webscout 8.3__py3-none-any.whl → 8.3.2__py3-none-any.whl

webscout/Provider/TTS/speechma.py

@@ -1,15 +1,12 @@
 ##################################################################################
 ##  Modified version of code written by t.me/infip1217                          ##
 ##################################################################################
-import time
 import requests
 import pathlib
 import tempfile
-from io import BytesIO
 from webscout import exceptions
 from webscout.litagent import LitAgent
-from concurrent.futures import ThreadPoolExecutor, as_completed
-from webscout.Provider.TTS import utils
+from webscout.Litlogger import Logger, LogLevel
 from webscout.Provider.TTS.base import BaseTTSProvider
 
 class SpeechMaTTS(BaseTTSProvider):
@@ -18,12 +15,11 @@ class SpeechMaTTS(BaseTTSProvider):
     """
     # Request headers
     headers = {
-        "accept": "*/*",
-        "accept-language": "en-IN,en-GB;q=0.9,en-US;q=0.8,en;q=0.7,en-AU;q=0.6",
-        "content-type": "application/json",
+        "authority": "speechma.com",
         "origin": "https://speechma.com",
-        "priority": "u=1, i",
-        "User-Agent": LitAgent().random()
+        "referer": "https://speechma.com/",
+        "content-type": "application/json",
+        **LitAgent().generate_fingerprint()
     }
 
     # Available voices with their IDs
@@ -496,8 +492,9 @@ class SpeechMaTTS(BaseTTSProvider):
         if proxies:
             self.session.proxies.update(proxies)
         self.timeout = timeout
+        self.logger = Logger(name="SpeechMaTTS", level=LogLevel.INFO)
 
-    def tts(self, text: str, voice: str = "Emma", pitch: int = 0, rate: int = 0) -> str:
+    def tts(self, text: str, voice: str = "Emma", pitch: int = 0, rate: int = 0, verbose: bool = False) -> str:
         """
         Converts text to speech using the SpeechMa API and saves it to a file.
 
@@ -506,6 +503,7 @@ class SpeechMaTTS(BaseTTSProvider):
             voice (str): The voice to use for TTS (default: "Emma")
             pitch (int): Voice pitch adjustment (-10 to 10, default: 0)
             rate (int): Voice rate/speed adjustment (-10 to 10, default: 0)
+            verbose (bool): Whether to print debug information (default: False)
 
         Returns:
             str: Path to the generated audio file
@@ -517,10 +515,13 @@ class SpeechMaTTS(BaseTTSProvider):
             voice in self.all_voices
         ), f"Voice '{voice}' not one of [{', '.join(self.all_voices.keys())}]"
 
+        if not text or text.strip() == '':
+            raise exceptions.FailedToGenerateResponseError("Text is empty")
+
         filename = pathlib.Path(tempfile.mktemp(suffix=".mp3", dir=self.temp_dir))
         voice_id = self.all_voices[voice]
 
-        # Prepare payload for the job-based API
+        # Prepare payload for the API
         payload = {
             "text": text,
             "voice": voice_id,
@@ -530,44 +531,49 @@ class SpeechMaTTS(BaseTTSProvider):
         }
 
         try:
+            # Set logger level based on verbose flag
+            if verbose:
+                self.logger.level = LogLevel.DEBUG
+                self.logger.debug(f"Generating audio for voice: {voice} ({voice_id})")
+                self.logger.debug(f"Text length: {len(text)} characters")
+            else:
+                self.logger.level = LogLevel.INFO
+
+            # Make the request to the SpeechMa API
             response = self.session.post(
                 self.api_url,
                 headers=self.headers,
                 json=payload,
                 timeout=self.timeout
             )
-            response.raise_for_status()
-            resp_json = response.json()
-            if not resp_json.get("success") or "data" not in resp_json or "job_id" not in resp_json["data"]:
-                raise exceptions.FailedToGenerateResponseError(f"SpeechMa API error: {resp_json}")
-            job_id = resp_json["data"]["job_id"]
 
-            # Poll for job completion
-            status_url = f"https://speechma.com/com.api/tts-api.php/status/{job_id}"
-            for _ in range(30):  # up to ~30 seconds
-                status_resp = self.session.get(
-                    status_url,
-                    headers=self.headers,
-                    timeout=self.timeout
-                )
-                status_resp.raise_for_status()
-                status_json = status_resp.json()
-                if status_json.get("success") and status_json.get("data", {}).get("status") == "completed":
-                    break
-                time.sleep(1)
-            else:
-                raise exceptions.FailedToGenerateResponseError("TTS job did not complete in time.")
+            if response.status_code != 200:
+                if verbose:
+                    self.logger.error(f"API error: Status {response.status_code}")
+                raise exceptions.FailedToGenerateResponseError(f"API returned status {response.status_code}: {response.text[:500]}")
 
-            # Download the audio file (API provides a URL in the status response)
-            data = status_json["data"]
-            audio_url = f"https://speechma.com/com.api/tts-api.php/audio/{job_id}"
-            audio_resp = self.session.get(audio_url, timeout=self.timeout)
-            audio_resp.raise_for_status()
-            with open(filename, 'wb') as f:
-                f.write(audio_resp.content)
-            return filename.as_posix()
+            # Check if response is audio data (content-type should be audio/mpeg)
+            content_type = response.headers.get('content-type', '').lower()
+            if verbose:
+                self.logger.debug(f"Response content type: {content_type}")
+                self.logger.debug(f"Response size: {len(response.content)} bytes")
+
+            if 'audio' in content_type or response.content.startswith(b'\xff\xfb') or response.content.startswith(b'ID3') or b'LAME' in response.content[:100]:
+                # This is audio data, save it directly
+                with open(filename, 'wb') as f:
+                    f.write(response.content)
+                if verbose:
+                    self.logger.debug(f"Audio saved to: {filename}")
+                return filename.as_posix()
+            else:
+                # Unexpected response format
+                if verbose:
+                    self.logger.error(f"Unexpected response format: {content_type}")
+                raise exceptions.FailedToGenerateResponseError(f"Unexpected response format. Content-Type: {content_type}, Content: {response.text[:200]}")
 
         except requests.exceptions.RequestException as e:
+            if verbose:
+                self.logger.error(f"Request failed: {e}")
             raise exceptions.FailedToGenerateResponseError(
                 f"Failed to perform the operation: {e}"
             )

webscout/Provider/TogetherAI.py

@@ -0,0 +1,366 @@
+from curl_cffi.requests import Session
+from curl_cffi import CurlError
+from typing import Any, Dict, Optional, Generator, Union
+
+from webscout.AIutel import Optimizers
+from webscout.AIutel import Conversation
+from webscout.AIutel import AwesomePrompts, sanitize_stream
+from webscout.AIbase import Provider
+from webscout import exceptions
+from webscout.litagent import LitAgent
+
+class TogetherAI(Provider):
+    """
+    A class to interact with the TogetherAI API.
+    """
+
+    AVAILABLE_MODELS = [
+        "Gryphe/MythoMax-L2-13b",
+        "Gryphe/MythoMax-L2-13b-Lite",
+        "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
+        "Qwen/QwQ-32B",
+        "Qwen/Qwen2-72B-Instruct",
+        "Qwen/Qwen2-VL-72B-Instruct",
+        "Qwen/Qwen2.5-72B-Instruct-Turbo",
+        "Qwen/Qwen2.5-7B-Instruct-Turbo",
+        "Qwen/Qwen2.5-Coder-32B-Instruct",
+        "Qwen/Qwen2.5-VL-72B-Instruct",
+        "Qwen/Qwen3-235B-A22B-fp8",
+        "Qwen/Qwen3-235B-A22B-fp8-tput",
+        "Rrrr/meta-llama/Llama-3-70b-chat-hf-6f9ad551",
+        "Rrrr/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo-03dc18e1",
+        "Rrrr/meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo-6c92f39d",
+        "arcee-ai/arcee-blitz",
+        "arcee-ai/caller",
+        "arcee-ai/coder-large",
+        "arcee-ai/maestro-reasoning",
+        "arcee-ai/virtuoso-large",
+        "arcee-ai/virtuoso-medium-v2",
+        "arcee_ai/arcee-spotlight",
+        "blackbox/meta-llama-3-1-8b",
+        "deepseek-ai/DeepSeek-R1",
+        "deepseek-ai/DeepSeek-R1-Distill-Llama-70B",
+        "deepseek-ai/DeepSeek-R1-Distill-Llama-70B-free",
+        "deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5B",
+        "deepseek-ai/DeepSeek-R1-Distill-Qwen-14B",
+        "deepseek-ai/DeepSeek-V3",
+        "deepseek-ai/DeepSeek-V3-p-dp",
+        "google/gemma-2-27b-it",
+        "google/gemma-2b-it",
+        "lgai/exaone-3-5-32b-instruct",
+        "lgai/exaone-deep-32b",
+        "marin-community/marin-8b-instruct",
+        "meta-llama/Llama-3-70b-chat-hf",
+        "meta-llama/Llama-3-8b-chat-hf",
+        "meta-llama/Llama-3.2-11B-Vision-Instruct-Turbo",
+        "meta-llama/Llama-3.2-3B-Instruct-Turbo",
+        "meta-llama/Llama-3.2-90B-Vision-Instruct-Turbo",
+        "meta-llama/Llama-3.3-70B-Instruct-Turbo",
+        "meta-llama/Llama-3.3-70B-Instruct-Turbo-Free",
+        "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8",
+        "meta-llama/Llama-4-Scout-17B-16E-Instruct",
+        "meta-llama/Llama-Vision-Free",
+        "meta-llama/Meta-Llama-3-70B-Instruct-Turbo",
+        "meta-llama/Meta-Llama-3-8B-Instruct-Lite",
+        "meta-llama/Meta-Llama-3.1-405B-Instruct-Turbo",
+        "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo",
+        "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo",
+        "mistralai/Mistral-7B-Instruct-v0.1",
+        "mistralai/Mistral-7B-Instruct-v0.2",
+        "mistralai/Mistral-7B-Instruct-v0.3",
+        "mistralai/Mistral-Small-24B-Instruct-2501",
+        "mistralai/Mixtral-8x7B-Instruct-v0.1",
+        "nvidia/Llama-3.1-Nemotron-70B-Instruct-HF",
+        "perplexity-ai/r1-1776",
+        "roberizk@gmail.com/meta-llama/Llama-3-70b-chat-hf-26ee936b",
+        "roberizk@gmail.com/meta-llama/Meta-Llama-3-70B-Instruct-6feb41f7",
+        "roberizk@gmail.com/meta-llama/Meta-Llama-3-8B-Instruct-8ced8839",
+        "scb10x/scb10x-llama3-1-typhoon2-70b-instruct",
+        "scb10x/scb10x-llama3-1-typhoon2-8b-instruct",
+        "togethercomputer/MoA-1",
+        "togethercomputer/MoA-1-Turbo",
+        "togethercomputer/Refuel-Llm-V2",
+        "togethercomputer/Refuel-Llm-V2-Small",
+    ]
+
+    @staticmethod
+    def _togetherai_extractor(chunk: Union[str, Dict[str, Any]]) -> Optional[str]:
+        """Extracts content from TogetherAI stream JSON objects."""
+        if isinstance(chunk, dict):
+            return chunk.get("choices", [{}])[0].get("delta", {}).get("content")
+        return None
+
+    def __init__(
+        self,
+        is_conversation: bool = True,
+        max_tokens: int = 2049,
+        timeout: int = 30,
+        intro: str = None,
+        filepath: str = None,
+        update_file: bool = True,
+        proxies: dict = {},
+        history_offset: int = 10250,
+        act: str = None,
+        model: str = "meta-llama/Llama-3.1-8B-Instruct-Turbo",
+        system_prompt: str = "You are a helpful assistant.",
+        browser: str = "chrome"
+    ):
+        """Initializes the TogetherAI API client."""
+        if model not in self.AVAILABLE_MODELS:
+            raise ValueError(f"Invalid model: {model}. Choose from: {self.AVAILABLE_MODELS}")
+
+        self.api_endpoint = "https://api.together.xyz/v1/chat/completions"
+        self.activation_endpoint = "https://www.codegeneration.ai/activate-v2"
+
+        # Initialize LitAgent
+        self.agent = LitAgent()
+        self.fingerprint = self.agent.generate_fingerprint(browser)
+
+        # Use the fingerprint for headers
+        self.headers = {
+            "Accept": self.fingerprint["accept"],
+            "Accept-Language": self.fingerprint["accept_language"],
+            "Content-Type": "application/json",
+            "Cache-Control": "no-cache",
+            "Origin": "https://www.codegeneration.ai",
+            "Pragma": "no-cache",
+            "Referer": "https://www.codegeneration.ai/",
+            "Sec-Fetch-Dest": "empty",
+            "Sec-Fetch-Mode": "cors",
+            "Sec-Fetch-Site": "same-site",
+            "User-Agent": self.fingerprint["user_agent"],
+        }
+
+        # Initialize curl_cffi Session
+        self.session = Session()
+        self.session.headers.update(self.headers)
+        self.session.proxies = proxies
+        self.system_prompt = system_prompt
+        self.is_conversation = is_conversation
+        self.max_tokens_to_sample = max_tokens
+        self.timeout = timeout
+        self.last_response = {}
+        self.model = model
+        self._api_key_cache = None
+
+        self.__available_optimizers = (
+            method
+            for method in dir(Optimizers)
+            if callable(getattr(Optimizers, method)) and not method.startswith("__")
+        )
+        Conversation.intro = (
+            AwesomePrompts().get_act(
+                act, raise_not_found=True, default=None, case_insensitive=True
+            )
+            if act
+            else intro or Conversation.intro
+        )
+
+        self.conversation = Conversation(
+            is_conversation, self.max_tokens_to_sample, filepath, update_file
+        )
+        self.conversation.history_offset = history_offset
+
+    def refresh_identity(self, browser: str = None):
+        """
+        Refreshes the browser identity fingerprint.
+
+        Args:
+            browser: Specific browser to use for the new fingerprint
+        """
+        browser = browser or self.fingerprint.get("browser_type", "chrome")
+        self.fingerprint = self.agent.generate_fingerprint(browser)
+
+        # Update headers with new fingerprint
+        self.headers.update({
+            "Accept": self.fingerprint["accept"],
+            "Accept-Language": self.fingerprint["accept_language"],
+            "User-Agent": self.fingerprint["user_agent"],
+        })
+
+        # Update session headers
+        self.session.headers.update(self.headers)
+
+        return self.fingerprint
+
+    def get_activation_key(self) -> str:
+        """Get API key from activation endpoint"""
+        if self._api_key_cache:
+            return self._api_key_cache
+
+        try:
+            response = self.session.get(
+                self.activation_endpoint,
+                headers={"Accept": "application/json"},
+                timeout=30
+            )
+            response.raise_for_status()
+            activation_data = response.json()
+            self._api_key_cache = activation_data["openAIParams"]["apiKey"]
+            return self._api_key_cache
+        except Exception as e:
+            raise exceptions.FailedToGenerateResponseError(f"Failed to get activation key: {e}")
+
+    def ask(
+        self,
+        prompt: str,
+        stream: bool = False,
+        raw: bool = False,
+        optimizer: str = None,
+        conversationally: bool = False,
+    ) -> Union[Dict[str, Any], Generator]:
+        """
+        Sends a prompt to the TogetherAI API and returns the response.
+        """
+        conversation_prompt = self.conversation.gen_complete_prompt(prompt)
+        if optimizer:
+            if optimizer in self.__available_optimizers:
+                conversation_prompt = getattr(Optimizers, optimizer)(
+                    conversation_prompt if conversationally else prompt
+                )
+            else:
+                raise Exception(f"Optimizer is not one of {self.__available_optimizers}")
+
+        # Get API key if not already set
+        if not self.headers.get("Authorization"):
+            api_key = self.get_activation_key()
+            self.headers["Authorization"] = f"Bearer {api_key}"
+            self.session.headers.update(self.headers)
+
+        # Payload construction
+        payload = {
+            "model": self.model,
+            "messages": [
+                {"role": "system", "content": self.system_prompt},
+                {"role": "user", "content": conversation_prompt},
+            ],
+            "stream": stream
+        }
+
+        def for_stream():
+            streaming_text = ""
+            try:
+                response = self.session.post(
+                    self.api_endpoint,
+                    json=payload,
+                    stream=True,
+                    timeout=self.timeout,
+                    impersonate="chrome110"
+                )
+                response.raise_for_status()
+
+                # Use sanitize_stream
+                processed_stream = sanitize_stream(
+                    data=response.iter_content(chunk_size=None),
+                    intro_value="data:",
+                    to_json=True,
+                    skip_markers=["[DONE]"],
+                    content_extractor=self._togetherai_extractor,
+                    yield_raw_on_error=False
+                )
+
+                for content_chunk in processed_stream:
+                    if content_chunk and isinstance(content_chunk, str):
+                        streaming_text += content_chunk
+                        resp = dict(text=content_chunk)
+                        yield resp if not raw else content_chunk
+
+            except CurlError as e:
+                raise exceptions.FailedToGenerateResponseError(f"Request failed (CurlError): {str(e)}") from e
+            except Exception as e:
+                raise exceptions.FailedToGenerateResponseError(f"Request failed ({type(e).__name__}): {str(e)}") from e
+            finally:
+                if streaming_text:
+                    self.last_response = {"text": streaming_text}
+                    self.conversation.update_chat_history(prompt, streaming_text)
+
+        def for_non_stream():
+            try:
+                response = self.session.post(
+                    self.api_endpoint,
+                    json=payload,
+                    timeout=self.timeout,
+                    impersonate="chrome110"
+                )
+                response.raise_for_status()
+
+                response_text = response.text
+
+                # Use sanitize_stream to parse the non-streaming JSON response
+                processed_stream = sanitize_stream(
+                    data=response_text,
+                    to_json=True,
+                    intro_value=None,
+                    content_extractor=lambda chunk: chunk.get("choices", [{}])[0].get("message", {}).get("content") if isinstance(chunk, dict) else None,
+                    yield_raw_on_error=False
+                )
+                content = next(processed_stream, None)
+                content = content if isinstance(content, str) else ""
+
+                self.last_response = {"text": content}
+                self.conversation.update_chat_history(prompt, content)
+                return self.last_response if not raw else content
+
+            except CurlError as e:
+                raise exceptions.FailedToGenerateResponseError(f"Request failed (CurlError): {e}") from e
+            except Exception as e:
+                err_text = getattr(e, 'response', None) and getattr(e.response, 'text', '')
+                raise exceptions.FailedToGenerateResponseError(f"Request failed ({type(e).__name__}): {e} - {err_text}") from e
+
+        return for_stream() if stream else for_non_stream()
+
+    def chat(
+        self,
+        prompt: str,
+        stream: bool = False,
+        optimizer: str = None,
+        conversationally: bool = False,
+    ) -> Union[str, Generator[str, None, None]]:
+        """Generate response `str`"""
+        def for_stream_chat():
+            gen = self.ask(
+                prompt, stream=True, raw=False,
+                optimizer=optimizer, conversationally=conversationally
+            )
+            for response_dict in gen:
+                yield self.get_message(response_dict)
+
+        def for_non_stream_chat():
+            response_data = self.ask(
+                prompt, stream=False, raw=False,
+                optimizer=optimizer, conversationally=conversationally
+            )
+            return self.get_message(response_data)
+
+        return for_stream_chat() if stream else for_non_stream_chat()
+
+    def get_message(self, response: dict) -> str:
+        """Retrieves message only from response"""
+        assert isinstance(response, dict), "Response should be of dict data-type only"
+        return response["text"]
+
+
+if __name__ == "__main__":
+    print("-" * 80)
+    print(f"{'Model':<50} {'Status':<10} {'Response'}")
+    print("-" * 80)
+
+    for model in TogetherAI.AVAILABLE_MODELS:
+        try:
+            test_ai = TogetherAI(model=model, timeout=60)
+            response = test_ai.chat("Say 'Hello' in one word", stream=True)
+            response_text = ""
+            for chunk in response:
+                response_text += chunk
+
+            if response_text and len(response_text.strip()) > 0:
+                status = "✓"
+                # Clean and truncate response
+                clean_text = response_text.strip().encode('utf-8', errors='ignore').decode('utf-8')
+                display_text = clean_text[:50] + "..." if len(clean_text) > 50 else clean_text
+            else:
+                status = "✗"
+                display_text = "Empty or invalid response"
+            print(f"\r{model:<50} {status:<10} {display_text}")
+        except Exception as e:
+            print(f"\r{model:<50} {'✗':<10} {str(e)}")