[FEATURE] support for lm studio

[pechaut78]

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

[jeffthorness]

I'm with you and frankly a little puzzled why this isn't already supported.

[SphaeroX]

@dev: just load openai llm like that (python)

llm = ChatOpenAI(base_url="http://localhost:1234/v1")

[Eigenwise]

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

[pechaut78]

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 GitHub<#1276 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AAE2VARKD6C6NUEBPCRLJ43YRTYUNAVCNFSM6AAAAAA7YQB5COVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSMRTHE3TQMZSHA>. You are receiving this because you authored the thread.Message ID: ***@***.***>

[shiloh92]

Is there a website where they share workflows?

[ArkMaster123]

@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

[new4u]

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!

[RicardoFernandez-UY]

@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!!

[rachdeg]

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.

[jan-wijman]

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.

[edvinPL]

Can you make this work by having the LM Studio locally, but flowise on Render and somehow connect the two?

[x85432]

Hey! I just started checking out Flowise and saw this issue marked as a good first issue.

Not sure if it’s still relevant since the workaround seems to be working already — just wanted to ask if it still makes sense to work on this?

Happy to help if it’s still needed, but no worries if it’s already covered!

[le5h]

if it still makes sense to work on this?

LM studio supports images, but this workaround is not