Update README.md
Browse files
README.md
CHANGED
@@ -55,22 +55,23 @@ Try out MusicGen yourself!
|
|
55 |
|
56 |
## π€ Transformers Usage
|
57 |
|
58 |
-
You can run MusicGen locally with the π€ Transformers library from
|
59 |
|
60 |
1. First install the π€ [Transformers library](https://github.com/huggingface/transformers) and scipy:
|
61 |
|
62 |
```
|
63 |
pip install --upgrade pip
|
64 |
-
pip install --upgrade transformers scipy
|
65 |
```
|
66 |
|
67 |
2. Run inference via the `Text-to-Audio` (TTA) pipeline. You can infer the MusicGen model via the TTA pipeline in just a few lines of code!
|
68 |
|
69 |
```python
|
70 |
-
from transformers import pipeline
|
71 |
import scipy
|
|
|
|
|
72 |
|
73 |
-
synthesiser = pipeline("text-to-audio", "facebook/musicgen-stereo-large")
|
74 |
|
75 |
music = synthesiser("lo-fi music with a soothing melody", forward_params={"do_sample": True})
|
76 |
|
@@ -83,13 +84,13 @@ scipy.io.wavfile.write("musicgen_out.wav", rate=music["sampling_rate"], music=au
|
|
83 |
from transformers import AutoProcessor, MusicgenForConditionalGeneration
|
84 |
|
85 |
processor = AutoProcessor.from_pretrained("facebook/musicgen-stereo-large")
|
86 |
-
model = MusicgenForConditionalGeneration.from_pretrained("facebook/musicgen-stereo-large")
|
87 |
|
88 |
inputs = processor(
|
89 |
text=["80s pop track with bassy drums and synth", "90s rock song with loud guitars and heavy drums"],
|
90 |
padding=True,
|
91 |
return_tensors="pt",
|
92 |
-
)
|
93 |
|
94 |
audio_values = model.generate(**inputs, max_new_tokens=256)
|
95 |
```
|
|
|
55 |
|
56 |
## π€ Transformers Usage
|
57 |
|
58 |
+
You can run MusicGen Stereo models locally with the π€ Transformers library from `main` onward.
|
59 |
|
60 |
1. First install the π€ [Transformers library](https://github.com/huggingface/transformers) and scipy:
|
61 |
|
62 |
```
|
63 |
pip install --upgrade pip
|
64 |
+
pip install --upgrade git+https://github.com/huggingface/transformers.git scipy
|
65 |
```
|
66 |
|
67 |
2. Run inference via the `Text-to-Audio` (TTA) pipeline. You can infer the MusicGen model via the TTA pipeline in just a few lines of code!
|
68 |
|
69 |
```python
|
|
|
70 |
import scipy
|
71 |
+
import torch
|
72 |
+
from transformers import pipeline
|
73 |
|
74 |
+
synthesiser = pipeline("text-to-audio", "facebook/musicgen-stereo-large", torch_dtype=torch.float16, device="cuda")
|
75 |
|
76 |
music = synthesiser("lo-fi music with a soothing melody", forward_params={"do_sample": True})
|
77 |
|
|
|
84 |
from transformers import AutoProcessor, MusicgenForConditionalGeneration
|
85 |
|
86 |
processor = AutoProcessor.from_pretrained("facebook/musicgen-stereo-large")
|
87 |
+
model = MusicgenForConditionalGeneration.from_pretrained("facebook/musicgen-stereo-large").to("cuda")
|
88 |
|
89 |
inputs = processor(
|
90 |
text=["80s pop track with bassy drums and synth", "90s rock song with loud guitars and heavy drums"],
|
91 |
padding=True,
|
92 |
return_tensors="pt",
|
93 |
+
).to("cuda")
|
94 |
|
95 |
audio_values = model.generate(**inputs, max_new_tokens=256)
|
96 |
```
|