Spaces:
Running
on
Zero
Running
on
Zero
Commit
·
313ab2a
1
Parent(s):
4ed25fe
debug tts
Browse files
app.py
CHANGED
|
@@ -2,7 +2,6 @@ import gradio as gr
|
|
| 2 |
import torch
|
| 3 |
import spaces
|
| 4 |
import torchaudio
|
| 5 |
-
from encodec import EncodecModel
|
| 6 |
from whisperspeech.vq_stoks import RQBottleneckTransformer
|
| 7 |
from encodec.utils import convert_audio
|
| 8 |
from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig, pipeline
|
|
@@ -64,8 +63,6 @@ def audio_to_sound_tokens_whisperspeech_transcribe(audio_path):
|
|
| 64 |
# print(tokenizer.encode("<|sound_0001|>", add_special_tokens=False))# return the audio tensor
|
| 65 |
# print(tokenizer.eos_token)
|
| 66 |
|
| 67 |
-
|
| 68 |
-
@spaces.GPU
|
| 69 |
def text_to_audio_file(text):
|
| 70 |
# gen a random id for the audio file
|
| 71 |
id = str(uuid.uuid4())
|
|
|
|
| 2 |
import torch
|
| 3 |
import spaces
|
| 4 |
import torchaudio
|
|
|
|
| 5 |
from whisperspeech.vq_stoks import RQBottleneckTransformer
|
| 6 |
from encodec.utils import convert_audio
|
| 7 |
from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig, pipeline
|
|
|
|
| 63 |
# print(tokenizer.encode("<|sound_0001|>", add_special_tokens=False))# return the audio tensor
|
| 64 |
# print(tokenizer.eos_token)
|
| 65 |
|
|
|
|
|
|
|
| 66 |
def text_to_audio_file(text):
|
| 67 |
# gen a random id for the audio file
|
| 68 |
id = str(uuid.uuid4())
|