Open pechaut78 opened 9 months ago
I'm with you and frankly a little puzzled why this isn't already supported.
@dev: just load openai llm like that (python)
llm = ChatOpenAI(base_url="http://localhost:1234/v1")
Works perfectly fine like this for me, I am using LM Studio, I just have API-key set to "none" in the credentials, Note the IP address for you will probably just be localhost:1234 but I am running the docker container so to find LM studio I needed to use the IP address of the vEthernet adapter
Excellent !!
Thanks a lot
Cordialement, CHAUT Pierre-Emmanuel
De : KennyVaneetvelde @.> Envoyé : vendredi 2 février 2024 15:18 À : FlowiseAI/Flowise @.> Cc : pechaut78 @.>; Author @.> Objet : Re: [FlowiseAI/Flowise] [FEATURE] support for lm studio (Issue #1276)
image.png (view on web)https://github.com/FlowiseAI/Flowise/assets/48944754/d356c9a8-6ab2-488c-b88b-0898e114f14c Works perfectly fine like this for me, I am using LM Studio, I just have API-key set to "none" in the credentials, Note the IP address for you will probably just be localhost:1234 but I am running the docker container so to find LM studio I needed to use the IP address of the vEthernet adapter
— Reply to this email directly, view it on GitHubhttps://github.com/FlowiseAI/Flowise/issues/1276#issuecomment-1923978328, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AAE2VARKD6C6NUEBPCRLJ43YRTYUNAVCNFSM6AAAAAA7YQB5COVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRTHE3TQMZSHA. You are receiving this because you authored the thread.Message ID: @.***>
Is there a website where they share workflows?
@KennyVaneetvelde I have tried this but didn't work. I have actually a flowise instance on DO but LMStudio on my local laptop. Do i have to have both locally for this to work? would love a DO deployment solution! Thanks in advance
在 Docker 中,要让容器内的应用访问宿主机的 localhost,您可以使用几种方法。以下是一些常见的方法:
使用特殊的 DNS 名称: Docker 为容器提供了一个特殊的 DNS 名称 host.docker.internal,它指向宿主机的 IP 地址。您可以在容器内部使用这个名称来访问宿主机的服务。
所以我用 http://host.docker.internal:1234/v1 替换了 http://localhost:1234/v1
It works!
@KennyVaneetvelde Hi, sounds great! Where did you get the docker image? I wasn't able to find one in dockerhub, and I wouldn't know how to build it. Thank you!!
在 Docker 中,要让容器内的应用访问宿主机的 localhost,您可以使用几种方法。以下是一些常见的方法:
使用特殊的 DNS 名称: Docker 为容器提供了一个特殊的 DNS 名称 host.docker.internal,它指向宿主机的 IP 地址。您可以在容器内部使用这个名称来访问宿主机的服务。
所以我用 http://host.docker.internal:1234/v1 替换了 http://localhost:1234/v1
It works!
I tried this but unfortunately it didn't work. What did work was changing the localhost url to http://172.17.0.1:1234/v1 and it worked like a charm! Note that I'm using flowise in Docker and LM studio locally.
Hi, Try tried the idea "KennyVaneetvelde commented on Feb 2 •", but the chat window is not rolling the message. The replay of the LLM in LLMStudio appears at ones. Although in LM studio I see the repons building up.
What can be wrong in my Flowise diagram.
Can you make this work by having the LM Studio locally, but flowise on Render and somehow connect the two?
Lm studio.is super easy to setup, and simpler than local ai.
It mimics openai api. Langchain supports it by passing a local base path..
Would bé wonderful to do thé same thing with flowise