Spaces:
Running
on
Zero
Running
on
Zero
Integrate OpenAI API for enhanced LLM functionality and update configuration settings
Browse files- app.py +413 -607
- requirements.txt +3 -1
app.py
CHANGED
@@ -6,13 +6,14 @@ import uuid
|
|
6 |
import json
|
7 |
import time
|
8 |
import traceback
|
9 |
-
import re
|
10 |
import gradio as gr
|
11 |
import torch
|
12 |
import numpy as np
|
13 |
import scipy.io.wavfile as wavfile
|
14 |
-
import
|
15 |
from dotenv import load_dotenv
|
|
|
|
|
16 |
|
17 |
# --- Whisper Import ---
|
18 |
try:
|
@@ -27,6 +28,7 @@ except ImportError:
|
|
27 |
# --- SNAC Import ---
|
28 |
try:
|
29 |
from snac import SNAC
|
|
|
30 |
except ImportError:
|
31 |
print("ERROR: SNAC library not found. Please install it:")
|
32 |
print("pip install git+https://github.com/hubertsiuzdak/snac.git")
|
@@ -35,680 +37,484 @@ except ImportError:
|
|
35 |
# --- Load Environment Variables ---
|
36 |
load_dotenv()
|
37 |
|
38 |
-
# ---
|
39 |
-
|
|
|
|
|
40 |
|
41 |
-
#
|
42 |
-
|
43 |
-
|
44 |
-
|
45 |
-
|
|
|
46 |
|
47 |
-
|
48 |
-
TTS_API_ENDPOINT = f"{SERVER_BASE_URL}/v1/completions"
|
49 |
-
TTS_MODEL = "mrrtmob/tts-khm-3"
|
50 |
|
51 |
# --- Device Setup ---
|
52 |
-
if torch.cuda.is_available()
|
53 |
-
|
54 |
-
|
55 |
-
print("SNAC vocoder and Whisper STT will use CUDA if possible.")
|
56 |
-
else:
|
57 |
-
tts_device = "cpu"
|
58 |
-
stt_device = "cpu"
|
59 |
-
print("CUDA not available. SNAC vocoder and Whisper STT will use CPU.")
|
60 |
|
61 |
# --- Model Loading ---
|
62 |
print("Loading SNAC vocoder model...")
|
63 |
-
snac_model = None
|
64 |
try:
|
65 |
snac_model = SNAC.from_pretrained("hubertsiuzdak/snac_24khz")
|
66 |
-
snac_model = snac_model.to(
|
67 |
snac_model.eval()
|
68 |
-
print(f"SNAC vocoder loaded to {
|
69 |
except Exception as e:
|
70 |
print(f"Error loading SNAC model: {e}")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
71 |
|
72 |
print("Loading Whisper STT model...")
|
73 |
WHISPER_MODEL_NAME = "base.en"
|
74 |
-
whisper_model = None
|
75 |
try:
|
76 |
whisper_model = whisper.load_model(WHISPER_MODEL_NAME, device=stt_device)
|
77 |
print(f"Whisper model '{WHISPER_MODEL_NAME}' loaded successfully.")
|
78 |
except Exception as e:
|
79 |
print(f"Error loading Whisper model: {e}")
|
|
|
80 |
|
81 |
# --- Constants ---
|
82 |
MAX_MAX_NEW_TOKENS = 4096
|
83 |
-
|
84 |
MAX_SEED = np.iinfo(np.int32).max
|
85 |
-
|
86 |
-
|
87 |
-
|
88 |
-
|
89 |
-
|
90 |
-
|
91 |
-
|
92 |
-
|
93 |
-
|
94 |
-
|
95 |
-
|
96 |
-
|
97 |
-
|
98 |
-
|
99 |
-
|
100 |
-
|
101 |
-
|
102 |
-
|
103 |
-
return
|
104 |
-
|
105 |
-
def clean_chat_history(limited_chat_history):
|
106 |
-
cleaned_ollama_format = []
|
107 |
-
if not limited_chat_history:
|
108 |
-
return []
|
109 |
-
for user_msg_display, bot_msg_display in limited_chat_history:
|
110 |
-
user_text = None
|
111 |
-
if isinstance(user_msg_display, str):
|
112 |
-
if user_msg_display.startswith("π€ (Audio Input): "):
|
113 |
-
user_text = user_msg_display.split("π€ (Audio Input): ", 1)[1]
|
114 |
-
elif user_msg_display.startswith(("@tara-tts ", "@tara-llm ")):
|
115 |
-
user_text = user_msg_display.split(" ", 1)[1]
|
116 |
-
else:
|
117 |
-
user_text = user_msg_display
|
118 |
-
elif isinstance(user_msg_display, tuple):
|
119 |
-
if len(user_msg_display) > 1 and isinstance(user_msg_display[1], str):
|
120 |
-
user_text = user_msg_display[1].replace("π€: ", "")
|
121 |
-
elif isinstance(user_msg_display[0], str) and not user_msg_display[0].endswith((".wav", ".mp3")):
|
122 |
-
user_text = user_msg_display[0]
|
123 |
-
|
124 |
-
bot_text = None
|
125 |
-
if isinstance(bot_msg_display, tuple):
|
126 |
-
if len(bot_msg_display) > 1 and isinstance(bot_msg_display[1], str):
|
127 |
-
bot_text = bot_msg_display[1]
|
128 |
-
elif isinstance(bot_msg_display, str):
|
129 |
-
if not bot_msg_display.startswith(("[Error", "(Error", "Sorry,", "(No input", "Processing", "(TTS failed")):
|
130 |
-
bot_text = bot_msg_display
|
131 |
-
|
132 |
-
if user_text and user_text.strip():
|
133 |
-
cleaned_ollama_format.append({"role": "user", "content": user_text})
|
134 |
-
if bot_text and bot_text.strip():
|
135 |
-
cleaned_ollama_format.append({"role": "assistant", "content": bot_text})
|
136 |
-
return cleaned_ollama_format
|
137 |
|
138 |
-
|
139 |
-
|
140 |
-
|
141 |
-
|
142 |
-
|
143 |
-
|
144 |
-
|
145 |
-
|
146 |
-
|
147 |
-
|
148 |
-
|
149 |
-
|
150 |
-
|
151 |
-
|
152 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
153 |
|
154 |
-
def redistribute_codes(
|
155 |
-
|
156 |
-
|
|
|
|
|
|
|
157 |
|
158 |
-
snac_device = next(target_snac_model.parameters()).device
|
159 |
layer_1, layer_2, layer_3 = [], [], []
|
160 |
-
num_tokens = len(absolute_code_list)
|
161 |
-
num_groups = num_tokens // ORPHEUS_N_LAYERS
|
162 |
|
163 |
-
|
164 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
165 |
|
166 |
-
|
|
|
167 |
|
168 |
-
|
169 |
-
|
170 |
-
|
171 |
-
|
172 |
-
|
173 |
-
group_codes = absolute_code_list[base_idx:base_idx + ORPHEUS_N_LAYERS]
|
174 |
-
processed_group = [None] * ORPHEUS_N_LAYERS
|
175 |
-
valid_group = True
|
176 |
-
|
177 |
-
for j, token_id in enumerate(group_codes):
|
178 |
-
if not (ORPHEUS_MIN_ID <= token_id < ORPHEUS_MAX_ID):
|
179 |
-
valid_group = False
|
180 |
-
break
|
181 |
-
|
182 |
-
layer_index = (token_id - ORPHEUS_MIN_ID) // ORPHEUS_TOKENS_PER_LAYER
|
183 |
-
code_index = (token_id - ORPHEUS_MIN_ID) % ORPHEUS_TOKENS_PER_LAYER
|
184 |
-
|
185 |
-
if layer_index != j:
|
186 |
-
valid_group = False
|
187 |
-
break
|
188 |
-
|
189 |
-
processed_group[j] = code_index
|
190 |
-
|
191 |
-
if not valid_group:
|
192 |
-
continue
|
193 |
-
|
194 |
-
try:
|
195 |
-
layer_1.append(processed_group[0])
|
196 |
-
layer_2.append(processed_group[1])
|
197 |
-
layer_3.append(processed_group[2])
|
198 |
-
layer_3.append(processed_group[3])
|
199 |
-
layer_2.append(processed_group[4])
|
200 |
-
layer_3.append(processed_group[5])
|
201 |
-
layer_3.append(processed_group[6])
|
202 |
-
except (IndexError, TypeError):
|
203 |
-
continue
|
204 |
|
205 |
-
|
206 |
-
|
207 |
-
|
208 |
-
|
209 |
-
print(f" - Final SNAC layer sizes: L1={len(layer_1)}, L2={len(layer_2)}, L3={len(layer_3)}")
|
210 |
-
|
211 |
-
codes = [
|
212 |
-
torch.tensor(layer_1, device=snac_device, dtype=torch.long).unsqueeze(0),
|
213 |
-
torch.tensor(layer_2, device=snac_device, dtype=torch.long).unsqueeze(0),
|
214 |
-
torch.tensor(layer_3, device=snac_device, dtype=torch.long).unsqueeze(0)
|
215 |
-
]
|
216 |
-
|
217 |
-
with torch.no_grad():
|
218 |
-
audio_hat = target_snac_model.decode(codes)
|
219 |
-
|
220 |
-
return audio_hat.detach().squeeze().cpu().numpy()
|
221 |
-
except Exception as e:
|
222 |
-
print(f"Error during tensor creation or SNAC decoding: {e}")
|
223 |
-
return None
|
224 |
|
225 |
-
def
|
226 |
-
|
|
|
227 |
return None
|
228 |
-
|
229 |
-
print(f"Generating speech via TTS server for: '{text[:50]}...'")
|
230 |
-
start_time = time.time()
|
231 |
-
|
232 |
-
payload = {
|
233 |
-
"model": TTS_MODEL,
|
234 |
-
"prompt": f"<|audio|>{voice}: {text}<|eot_id|>",
|
235 |
-
"temperature": tts_temperature,
|
236 |
-
"top_p": tts_top_p,
|
237 |
-
"repeat_penalty": tts_repetition_penalty,
|
238 |
-
"max_tokens": max_new_tokens_audio,
|
239 |
-
"stop": ["<|eot_id|>", "<|audio|>"],
|
240 |
-
"stream": False
|
241 |
-
}
|
242 |
-
|
243 |
-
print(f" - Sending payload to {TTS_API_ENDPOINT} (Model: {TTS_MODEL})")
|
244 |
|
245 |
try:
|
246 |
-
|
247 |
-
response = requests.post(
|
248 |
-
TTS_API_ENDPOINT,
|
249 |
-
json=payload,
|
250 |
-
headers=headers,
|
251 |
-
timeout=180
|
252 |
-
)
|
253 |
-
response.raise_for_status()
|
254 |
-
response_json = response.json()
|
255 |
|
256 |
-
|
257 |
|
258 |
-
|
259 |
-
|
260 |
-
|
261 |
-
|
262 |
-
|
263 |
-
|
264 |
-
|
265 |
-
|
266 |
-
|
267 |
-
|
268 |
-
|
269 |
-
|
270 |
-
|
271 |
-
|
272 |
-
|
273 |
-
|
274 |
-
|
275 |
-
|
276 |
-
|
277 |
-
snac_time = time.time()
|
278 |
-
print(f" - Generated audio samples via SNAC, shape: {audio_samples.shape}")
|
279 |
-
print(f" - Total TTS generation time: {snac_time - start_time:.2f}s")
|
280 |
return (24000, audio_samples)
|
281 |
-
|
282 |
else:
|
283 |
-
print(
|
284 |
return None
|
285 |
|
286 |
-
except requests.exceptions.RequestException as e:
|
287 |
-
print(f"Error during request to TTS server: {e}")
|
288 |
-
return None
|
289 |
except Exception as e:
|
290 |
-
print(f"Error during
|
291 |
traceback.print_exc()
|
292 |
return None
|
293 |
|
294 |
-
# ---
|
295 |
-
def
|
296 |
-
|
|
|
|
|
|
|
297 |
try:
|
298 |
-
|
299 |
-
|
300 |
-
|
301 |
-
|
302 |
-
|
303 |
-
|
304 |
-
|
305 |
-
|
306 |
-
|
307 |
-
|
308 |
-
|
|
|
|
|
|
|
309 |
|
310 |
-
|
311 |
start_time = time.time()
|
312 |
-
|
313 |
-
|
314 |
-
|
315 |
-
|
316 |
-
|
|
|
|
|
|
|
317 |
)
|
318 |
-
response.raise_for_status()
|
319 |
-
response_json = response.json()
|
320 |
-
end_time = time.time()
|
321 |
|
322 |
-
|
|
|
323 |
|
324 |
-
if
|
325 |
-
|
326 |
-
|
327 |
-
|
328 |
-
elif "text" in choice:
|
329 |
-
final_response = choice["text"].strip()
|
330 |
-
else:
|
331 |
-
final_response = "[Error: Unexpected response format]"
|
332 |
else:
|
333 |
-
|
334 |
|
335 |
-
except
|
336 |
-
|
337 |
except Exception as e:
|
338 |
-
final_response = f"[Unexpected Error: {e}]"
|
339 |
traceback.print_exc()
|
340 |
-
|
341 |
-
print(f" - LLM response: '{final_response[:100]}...'")
|
342 |
-
return final_response
|
343 |
|
344 |
-
# ---
|
345 |
-
def
|
346 |
-
|
347 |
-
|
348 |
-
|
349 |
-
|
350 |
-
|
351 |
-
|
352 |
-
|
353 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
354 |
):
|
355 |
-
|
356 |
-
|
357 |
-
|
358 |
-
|
359 |
-
|
360 |
-
|
361 |
-
|
362 |
-
|
363 |
-
|
364 |
-
|
365 |
-
|
366 |
-
|
367 |
-
# Handle Audio Input
|
368 |
-
if audio_input_path and whisper_model:
|
369 |
-
if os.path.isfile(audio_input_path):
|
370 |
-
audio_filepath_to_clean = audio_input_path
|
371 |
-
transcription_source = "voice"
|
372 |
-
print(f"Processing audio input: {audio_input_path}")
|
373 |
-
try:
|
374 |
-
stt_start_time = time.time()
|
375 |
-
result = whisper_model.transcribe(audio_input_path, fp16=(stt_device == 'cuda'))
|
376 |
-
original_user_input_text = result["text"].strip()
|
377 |
-
stt_end_time = time.time()
|
378 |
-
print(f" - Whisper transcription: '{original_user_input_text}' (took {stt_end_time - stt_start_time:.2f}s)")
|
379 |
-
user_display_input = f"π€ (Audio Input): {original_user_input_text}"
|
380 |
-
text_to_process = original_user_input_text
|
381 |
-
|
382 |
-
# Check if transcription is already a command
|
383 |
-
known_prefixes = ["@tara-tts", "@jess-tts", "@leo-tts", "@leah-tts", "@dan-tts", "@mia-tts", "@zac-tts", "@zoe-tts",
|
384 |
-
"@tara-llm", "@jess-llm", "@leo-llm", "@leah-llm", "@dan-llm", "@mia-llm", "@zac-llm", "@zoe-llm"]
|
385 |
-
is_already_command = any(original_user_input_text.lower().startswith(p) for p in known_prefixes)
|
386 |
-
|
387 |
-
if not is_already_command:
|
388 |
-
if plain_llm_checkbox:
|
389 |
-
prefix_to_add = None
|
390 |
-
force_plain_llm = True
|
391 |
-
print(f" - Plain LLM checked. Processing audio as text input for LLM.")
|
392 |
-
elif auto_prefix_tts_checkbox:
|
393 |
-
prefix_to_add = "@tara-tts"
|
394 |
-
print(f" - Auto-prefix TTS checked. Applying to audio.")
|
395 |
-
elif auto_prefix_llm_checkbox:
|
396 |
-
prefix_to_add = "@tara-llm"
|
397 |
-
print(f" - Auto-prefix LLM checked. Applying to audio.")
|
398 |
-
else:
|
399 |
-
print(f" - No default prefix checkbox checked for audio. Processing as text for LLM.")
|
400 |
-
|
401 |
-
if prefix_to_add:
|
402 |
-
text_to_process = f"{prefix_to_add} {original_user_input_text}"
|
403 |
-
else:
|
404 |
-
print(f" - Transcribed audio is already a command '{original_user_input_text[:20]}...'.")
|
405 |
-
text_to_process = original_user_input_text
|
406 |
-
|
407 |
-
except Exception as e:
|
408 |
-
print(f"Error during Whisper transcription: {e}")
|
409 |
-
traceback.print_exc()
|
410 |
-
error_msg = f"[Error during local transcription: {e}]"
|
411 |
-
chat_history.append((f"π€ (Audio Input Error: {audio_input_path})", error_msg))
|
412 |
-
if audio_filepath_to_clean and os.path.exists(audio_filepath_to_clean):
|
413 |
-
try:
|
414 |
-
os.remove(audio_filepath_to_clean)
|
415 |
-
except Exception as e_clean:
|
416 |
-
print(f"Warning: Could not clean up STT temp file {audio_filepath_to_clean}: {e_clean}")
|
417 |
-
return chat_history, None, None
|
418 |
else:
|
419 |
-
|
420 |
-
|
421 |
-
|
422 |
-
|
423 |
-
|
424 |
-
|
425 |
-
|
426 |
-
|
427 |
-
text_to_process = original_user_input_text
|
428 |
-
|
429 |
-
known_prefixes = ["@tara-tts", "@jess-tts", "@leo-tts", "@leah-tts", "@dan-tts", "@mia-tts", "@zac-tts", "@zoe-tts",
|
430 |
-
"@tara-llm", "@jess-llm", "@leo-llm", "@leah-llm", "@dan-llm", "@mia-llm", "@zac-llm", "@zoe-llm"]
|
431 |
-
is_already_command = any(original_user_input_text.lower().startswith(p) for p in known_prefixes)
|
432 |
-
|
433 |
-
if not is_already_command:
|
434 |
-
if plain_llm_checkbox:
|
435 |
-
prefix_to_add = None
|
436 |
-
force_plain_llm = True
|
437 |
-
print(f" - Plain LLM checked. Processing text input for LLM.")
|
438 |
-
elif auto_prefix_tts_checkbox:
|
439 |
-
prefix_to_add = "@tara-tts"
|
440 |
-
print(f" - Auto-prefix TTS checked. Applying to text.")
|
441 |
-
elif auto_prefix_llm_checkbox:
|
442 |
-
prefix_to_add = "@tara-llm"
|
443 |
-
print(f" - Auto-prefix LLM checked. Applying to text.")
|
444 |
-
else:
|
445 |
-
print(f" - No default prefix checkbox enabled for text input.")
|
446 |
else:
|
447 |
-
|
448 |
-
|
449 |
-
if prefix_to_add:
|
450 |
-
text_to_process = f"{prefix_to_add} {original_user_input_text}"
|
451 |
-
|
452 |
-
# Cleanup audio file
|
453 |
-
if audio_filepath_to_clean and os.path.exists(audio_filepath_to_clean):
|
454 |
-
try:
|
455 |
-
os.remove(audio_filepath_to_clean)
|
456 |
-
print(f" - Cleaned up temporary STT audio file: {audio_filepath_to_clean}")
|
457 |
-
except Exception as e_clean:
|
458 |
-
print(f"Warning: Could not clean up temp STT audio file {audio_filepath_to_clean}: {e_clean}")
|
459 |
-
|
460 |
-
if not text_to_process:
|
461 |
-
print("No valid text or audio input to process.")
|
462 |
-
return chat_history, None, None
|
463 |
-
|
464 |
-
chat_history.append((user_display_input, None))
|
465 |
-
|
466 |
-
# Process Input Text
|
467 |
-
lower_text = text_to_process.lower()
|
468 |
-
print(f" - Routing query ({transcription_source}): '{text_to_process[:100]}...'")
|
469 |
|
470 |
-
|
471 |
-
|
472 |
-
|
|
|
|
|
473 |
|
474 |
-
final_bot_message = None
|
475 |
-
|
476 |
try:
|
477 |
-
|
478 |
-
|
479 |
-
|
480 |
-
|
481 |
-
|
482 |
-
|
483 |
-
|
484 |
-
|
485 |
-
|
486 |
-
|
487 |
-
|
488 |
-
|
489 |
-
|
490 |
-
|
491 |
-
|
492 |
-
|
493 |
-
|
494 |
-
|
495 |
-
|
496 |
-
|
497 |
-
|
498 |
-
|
499 |
-
|
500 |
-
|
501 |
-
|
502 |
-
|
503 |
-
temp_dir = "temp_audio_files"
|
504 |
-
os.makedirs(temp_dir, exist_ok=True)
|
505 |
-
temp_audio_path = os.path.join(temp_dir, f"temp_audio_{uuid.uuid4().hex}.wav")
|
506 |
-
wavfile.write(temp_audio_path, sample_rate, audio_data)
|
507 |
-
print(f" - Saved TTS audio: {temp_audio_path}")
|
508 |
-
final_bot_message = (temp_audio_path, None)
|
509 |
-
else:
|
510 |
-
final_bot_message = f"Sorry, couldn't generate speech for '{text_to_speak[:50]}...'."
|
511 |
-
break
|
512 |
-
|
513 |
-
# Branch 2: LLM + TTS
|
514 |
-
if not matched_tts:
|
515 |
-
for tag, voice in llm_tags.items():
|
516 |
-
if lower_text.startswith(tag):
|
517 |
-
matched_llm_tts = True
|
518 |
-
prompt_for_llm = text_to_process[len(tag):].strip()
|
519 |
-
print(f" - LLM+TTS request for voice '{voice}': '{prompt_for_llm[:75]}...'")
|
520 |
-
if snac_model is None:
|
521 |
-
raise ValueError("SNAC vocoder not loaded.")
|
522 |
-
|
523 |
-
history_before_current = chat_history[:-1]
|
524 |
-
limited_history_turns = history_before_current[-CONTEXT_TURN_LIMIT:]
|
525 |
-
cleaned_hist_for_llm = clean_chat_history(limited_history_turns)
|
526 |
-
|
527 |
-
messages = [
|
528 |
-
{"role": "system", "content": OLLAMA_SYSTEM_PROMPT}
|
529 |
-
] + cleaned_hist_for_llm + [
|
530 |
-
{"role": "user", "content": prompt_for_llm}
|
531 |
-
]
|
532 |
-
|
533 |
-
llm_params = {
|
534 |
-
'ollama_temperature': ollama_temperature,
|
535 |
-
'ollama_top_p': ollama_top_p,
|
536 |
-
'ollama_top_k': ollama_top_k,
|
537 |
-
'ollama_max_new_tokens': ollama_max_new_tokens,
|
538 |
-
'ollama_repetition_penalty': ollama_repetition_penalty
|
539 |
-
}
|
540 |
-
|
541 |
-
llm_response_text = call_ollama_non_streaming(
|
542 |
-
{"messages": messages},
|
543 |
-
llm_params
|
544 |
-
)
|
545 |
-
|
546 |
-
if llm_response_text and not llm_response_text.startswith("[Error"):
|
547 |
-
audio_output = generate_speech_gguf(
|
548 |
-
llm_response_text, voice,
|
549 |
-
tts_temperature, tts_top_p, tts_repetition_penalty,
|
550 |
-
MAX_MAX_NEW_TOKENS
|
551 |
-
)
|
552 |
-
if audio_output:
|
553 |
-
sample_rate, audio_data = audio_output
|
554 |
-
if audio_data.dtype != np.int16:
|
555 |
-
if np.issubdtype(audio_data.dtype, np.floating):
|
556 |
-
max_val = np.max(np.abs(audio_data))
|
557 |
-
audio_data = np.int16(audio_data/max_val*32767) if max_val > 1e-6 else np.zeros_like(audio_data, dtype=np.int16)
|
558 |
-
else:
|
559 |
-
audio_data = audio_data.astype(np.int16)
|
560 |
-
temp_dir = "temp_audio_files"
|
561 |
-
os.makedirs(temp_dir, exist_ok=True)
|
562 |
-
temp_audio_path = os.path.join(temp_dir, f"temp_audio_{uuid.uuid4().hex}.wav")
|
563 |
-
wavfile.write(temp_audio_path, sample_rate, audio_data)
|
564 |
-
print(f" - Saved LLM+TTS audio: {temp_audio_path}")
|
565 |
-
final_bot_message = (temp_audio_path, llm_response_text)
|
566 |
-
else:
|
567 |
-
print("Warning: TTS generation failed...")
|
568 |
-
final_bot_message = f"{llm_response_text}\n\n(TTS failed...)"
|
569 |
-
else:
|
570 |
-
final_bot_message = llm_response_text
|
571 |
-
break
|
572 |
-
|
573 |
-
# Branch 3: Plain LLM
|
574 |
-
if force_plain_llm or (not matched_tts and not matched_llm_tts):
|
575 |
-
if force_plain_llm:
|
576 |
-
print(f" - Plain LLM chat mode forced by checkbox...")
|
577 |
-
else:
|
578 |
-
print(f" - Default text chat (no command prefix detected/added)...")
|
579 |
-
|
580 |
-
history_before_current = chat_history[:-1]
|
581 |
-
limited_history_turns = history_before_current[-CONTEXT_TURN_LIMIT:]
|
582 |
-
cleaned_hist_for_llm = clean_chat_history(limited_history_turns)
|
583 |
-
|
584 |
-
messages = [
|
585 |
-
{"role": "system", "content": OLLAMA_SYSTEM_PROMPT}
|
586 |
-
] + cleaned_hist_for_llm + [
|
587 |
-
{"role": "user", "content": original_user_input_text}
|
588 |
-
]
|
589 |
-
|
590 |
-
llm_params = {
|
591 |
-
'ollama_temperature': ollama_temperature,
|
592 |
-
'ollama_top_p': ollama_top_p,
|
593 |
-
'ollama_top_k': ollama_top_k,
|
594 |
-
'ollama_max_new_tokens': ollama_max_new_tokens,
|
595 |
-
'ollama_repetition_penalty': ollama_repetition_penalty
|
596 |
-
}
|
597 |
-
|
598 |
-
final_bot_message = call_ollama_non_streaming(
|
599 |
-
{"messages": messages},
|
600 |
-
llm_params
|
601 |
)
|
602 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
603 |
except Exception as e:
|
604 |
-
|
|
|
605 |
traceback.print_exc()
|
606 |
-
|
607 |
-
|
608 |
-
chat_history[-1] = (user_display_input, final_bot_message)
|
609 |
-
return chat_history, None, None
|
610 |
|
611 |
# --- Gradio Interface ---
|
612 |
-
def
|
613 |
-
|
614 |
-
|
615 |
-
|
616 |
-
return gr.update(value=False), gr.update(value=True), gr.update(value=False)
|
617 |
-
elif selected_checkbox_label == "plain":
|
618 |
-
return gr.update(value=False), gr.update(value=False), gr.update(value=True)
|
619 |
-
else:
|
620 |
-
return gr.update(), gr.update(), gr.update()
|
621 |
-
|
622 |
-
print("Setting up Gradio Interface with gr.Blocks...")
|
623 |
-
theme_to_use = None
|
624 |
-
|
625 |
-
with gr.Blocks(theme=theme_to_use) as demo:
|
626 |
-
gr.Markdown(f"# Orpheus Edge π€ ({OLLAMA_MODEL}) Chat & TTS")
|
627 |
-
|
628 |
-
chatbot = gr.Chatbot(label="Chat History", height=500)
|
629 |
-
|
630 |
-
with gr.Row():
|
631 |
-
with gr.Column(scale=3):
|
632 |
-
text_input_box = gr.Textbox(label="Type your message or use microphone", lines=2)
|
633 |
-
with gr.Column(scale=1):
|
634 |
-
audio_input_mic = gr.Audio(label="Record Audio Input", type="filepath")
|
635 |
-
|
636 |
-
with gr.Row():
|
637 |
-
auto_prefix_tts_checkbox = gr.Checkbox(label="Default to TTS (@tara-tts)", value=True, elem_id="cb_tts")
|
638 |
-
auto_prefix_llm_checkbox = gr.Checkbox(label="Default to LLM+TTS (@tara-llm)", value=False, elem_id="cb_llm")
|
639 |
-
plain_llm_checkbox = gr.Checkbox(label="Plain LLM Chat (Text Out)", value=False, elem_id="cb_plain")
|
640 |
-
|
641 |
-
with gr.Row():
|
642 |
-
submit_button = gr.Button("Send / Submit")
|
643 |
-
clear_button = gr.ClearButton([text_input_box, audio_input_mic, chatbot])
|
644 |
-
|
645 |
-
with gr.Accordion("Generation Parameters", open=False):
|
646 |
-
gr.Markdown("### LLM Parameters")
|
647 |
-
ollama_max_new_tokens_slider = gr.Slider(label="Max New Tokens", minimum=32, maximum=4096, step=32, value=DEFAULT_OLLAMA_MAX_TOKENS)
|
648 |
-
ollama_temperature_slider = gr.Slider(label="Temperature", minimum=0.0, maximum=2.0, step=0.05, value=DEFAULT_OLLAMA_TEMP)
|
649 |
-
ollama_top_p_slider = gr.Slider(label="Top-p", minimum=0.05, maximum=1.0, step=0.05, value=DEFAULT_OLLAMA_TOP_P)
|
650 |
-
ollama_top_k_slider = gr.Slider(label="Top-k", minimum=1, maximum=100, step=1, value=DEFAULT_OLLAMA_TOP_K)
|
651 |
-
ollama_repetition_penalty_slider = gr.Slider(label="Repetition Penalty", minimum=1.0, maximum=2.0, step=0.05, value=DEFAULT_OLLAMA_REP_PENALTY)
|
652 |
|
653 |
-
|
654 |
-
|
655 |
-
|
656 |
-
|
657 |
-
|
658 |
-
|
659 |
-
|
660 |
-
|
661 |
-
|
662 |
-
|
663 |
-
|
664 |
-
|
665 |
-
|
666 |
-
|
667 |
-
|
668 |
-
|
669 |
-
|
670 |
-
|
671 |
-
|
672 |
-
|
673 |
-
|
674 |
-
|
675 |
-
|
676 |
-
|
677 |
-
|
678 |
-
|
679 |
-
|
680 |
-
|
681 |
-
|
682 |
-
|
683 |
-
|
684 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
685 |
|
686 |
-
|
687 |
-
fn=process_input_blocks,
|
688 |
-
inputs=all_inputs,
|
689 |
-
outputs=[chatbot, text_input_box, audio_input_mic]
|
690 |
-
)
|
691 |
-
text_input_box.submit(
|
692 |
-
fn=process_input_blocks,
|
693 |
-
inputs=all_inputs,
|
694 |
-
outputs=[chatbot, text_input_box, audio_input_mic]
|
695 |
-
)
|
696 |
|
697 |
-
# --- Application
|
698 |
if __name__ == "__main__":
|
699 |
-
print("-" *
|
700 |
-
print(
|
701 |
-
print(f"Whisper STT
|
702 |
-
print(f"
|
703 |
-
print(f"
|
704 |
-
print(f"
|
705 |
-
print(
|
706 |
-
|
707 |
-
|
708 |
-
|
709 |
-
|
710 |
-
|
711 |
-
|
712 |
-
|
713 |
-
demo.launch(share=False)
|
714 |
-
print("Gradio Interface launched. Press Ctrl+C to stop.")
|
|
|
6 |
import json
|
7 |
import time
|
8 |
import traceback
|
|
|
9 |
import gradio as gr
|
10 |
import torch
|
11 |
import numpy as np
|
12 |
import scipy.io.wavfile as wavfile
|
13 |
+
import openai
|
14 |
from dotenv import load_dotenv
|
15 |
+
from huggingface_hub import snapshot_download
|
16 |
+
from transformers import AutoModelForCausalLM, AutoTokenizer
|
17 |
|
18 |
# --- Whisper Import ---
|
19 |
try:
|
|
|
28 |
# --- SNAC Import ---
|
29 |
try:
|
30 |
from snac import SNAC
|
31 |
+
print("SNAC library imported successfully.")
|
32 |
except ImportError:
|
33 |
print("ERROR: SNAC library not found. Please install it:")
|
34 |
print("pip install git+https://github.com/hubertsiuzdak/snac.git")
|
|
|
37 |
# --- Load Environment Variables ---
|
38 |
load_dotenv()
|
39 |
|
40 |
+
# --- OpenAI Configuration ---
|
41 |
+
OPENAI_API_KEY = os.getenv("OPENAI_API_KEY","sk-proj-B97S8pFXA6YhSSIFileVT3BlbkFJWQvPI0PON1KZReYYRGge")
|
42 |
+
OPENAI_API_BASE = os.getenv("OPENAI_API_BASE", "https://api.openai.com/v1")
|
43 |
+
OPENAI_MODEL = os.getenv("OPENAI_MODEL", "gpt-4o")
|
44 |
|
45 |
+
# Initialize OpenAI Client
|
46 |
+
if OPENAI_API_KEY:
|
47 |
+
print("OpenAI API key found.")
|
48 |
+
else:
|
49 |
+
print("ERROR: OPENAI_API_KEY environment variable not set.")
|
50 |
+
exit(1)
|
51 |
|
52 |
+
SYSTEM_PROMPT = "You are a helpful AI assistant. Respond in a conversational and friendly manner."
|
|
|
|
|
53 |
|
54 |
# --- Device Setup ---
|
55 |
+
device = "cuda" if torch.cuda.is_available() else "cpu"
|
56 |
+
stt_device = device
|
57 |
+
print(f"Using device: {device}")
|
|
|
|
|
|
|
|
|
|
|
58 |
|
59 |
# --- Model Loading ---
|
60 |
print("Loading SNAC vocoder model...")
|
|
|
61 |
try:
|
62 |
snac_model = SNAC.from_pretrained("hubertsiuzdak/snac_24khz")
|
63 |
+
snac_model = snac_model.to(device)
|
64 |
snac_model.eval()
|
65 |
+
print(f"SNAC vocoder loaded to {device}")
|
66 |
except Exception as e:
|
67 |
print(f"Error loading SNAC model: {e}")
|
68 |
+
exit(1)
|
69 |
+
|
70 |
+
print("Loading TTS model...")
|
71 |
+
tts_model_name = "mrrtmob/tts-khm-3"
|
72 |
+
try:
|
73 |
+
snapshot_download(
|
74 |
+
repo_id=tts_model_name,
|
75 |
+
allow_patterns=["config.json", "*.safetensors", "model.safetensors.index.json"],
|
76 |
+
ignore_patterns=["optimizer.pt", "pytorch_model.bin", "training_args.bin", "scheduler.pt", "tokenizer.json", "tokenizer_config.json", "special_tokens_map.json", "vocab.json", "merges.txt", "tokenizer.*"]
|
77 |
+
)
|
78 |
+
|
79 |
+
tts_model = AutoModelForCausalLM.from_pretrained(tts_model_name, torch_dtype=torch.bfloat16)
|
80 |
+
tts_model.to(device)
|
81 |
+
tts_tokenizer = AutoTokenizer.from_pretrained(tts_model_name)
|
82 |
+
print(f"TTS model '{tts_model_name}' loaded to {device}")
|
83 |
+
except Exception as e:
|
84 |
+
print(f"Error loading TTS model: {e}")
|
85 |
+
exit(1)
|
86 |
|
87 |
print("Loading Whisper STT model...")
|
88 |
WHISPER_MODEL_NAME = "base.en"
|
|
|
89 |
try:
|
90 |
whisper_model = whisper.load_model(WHISPER_MODEL_NAME, device=stt_device)
|
91 |
print(f"Whisper model '{WHISPER_MODEL_NAME}' loaded successfully.")
|
92 |
except Exception as e:
|
93 |
print(f"Error loading Whisper model: {e}")
|
94 |
+
exit(1)
|
95 |
|
96 |
# --- Constants ---
|
97 |
MAX_MAX_NEW_TOKENS = 4096
|
98 |
+
DEFAULT_OPENAI_MAX_TOKENS = 512
|
99 |
MAX_SEED = np.iinfo(np.int32).max
|
100 |
+
DEFAULT_OPENAI_TEMP = 0.7
|
101 |
+
DEFAULT_OPENAI_TOP_P = 0.9
|
102 |
+
DEFAULT_TTS_TEMP = 0.6
|
103 |
+
DEFAULT_TTS_TOP_P = 0.95
|
104 |
+
CONTEXT_TURN_LIMIT = 5
|
105 |
+
|
106 |
+
# --- TTS Functions ---
|
107 |
+
def process_prompt(prompt, voice, tokenizer, device):
|
108 |
+
"""Process text prompt for TTS model"""
|
109 |
+
prompt = f"{voice}: {prompt}"
|
110 |
+
input_ids = tokenizer(prompt, return_tensors="pt").input_ids
|
111 |
+
|
112 |
+
start_token = torch.tensor([[128259]], dtype=torch.int64)
|
113 |
+
end_tokens = torch.tensor([[128009, 128260]], dtype=torch.int64)
|
114 |
+
|
115 |
+
modified_input_ids = torch.cat([start_token, input_ids, end_tokens], dim=1)
|
116 |
+
attention_mask = torch.ones_like(modified_input_ids)
|
117 |
+
|
118 |
+
return modified_input_ids.to(device), attention_mask.to(device)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
119 |
|
120 |
+
def parse_output(generated_ids):
|
121 |
+
"""Parse TTS model output tokens"""
|
122 |
+
token_to_find = 128257
|
123 |
+
token_to_remove = 128258
|
124 |
+
|
125 |
+
token_indices = (generated_ids == token_to_find).nonzero(as_tuple=True)
|
126 |
+
if len(token_indices[1]) > 0:
|
127 |
+
last_occurrence_idx = token_indices[1][-1].item()
|
128 |
+
cropped_tensor = generated_ids[:, last_occurrence_idx+1:]
|
129 |
+
else:
|
130 |
+
cropped_tensor = generated_ids
|
131 |
+
|
132 |
+
processed_rows = []
|
133 |
+
for row in cropped_tensor:
|
134 |
+
masked_row = row[row != token_to_remove]
|
135 |
+
processed_rows.append(masked_row)
|
136 |
+
|
137 |
+
code_lists = []
|
138 |
+
for row in processed_rows:
|
139 |
+
row_length = row.size(0)
|
140 |
+
new_length = (row_length // 7) * 7
|
141 |
+
trimmed_row = row[:new_length]
|
142 |
+
trimmed_row = [t - 128266 for t in trimmed_row]
|
143 |
+
code_lists.append(trimmed_row)
|
144 |
+
|
145 |
+
return code_lists[0] if code_lists else []
|
146 |
|
147 |
+
def redistribute_codes(code_list, snac_model):
|
148 |
+
"""Convert codes to audio using SNAC"""
|
149 |
+
if not code_list:
|
150 |
+
return np.array([])
|
151 |
+
|
152 |
+
snac_device = next(snac_model.parameters()).device
|
153 |
|
|
|
154 |
layer_1, layer_2, layer_3 = [], [], []
|
|
|
|
|
155 |
|
156 |
+
for i in range(len(code_list) // 7):
|
157 |
+
base = 7 * i
|
158 |
+
if base + 6 < len(code_list):
|
159 |
+
layer_1.append(code_list[base])
|
160 |
+
layer_2.append(code_list[base+1] - 4096)
|
161 |
+
layer_3.append(code_list[base+2] - (2*4096))
|
162 |
+
layer_3.append(code_list[base+3] - (3*4096))
|
163 |
+
layer_2.append(code_list[base+4] - (4*4096))
|
164 |
+
layer_3.append(code_list[base+5] - (5*4096))
|
165 |
+
layer_3.append(code_list[base+6] - (6*4096))
|
166 |
|
167 |
+
if not layer_1:
|
168 |
+
return np.array([])
|
169 |
|
170 |
+
codes = [
|
171 |
+
torch.tensor(layer_1, device=snac_device).unsqueeze(0),
|
172 |
+
torch.tensor(layer_2, device=snac_device).unsqueeze(0),
|
173 |
+
torch.tensor(layer_3, device=snac_device).unsqueeze(0)
|
174 |
+
]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
175 |
|
176 |
+
with torch.no_grad():
|
177 |
+
audio_hat = snac_model.decode(codes)
|
178 |
+
|
179 |
+
return audio_hat.detach().squeeze().cpu().numpy()
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
180 |
|
181 |
+
def generate_speech(text, voice="tara", temperature=0.6, top_p=0.95, max_new_tokens=1200):
|
182 |
+
"""Generate speech from text"""
|
183 |
+
if not text.strip():
|
184 |
return None
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
185 |
|
186 |
try:
|
187 |
+
print(f"Generating speech for: '{text[:50]}...'")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
188 |
|
189 |
+
input_ids, attention_mask = process_prompt(text, voice, tts_tokenizer, device)
|
190 |
|
191 |
+
with torch.no_grad():
|
192 |
+
generated_ids = tts_model.generate(
|
193 |
+
input_ids=input_ids,
|
194 |
+
attention_mask=attention_mask,
|
195 |
+
max_new_tokens=max_new_tokens,
|
196 |
+
do_sample=True,
|
197 |
+
temperature=temperature,
|
198 |
+
top_p=top_p,
|
199 |
+
repetition_penalty=1.1,
|
200 |
+
num_return_sequences=1,
|
201 |
+
eos_token_id=128258,
|
202 |
+
)
|
203 |
+
|
204 |
+
code_list = parse_output(generated_ids)
|
205 |
+
audio_samples = redistribute_codes(code_list, snac_model)
|
206 |
+
|
207 |
+
if len(audio_samples) > 0:
|
208 |
+
print(f"Generated audio with shape: {audio_samples.shape}")
|
|
|
|
|
|
|
|
|
209 |
return (24000, audio_samples)
|
|
|
210 |
else:
|
211 |
+
print("No audio generated")
|
212 |
return None
|
213 |
|
|
|
|
|
|
|
214 |
except Exception as e:
|
215 |
+
print(f"Error during speech generation: {e}")
|
216 |
traceback.print_exc()
|
217 |
return None
|
218 |
|
219 |
+
# --- STT Function ---
|
220 |
+
def transcribe_audio(audio_path):
|
221 |
+
"""Transcribe audio to text using Whisper"""
|
222 |
+
if not audio_path:
|
223 |
+
return ""
|
224 |
+
|
225 |
try:
|
226 |
+
print(f"Transcribing audio: {audio_path}")
|
227 |
+
result = whisper_model.transcribe(audio_path, fp16=(stt_device == 'cuda'))
|
228 |
+
transcribed_text = result["text"].strip()
|
229 |
+
print(f"Transcribed: '{transcribed_text}'")
|
230 |
+
return transcribed_text
|
231 |
+
except Exception as e:
|
232 |
+
print(f"Error during transcription: {e}")
|
233 |
+
return f"[Transcription Error: {e}]"
|
234 |
+
|
235 |
+
# --- OpenAI LLM Function ---
|
236 |
+
def call_openai_llm(messages, temperature=0.7, top_p=0.9, max_tokens=512):
|
237 |
+
"""Call OpenAI API for text generation"""
|
238 |
+
try:
|
239 |
+
client = openai.OpenAI(api_key=OPENAI_API_KEY, base_url=OPENAI_API_BASE)
|
240 |
|
241 |
+
print(f"Sending to OpenAI with model {OPENAI_MODEL}")
|
242 |
start_time = time.time()
|
243 |
+
|
244 |
+
response = client.chat.completions.create(
|
245 |
+
model=OPENAI_MODEL,
|
246 |
+
messages=messages,
|
247 |
+
temperature=temperature,
|
248 |
+
top_p=top_p,
|
249 |
+
max_tokens=max_tokens,
|
250 |
+
stream=False
|
251 |
)
|
|
|
|
|
|
|
252 |
|
253 |
+
end_time = time.time()
|
254 |
+
print(f"LLM request took {end_time - start_time:.2f}s")
|
255 |
|
256 |
+
if response.choices:
|
257 |
+
response_text = response.choices[0].message.content.strip()
|
258 |
+
print(f"LLM response: '{response_text[:100]}...'")
|
259 |
+
return response_text
|
|
|
|
|
|
|
|
|
260 |
else:
|
261 |
+
return "[Error: No response from model]"
|
262 |
|
263 |
+
except openai.APIError as e:
|
264 |
+
return f"[OpenAI API Error: {e}]"
|
265 |
except Exception as e:
|
|
|
266 |
traceback.print_exc()
|
267 |
+
return f"[Unexpected Error: {e}]"
|
|
|
|
|
268 |
|
269 |
+
# --- Utility Functions ---
|
270 |
+
def clean_chat_history(chat_history, limit=CONTEXT_TURN_LIMIT):
|
271 |
+
"""Clean and format chat history for OpenAI API"""
|
272 |
+
if not chat_history:
|
273 |
+
return []
|
274 |
+
|
275 |
+
messages = []
|
276 |
+
recent_history = chat_history[-limit:] if len(chat_history) > limit else chat_history
|
277 |
+
|
278 |
+
for user_msg, bot_msg in recent_history:
|
279 |
+
if user_msg and isinstance(user_msg, str):
|
280 |
+
# Handle audio input display format
|
281 |
+
if user_msg.startswith("π€ Audio: "):
|
282 |
+
user_text = user_msg.replace("π€ Audio: ", "")
|
283 |
+
else:
|
284 |
+
user_text = user_msg
|
285 |
+
|
286 |
+
if user_text.strip():
|
287 |
+
messages.append({"role": "user", "content": user_text.strip()})
|
288 |
+
|
289 |
+
if bot_msg and isinstance(bot_msg, str) and not bot_msg.startswith("[Error"):
|
290 |
+
messages.append({"role": "assistant", "content": bot_msg.strip()})
|
291 |
+
|
292 |
+
return messages
|
293 |
+
|
294 |
+
# --- Main Processing Function ---
|
295 |
+
def process_conversation(
|
296 |
+
text_input, audio_input,
|
297 |
+
mode, enable_tts,
|
298 |
+
openai_temp, openai_top_p, openai_max_tokens,
|
299 |
+
tts_temp, tts_top_p,
|
300 |
+
chat_history
|
301 |
):
|
302 |
+
"""Main function to process user input and generate responses"""
|
303 |
+
|
304 |
+
user_text = ""
|
305 |
+
display_input = ""
|
306 |
+
|
307 |
+
# Handle audio input
|
308 |
+
if audio_input and mode in ["audio_only", "audio_text"]:
|
309 |
+
transcribed = transcribe_audio(audio_input)
|
310 |
+
if transcribed and not transcribed.startswith("["):
|
311 |
+
user_text = transcribed
|
312 |
+
display_input = f"π€ Audio: {transcribed}"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
313 |
else:
|
314 |
+
chat_history.append((f"π€ Audio Input", transcribed))
|
315 |
+
return chat_history, "", None
|
316 |
+
|
317 |
+
# Handle text input
|
318 |
+
if text_input and mode in ["text_only", "audio_text"]:
|
319 |
+
if user_text: # If we already have audio transcription
|
320 |
+
user_text += f" {text_input}"
|
321 |
+
display_input = f"π€ Audio: {transcribed} + Text: {text_input}"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
322 |
else:
|
323 |
+
user_text = text_input
|
324 |
+
display_input = text_input
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
325 |
|
326 |
+
if not user_text.strip():
|
327 |
+
return chat_history, "", None
|
328 |
+
|
329 |
+
# Add user message to chat
|
330 |
+
chat_history.append((display_input, "Thinking..."))
|
331 |
|
|
|
|
|
332 |
try:
|
333 |
+
# Prepare messages for OpenAI
|
334 |
+
cleaned_history = clean_chat_history(chat_history[:-1]) # Exclude current "Thinking..." entry
|
335 |
+
messages = [{"role": "system", "content": SYSTEM_PROMPT}] + cleaned_history
|
336 |
+
messages.append({"role": "user", "content": user_text})
|
337 |
+
|
338 |
+
# Get LLM response
|
339 |
+
llm_response = call_openai_llm(
|
340 |
+
messages=messages,
|
341 |
+
temperature=openai_temp,
|
342 |
+
top_p=openai_top_p,
|
343 |
+
max_tokens=openai_max_tokens
|
344 |
+
)
|
345 |
+
|
346 |
+
if llm_response.startswith("[Error"):
|
347 |
+
chat_history[-1] = (display_input, llm_response)
|
348 |
+
return chat_history, "", None
|
349 |
+
|
350 |
+
# Generate TTS if enabled
|
351 |
+
bot_response = llm_response
|
352 |
+
audio_output = None
|
353 |
+
|
354 |
+
if enable_tts and llm_response:
|
355 |
+
audio_result = generate_speech(
|
356 |
+
text=llm_response,
|
357 |
+
temperature=tts_temp,
|
358 |
+
top_p=tts_top_p
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
359 |
)
|
360 |
+
|
361 |
+
if audio_result:
|
362 |
+
audio_output = audio_result
|
363 |
+
# For display purposes, we'll show text + indicate audio is available
|
364 |
+
bot_response = llm_response
|
365 |
+
|
366 |
+
# Update chat history
|
367 |
+
chat_history[-1] = (display_input, bot_response)
|
368 |
+
|
369 |
+
return chat_history, "", audio_output
|
370 |
+
|
371 |
except Exception as e:
|
372 |
+
error_msg = f"[Processing Error: {e}]"
|
373 |
+
chat_history[-1] = (display_input, error_msg)
|
374 |
traceback.print_exc()
|
375 |
+
return chat_history, "", None
|
|
|
|
|
|
|
376 |
|
377 |
# --- Gradio Interface ---
|
378 |
+
def create_interface():
|
379 |
+
with gr.Blocks(title="STT + OpenAI + TTS Demo") as demo:
|
380 |
+
gr.Markdown("""
|
381 |
+
# π€ Complete AI Assistant: Speech-to-Text + OpenAI + Text-to-Speech
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
382 |
|
383 |
+
This demo combines:
|
384 |
+
- **Whisper STT**: Convert your speech to text
|
385 |
+
- **OpenAI LLM**: Generate intelligent responses
|
386 |
+
- **Local TTS**: Convert responses back to speech
|
387 |
+
""")
|
388 |
+
|
389 |
+
with gr.Row():
|
390 |
+
chatbot = gr.Chatbot(
|
391 |
+
label="Conversation",
|
392 |
+
height=400,
|
393 |
+
show_copy_button=True
|
394 |
+
)
|
395 |
+
|
396 |
+
with gr.Row():
|
397 |
+
with gr.Column(scale=3):
|
398 |
+
text_input = gr.Textbox(
|
399 |
+
label="Type your message",
|
400 |
+
placeholder="Enter your message here...",
|
401 |
+
lines=2
|
402 |
+
)
|
403 |
+
with gr.Column(scale=2):
|
404 |
+
audio_input = gr.Audio(
|
405 |
+
label="Record Audio",
|
406 |
+
type="filepath"
|
407 |
+
)
|
408 |
+
|
409 |
+
with gr.Row():
|
410 |
+
mode = gr.Radio(
|
411 |
+
choices=["text_only", "audio_only", "audio_text"],
|
412 |
+
value="text_only",
|
413 |
+
label="Input Mode",
|
414 |
+
info="How do you want to provide input?"
|
415 |
+
)
|
416 |
+
enable_tts = gr.Checkbox(
|
417 |
+
label="Enable Text-to-Speech Output",
|
418 |
+
value=True,
|
419 |
+
info="Convert responses to speech"
|
420 |
+
)
|
421 |
+
|
422 |
+
with gr.Row():
|
423 |
+
submit_btn = gr.Button("Send", variant="primary", size="lg")
|
424 |
+
clear_btn = gr.Button("Clear Conversation", size="lg")
|
425 |
+
|
426 |
+
audio_output = gr.Audio(
|
427 |
+
label="AI Response (Audio)",
|
428 |
+
type="numpy",
|
429 |
+
autoplay=False
|
430 |
+
)
|
431 |
+
|
432 |
+
with gr.Accordion("βοΈ Advanced Settings", open=False):
|
433 |
+
gr.Markdown("### OpenAI Settings")
|
434 |
+
with gr.Row():
|
435 |
+
openai_temp = gr.Slider(
|
436 |
+
minimum=0.1, maximum=2.0, value=DEFAULT_OPENAI_TEMP, step=0.1,
|
437 |
+
label="Temperature",
|
438 |
+
info="Higher = more creative"
|
439 |
+
)
|
440 |
+
openai_top_p = gr.Slider(
|
441 |
+
minimum=0.1, maximum=1.0, value=DEFAULT_OPENAI_TOP_P, step=0.05,
|
442 |
+
label="Top-p",
|
443 |
+
info="Nucleus sampling"
|
444 |
+
)
|
445 |
+
openai_max_tokens = gr.Slider(
|
446 |
+
minimum=50, maximum=2048, value=DEFAULT_OPENAI_MAX_TOKENS, step=50,
|
447 |
+
label="Max Tokens",
|
448 |
+
info="Maximum response length"
|
449 |
+
)
|
450 |
+
|
451 |
+
gr.Markdown("### TTS Settings")
|
452 |
+
with gr.Row():
|
453 |
+
tts_temp = gr.Slider(
|
454 |
+
minimum=0.1, maximum=1.5, value=DEFAULT_TTS_TEMP, step=0.05,
|
455 |
+
label="TTS Temperature",
|
456 |
+
info="Speech expressiveness"
|
457 |
+
)
|
458 |
+
tts_top_p = gr.Slider(
|
459 |
+
minimum=0.1, maximum=1.0, value=DEFAULT_TTS_TOP_P, step=0.05,
|
460 |
+
label="TTS Top-p",
|
461 |
+
info="Speech variation"
|
462 |
+
)
|
463 |
+
|
464 |
+
# Event handlers
|
465 |
+
inputs = [
|
466 |
+
text_input, audio_input,
|
467 |
+
mode, enable_tts,
|
468 |
+
openai_temp, openai_top_p, openai_max_tokens,
|
469 |
+
tts_temp, tts_top_p,
|
470 |
+
chatbot
|
471 |
+
]
|
472 |
+
|
473 |
+
outputs = [chatbot, text_input, audio_output]
|
474 |
+
|
475 |
+
submit_btn.click(
|
476 |
+
fn=process_conversation,
|
477 |
+
inputs=inputs,
|
478 |
+
outputs=outputs
|
479 |
+
)
|
480 |
+
|
481 |
+
text_input.submit(
|
482 |
+
fn=process_conversation,
|
483 |
+
inputs=inputs,
|
484 |
+
outputs=outputs
|
485 |
+
)
|
486 |
+
|
487 |
+
clear_btn.click(
|
488 |
+
fn=lambda: ([], "", None),
|
489 |
+
outputs=[chatbot, text_input, audio_output]
|
490 |
+
)
|
491 |
+
|
492 |
+
# Examples
|
493 |
+
gr.Examples(
|
494 |
+
examples=[
|
495 |
+
["Hello, how are you today?"],
|
496 |
+
["Can you explain quantum computing in simple terms?"],
|
497 |
+
["Tell me a short joke"],
|
498 |
+
["What's the weather like?"]
|
499 |
+
],
|
500 |
+
inputs=text_input,
|
501 |
+
)
|
502 |
|
503 |
+
return demo
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
504 |
|
505 |
+
# --- Launch Application ---
|
506 |
if __name__ == "__main__":
|
507 |
+
print("-" * 60)
|
508 |
+
print("π Initializing Complete AI Assistant")
|
509 |
+
print(f"π± Whisper STT: {WHISPER_MODEL_NAME} on {stt_device}")
|
510 |
+
print(f"π€ OpenAI Model: {OPENAI_MODEL}")
|
511 |
+
print(f"π TTS Model: {tts_model_name}")
|
512 |
+
print(f"πΎ SNAC Vocoder on {device}")
|
513 |
+
print("-" * 60)
|
514 |
+
|
515 |
+
demo = create_interface()
|
516 |
+
demo.launch(
|
517 |
+
share=False,
|
518 |
+
server_name="0.0.0.0",
|
519 |
+
server_port=7860
|
520 |
+
)
|
|
|
|
requirements.txt
CHANGED
@@ -6,4 +6,6 @@ spaces
|
|
6 |
openai-whisper
|
7 |
requests
|
8 |
gradio
|
9 |
-
scipy
|
|
|
|
|
|
6 |
openai-whisper
|
7 |
requests
|
8 |
gradio
|
9 |
+
scipy
|
10 |
+
openai
|
11 |
+
huggingface-hub
|