用本地模型

[ggtogether]

安装确实简单，但用本地模型后，在世界观那里直接报错503，不知道是不是和python run_tests.py里的那堆报错有关

[ydsgangge-ux]

服务器里面有错误代码吗？

[ggtogether]

NFO: 127.0.0.1:7706 - "GET /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/hooks HTTP/1.1" 200 OK
2026-04-01 15:40:51,001 [INFO] --> GET /api/books/火影之我是宇智波/relationships
2026-04-01 15:40:51,002 [INFO] <-- GET /api/books/火影之我是宇智波/relationships 200
INFO: 127.0.0.1:7706 - "GET /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/relationships HTTP/1.1" 200 OK
2026-04-01 15:40:51,003 [INFO] --> GET /api/books/火影之我是宇智波/outline
2026-04-01 15:40:51,004 [INFO] <-- GET /api/books/火影之我是宇智波/outline 404
INFO: 127.0.0.1:7706 - "GET /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/outline HTTP/1.1" 404 Not Found
2026-04-01 15:40:51,005 [INFO] --> GET /api/books/火影之我是宇智波/chapter-outlines
2026-04-01 15:40:51,006 [INFO] <-- GET /api/books/火影之我是宇智波/chapter-outlines 404
INFO: 127.0.0.1:7706 - "GET /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/chapter-outlines HTTP/1.1" 404 Not Found
2026-04-01 15:41:10,040 [INFO] --> GET /api/books/火影之我是宇智波/setup/status
2026-04-01 15:41:10,043 [INFO] <-- GET /api/books/火影之我是宇智波/setup/status 200
INFO: 127.0.0.1:7705 - "GET /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/setup/status HTTP/1.1" 200 OK
2026-04-01 15:41:10,046 [INFO] --> POST /api/books/火影之我是宇智波/setup/init

已存在：characters.json
已存在：events.json
已存在：world.json
模板已存在。
请编辑：G:\小说\dramatica-flow\books\火影之我是宇智波\setup
完成后运行：df setup load <book_id>
2026-04-01 15:41:10,048 [INFO] <-- POST /api/books/火影之我是宇智波/setup/init 200
INFO: 127.0.0.1:7705 - "POST /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/setup/init HTTP/1.1" 200 OK
2026-04-01 15:41:22,408 [INFO] --> POST /api/books/火影之我是宇智波/ai-generate/setup
2026-04-01 15:41:24,951 [INFO] HTTP Request: POST http://localhost:11434/v1/chat/completions "HTTP/1.1 503 Service Unavailable"
2026-04-01 15:41:24,952 [INFO] Retrying request to /chat/completions in 0.472236 seconds
2026-04-01 15:41:26,902 [INFO] HTTP Request: POST http://localhost:11434/v1/chat/completions "HTTP/1.1 503 Service Unavailable"
2026-04-01 15:41:26,902 [INFO] Retrying request to /chat/completions in 0.950688 seconds
2026-04-01 15:41:29,316 [INFO] HTTP Request: POST http://localhost:11434/v1/chat/completions "HTTP/1.1 503 Service Unavailable"
2026-04-01 15:41:29,317 [INFO] <-- POST /api/books/火影之我是宇智波/ai-generate/setup 500
INFO: 127.0.0.1:7704 - "POST /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/ai-generate/setup HTTP/1.1" 500 Internal Server Error

.env里的是

Ollama local model (free)

LLM_PROVIDER=ollama
OLLAMA_BASE_URL=http://localhost:11434/v1
OLLAMA_MODEL=qwen2.5

[ydsgangge-ux]

没看懂你想要说什么？本地模型配置，你再网页上可以直接去配置的 ggtogether ***@***.***> 于2026年4月1日周三 15:44写道：

…

*ggtogether* left a comment (ydsgangge-ux/dramatica-flow#2) <#2 (comment)> INFO: 127.0.0.1:7706 - "GET /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/hooks HTTP/1.1" 200 OK 2026-04-01 15:40:51,001 [INFO] --> GET /api/books/火影之我是宇智波/relationships 2026-04-01 15:40:51,002 [INFO] <-- GET /api/books/火影之我是宇智波/relationships 200 INFO: 127.0.0.1:7706 - "GET /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/relationships HTTP/1.1" 200 OK 2026-04-01 15:40:51,003 [INFO] --> GET /api/books/火影之我是宇智波/outline 2026-04-01 15:40:51,004 [INFO] <-- GET /api/books/火影之我是宇智波/outline 404 INFO: 127.0.0.1:7706 - "GET /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/outline HTTP/1.1" 404 Not Found 2026-04-01 15:40:51,005 [INFO] --> GET /api/books/火影之我是宇智波/chapter-outlines 2026-04-01 15:40:51,006 [INFO] <-- GET /api/books/火影之我是宇智波/chapter-outlines 404 INFO: 127.0.0.1:7706 - "GET /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/chapter-outlines HTTP/1.1" 404 Not Found 2026-04-01 15:41:10,040 [INFO] --> GET /api/books/火影之我是宇智波/setup/status 2026-04-01 15:41:10,043 [INFO] <-- GET /api/books/火影之我是宇智波/setup/status 200 INFO: 127.0.0.1:7705 - "GET /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/setup/status HTTP/1.1" 200 OK 2026-04-01 15:41:10,046 [INFO] --> POST /api/books/火影之我是宇智波/setup/init - 已存在：characters.json - 已存在：events.json - 已存在：world.json 模板已存在。 请编辑：G:\小说\dramatica-flow\books\火影之我是宇智波\setup 完成后运行：df setup load <book_id> 2026-04-01 15:41:10,048 [INFO] <-- POST /api/books/火影之我是宇智波/setup/init 200 INFO: 127.0.0.1:7705 - "POST /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/setup/init HTTP/1.1" 200 OK 2026-04-01 15:41:22,408 [INFO] --> POST /api/books/火影之我是宇智波/ai-generate/setup 2026-04-01 15:41:24,951 [INFO] HTTP Request: POST http://localhost:11434/v1/chat/completions "HTTP/1.1 503 Service Unavailable" 2026-04-01 15:41:24,952 [INFO] Retrying request to /chat/completions in 0.472236 seconds 2026-04-01 15:41:26,902 [INFO] HTTP Request: POST http://localhost:11434/v1/chat/completions "HTTP/1.1 503 Service Unavailable" 2026-04-01 15:41:26,902 [INFO] Retrying request to /chat/completions in 0.950688 seconds 2026-04-01 15:41:29,316 [INFO] HTTP Request: POST http://localhost:11434/v1/chat/completions "HTTP/1.1 503 Service Unavailable" 2026-04-01 15:41:29,317 [INFO] <-- POST /api/books/火影之我是宇智波/ai-generate/setup 500 INFO: 127.0.0.1:7704 - "POST /api/books/%E7%81%AB%E5%BD%B1%E4%B9%8B%E6%88%91%E6%98%AF%E5%AE%87%E6%99%BA%E6%B3%A2/ai-generate/setup HTTP/1.1" 500 Internal Server Error .env文件 ── LLM 配置（DeepSeek 默认） ────────────────────────────────────────────────── #LLM_PROVIDER=deepseek #DEEPSEEK_API_KEY=sk-xxx #DEEPSEEK_BASE_URL=https://api.deepseek.com/v1 #DEEPSEEK_MODEL=deepseek-chat ── Ollama 本地模型（免费，零成本） ────────────────────────────────────────────── 要使用 Ollama，将上面的 LLM_PROVIDER 注释掉，取消下面这行注释： LLM_PROVIDER=ollama OLLAMA_BASE_URL=http://localhost:11434/v1 OLLAMA_MODEL=llama3.1 ── 各 Agent 可单独覆盖模型（可选） ─────────────────────────────────────────── ARCHITECT_MODEL=deepseek-chat WRITER_MODEL=deepseek-chat AUDITOR_MODEL=deepseek-reasoner # 审计用推理模型效果更好 REVISER_MODEL=deepseek-chat ── 写作参数 ─────────────────────────────────────────────────────────────────── DEFAULT_WORDS_PER_CHAPTER=4000 DEFAULT_TEMPERATURE=0.7 AUDITOR_TEMPERATURE=0.0 # 审计员锁温度 0，结果稳定 ── 项目根目录 ───────────────────────────────────────────────────────────────── BOOKS_DIR=./books #DEEPSEEK_API_KEY=sk-xxx #DEEPSEEK_BASE_URL=https://api.deepseek.com/v1 #DEEPSEEK_MODEL=deepseek-chat Ollama local model (free) LLM_PROVIDER=ollama OLLAMA_BASE_URL=http://localhost:11434/v1 OLLAMA_MODEL=qwen2.5 DEEPSEEK_API_KEY=sk-xxx DEEPSEEK_BASE_URL=https://api.deepseek.com/v1 DEEPSEEK_MODEL=deepseek-chat Ollama local model (free) LLM_PROVIDER=ollama OLLAMA_BASE_URL=http://localhost:11434/v1 OLLAMA_MODEL=llama3.1 — Reply to this email directly, view it on GitHub <#2?email_source=notifications&email_token=B223M522TIIMRBZXXQ3LE634TTCHBA5CNFSNUABFM5UWIORPF5TWS5BNNB2WEL2JONZXKZKDN5WW2ZLOOQXTIMJWHAYTSMZWGE3KM4TFMFZW63VHMNXW23LFNZ2KKZLWMVXHJLDGN5XXIZLSL5RWY2LDNM#issuecomment-4168193616>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/B223M5YEZN3HJEMB73FVYEL4TTCHBAVCNFSM6AAAAACXH6WJ62VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DCNRYGE4TGNRRGY> . You are receiving this because you commented.Message ID: ***@***.***>