Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -7,7 +7,6 @@ from huggingface_hub import snapshot_download
|
|
7 |
from dotenv import load_dotenv
|
8 |
load_dotenv()
|
9 |
|
10 |
-
|
11 |
# Check if CUDA is available
|
12 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
13 |
|
@@ -59,23 +58,6 @@ def process_prompt(prompt, voice, tokenizer, device):
|
|
59 |
|
60 |
return modified_input_ids.to(device), attention_mask.to(device)
|
61 |
|
62 |
-
# Generate speech tokens
|
63 |
-
@spaces.GPU()
|
64 |
-
def generate_speech_tokens(input_ids, attention_mask, model, params):
|
65 |
-
with torch.no_grad():
|
66 |
-
generated_ids = model.generate(
|
67 |
-
input_ids=input_ids,
|
68 |
-
attention_mask=attention_mask,
|
69 |
-
max_new_tokens=params["max_new_tokens"],
|
70 |
-
do_sample=True,
|
71 |
-
temperature=params["temperature"],
|
72 |
-
top_p=params["top_p"],
|
73 |
-
repetition_penalty=params["repetition_penalty"],
|
74 |
-
num_return_sequences=1,
|
75 |
-
eos_token_id=128258,
|
76 |
-
)
|
77 |
-
return generated_ids
|
78 |
-
|
79 |
# Parse output tokens to audio
|
80 |
def parse_output(generated_ids):
|
81 |
token_to_find = 128257
|
@@ -131,6 +113,7 @@ def redistribute_codes(code_list, snac_model):
|
|
131 |
return audio_hat.detach().squeeze().cpu().numpy() # Always return CPU numpy array
|
132 |
|
133 |
# Main generation function
|
|
|
134 |
def generate_speech(text, voice, temperature, top_p, repetition_penalty, max_new_tokens, progress=gr.Progress()):
|
135 |
if not text.strip():
|
136 |
return None
|
@@ -140,13 +123,18 @@ def generate_speech(text, voice, temperature, top_p, repetition_penalty, max_new
|
|
140 |
input_ids, attention_mask = process_prompt(text, voice, tokenizer, device)
|
141 |
|
142 |
progress(0.3, "Generating speech tokens...")
|
143 |
-
|
144 |
-
|
145 |
-
|
146 |
-
|
147 |
-
|
148 |
-
|
149 |
-
|
|
|
|
|
|
|
|
|
|
|
150 |
|
151 |
progress(0.6, "Processing speech tokens...")
|
152 |
code_list = parse_output(generated_ids)
|
|
|
7 |
from dotenv import load_dotenv
|
8 |
load_dotenv()
|
9 |
|
|
|
10 |
# Check if CUDA is available
|
11 |
device = "cuda" if torch.cuda.is_available() else "cpu"
|
12 |
|
|
|
58 |
|
59 |
return modified_input_ids.to(device), attention_mask.to(device)
|
60 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
61 |
# Parse output tokens to audio
|
62 |
def parse_output(generated_ids):
|
63 |
token_to_find = 128257
|
|
|
113 |
return audio_hat.detach().squeeze().cpu().numpy() # Always return CPU numpy array
|
114 |
|
115 |
# Main generation function
|
116 |
+
@spaces.GPU()
|
117 |
def generate_speech(text, voice, temperature, top_p, repetition_penalty, max_new_tokens, progress=gr.Progress()):
|
118 |
if not text.strip():
|
119 |
return None
|
|
|
123 |
input_ids, attention_mask = process_prompt(text, voice, tokenizer, device)
|
124 |
|
125 |
progress(0.3, "Generating speech tokens...")
|
126 |
+
with torch.no_grad():
|
127 |
+
generated_ids = model.generate(
|
128 |
+
input_ids=input_ids,
|
129 |
+
attention_mask=attention_mask,
|
130 |
+
max_new_tokens=max_new_tokens,
|
131 |
+
do_sample=True,
|
132 |
+
temperature=temperature,
|
133 |
+
top_p=top_p,
|
134 |
+
repetition_penalty=repetition_penalty,
|
135 |
+
num_return_sequences=1,
|
136 |
+
eos_token_id=128258,
|
137 |
+
)
|
138 |
|
139 |
progress(0.6, "Processing speech tokens...")
|
140 |
code_list = parse_output(generated_ids)
|