lollms-webui
lollms-webui copied to clipboard
ParisNeo
Prompt for sending messages gone in "Discussions" after update
Expected Behavior
The prompt for typing and interacting with the AI should be there
Current Behavior
The prompt is gone, including the sidebar for changing models and personalities.
Steps to Reproduce
Cannot say detailed.
Updated via UI and now its gone. Maybe its a problem on my side but nevertheless: Better open the issue.
Screenshots
Hi, same problem here. no error in console, restarting does not solve the issue. Prompt just suddenly disappeared while i was trying to get a model running. have tried a lot. mount/unmount personalities, binding is allGPT in CPU mode now. but now the Prompt is gone ;-) Going to cry. goodnight. ;-)
found a workaround to fix this. I selected the binding Lord of llms Remote nodes, suddenly the chat prompt appeared after reload. after switching back to allgpt binding and a new persona (john doe) everything is back to normal.
Yesterday I changed the main upgrade script which messed up things for old users. Now you need to do two upgrades, the first one will upgrade your upgrade screipt, then the second one will execute the new upgrade script.
I know that sucks, but we are converging towards a good stable version. (I hope)
Sorry for my absence. The error still persists, but the workaround from @ediweber works most of the time. Its very curious actually. Sometimes the workaround doesnt work and I have to restart the program and then it works.
Unfortunately no console output but I can try debug-mode if wanted.
Morning. I have the same issue. I have tried the workarounds as mentioned by @ediweber but it did not work for me at all. I also tried to update twice as suggested by @ParisNeo. No errors in the console.
Hi. Which binding are you using? Next solution would be to try to reinstall the application from scratch. You can reuse the same personal folder so that you won't loose your data. Just remove the whole install folder, reinstall using the install script from the github release page then after install, reinstall the binding as the update only updates the app, not the bindings.
Afternoon. I am using the "CTransformers" binding with codellama-7b-instruct.Q5_K_M.gguf model.
I did try it with GPT4ALL binding as well but i wasn't sure which models were supported.
I will do the install from scratch and let you know..
GPT4ALL only supports Q4 models unfortunately. CTransformers is the best in terms of versatility. it supports many quantization levels down to Q2