Spaces:
Runtime error
Runtime error
update README
Browse files
README.md
CHANGED
|
@@ -28,7 +28,7 @@ You can play around with the model I trained on about 500 songs from my Spotify
|
|
| 28 |
#### Training can be run with Mel spectrograms of resolution 64x64 on a single commercial grade GPU (e.g. RTX 2080 Ti). The `hop_length` should be set to 1024 for better results.
|
| 29 |
|
| 30 |
```bash
|
| 31 |
-
python
|
| 32 |
--resolution 64 \
|
| 33 |
--hop_length 1024\
|
| 34 |
--input_dir path-to-audio-files \
|
|
@@ -38,7 +38,7 @@ python audiodiffusion/audio_to_images.py \
|
|
| 38 |
#### Generate dataset of 256x256 Mel spectrograms and push to hub (you will need to be authenticated with `huggingface-cli login`).
|
| 39 |
|
| 40 |
```bash
|
| 41 |
-
python
|
| 42 |
--resolution 256 \
|
| 43 |
--input_dir path-to-audio-files \
|
| 44 |
--output_dir data-256 \
|
|
@@ -49,7 +49,7 @@ python audiodiffusion/audio_to_images.py \
|
|
| 49 |
|
| 50 |
```bash
|
| 51 |
accelerate launch --config_file accelerate_local.yaml \
|
| 52 |
-
|
| 53 |
--dataset_name data-64 \
|
| 54 |
--resolution 64 \
|
| 55 |
--hop_length 1024 \
|
|
@@ -66,7 +66,7 @@ accelerate launch --config_file accelerate_local.yaml \
|
|
| 66 |
|
| 67 |
```bash
|
| 68 |
accelerate launch --config_file accelerate_local.yaml \
|
| 69 |
-
|
| 70 |
--dataset_name teticio/audio-diffusion-256 \
|
| 71 |
--resolution 256 \
|
| 72 |
--output_dir ddpm-ema-audio-256 \
|
|
@@ -86,7 +86,7 @@ accelerate launch --config_file accelerate_local.yaml \
|
|
| 86 |
|
| 87 |
```bash
|
| 88 |
accelerate launch --config_file accelerate_sagemaker.yaml \
|
| 89 |
-
|
| 90 |
--dataset_name teticio/audio-diffusion-256 \
|
| 91 |
--resolution 256 \
|
| 92 |
--output_dir ddpm-ema-audio-256 \
|
|
|
|
| 28 |
#### Training can be run with Mel spectrograms of resolution 64x64 on a single commercial grade GPU (e.g. RTX 2080 Ti). The `hop_length` should be set to 1024 for better results.
|
| 29 |
|
| 30 |
```bash
|
| 31 |
+
python audio_to_images.py \
|
| 32 |
--resolution 64 \
|
| 33 |
--hop_length 1024\
|
| 34 |
--input_dir path-to-audio-files \
|
|
|
|
| 38 |
#### Generate dataset of 256x256 Mel spectrograms and push to hub (you will need to be authenticated with `huggingface-cli login`).
|
| 39 |
|
| 40 |
```bash
|
| 41 |
+
python audio_to_images.py \
|
| 42 |
--resolution 256 \
|
| 43 |
--input_dir path-to-audio-files \
|
| 44 |
--output_dir data-256 \
|
|
|
|
| 49 |
|
| 50 |
```bash
|
| 51 |
accelerate launch --config_file accelerate_local.yaml \
|
| 52 |
+
train_unconditional.py \
|
| 53 |
--dataset_name data-64 \
|
| 54 |
--resolution 64 \
|
| 55 |
--hop_length 1024 \
|
|
|
|
| 66 |
|
| 67 |
```bash
|
| 68 |
accelerate launch --config_file accelerate_local.yaml \
|
| 69 |
+
train_unconditional.py \
|
| 70 |
--dataset_name teticio/audio-diffusion-256 \
|
| 71 |
--resolution 256 \
|
| 72 |
--output_dir ddpm-ema-audio-256 \
|
|
|
|
| 86 |
|
| 87 |
```bash
|
| 88 |
accelerate launch --config_file accelerate_sagemaker.yaml \
|
| 89 |
+
strain_unconditional.py \
|
| 90 |
--dataset_name teticio/audio-diffusion-256 \
|
| 91 |
--resolution 256 \
|
| 92 |
--output_dir ddpm-ema-audio-256 \
|