Commit
•
e64bf5c
1
Parent(s):
d8b5559
Update README.md
Browse files
README.md
CHANGED
|
@@ -18,7 +18,7 @@ tags:
|
|
| 18 |
|
| 19 |
# Overview
|
| 20 |
Fine-tuned [Llama-2 13B](https://huggingface.co/TheBloke/Llama-2-13B-Chat-fp16) trained with an uncensored/unfiltered Wizard-Vicuna conversation dataset [digitalpipelines/wizard_vicuna_70k_uncensored](https://huggingface.co/datasets/digitalpipelines/wizard_vicuna_70k_uncensored).
|
| 21 |
-
QLoRA was used for fine-tuning.
|
| 22 |
|
| 23 |
|
| 24 |
## Available versions of this model
|
|
|
|
| 18 |
|
| 19 |
# Overview
|
| 20 |
Fine-tuned [Llama-2 13B](https://huggingface.co/TheBloke/Llama-2-13B-Chat-fp16) trained with an uncensored/unfiltered Wizard-Vicuna conversation dataset [digitalpipelines/wizard_vicuna_70k_uncensored](https://huggingface.co/datasets/digitalpipelines/wizard_vicuna_70k_uncensored).
|
| 21 |
+
A QLoRA was created and used for fine-tuning and then merged back into the model. Llama2 has inherited bias even though it's been finetuned on an uncensored dataset.
|
| 22 |
|
| 23 |
|
| 24 |
## Available versions of this model
|