Skip to content

Fix Nesa docs links and demo instructions #26

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
giwaov wants to merge 1 commit into
base: main
Choose a base branch
from
+10 −11
Open

Fix Nesa docs links and demo instructions #26

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
18 changes: 9 additions & 9 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -20,8 +20,8 @@ Nesa: Run AI models end-to-end encrypted.
</h3>

<!-- GitHub Repo Stats -->
[![GitHub Repo stars](https://img.shields.io/github/stars/nesaorg/Equivariant-Encryption-for-AI)](https://github.com/nesaorg/Equivariant-Encryption-for-AI/stargazers)
[![GitHub Repo forks](https://img.shields.io/github/forks/nesaorg/Equivariant-Encryption-for-AI)](https://github.com/nesaorg/Equivariant-Encryption-for-AI/network/members)
[![GitHub Repo stars](https://img.shields.io/github/stars/nesaorg/nesa)](https://github.com/nesaorg/nesa/stargazers)
[![GitHub Repo forks](https://img.shields.io/github/forks/nesaorg/nesa)](https://github.com/nesaorg/nesa/network/members)
[![GitHub followers](https://img.shields.io/github/followers/nesaorg?label=Follow)](https://github.com/nesaorg)
</div>

Expand All @@ -36,10 +36,10 @@ Forget multi-million dollar on-prem infrastructure, get the same privacy guarant
<table border="0">
<tr>
<td><b style="font-size:30px">Full Privacy</b></td>
<td>nesa serves AI with zero visibility on underlying data and full blindness on query</b></td>
<td>nesa serves AI with zero visibility on underlying data and full blindness on query</td>
</tr>
<tr>
<td><b style="font-size:30px">Speedy </td>
<td><b style="font-size:30px">Speedy</b></td>
<td>nesa delivers no latency on encrypted inference (<0.1% original execution time)</td>
</tr>
<tr>
Expand All @@ -53,11 +53,11 @@ Forget multi-million dollar on-prem infrastructure, get the same privacy guarant
</tr>
<tr>
<td><b style="font-size:30px">ChatGPT Compatible</b></td>
<td>nesa has a ChatGPT-compatible API for running encrypted AI with a one line change</td>
<td>nesa has a ChatGPT-compatible API for running encrypted AI with a one-line change</td>
</tr>
<tr>
<td><b style="font-size:30px">Quick Set-up</b></td>
<td>nesa is one click install and go. See documentation</td>
<td>nesa is one-click install and go. See documentation</td>
</tr>
</table>

Expand Down Expand Up @@ -141,7 +141,7 @@ Using a state-of-the-art large language model such as GPT-4o to evaluate whether

### Linguistic Domain Knowledge Attack

Using domain knowledge to design the loss function L, so that the loss L can capture the semantic meaning in the (decrypted) input, output and between.
Using domain knowledge to design the loss function L, so that the loss L can capture the semantic meaning in the (decrypted) input, output, and the relationship between them.

### Brute-force Algorithm Attack

Expand Down Expand Up @@ -180,7 +180,7 @@ Under the hood, the text you type is turned into encrypted tokens, the model pro

If you’d like to peek under the hood, below are quick examples demonstrating how to load the models directly from Hugging Face and run basic inferences.

##### Distillbert
##### DistilBERT

```python
import torch
Expand Down Expand Up @@ -245,4 +245,4 @@ Decoded Text: I'm super excited to join Nesa's Equivariant Encryption initiative
We invite the community to examine and test the security claims of Equivariant Encryption. As part of our commitment to transparency and continual refinement, we have organized a competition encouraging participants to probe for weaknesses and demonstrate potential exploits.

For details, please visit:
[https://github.com/nesaorg/Equivariant-Encryption-for-AI/blob/main/CONTEST.md](https://github.com/nesaorg/Equivariant-Encryption-for-AI/blob/main/CONTEST.md)
[https://github.com/nesaorg/nesa/blob/main/CONTEST.md](https://github.com/nesaorg/nesa/blob/main/CONTEST.md)
2 changes: 1 addition & 1 deletion demo/docs/tokenize.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@ from transformers import AutoTokenizer

hf_token = "<HF TOKEN>" # Replace with your token
model_id = "nesaorg/Llama-3.2-1B-Instruct-Encrypted"
tokenizer = AutoTokenizer.from_pretrained(model_id, token=hf_token, local_files_only=True)
tokenizer = AutoTokenizer.from_pretrained(model_id, token=hf_token, local_files_only=False)
```

### Tokenize and Decode Text
Expand Down
1 change: 0 additions & 1 deletion demo/readme.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -33,7 +33,6 @@ This folder contains a modified version of [oobabooga/text-generation-webui](htt

4. After installation completes, it should automatically launch a local client, or you can manually navigate to http://localhost:7860.

Usage
## Usage
We automatically download and select a model for the user, so they can start using right away after the local client loads. We've included information about switching between both demo models below:

Expand Down