Update utils.py

utils.py

@@ -1,18 +1,65 @@
-runtime error
-Exit code: 1. Reason: Traceback (most recent call last):
-  File "/home/user/app/app.py", line 2, in <module>
-    from utils import generate_script, generate_audio, truncate_text
-  File "/home/user/app/utils.py", line 17, in <module>
-    tts_male = pipeline("text-to-speech", model="microsoft/speecht5_tts", device="cpu")
-  File "/usr/local/lib/python3.10/site-packages/transformers/pipelines/__init__.py", line 999, in pipeline
-    tokenizer = AutoTokenizer.from_pretrained(
-  File "/usr/local/lib/python3.10/site-packages/transformers/models/auto/tokenization_auto.py", line 907, in from_pretrained
-    return tokenizer_class.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)
-  File "/usr/local/lib/python3.10/site-packages/transformers/utils/import_utils.py", line 1637, in __getattribute__
-    requires_backends(cls, cls._backends)
-  File "/usr/local/lib/python3.10/site-packages/transformers/utils/import_utils.py", line 1625, in requires_backends
-    raise ImportError("".join(failed))
-ImportError: 
-SpeechT5Tokenizer requires the SentencePiece library but it was not found in your environment. Checkout the instructions on the
-installation page of its repo: https://github.com/google/sentencepiece#installation and follow the ones
-that match your environment. Please note that you may need to restart your runtime after installation.
+from groq import Groq
+from pydantic import BaseModel, ValidationError
+from typing import List, Literal
+import os
+import tiktoken
+import json
+import re
+from gtts import gTTS
+import tempfile
+
+groq_client = Groq(api_key=os.environ["GROQ_API_KEY"])
+tokenizer = tiktoken.get_encoding("cl100k_base")
+
+class DialogueItem(BaseModel):
+    speaker: Literal["John", "Sarah"]
+    text: str
+
+class Dialogue(BaseModel):
+    dialogue: List[DialogueItem]
+
+def truncate_text(text, max_tokens=2048):
+    tokens = tokenizer.encode(text)
+    if len(tokens) > max_tokens:
+        return tokenizer.decode(tokens[:max_tokens])
+    return text
+
+def generate_script(system_prompt: str, input_text: str, tone: str):
+    input_text = truncate_text(input_text)
+    prompt = f"{system_prompt}\nTONE: {tone}\nINPUT TEXT: {input_text}"
+    
+    response = groq_client.chat.completions.create(
+        messages=[
+            {"role": "system", "content": prompt},
+        ],
+        model="llama-3.1-70b-versatile",
+        max_tokens=2048,
+        temperature=0.7
+    )
+    
+    content = response.choices[0].message.content
+    content = re.sub(r'```json\s*|\s*```', '', content)
+    
+    try:
+        json_data = json.loads(content)
+        dialogue = Dialogue.model_validate(json_data)
+    except json.JSONDecodeError as json_error:
+        match = re.search(r'\{.*\}', content, re.DOTALL)
+        if match:
+            try:
+                json_data = json.loads(match.group())
+                dialogue = Dialogue.model_validate(json_data)
+            except (json.JSONDecodeError, ValidationError) as e:
+                raise ValueError(f"Failed to parse dialogue JSON: {e}\nContent: {content}")
+        else:
+            raise ValueError(f"Failed to find valid JSON in the response: {content}")
+    except ValidationError as e:
+        raise ValueError(f"Failed to validate dialogue structure: {e}\nContent: {content}")
+    
+    return dialogue
+
+def generate_audio(text: str, speaker: str) -> str:
+    tts = gTTS(text=text, lang='en', tld='com' if speaker == "John" else 'co.uk')
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as temp_audio:
+        tts.save(temp_audio.name)
+        return temp_audio.name