BarfingLemurs

Attention Emphasis/Hacks

2

**Description** I would be so cool to see this in whatever syntax is possible. ``` ### Instruction: Please make a formal annoucement that the pool's closed, due to ((AIDS)), specifically....

Is FlexGen+GPTQ 4bit possible?

1

Just a curious question I suppose! GPTQ 4bit - https://github.com/qwopqwop200/GPTQ-for-LLaMa Suppose someone eventually finetunes 175B OPT model, with loras or regular finetunng. or perhaps the BLOOM or BLOOMZ model, would...

[Feature Request] Raw txt file training

6

It would be great to include an example for training with a simple unformatted text file, in the readme!

[Feature Request]: Full transcript summary for videos

2

**Is your feature request related to a problem? Please describe.** **新功能是否与解决某个问题相关, 请描述** The video summary is often truncated, especially long videos. I tried 2hr YouTube videos, but the tokens seen...

local exllamav2 (TabbyAPI) KeyError: 'stop'

5

Thanks for sharing the project! The interrupt feature is really impressive! :) I'm getting an error on Ubuntu 22.04 when trying a different backend, with a fresh install of tabbyAPI:...

it looks like eng and other cleaners isn't supported on the Feb release, will there be an update to the apk releases?

[Question]: Back button?

2

### Question Is there a way to go back to my previous dialog? A regeneration feature might not be useful since aider adopts a temp=0, but I often find myself...

I posted a very long bunch of text through the gui. After generating 800 or so, I get an AssertionError. Linux, 3090 ``` AssertionError: Every "[" character must be paired...

 2409.13588v1 is there a plugin? thanks.

Hi @cg123, would there be a way to swap the original distilled deepseek model's tokenizers to make them somewhat compatible with llama.cpp speculative decoding? Another case, could deepseekvl2, which has...