AI Assistant answers blank response after fresh OpenOps install

Copilot & AI
1

Hi everyone,
Thank you so much for all your work around OpenOps.
I try to experiment OpenOps on a self-hosted machine.
I followed exactely the install instructions using localai backend.

I don’t know how to check OpenOps version but
I cloned GitHub - mattermost/openops: Open source stack for applying AI to workflows in secure environments on master branch currently on 5eee75a.

I succeeded to connect to the Mattermost OpenOps instance.
When I try to initiate a discussion with AI Assistant, the latter responds instantly with blank answers.

image(1)
image(1)1917×993 116 KB

Thank you in advance for your help, wish you best for the end of the year.

2

When I examin openops-localai docker container stdout, with docker logs, I see this error:

12:34PM DBG Request received: {"model":"gpt4all-j","file":"","language":"","response_format":"","size":"","prompt":null,"instruction":"","input":null,"stop":null,"messages":[{"role":"system","content":"You are a helpful assistant."},{"role":"user","content":"Hello."}],"stream":true,"echo":false,"top_p":1,"top_k":0,"temperature":1,"max_tokens":0,"n":0,"batch":0,"f16":false,"ignore_eos":false,"repeat_penalty":0,"n_keep":0,"mirostat_eta":0,"mirostat_tau":0,"mirostat":0,"frequency_penalty":0,"tfz":0,"seed":0,"mode":0,"step":0,"typical_p":0}
12:34PM DBG Parameter Config: &{OpenAIRequest:{Model:gpt4all-j File: Language: ResponseFormat: Size: Prompt:<nil> Instruction: Input:<nil> Stop:<nil> Messages:[]
Stream:false Echo:false TopP:1 TopK:80 Temperature:1 Maxtokens:512 N:0 Batch:0 F16:false IgnoreEOS:false RepeatPenalty:0 Keep:0 MirostatETA:0 MirostatTAU:0 Mirostat:0 FrequencyPenalty:0 TFZ:0 Seed:0 Mode:0 Step:0 TypicalP:0} Name: StopWords:[] Cutstrings:[] TrimSpace:[] ContextSize:512 F16:false Threads:4 Debug:true Roles:map[] Embeddings:false Backend: TemplateConfig:{Completion: Chat: Edit:} MirostatETA:0 MirostatTAU:0 Mirostat:0 NGPULayers:0 MMap:false MMlock:false TensorSplit:
MainGPU: ImageGenerationAssets: PromptCachePath: PromptCacheAll:false PromptCacheRO:false PromptStrings:[] InputStrings:[] InputToken:[]}
12:34PM DBG Stream request received
12:34PM DBG Template found, input modified to: The prompt below is a question to answer, a task to complete, or a conversation to respond to; decide which and write an appropriate response
### Prompt:
You are a helpful assistant.
Hello.
### Response:
12:34PM DBG Loading model 'gpt4all-j' greedly
12:34PM DBG [llama] Attempting to load
12:34PM DBG Loading model llama from gpt4all-j
12:34PM DBG Sending chunk: {"object":"chat.completion.chunk","model":"gpt4all-j","choices":[{"delta":{"role":"assistant"}}],"usage":{"prompt_tokens":0,"completion_tokens":0,"total_tokens":0}}

12:34PM DBG Loading model in memory from file: /models/gpt4all-j
llama.cpp: loading model from /models/gpt4all-j
error loading model: unexpectedly reached end of file
llama_init_from_file: failed to load model
12:34PM DBG [llama] Fails: failed loading model
12:34PM DBG [gpt4all] Attempting to load
12:34PM DBG Loading model gpt4all from gpt4all-j
12:34PM DBG Loading model in memory from file: /models/gpt4all-j
12:34PM DBG [gpt4all] Loads OK

which indicates the model was not correctly loaded.

Would you know why?

3

Checking the OpenOps version can be a bit tricky, but since you cloned from the master branch, you’re likely up-to-date.

4

As for the AI Assistant giving blank responses, I’ve encountered a similar situation before. One time, while testing out an AI for sales, I faced a similar issue where it seemed like the AI wasn’t engaging properly. What helped me was tweaking the settings and giving it more context. Sometimes, these AIs need a bit of guidance to understand what we’re looking for. Maybe try phrasing your questions a bit differently or providing more details to see if that helps spark some meaningful responses.

5

I ran into a similar issue after a clean install of OpenOps a while back — in my case, the AI Assistant was returning blank responses because the API key wasn’t properly linked, even though it looked like everything was set up correctly. Double-checking the environment variables and ensuring the model endpoint was reachable solved it for me.

Another thing worth checking is the console/log output when the assistant runs — sometimes there are silent errors related to permissions or token limits that don’t show up in the UI.

To help troubleshoot things like this more efficiently, I’ve started using Overchat AI — it keeps my prompt history, results, and context all in one place while I work in the browser. Makes it easier to compare outputs, tweak inputs, and see what changed after system updates like this.

Let me know if you get it working or need help comparing configs — happy to share what worked on my end.