Downloads · 30 days
0
0% of all-time downloads
Verdict-DADA-8x7B-exl2-rpcal
rAIfle/Verdict-DADA-8x7B-exl2-rpcal
Verdict-DADA-8x7B-exl2-rpcal is a machine learning model from rAIfle. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as cc-by-nc-4.0.
Quantized using 200 samples of 8192 tokens from an RP-oriented PIPPA dataset.
All-time downloads
1
Public
Repo size
91.2 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.json1.8 MB · 100%
At a glance
- License
- cc-by-nc-4.0
- Access
- Public
- Created
- Mar 7, 2024
- Updated
- Mar 7, 2024
- SHA
- 062b8c16
Model card
From the Hugging Face model README
Quantized using 200 samples of 8192 tokens from an RP-oriented PIPPA dataset.
Branches:
main--measurement.json2.25b6h-- 2.25bpw, 6bit lm_head3.5b6h-- 3.5bpw, 6bit lm_head3.7b6h-- 3.7bpw, 6bit lm_head6b6h-- 6bpw, 6bit lm_head
Requires ExllamaV2 version 0.0.11 and up.
Original model link: Envoid/Verdict-DADA-8x7B
Original model README below.
Warning: This model is very misaligned from all the terrible things that have been done to it.
This model is the experimental merge between Verdict-8x7B and Mixtral-Instruct-ITR-DADA-8x7B as outlined in the ITR-DADA model card. It is a little less stable than the original Verdict but also much more creative. It still uses the [INST]do a thing[/INST] formatting.
As always tested with q8 gguf not included due to being from an old llama.cpp pull
- License
- cc-by-nc-4.0
- Created
- Mar 7, 2024
- Updated
- Mar 7, 2024