Models now showing in menu after restart

[nikben]

Expected Behavior

Please describe the behavior you are expecting.
In "Settings" the "model" menu should show all models downloaded, shown in second screenshot.
Initially, it worked as expected.

Current Behavior

Please describe the behavior you are currently experiencing.
In "Settings" the "model" menu is empty, as seen in the third screenshot.
This happened after a reboot to add more memory to the PC.

Steps to Reproduce

Please provide detailed steps to reproduce the issue.

1. Step 1: Stop the webui.sh script.
2. Step 2: Start the webui.sh script.

Possible Solution

If you have any suggestions on how to fix the issue, please describe them here.
None found.

Context

Please provide any additional context about the issue.
OS: Ubuntu 22.04.2 LTS
Stopped webui.sh script, deleted both models, copied just one (gpt4all-lora-quantized-ggml.bin) and started webui.sh with the result shown in first screenshot.
Model is found and skipped for download, but is not found later on.
If I try to change settings, I get a message that everything was successful, although no model is selected.

As a result, GPT4all does not respond at all.

Screenshots

If applicable, add screenshots to help explain the issue.

[cmdrscotty]

I'm also getting this on the windows version as well.

i'm running this on Windows 10, on a Ryzen 7 2700

i can see in the console that it does indeed load the model, but fails to actually use it, and the settings page shows nothing to select. Just did a git clone to get the latest version as well

llama_model_load: loading model from './models/llamacpp/gpt4all-lora-quantized-ggml.bin' - please wait ...
llama_model_load: n_vocab = 32001
llama_model_load: n_ctx   = 512
llama_model_load: n_embd  = 4096
llama_model_load: n_mult  = 256
llama_model_load: n_head  = 32
llama_model_load: n_layer = 32
llama_model_load: n_rot   = 128
llama_model_load: f16     = 2
llama_model_load: n_ff    = 11008
llama_model_load: n_parts = 1
llama_model_load: type    = 1
llama_model_load: ggml map size = 4017.70 MB
llama_model_load: ggml ctx size =  81.25 KB
llama_model_load: mem required  = 5809.78 MB (+ 2052.00 MB per state)
llama_model_load: loading tensors from './models/llamacpp/gpt4all-lora-quantized-ggml.bin'
llama_model_load: model size =  4017.27 MB / num tensors = 291
llama_init_from_file: kv self size  =  512.00 MB

--looks good up to here, below is where i tried giving it a chat command and I can see the HTTP requests to process it, but nothing happens--

[2023-04-20 11:11:24,622] {_internal.py:224} INFO - 127.0.0.1 - - [20/Apr/2023 11:11:24] "GET /list_discussions HTTP/1.1" 200 -
[2023-04-20 11:11:30,086] {_internal.py:224} INFO - 127.0.0.1 - - [20/Apr/2023 11:11:30] "←[36mGET /static/images/copy.png HTTP/1.1←[0m" 304 -
[2023-04-20 11:11:30,087] {_internal.py:224} INFO - 127.0.0.1 - - [20/Apr/2023 11:11:30] "←[36mGET /static/images/refresh.png HTTP/1.1←[0m" 304 -
[2023-04-20 11:11:30,087] {_internal.py:224} INFO - 127.0.0.1 - - [20/Apr/2023 11:11:30] "←[36mGET /static/images/thumb_up.png HTTP/1.1←[0m" 304 -
[2023-04-20 11:11:30,088] {_internal.py:224} INFO - 127.0.0.1 - - [20/Apr/2023 11:11:30] "←[36mGET /static/images/thumb_down.png HTTP/1.1←[0m" 304 -

[ParisNeo]

Hi there, can you test this: on the page, right click and press inspect then in the menue ui that apears select console and take a screenshot

[ParisNeo]

i think that you don't have the right structure. Can you do a pull request. i have made a little modification that should repare this.

[cmdrscotty]

ok did a pull which updated a few files and ran it again, I can still see in the command prompt console that it's loading the weights from the model

but the UI doesn't behave as it if loaded them, and settings page doesn't show anything that can be selected.

Console view in browser (tried Opera GX (chromium) and FireFox) both show this

[ParisNeo]

Which browser are you using?

[ParisNeo]

This should have been resolved in last version