Playground seems broken after today's Pull

[cainezen]

Anyone else see that the Playground extension broke after a pull today? Yesterday it was working fine but when I just did a Pull, I'm getting errors. Just curious if this is a known issue. Error:
Closing server running on port: 7860
18:47:59-142431 INFO Loading the extension "gallery"
18:47:59-144431 INFO Loading the extension "Playground"
18:47:59-147432 ERROR Failed to load the extension "Playground".
Traceback (most recent call last):
File "E:\StableLLM\Dec23\text-generation-webui\modules\extensions.py", line 37, in load_extensions
exec(f"import extensions.{name}.script")
File "", line 1, in
File "E:\StableLLM\Dec23\text-generation-webui\extensions\Playground\script.py", line 12, in
from modules.LoRA import add_lora_autogptq, add_lora_exllama, add_lora_exllamav2
ImportError: cannot import name 'add_lora_exllama' from 'modules.LoRA' (E:\StableLLM\Dec23\text-generation-webui\modules\LoRA.py)
--Update ---
I was able to at least suppress this error by removing the reference to "add_lora_exllama" in the script.py of Playground.

[TheLounger]

ExLlama (v1) was removed recently, so the extension would need to be updated.

[cainezen]

I figured it was some kind of deprecated feature. I use Playground religiously, so hopefully they update that extension.

[DelteRT]

I removed the lines referring to ExLlama (v1) and it is working again, if someone needs a quick and temporary solution, the lines to modify are (In the main script.py file):

Before: from modules.LoRA import add_lora_autogptq, add_lora_exllama, add_lora_exllamav2
After: from modules.LoRA import add_lora_autogptq, add_lora_exllamav2

Before:

    def add_lora_to_model(lora_name):
        if 'GPTQForCausalLM' in shared.model.__class__.__name__ or shared.args.loader == 'AutoGPTQ':
            add_lora_autogptq([lora_name])
        elif shared.model.__class__.__name__ in ['ExllamaModel', 'ExllamaHF'] or shared.args.loader == 'ExLlama':
            add_lora_exllama([lora_name])
        elif shared.model.__class__.__name__ in ['Exllamav2Model', 'Exllamav2HF'] or shared.args.loader == ['ExLlamav2', 'ExLlamav2_HF']:
            add_lora_exllamav2([lora_name])
        else:

After:

    def add_lora_to_model(lora_name):
        if 'GPTQForCausalLM' in shared.model.__class__.__name__ or shared.args.loader == 'AutoGPTQ':
            add_lora_autogptq([lora_name])
        elif shared.model.__class__.__name__ in ['Exllamav2Model', 'Exllamav2HF'] or shared.args.loader == ['ExLlamav2', 'ExLlamav2_HF']:
            add_lora_exllamav2([lora_name])
        else:

I'm not a programmer, and this will most likely break something related to the lora models, but if someone like me uses LLMs for work, and can't wait a couple of days, it's a temporary solution that seems to work

[cainezen]

Looks like it's broken again, this time on "AutoGPTQ". I had to remove the first two lines and just leave the "if" statement with ExLlamaV2.