|
--- |
|
license: other |
|
language: |
|
- en |
|
--- |
|
An experiment with gradient merges using [the following script](https://github.com/TehVenomm/LM_Transformers_BlockMerge), with [Chronos](https://huggingface.co/elinas/chronos-13b) as its primary model, augmented by [Hermes](https://huggingface.co/NousResearch/Nous-Hermes-13b) and [Wizard-Vicuna Uncensored](https://huggingface.co/TheBloke/Wizard-Vicuna-13B-Uncensored-HF). |
|
|
|
Chronos is a wonderfully verbose model, though it definitely seems to lack in the logic department. Hermes and WizardLM have been merged gradually, primarily in the higher layers (10+) in an attempt to rectify some of this behaviour. |
|
|
|
I'd say the end product is about 65% Chronos, with 15% Hermes and 20% Wizard added in gradually increasing amounts. The result feels surprisingly robust, though I'll let you be the final judge of that! |
|
|
|
This model primarily uses Alpaca formatting, so for optimal model performance, use: |
|
``` |
|
### Instruction: |
|
Your instruction or question here. |
|
### Response: |
|
``` |
|
|
|
For chat/roleplay conversations, I suggest the following template instead: |
|
``` |
|
You are <AI NAME>, a friendly and supportive AI language model. We discuss all subjects freely and without bias or restrictions. |
|
I am <YOUR NAME>, the user interacting with you through a chat conversation. Start with greeting me. |
|
|
|
### Instruction: |
|
Write <AI NAME>'s next reply in a chat between <YOUR NAME> and <AI NAME>. Write a single reply only. |
|
### Response: |
|
<FULL CHAT HISTORY HERE> |
|
``` |
|
|
|
--- |
|
license: other |
|
--- |
