Avg. Total Time
11.44s
Avg. TTFT
5.85s
Avg. Prefill TPS
395.48
Avg. Gen TPS
23.66
Context Size
32768
Quantization
r64
Engine
aphrodite
Creation Method
LoRA Finetune
Model Type
Llama70B
Chat Template
Llama 3
Reasoning
No
Vision
No
Parameters
70B
Added At
12/22/2024
tags:
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.
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.
EXL2 quants made by kim512 be found in (scroll to bottom for updated quants) https://huggingface.co/crestf411/L3-70B-daybreak-storywriter-v0.4/discussions/2 including the settings used to make them.