variables > config/llm.json > built-in presets.
Copy `config.example.json` to `config/llm.json` and fill in your key, or just export env vars.
"""
import json
import os
import urllib.error
import urllib.request
from pathlib import Path
ROOT = Path(__file__).resolve().parent.parent
# Built-in presets for common OpenAI-compatible endpoints (China-friendly first).
PRESETS = {
"deepseek": {
"base_url": "https://api.deepseek.com/v1",
"model": "deepseek-chat",
"embedding_model": None,
},
"siliconflow": {
"base_url": "https://api.siliconflow.cn/v1",
"model": "deepseek-ai/DeepSeek-V3",
"embedding_model": "BAAI/bge-m3",
},
"moonshot": {
"base_url": "https://api.moonshot.cn/v1",
"model": "moonshot-v1-8k",
"embedding_model": None,
},
"zhipu": {
"base_url": "https://open.bigmodel.cn/api/paas/v4",
"model": "glm-4-flash",
"embedding_model": "embedding-3",
},
"qwen": {
"base_url": "https://dashscope.aliyuncs.com/compatible-mode/v1",
"model": "qwen-plus",
"embedding_model": "text-embedding-v3",
},
"volcengine": {
"base_url": "https://ark.cn-beijing.volces.com/api/v3",
"model": "<your-endpoint-id>",
"embedding_model": None,
},
"openai": {
"base_url": "https://api.openai.com/v1",
"model": "gpt-4o-mini",
"embedding_model": "text-embedding-3-small",
},
"ollama": {
"base_url": "http://localhost:11434/v1",
"model": "qwen2.5:7b",
"embedding_model": "bge-m3",
},
}
def load_config():
p = ROOT / "config" / "llm.json"
if p.exists():
try:
return json.loads(p.read_text(encoding="utf-8"))
except Exception: # noqa: BLE001
return {}
return {}
def available_presets():
return sorted(PRESETS)
def resolve_llm(provider=None, base_url=None, api_key=None, model=None):
cfg = load_config()
name = provider or os.environ.get("LLM_PROVIDER") or cfg.get("provider")
preset = PRESETS.get(name, {}) if name else {}
merged = {**preset, **(cfg.get("llm") or {})}
return {
"provider": name,
"base_url": (
base_url or os.environ.get("OPENAI_BASE_URL") or merged.get("base_url") or "https://api.openai.com/v1"
).rstrip("/"),
"api_key": api_key or os.environ.get("OPENAI_API_KEY") or merged.get("api_key") or "",
"model": model or os.environ.get("OPENAI_MODEL") or merged.get("model") or "gpt-4o-mini",
}
def resolve_embedding(provider=None, base_url=None, api_key=None, model=None):
cfg = load_config()
name = provider or os.environ.get("EMBEDDING_PROVIDER") or cfg.get("provider")
preset = PRESETS.get(name, {}) if name else {}
merged = {**preset, **(cfg.get("embedding") or {})}
return {
"provider": name,
"base_url": (
base_url or os.environ.get("EMBEDDING_BASE_URL") or merged.get("base_url") or "https://api.openai.com/v1"
).rstrip("/"),
"api_key": api_key or os.environ.get("EMBEDDING_API_KEY") or os.environ.get("OPENAI_API_KEY") or merged.get("api_key") or "",
"model": model or os.environ.get("EMBEDDING_MODEL") or merged.get("embedding_model") or "text-embedding-3-small",
}
def chat(messages, temperature=0.2, timeout=180, provider=None, base_url=None, api_key=None, model=None):
"""Call any OpenAI-compatible /chat/completions endpoint."""
conf = resolve_llm(provider, base_url, api_key, model)
if not conf["api_key"]:
raise RuntimeError(
"未配置 API Key。请设置 OPENAI_API_KEY 环境变量，或把 config.example.json 复制为 config/llm.json 并填入 key。"
)
body = {"model": conf["model"], "messages": messages, "temperature": temperature}
req = urllib.request.Request(
f"{conf['base_url']}/chat/completions",
data=json.dumps(body).encode("utf-8"),
headers={"Content-Type": "application/json", "Authorization": f"Bearer {conf['api_key']}"},
)
with urllib.request.urlopen(req, timeout=timeout) as resp:
data = json.loads(resp.read().decode("utf-8"))
return data["choices"][0]["message"]["content"], conf
def http_error_hint(exc):
if isinstance(exc, urllib.error.HTTPError):
try:
detail = exc.read().decode("utf-8")[:300]
except Exception: # noqa: BLE001
detail = ""
return f"HTTP {exc.code}: {detail or exc.reason}"
return str(exc)
|