agentrun-mem0ai 0.0.11__py3-none-any.whl

agentrun_mem0/llms/ollama.py

@@ -0,0 +1,117 @@
+from typing import Dict, List, Optional, Union
+
+try:
+    from ollama import Client
+except ImportError:
+    raise ImportError("The 'ollama' library is required. Please install it using 'pip install ollama'.")
+
+from agentrun_mem0.configs.llms.base import BaseLlmConfig
+from agentrun_mem0.configs.llms.ollama import OllamaConfig
+from agentrun_mem0.llms.base import LLMBase
+
+
+class OllamaLLM(LLMBase):
+    def __init__(self, config: Optional[Union[BaseLlmConfig, OllamaConfig, Dict]] = None):
+        # Convert to OllamaConfig if needed
+        if config is None:
+            config = OllamaConfig()
+        elif isinstance(config, dict):
+            config = OllamaConfig(**config)
+        elif isinstance(config, BaseLlmConfig) and not isinstance(config, OllamaConfig):
+            # Convert BaseLlmConfig to OllamaConfig
+            config = OllamaConfig(
+                model=config.model,
+                temperature=config.temperature,
+                api_key=config.api_key,
+                max_tokens=config.max_tokens,
+                top_p=config.top_p,
+                top_k=config.top_k,
+                enable_vision=config.enable_vision,
+                vision_details=config.vision_details,
+                http_client_proxies=config.http_client,
+            )
+
+        super().__init__(config)
+
+        if not self.config.model:
+            self.config.model = "llama3.1:70b"
+
+        self.client = Client(host=self.config.ollama_base_url)
+
+    def _parse_response(self, response, tools):
+        """
+        Process the response based on whether tools are used or not.
+
+        Args:
+            response: The raw response from API.
+            tools: The list of tools provided in the request.
+
+        Returns:
+            str or dict: The processed response.
+        """
+        # Get the content from response
+        if isinstance(response, dict):
+            content = response["message"]["content"]
+        else:
+            content = response.message.content
+
+        if tools:
+            processed_response = {
+                "content": content,
+                "tool_calls": [],
+            }
+
+            # Ollama doesn't support tool calls in the same way, so we return the content
+            return processed_response
+        else:
+            return content
+
+    def generate_response(
+        self,
+        messages: List[Dict[str, str]],
+        response_format=None,
+        tools: Optional[List[Dict]] = None,
+        tool_choice: str = "auto",
+        **kwargs,
+    ):
+        """
+        Generate a response based on the given messages using Ollama.
+
+        Args:
+            messages (list): List of message dicts containing 'role' and 'content'.
+            response_format (str or object, optional): Format of the response. Defaults to "text".
+            tools (list, optional): List of tools that the model can call. Defaults to None.
+            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            **kwargs: Additional Ollama-specific parameters.
+
+        Returns:
+            str: The generated response.
+        """
+        # Build parameters for Ollama
+        params = {
+            "model": self.config.model,
+            "messages": messages,
+        }
+
+        # Handle JSON response format by using Ollama's native format parameter
+        if response_format and response_format.get("type") == "json_object":
+            params["format"] = "json"
+            # Also add JSON format instruction to the last message as a fallback
+            if messages and messages[-1]["role"] == "user":
+                messages[-1]["content"] += "\n\nPlease respond with valid JSON only."
+            else:
+                messages.append({"role": "user", "content": "Please respond with valid JSON only."})
+
+        # Add options for Ollama (temperature, num_predict, top_p)
+        options = {
+            "temperature": self.config.temperature,
+            "num_predict": self.config.max_tokens,
+            "top_p": self.config.top_p,
+        }
+        params["options"] = options
+
+        # Remove OpenAI-specific parameters that Ollama doesn't support
+        params.pop("max_tokens", None)  # Ollama uses different parameter names
+
+        response = self.client.chat(**params)
+        return self._parse_response(response, tools)

agentrun_mem0/llms/openai.py

@@ -0,0 +1,147 @@
+import json
+import logging
+import os
+from typing import Dict, List, Optional, Union
+
+from openai import OpenAI
+
+from agentrun_mem0.configs.llms.base import BaseLlmConfig
+from agentrun_mem0.configs.llms.openai import OpenAIConfig
+from agentrun_mem0.llms.base import LLMBase
+from agentrun_mem0.memory.utils import extract_json
+
+
+class OpenAILLM(LLMBase):
+    def __init__(self, config: Optional[Union[BaseLlmConfig, OpenAIConfig, Dict]] = None):
+        # Convert to OpenAIConfig if needed
+        if config is None:
+            config = OpenAIConfig()
+        elif isinstance(config, dict):
+            config = OpenAIConfig(**config)
+        elif isinstance(config, BaseLlmConfig) and not isinstance(config, OpenAIConfig):
+            # Convert BaseLlmConfig to OpenAIConfig
+            config = OpenAIConfig(
+                model=config.model,
+                temperature=config.temperature,
+                api_key=config.api_key,
+                max_tokens=config.max_tokens,
+                top_p=config.top_p,
+                top_k=config.top_k,
+                enable_vision=config.enable_vision,
+                vision_details=config.vision_details,
+                http_client_proxies=config.http_client,
+            )
+
+        super().__init__(config)
+
+        if not self.config.model:
+            self.config.model = "gpt-4.1-nano-2025-04-14"
+
+        if os.environ.get("OPENROUTER_API_KEY"):  # Use OpenRouter
+            self.client = OpenAI(
+                api_key=os.environ.get("OPENROUTER_API_KEY"),
+                base_url=self.config.openrouter_base_url
+                or os.getenv("OPENROUTER_API_BASE")
+                or "https://openrouter.ai/api/v1",
+            )
+        else:
+            api_key = self.config.api_key or os.getenv("OPENAI_API_KEY")
+            base_url = self.config.openai_base_url or os.getenv("OPENAI_BASE_URL") or "https://api.openai.com/v1"
+
+            self.client = OpenAI(api_key=api_key, base_url=base_url)
+
+    def _parse_response(self, response, tools):
+        """
+        Process the response based on whether tools are used or not.
+
+        Args:
+            response: The raw response from API.
+            tools: The list of tools provided in the request.
+
+        Returns:
+            str or dict: The processed response.
+        """
+        if tools:
+            processed_response = {
+                "content": response.choices[0].message.content,
+                "tool_calls": [],
+            }
+
+            if response.choices[0].message.tool_calls:
+                for tool_call in response.choices[0].message.tool_calls:
+                    processed_response["tool_calls"].append(
+                        {
+                            "name": tool_call.function.name,
+                            "arguments": json.loads(extract_json(tool_call.function.arguments)),
+                        }
+                    )
+
+            return processed_response
+        else:
+            return response.choices[0].message.content
+
+    def generate_response(
+        self,
+        messages: List[Dict[str, str]],
+        response_format=None,
+        tools: Optional[List[Dict]] = None,
+        tool_choice: str = "auto",
+        **kwargs,
+    ):
+        """
+        Generate a JSON response based on the given messages using OpenAI.
+
+        Args:
+            messages (list): List of message dicts containing 'role' and 'content'.
+            response_format (str or object, optional): Format of the response. Defaults to "text".
+            tools (list, optional): List of tools that the model can call. Defaults to None.
+            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            **kwargs: Additional OpenAI-specific parameters.
+
+        Returns:
+            json: The generated response.
+        """
+        params = self._get_supported_params(messages=messages, **kwargs)
+        
+        params.update({
+            "model": self.config.model,
+            "messages": messages,
+        })
+
+        if os.getenv("OPENROUTER_API_KEY"):
+            openrouter_params = {}
+            if self.config.models:
+                openrouter_params["models"] = self.config.models
+                openrouter_params["route"] = self.config.route
+                params.pop("model")
+
+            if self.config.site_url and self.config.app_name:
+                extra_headers = {
+                    "HTTP-Referer": self.config.site_url,
+                    "X-Title": self.config.app_name,
+                }
+                openrouter_params["extra_headers"] = extra_headers
+
+            params.update(**openrouter_params)
+        
+        else:
+            openai_specific_generation_params = ["store"]
+            for param in openai_specific_generation_params:
+                if hasattr(self.config, param):
+                    params[param] = getattr(self.config, param)
+            
+        if response_format:
+            params["response_format"] = response_format
+        if tools:  # TODO: Remove tools if no issues found with new memory addition logic
+            params["tools"] = tools
+            params["tool_choice"] = tool_choice
+        response = self.client.chat.completions.create(**params)
+        parsed_response = self._parse_response(response, tools)
+        if self.config.response_callback:
+            try:
+                self.config.response_callback(self, response, params)
+            except Exception as e:
+                # Log error but don't propagate
+                logging.error(f"Error due to callback: {e}")
+                pass
+        return parsed_response

agentrun_mem0/llms/openai_structured.py

@@ -0,0 +1,52 @@
+import os
+from typing import Dict, List, Optional
+
+from openai import OpenAI
+
+from agentrun_mem0.configs.llms.base import BaseLlmConfig
+from agentrun_mem0.llms.base import LLMBase
+
+
+class OpenAIStructuredLLM(LLMBase):
+    def __init__(self, config: Optional[BaseLlmConfig] = None):
+        super().__init__(config)
+
+        if not self.config.model:
+            self.config.model = "gpt-4o-2024-08-06"
+
+        api_key = self.config.api_key or os.getenv("OPENAI_API_KEY")
+        base_url = self.config.openai_base_url or os.getenv("OPENAI_API_BASE") or "https://api.openai.com/v1"
+        self.client = OpenAI(api_key=api_key, base_url=base_url)
+
+    def generate_response(
+        self,
+        messages: List[Dict[str, str]],
+        response_format: Optional[str] = None,
+        tools: Optional[List[Dict]] = None,
+        tool_choice: str = "auto",
+    ) -> str:
+        """
+        Generate a response based on the given messages using OpenAI.
+
+        Args:
+            messages (List[Dict[str, str]]): A list of dictionaries, each containing a 'role' and 'content' key.
+            response_format (Optional[str]): The desired format of the response. Defaults to None.
+
+
+        Returns:
+            str: The generated response.
+        """
+        params = {
+            "model": self.config.model,
+            "messages": messages,
+            "temperature": self.config.temperature,
+        }
+
+        if response_format:
+            params["response_format"] = response_format
+        if tools:
+            params["tools"] = tools
+            params["tool_choice"] = tool_choice
+
+        response = self.client.beta.chat.completions.parse(**params)
+        return response.choices[0].message.content

agentrun_mem0/llms/sarvam.py

@@ -0,0 +1,89 @@
+import os
+from typing import Dict, List, Optional
+
+import requests
+
+from agentrun_mem0.configs.llms.base import BaseLlmConfig
+from agentrun_mem0.llms.base import LLMBase
+
+
+class SarvamLLM(LLMBase):
+    def __init__(self, config: Optional[BaseLlmConfig] = None):
+        super().__init__(config)
+
+        # Set default model if not provided
+        if not self.config.model:
+            self.config.model = "sarvam-m"
+
+        # Get API key from config or environment variable
+        self.api_key = self.config.api_key or os.getenv("SARVAM_API_KEY")
+
+        if not self.api_key:
+            raise ValueError(
+                "Sarvam API key is required. Set SARVAM_API_KEY environment variable or provide api_key in config."
+            )
+
+        # Set base URL - use config value or environment or default
+        self.base_url = (
+            getattr(self.config, "sarvam_base_url", None) or os.getenv("SARVAM_API_BASE") or "https://api.sarvam.ai/v1"
+        )
+
+    def generate_response(self, messages: List[Dict[str, str]], response_format=None) -> str:
+        """
+        Generate a response based on the given messages using Sarvam-M.
+
+        Args:
+            messages (list): List of message dicts containing 'role' and 'content'.
+            response_format (str or object, optional): Format of the response.
+                                                     Currently not used by Sarvam API.
+
+        Returns:
+            str: The generated response.
+        """
+        url = f"{self.base_url}/chat/completions"
+
+        headers = {"Authorization": f"Bearer {self.api_key}", "Content-Type": "application/json"}
+
+        # Prepare the request payload
+        params = {
+            "messages": messages,
+            "model": self.config.model if isinstance(self.config.model, str) else "sarvam-m",
+        }
+
+        # Add standard parameters that already exist in BaseLlmConfig
+        if self.config.temperature is not None:
+            params["temperature"] = self.config.temperature
+
+        if self.config.max_tokens is not None:
+            params["max_tokens"] = self.config.max_tokens
+
+        if self.config.top_p is not None:
+            params["top_p"] = self.config.top_p
+
+        # Handle Sarvam-specific parameters if model is passed as dict
+        if isinstance(self.config.model, dict):
+            # Extract model name
+            params["model"] = self.config.model.get("name", "sarvam-m")
+
+            # Add Sarvam-specific parameters
+            sarvam_specific_params = ["reasoning_effort", "frequency_penalty", "presence_penalty", "seed", "stop", "n"]
+
+            for param in sarvam_specific_params:
+                if param in self.config.model:
+                    params[param] = self.config.model[param]
+
+        try:
+            response = requests.post(url, headers=headers, json=params, timeout=30)
+            response.raise_for_status()
+
+            result = response.json()
+
+            if "choices" in result and len(result["choices"]) > 0:
+                return result["choices"][0]["message"]["content"]
+            else:
+                raise ValueError("No response choices found in Sarvam API response")
+
+        except requests.exceptions.RequestException as e:
+            raise RuntimeError(f"Sarvam API request failed: {e}")
+        except KeyError as e:
+            raise ValueError(f"Unexpected response format from Sarvam API: {e}")

agentrun_mem0/llms/together.py

@@ -0,0 +1,88 @@
+import json
+import os
+from typing import Dict, List, Optional
+
+try:
+    from together import Together
+except ImportError:
+    raise ImportError("The 'together' library is required. Please install it using 'pip install together'.")
+
+from agentrun_mem0.configs.llms.base import BaseLlmConfig
+from agentrun_mem0.llms.base import LLMBase
+from agentrun_mem0.memory.utils import extract_json
+
+
+class TogetherLLM(LLMBase):
+    def __init__(self, config: Optional[BaseLlmConfig] = None):
+        super().__init__(config)
+
+        if not self.config.model:
+            self.config.model = "mistralai/Mixtral-8x7B-Instruct-v0.1"
+
+        api_key = self.config.api_key or os.getenv("TOGETHER_API_KEY")
+        self.client = Together(api_key=api_key)
+
+    def _parse_response(self, response, tools):
+        """
+        Process the response based on whether tools are used or not.
+
+        Args:
+            response: The raw response from API.
+            tools: The list of tools provided in the request.
+
+        Returns:
+            str or dict: The processed response.
+        """
+        if tools:
+            processed_response = {
+                "content": response.choices[0].message.content,
+                "tool_calls": [],
+            }
+
+            if response.choices[0].message.tool_calls:
+                for tool_call in response.choices[0].message.tool_calls:
+                    processed_response["tool_calls"].append(
+                        {
+                            "name": tool_call.function.name,
+                            "arguments": json.loads(extract_json(tool_call.function.arguments)),
+                        }
+                    )
+
+            return processed_response
+        else:
+            return response.choices[0].message.content
+
+    def generate_response(
+        self,
+        messages: List[Dict[str, str]],
+        response_format=None,
+        tools: Optional[List[Dict]] = None,
+        tool_choice: str = "auto",
+    ):
+        """
+        Generate a response based on the given messages using TogetherAI.
+
+        Args:
+            messages (list): List of message dicts containing 'role' and 'content'.
+            response_format (str or object, optional): Format of the response. Defaults to "text".
+            tools (list, optional): List of tools that the model can call. Defaults to None.
+            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+
+        Returns:
+            str: The generated response.
+        """
+        params = {
+            "model": self.config.model,
+            "messages": messages,
+            "temperature": self.config.temperature,
+            "max_tokens": self.config.max_tokens,
+            "top_p": self.config.top_p,
+        }
+        if response_format:
+            params["response_format"] = response_format
+        if tools:  # TODO: Remove tools if no issues found with new memory addition logic
+            params["tools"] = tools
+            params["tool_choice"] = tool_choice
+
+        response = self.client.chat.completions.create(**params)
+        return self._parse_response(response, tools)

agentrun_mem0/llms/vllm.py

@@ -0,0 +1,107 @@
+import json
+import os
+from typing import Dict, List, Optional, Union
+
+from openai import OpenAI
+
+from agentrun_mem0.configs.llms.base import BaseLlmConfig
+from agentrun_mem0.configs.llms.vllm import VllmConfig
+from agentrun_mem0.llms.base import LLMBase
+from agentrun_mem0.memory.utils import extract_json
+
+
+class VllmLLM(LLMBase):
+    def __init__(self, config: Optional[Union[BaseLlmConfig, VllmConfig, Dict]] = None):
+        # Convert to VllmConfig if needed
+        if config is None:
+            config = VllmConfig()
+        elif isinstance(config, dict):
+            config = VllmConfig(**config)
+        elif isinstance(config, BaseLlmConfig) and not isinstance(config, VllmConfig):
+            # Convert BaseLlmConfig to VllmConfig
+            config = VllmConfig(
+                model=config.model,
+                temperature=config.temperature,
+                api_key=config.api_key,
+                max_tokens=config.max_tokens,
+                top_p=config.top_p,
+                top_k=config.top_k,
+                enable_vision=config.enable_vision,
+                vision_details=config.vision_details,
+                http_client_proxies=config.http_client,
+            )
+
+        super().__init__(config)
+
+        if not self.config.model:
+            self.config.model = "Qwen/Qwen2.5-32B-Instruct"
+
+        self.config.api_key = self.config.api_key or os.getenv("VLLM_API_KEY") or "vllm-api-key"
+        base_url = self.config.vllm_base_url or os.getenv("VLLM_BASE_URL")
+        self.client = OpenAI(api_key=self.config.api_key, base_url=base_url)
+
+    def _parse_response(self, response, tools):
+        """
+        Process the response based on whether tools are used or not.
+
+        Args:
+            response: The raw response from API.
+            tools: The list of tools provided in the request.
+
+        Returns:
+            str or dict: The processed response.
+        """
+        if tools:
+            processed_response = {
+                "content": response.choices[0].message.content,
+                "tool_calls": [],
+            }
+
+            if response.choices[0].message.tool_calls:
+                for tool_call in response.choices[0].message.tool_calls:
+                    processed_response["tool_calls"].append(
+                        {
+                            "name": tool_call.function.name,
+                            "arguments": json.loads(extract_json(tool_call.function.arguments)),
+                        }
+                    )
+
+            return processed_response
+        else:
+            return response.choices[0].message.content
+
+    def generate_response(
+        self,
+        messages: List[Dict[str, str]],
+        response_format=None,
+        tools: Optional[List[Dict]] = None,
+        tool_choice: str = "auto",
+        **kwargs,
+    ):
+        """
+        Generate a response based on the given messages using vLLM.
+
+        Args:
+            messages (list): List of message dicts containing 'role' and 'content'.
+            response_format (str or object, optional): Format of the response. Defaults to "text".
+            tools (list, optional): List of tools that the model can call. Defaults to None.
+            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+            **kwargs: Additional vLLM-specific parameters.
+
+        Returns:
+            str: The generated response.
+        """
+        params = self._get_supported_params(messages=messages, **kwargs)
+        params.update(
+            {
+                "model": self.config.model,
+                "messages": messages,
+            }
+        )
+
+        if tools:
+            params["tools"] = tools
+            params["tool_choice"] = tool_choice
+
+        response = self.client.chat.completions.create(**params)
+        return self._parse_response(response, tools)

agentrun_mem0/llms/xai.py

@@ -0,0 +1,52 @@
+import os
+from typing import Dict, List, Optional
+
+from openai import OpenAI
+
+from agentrun_mem0.configs.llms.base import BaseLlmConfig
+from agentrun_mem0.llms.base import LLMBase
+
+
+class XAILLM(LLMBase):
+    def __init__(self, config: Optional[BaseLlmConfig] = None):
+        super().__init__(config)
+
+        if not self.config.model:
+            self.config.model = "grok-2-latest"
+
+        api_key = self.config.api_key or os.getenv("XAI_API_KEY")
+        base_url = self.config.xai_base_url or os.getenv("XAI_API_BASE") or "https://api.x.ai/v1"
+        self.client = OpenAI(api_key=api_key, base_url=base_url)
+
+    def generate_response(
+        self,
+        messages: List[Dict[str, str]],
+        response_format=None,
+        tools: Optional[List[Dict]] = None,
+        tool_choice: str = "auto",
+    ):
+        """
+        Generate a response based on the given messages using XAI.
+
+        Args:
+            messages (list): List of message dicts containing 'role' and 'content'.
+            response_format (str or object, optional): Format of the response. Defaults to "text".
+            tools (list, optional): List of tools that the model can call. Defaults to None.
+            tool_choice (str, optional): Tool choice method. Defaults to "auto".
+
+        Returns:
+            str: The generated response.
+        """
+        params = {
+            "model": self.config.model,
+            "messages": messages,
+            "temperature": self.config.temperature,
+            "max_tokens": self.config.max_tokens,
+            "top_p": self.config.top_p,
+        }
+
+        if response_format:
+            params["response_format"] = response_format
+
+        response = self.client.chat.completions.create(**params)
+        return response.choices[0].message.content