Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: llama2
|
| 3 |
+
language:
|
| 4 |
+
- en
|
| 5 |
+
pipeline_tag: conversational
|
| 6 |
+
tags:
|
| 7 |
+
- Platypus2
|
| 8 |
+
- WinterGoddess
|
| 9 |
+
- frankenmerge
|
| 10 |
+
- dare
|
| 11 |
+
- ties
|
| 12 |
+
- 90b
|
| 13 |
+
---
|
| 14 |
+
# BigWeave v8 90B
|
| 15 |
+
|
| 16 |
+
<img src="https://cdn-uploads.huggingface.co/production/uploads/65a6db055c58475cf9e6def1/4CbbAN-X7ZWj702JrcCGH.png" width=600>
|
| 17 |
+
|
| 18 |
+
The BigWeave models aim to identify merge settings equaling or surpassing the performance of Goliath-120b. The version number merely tracks various attempts and is not a quality indicator. Only results demonstrating good performance are retained and shared.
|
| 19 |
+
|
| 20 |
+
This version is a passthrough merge of Platypus2-70b-instruct + WinterGoddess-1.4x-70b.
|
| 21 |
+
|
| 22 |
+
The 90b size allows for 4bit quants to fit into 48GB of VRAM.
|
| 23 |
+
|
| 24 |
+
# Prompting Format
|
| 25 |
+
Vicuna and Alpaca.
|
| 26 |
+
|
| 27 |
+
# Merge process
|
| 28 |
+
The models used in the merge are [Platypus2-70b-instruct](https://huggingface.co/garage-bAInd/Platypus2-70B-instruct) and [WinterGoddess-1.4x-70b](https://huggingface.co/Sao10K/WinterGoddess-1.4x-70B-L2).
|
| 29 |
+
|
| 30 |
+
# Acknowledgements
|
| 31 |
+
[@garage-bAInd](https://huggingface.co/garage-bAInd) For creating Platypus2
|
| 32 |
+
|
| 33 |
+
[@Sao10K](https://huggingface.co/Sao10K) For creating WinterGoddess
|
| 34 |
+
|
| 35 |
+
[@alpindale](https://huggingface.co/alpindale) For creating the original Goliath
|
| 36 |
+
|
| 37 |
+
[@chargoddard](https://huggingface.co/chargoddard) For developing [mergekit](https://github.com/cg123/mergekit).
|
| 38 |
+
|