---
tags:
- not-for-all-audiences
---

# EXL2 quants of [crestf411/L3-70B-daybreak-storywriter-v0.4](https://huggingface.co/crestf411/L3-70B-daybreak-storywriter-v0.4)

[3.00 bits per weight](https://huggingface.co/kim512/L3-70B-daybreak-storywriter-v0.4-3.0bpw-h6-exl2)  
[3.50 bits per weight](https://huggingface.co/kim512/L3-70B-daybreak-storywriter-v0.4-3.5bpw-h6-exl2)  
[4.00 bits per weight](https://huggingface.co/kim512/L3-70B-daybreak-storywriter-v0.4-4.0bpw-h6-exl2)  
[4.50 bits per weight](https://huggingface.co/kim512/L3-70B-daybreak-storywriter-v0.4-4.5bpw-h6-exl2)  
[5.00 bits per weight](https://huggingface.co/kim512/L3-70B-daybreak-storywriter-v0.4-5.0bpw-h6-exl2)  
[6.00 bits per weight](https://huggingface.co/kim512/L3-70B-daybreak-storywriter-v0.4-6.0bpw-h6-exl2)  
[8.00 bits per weight](https://huggingface.co/kim512/L3-70B-daybreak-storywriter-v0.4-8.0bpw-h8-exl2)  

Created using the defaults from exllamav2 0.0.21 convert.py  
3.0bpw to 6.0bpw head bits = 6  
8.0bpw head bits = 8  
length = 8192  
dataset rows = 200  
measurement rows = 32  
measurement length = 8192  
  
# L3-70B-daybreak-storywriter-v0.4

Daybreak (2024 May 24) v0.4 LoRA on top of https://huggingface.co/tdrussell/Llama-3-70B-Instruct-Storywriter

Dataset curation to remove slop-perceived expressions continues.

The below regexes return 0 matches. **Bold** entries are new since v0.3.

* 'barely above a whisper',
* **'barely audible',**
* 'shiver([s]?) down',
* ' ministration',
* 'audible (["\'"]?)p[l]?op',
* 'can\'t help but',
* 'buck([s]?) my ',
* 'buck([s]?) h[ei][rs] ',
* '[Dd]espite h[ie][mr]self',
* 'slick slit',
* 'whatever it takes',
* 'unlike anything (s?)he',
* **'a mix([a-z]*) of',**
* 'wave after wave',
* 'reckless abandon',
* '[Mm]aybe, just maybe',
* **'eyes gleaming',**
* **'mischievously',**
* **"couldn't help but",**

From testing so far, it feels like temperature 0.8-0.9 is a good starting point. I have mostly tested with everything neutralized. Please give feedback on which parameters work good for you.