webscout 8.2.2__py3-none-any.whl → 8.2.7__py3-none-any.whl

webscout/Provider/OPENAI/deepinfra.py

@@ -1,272 +0,0 @@
-import requests
-import json
-import time
-import uuid
-from typing import List, Dict, Optional, Union, Generator, Any
-
-# Import base classes and utility structures
-from .base import OpenAICompatibleProvider, BaseChat, BaseCompletions
-from .utils import (
-    ChatCompletionChunk, ChatCompletion, Choice, ChoiceDelta,
-    ChatCompletionMessage, CompletionUsage
-)
-
-# Attempt to import LitAgent, fallback if not available
-try:
-    from webscout.litagent import LitAgent
-except ImportError:
-    pass
-
-# --- DeepInfra Client ---
-
-class Completions(BaseCompletions):
-    def __init__(self, client: 'DeepInfra'):
-        self._client = client
-
-    def create(
-        self,
-        *,
-        model: str,
-        messages: List[Dict[str, str]],
-        max_tokens: Optional[int] = 2049,
-        stream: bool = False,
-        temperature: Optional[float] = None,
-        top_p: Optional[float] = None,
-        **kwargs: Any
-    ) -> Union[ChatCompletion, Generator[ChatCompletionChunk, None, None]]:
-        """
-        Creates a model response for the given chat conversation.
-        Mimics openai.chat.completions.create
-        """
-        payload = {
-            "model": model,
-            "messages": messages,
-            "max_tokens": max_tokens,
-            "stream": stream,
-        }
-        if temperature is not None:
-            payload["temperature"] = temperature
-        if top_p is not None:
-            payload["top_p"] = top_p
-
-        payload.update(kwargs)
-
-        request_id = f"chatcmpl-{uuid.uuid4()}"
-        created_time = int(time.time())
-
-        if stream:
-            return self._create_stream(request_id, created_time, model, payload)
-        else:
-            return self._create_non_stream(request_id, created_time, model, payload)
-
-    def _create_stream(
-        self, request_id: str, created_time: int, model: str, payload: Dict[str, Any]
-    ) -> Generator[ChatCompletionChunk, None, None]:
-        try:
-            response = self._client.session.post(
-                self._client.base_url,
-                headers=self._client.headers,
-                json=payload,
-                stream=True,
-                timeout=self._client.timeout
-            )
-            response.raise_for_status()
-
-            # Track token usage across chunks
-            prompt_tokens = 0
-            completion_tokens = 0
-            total_tokens = 0
-
-            for line in response.iter_lines():
-                if line:
-                    decoded_line = line.decode('utf-8').strip()
-
-                    if decoded_line.startswith("data: "):
-                        json_str = decoded_line[6:]
-                        if json_str == "[DONE]":
-                            # Format the final [DONE] marker in OpenAI format
-                            # print("data: [DONE]")
-                            break
-
-                        try:
-                            data = json.loads(json_str)
-                            choice_data = data.get('choices', [{}])[0]
-                            delta_data = choice_data.get('delta', {})
-                            finish_reason = choice_data.get('finish_reason')
-
-                            # Update token counts if available
-                            usage_data = data.get('usage', {})
-                            if usage_data:
-                                prompt_tokens = usage_data.get('prompt_tokens', prompt_tokens)
-                                completion_tokens = usage_data.get('completion_tokens', completion_tokens)
-                                total_tokens = usage_data.get('total_tokens', total_tokens)
-
-                            # Create the delta object
-                            delta = ChoiceDelta(
-                                content=delta_data.get('content'),
-                                role=delta_data.get('role'),
-                                tool_calls=delta_data.get('tool_calls')
-                            )
-
-                            # Create the choice object
-                            choice = Choice(
-                                index=choice_data.get('index', 0),
-                                delta=delta,
-                                finish_reason=finish_reason,
-                                logprobs=choice_data.get('logprobs')
-                            )
-
-                            # Create the chunk object
-                            chunk = ChatCompletionChunk(
-                                id=request_id,
-                                choices=[choice],
-                                created=created_time,
-                                model=model,
-                                system_fingerprint=data.get('system_fingerprint')
-                            )
-
-                            # Convert to dict for proper formatting
-                            chunk_dict = chunk.to_dict()
-
-                            # Add usage information to match OpenAI format
-                            # Even if we don't have real token counts, include estimated usage
-                            # This matches the format in the examples
-                            usage_dict = {
-                                "prompt_tokens": prompt_tokens or 10,
-                                "completion_tokens": completion_tokens or (len(delta_data.get('content', '')) if delta_data.get('content') else 0),
-                                "total_tokens": total_tokens or (10 + (len(delta_data.get('content', '')) if delta_data.get('content') else 0)),
-                                "estimated_cost": None
-                            }
-
-                            # Update completion_tokens and total_tokens as we receive more content
-                            if delta_data.get('content'):
-                                completion_tokens += 1
-                                total_tokens = prompt_tokens + completion_tokens
-                                usage_dict["completion_tokens"] = completion_tokens
-                                usage_dict["total_tokens"] = total_tokens
-
-                            chunk_dict["usage"] = usage_dict
-
-                            # Format the response in OpenAI format exactly as requested
-                            # We need to print the raw string and also yield the chunk object
-                            # This ensures both the console output and the returned object are correct
-                            # print(f"data: {json.dumps(chunk_dict)}")
-
-                            # Return the chunk object for internal processing
-                            yield chunk
-                        except json.JSONDecodeError:
-                            print(f"Warning: Could not decode JSON line: {json_str}")
-                            continue
-        except requests.exceptions.RequestException as e:
-            print(f"Error during DeepInfra stream request: {e}")
-            raise IOError(f"DeepInfra request failed: {e}") from e
-        except Exception as e:
-            print(f"Error processing DeepInfra stream: {e}")
-            raise
-
-    def _create_non_stream(
-        self, request_id: str, created_time: int, model: str, payload: Dict[str, Any]
-    ) -> ChatCompletion:
-        try:
-            response = self._client.session.post(
-                self._client.base_url,
-                headers=self._client.headers,
-                json=payload,
-                timeout=self._client.timeout
-            )
-            response.raise_for_status()
-            data = response.json()
-
-            choices_data = data.get('choices', [])
-            usage_data = data.get('usage', {})
-
-            choices = []
-            for choice_d in choices_data:
-                message_d = choice_d.get('message', {})
-                message = ChatCompletionMessage(
-                    role=message_d.get('role', 'assistant'),
-                    content=message_d.get('content', '')
-                )
-                choice = Choice(
-                    index=choice_d.get('index', 0),
-                    message=message,
-                    finish_reason=choice_d.get('finish_reason', 'stop')
-                )
-                choices.append(choice)
-
-            usage = CompletionUsage(
-                prompt_tokens=usage_data.get('prompt_tokens', 0),
-                completion_tokens=usage_data.get('completion_tokens', 0),
-                total_tokens=usage_data.get('total_tokens', 0)
-            )
-
-            completion = ChatCompletion(
-                id=request_id,
-                choices=choices,
-                created=created_time,
-                model=data.get('model', model),
-                usage=usage,
-            )
-            return completion
-
-        except requests.exceptions.RequestException as e:
-            print(f"Error during DeepInfra non-stream request: {e}")
-            raise IOError(f"DeepInfra request failed: {e}") from e
-        except Exception as e:
-            print(f"Error processing DeepInfra response: {e}")
-            raise
-
-class Chat(BaseChat):
-    def __init__(self, client: 'DeepInfra'):
-        self.completions = Completions(client)
-
-class DeepInfra(OpenAICompatibleProvider):
-    AVAILABLE_MODELS = [
-        "deepseek-ai/DeepSeek-V3",
-        "google/gemma-2-27b-it",
-        "google/gemma-2-9b-it",
-        "meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8",
-        "meta-llama/Llama-4-Scout-17B-16E-Instruct",
-        "meta-llama/Llama-3.3-70B-Instruct",
-        "meta-llama/Llama-3.3-70B-Instruct-Turbo",
-        "meta-llama/Meta-Llama-3.1-8B-Instruct",
-        "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo",
-        "microsoft/phi-4",
-        "microsoft/Phi-4-multimodal-instruct",
-        "microsoft/WizardLM-2-8x22B",
-        "mistralai/Mistral-Small-24B-Instruct-2501",
-        "nvidia/Llama-3.1-Nemotron-70B-Instruct",
-        "Qwen/QwQ-32B",
-        "Qwen/Qwen2.5-72B-Instruct",
-        "Qwen/Qwen2.5-Coder-32B-Instruct",
-    ]
-
-    def __init__(self, timeout: Optional[int] = None, browser: str = "chrome"):
-        self.timeout = timeout
-        self.base_url = "https://api.deepinfra.com/v1/openai/chat/completions"
-        self.session = requests.Session()
-
-        agent = LitAgent()
-        fingerprint = agent.generate_fingerprint(browser)
-
-        self.headers = {
-            "Accept": fingerprint["accept"],
-            "Accept-Encoding": "gzip, deflate, br, zstd",
-            "Accept-Language": fingerprint["accept_language"],
-            "Content-Type": "application/json",
-            "Cache-Control": "no-cache",
-            "Connection": "keep-alive",
-            "Origin": "https://deepinfra.com",
-            "Pragma": "no-cache",
-            "Referer": "https://deepinfra.com/",
-            "Sec-Fetch-Dest": "empty",
-            "Sec-Fetch-Mode": "cors",
-            "Sec-Fetch-Site": "same-site",
-            "X-Deepinfra-Source": "web-embed",
-            "Sec-CH-UA": fingerprint["sec_ch_ua"] or '"Not)A;Brand";v="99", "Microsoft Edge";v="127", "Chromium";v="127"',
-            "Sec-CH-UA-Mobile": "?0",
-            "Sec-CH-UA-Platform": f'"{fingerprint["platform"]}"',
-            "User-Agent": fingerprint["user_agent"],
-        }
-        self.session.headers.update(self.headers)
-        self.chat = Chat(self)