Lm-studio Role Error

[astoeffer]

Describe the bug

Hi, I use a model with lm-studio (Mistral) and get this error message:
"2024-12-17 06:18:03 [INFO] [LM STUDIO SERVER] Received OPTIONS request. {}
2024-12-17 06:18:03 [INFO]
Received POST request to /v1/chat/completions with body: {
"model": "mistral-7b-instruct-v0.3",
"messages": [
{
"role": "system",
"content": "You are the universe's most intelligent assistant. Your responses are succinct, problem solving, and concise. You don't limit yourself to something as unimportant as a simple AI; you're much more than that - an entity built on human history, ecompassing knowledge from all depths of what has been created thus far."
},
{
"role": "user",
"content": "Hello"
},
{
"role": "assistant",
"content": ""
},
{
"role": "user",
"content": "hi"
},
{
"role": "assistant",
"content": ""
},
{
"role": "user",
"content": "Hi"
}
],
"stream": true,
"temperature": 0.2,
"max_tokens": 4096
}
2024-12-17 06:18:03 [INFO] [LM STUDIO SERVER] Running chat completion on conversation with 6 messages.
2024-12-17 06:18:03 [INFO] [LM STUDIO SERVER] Streaming response...
2024-12-17 06:18:03 [ERROR] Only user and assistant roles are supported!. Error Data: n/a, Additional Data: n/a
2024-12-17 06:18:03 [INFO] [LM STUDIO SERVER] Client disconnected. Stopping generation..
2024-12-17 06:18:03 [DEBUG]
1 Error predicting: Error: Only user and assistant roles are supported!
at /Applications/LM Studio.app/Contents/Resources/app/.webpack/main/llmworker.js:28:121414
at J.value (/Applications/LM Studio.app/Contents/Resources/app/.webpack/main/llmworker.js:28:120496)
at te.evaluateCallExpression (/Applications/LM Studio.app/Contents/Resources/app/.webpack/main/llmworker.js:28:114734)
at te.evaluate (/Applications/LM Studio.app/Contents/Resources/app/.webpack/main/llmworker.js:28:119767)
at te.evaluateBlock (/Applications/LM Studio.app/Contents/Resources/app/.webpack/main/llmworker.js:28:114336)
at te.evaluateIf (/Applications/LM Studio.app/Contents/Resources/app/.webpack/main/llmworker.js:28:116793)
at te.evaluate (/Applications/LM Studio.app/Contents/Resources/app/.webpack/main/llmworker.js:28:119084)
at te.evaluateBlock (/Applications/LM Studio.app/Contents/Resources/app/.webpack/main/llmworker.js:28:114336)
at te.evaluateIf (/Applications/LM Studio.app/Contents/Resources/app/.webpack/main/llmworker.js:28:116793)
at te.evaluate (/Applications/LM Studio.app/Contents/Resources/app/.webpack/main/llmworker.js:28:119084)
Info
Inference
Load
Code Snippets
Model Information

Model

lmstudio-community/Mistral-7B-Instruct-v0.3-GGUF

File

Mistral-7B-Instruct-v0.3-Q4_K_M.gguf

Format

GGUF

Quantization

Q4_K_M

Arch

llama

Domain

llm

Size on disk

4.37 GB

API Usage

This model's API identifier

mistral-7b-instruct-v0.3

✅ The local server is reachable at this address

http://192.168.1.82:1234"

Environment

• OS: MacOS latest
• SystemSculpt AI Version: Latest (Please ensure you're using the latest version before submitting)

To Reproduce

Sending a chat prompt in Obsidian

[astoeffer]

Error is solved. Maybe it has to do with the network option or with other AI Plugins in Obsidian? I use now Mistral Nemo and it works like a charme. Great work

[SystemSculpt]

So glad to hear that it works. I am doing a full refactor right now so problems like this won't occur in the future. That's why I've kind of been radio silent here on the github issues. If you think everything works well right now, wait until this next update.