Update README.md
Browse files
README.md
CHANGED
@@ -15,38 +15,57 @@ license: llama2
|
|
15 |
- NVIDIA GPU + [CUDA](https://developer.nvidia.com/cuda-downloads)
|
16 |
|
17 |
### Installation
|
18 |
-
|
19 |
|
20 |
-
|
21 |
-
|
22 |
-
|
23 |
-
|
|
|
24 |
|
25 |
-
2. Install dependent packages
|
26 |
-
|
27 |
-
```bash
|
28 |
-
pip install -r requirements.txt
|
29 |
-
```
|
30 |
|
31 |
### Model Weights
|
32 |
-
We
|
33 |
Please download the checkpoints and save under the folder `./pretrained`.
|
34 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
35 |
To reconstruct the image from the SEED visual codes using unCLIP SD-UNet, please download the pretrained [unCLIP SD](https://huggingface.co/stabilityai/stable-diffusion-2-1-unclip).
|
36 |
Rename the checkpoint directory to **"diffusion_model"** and create a soft link to the "pretrained/seed_tokenizer" directory.
|
37 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
38 |
|
39 |
### Inference for visual tokenization and de-tokenization
|
40 |
To discretize an image to 1D visual codes with causal dependency, and reconstruct the image from the visual codes using the off-the-shelf unCLIP SD-UNet:
|
|
|
41 |
```bash
|
|
|
42 |
python scripts/seed_tokenizer_inference.py
|
43 |
```
|
44 |
|
45 |
-
### Launching Demo of SEED-LLaMA Locally
|
|
|
|
|
46 |
```bash
|
47 |
-
|
48 |
-
|
|
|
|
|
|
|
49 |
```
|
|
|
|
|
50 |
|
51 |
## Citation
|
52 |
If you find the work helpful, please consider citing:
|
|
|
15 |
- NVIDIA GPU + [CUDA](https://developer.nvidia.com/cuda-downloads)
|
16 |
|
17 |
### Installation
|
18 |
+
Clone the repo and install dependent packages
|
19 |
|
20 |
+
```bash
|
21 |
+
git clone https://github.com/AILab-CVC/SEED.git
|
22 |
+
cd SEED
|
23 |
+
pip install -r requirements.txt
|
24 |
+
```
|
25 |
|
|
|
|
|
|
|
|
|
|
|
26 |
|
27 |
### Model Weights
|
28 |
+
We release the pretrained SEED Tokenizer and De-Tokenizer, instruction tuned SEED-LLaMA-8B and SEED-LLaMA-14B in [SEED Hugging Face](https://huggingface.co/AILab-CVC/SEED).
|
29 |
Please download the checkpoints and save under the folder `./pretrained`.
|
30 |
|
31 |
+
```bash
|
32 |
+
cd pretrained # SEED/pretrained
|
33 |
+
git lfs install
|
34 |
+
git clone https://huggingface.co/AILab-CVC/SEED
|
35 |
+
mv SEED/* ./
|
36 |
+
```
|
37 |
+
|
38 |
To reconstruct the image from the SEED visual codes using unCLIP SD-UNet, please download the pretrained [unCLIP SD](https://huggingface.co/stabilityai/stable-diffusion-2-1-unclip).
|
39 |
Rename the checkpoint directory to **"diffusion_model"** and create a soft link to the "pretrained/seed_tokenizer" directory.
|
40 |
|
41 |
+
```bash
|
42 |
+
# SEED/pretrained
|
43 |
+
git lfs install
|
44 |
+
git clone https://huggingface.co/stabilityai/stable-diffusion-2-1-unclip
|
45 |
+
mv stable-diffusion-2-1-unclip seed_tokenizer/diffusion_model
|
46 |
+
```
|
47 |
+
|
48 |
|
49 |
### Inference for visual tokenization and de-tokenization
|
50 |
To discretize an image to 1D visual codes with causal dependency, and reconstruct the image from the visual codes using the off-the-shelf unCLIP SD-UNet:
|
51 |
+
|
52 |
```bash
|
53 |
+
cd .. # SEED/
|
54 |
python scripts/seed_tokenizer_inference.py
|
55 |
```
|
56 |
|
57 |
+
### Launching Gradio Demo of SEED-LLaMA-14B Locally
|
58 |
+
Building the local demo of SEED-LLaMA-14B currently requires 2*32GB devices.
|
59 |
+
|
60 |
```bash
|
61 |
+
# SEED/
|
62 |
+
# in first terminal
|
63 |
+
sh scripts/start_backend.sh
|
64 |
+
# in second terminal
|
65 |
+
sh scripts/start_frontend.sh
|
66 |
```
|
67 |
+
Then the demo can be accessed through http://127.0.0.1:80
|
68 |
+
|
69 |
|
70 |
## Citation
|
71 |
If you find the work helpful, please consider citing:
|