请问，本地ollama的API的话，.env要怎么配置呢？

[jianjunLeung]

如题。我试了PRIMARY_MODEL填id和name都是会有下面的这个问题，本地ollama的地址确定是没问题的：

2025-02-24 10:07:53.188 | DEBUG | utils.pb_api:__init__:12 - initializing pocketbase client: http://127.0.0.1:8090 2025-02-24 10:07:53.471 | INFO | utils.pb_api:__init__:22 - pocketbase ready authenticated as admin - 13143398903@163.com 2025-02-24 10:07:53.474 | INFO | __main__:schedule_task:19 - task execute loop 1 2025-02-24 10:07:53.557 | DEBUG | general_process:main_process:54 - new task initializing... 2025-02-24 10:07:53.558 | DEBUG | general_process:main_process:58 - focus_id: m4ptje6n1bkm0k3, focus_point: 实时新闻 消息, explanation: 获取批量的实时的新闻消息，包括新闻的标题，新闻的内容，新闻的链接，新闻的时间。获取的时候获取多条。, search_engine: False [INIT].... → Crawl4AI 0.4.247 2025-02-24 10:07:54.207 | DEBUG | general_process:main_process:140 - process new url, still 0 urls in working list [FETCH]... ↓ https://www.jin10.com/... | Status: True | Time: 0.02s [COMPLETE] ● https://www.jin10.com/... | Status: True | Total: 0.02s INFO:openai._base_client:Retrying request to /chat/completions in 0.425538 seconds INFO:openai._base_client:Retrying request to /chat/completions in 0.848044 seconds Error code: 503 INFO:openai._base_client:Retrying request to /chat/completions in 0.430297 seconds INFO:openai._base_client:Retrying request to /chat/completions in 0.810780 seconds Error code: 503 INFO:openai._base_client:Retrying request to /chat/completions in 0.389863 seconds INFO:openai._base_client:Retrying request to /chat/completions in 0.857039 seconds Error code: 503 INFO:openai._base_client:Retrying request to /chat/completions in 0.406839 seconds INFO:openai._base_client:Retrying request to /chat/completions in 0.966001 seconds Error code: 503 INFO:openai._base_client:Retrying request to /chat/completions in 0.428617 seconds INFO:openai._base_client:Retrying request to /chat/completions in 0.927592 seconds Error code: 503 INFO:openai._base_client:Retrying request to /chat/completions in 0.390710 seconds INFO:openai._base_client:Retrying request to /chat/completions in 0.849299 seconds Error code: 503 INFO:openai._base_client:Retrying request to /chat/completions in 0.458106 seconds INFO:openai._base_client:Retrying request to /chat/completions in 0.784097 seconds Error code: 503 INFO:openai._base_client:Retrying request to /chat/completions in 0.384003 seconds INFO:openai._base_client:Retrying request to /chat/completions in 0.792511 seconds Error code: 503 INFO:openai._base_client:Retrying request to /chat/completions in 0.454747 seconds INFO:openai._base_client:Retrying request to /chat/completions in 0.914861 seconds Error code: 503 2025-02-24 10:08:53.689 | WARNING | agents.get_info:get_author_and_publish_date:239 - failed to parse from llm output: INFO:openai._base_client:Retrying request to /chat/completions in 0.448296 seconds INFO:openai._base_client:Retrying request to /chat/completions in 0.952611 seconds Error code: 503 2025-02-24 10:09:00.213 | WARNING | agents.get_info:get_info:321 - model hallucination: [] contains no summary tag 2025-02-24 10:09:00.359 | INFO | __main__:schedule_task:33 - task execute loop finished, work after 3600 seconds

现在我的.env的内容是
`# LLM_API_KEY=""
LLM_API_BASE="http://localhost:11434/"
ZHIPU_API_KEY="xxxx" #for the search tool
PRIMARY_MODEL="qwen2.5:3b"
SECONDARY_MODEL="deepseek-r1:7b"
#use a secondary model to excute the filtering task for the cost saving
#if not set, will use the primary model to excute the filtering task
VL_MODEL="qwen2.5:3b"
PB_API_AUTH="13143398903@163.com|xxxxxx" ##your pb superuser account and password

##belowing is optional, go as you need
VERBOSE="true" ##for detail log info. If not need, remove this item.
PROJECT_DIR="work_dir"
#PB_API_BASE="" ##only use if your pb not run on 127.0.0.1:8090
#LLM_CONCURRENT_NUMBER=8 ##for concurrent llm requests, make sure your llm provider supports it(leave default is 1)`

[jianjunLeung]

[tonyzhu2021]

http://localhost:11434/v1

[jianjunLeung]

http://localhost:11434/v1

这个也试过了，还是一样的问题

[bigbrother666sh]

你查下 ollama 使用openai 格式接口怎么设置啊……
看下 ollama 文档，或者随便找个大模型问问……

[McRonnie]

LLM_API_BASE="http://127.0.0.1:11434/v1/"
我这样设置是可以用的，是多了个斜杠的问题吗？