Added New LLM Cohere

src/beyondllm/llms/cohere.py

@@ -0,0 +1,73 @@
+from beyondllm.llms.base import BaseLLMModel, ModelConfig
+from typing import Any, Dict
+from dataclasses import dataclass, field
+import os
+import cohere
+
+@dataclass
+class CohereModel:
+    """
+    Class representing a Language Model (LLM) model using Cohere.
+
+    Example:
+    ```
+    >>> llm = CohereModel(api_key="<your_api_key>", model_kwargs={"temperature": 0.5})
+    ```
+    or
+    ```
+    >>> import os
+    >>> os.environ['COHERE_API_KEY'] = "***********" #replace with your key
+    >>> llm = CohereModel()
+    ```
+    """
+    api_key: str =" "
+    model_kwargs: dict = field(default_factory=lambda: {
+        "temperature": 0.5,
+        "top_p": 1,
+        "max_tokens": 2048,
+    })
+
+    def __post_init__(self):
+        if not self.api_key:
+            self.api_key = os.getenv('COHERE_API_KEY')
+            if not self.api_key:
+                raise ValueError("COHERE_API_KEY is not provided and not found in environment variables.")
+        self.load_llm()
+
+    def load_llm(self):
+        """Load the Cohere client."""
+        try:
+            self.client = cohere.ClientV2(api_key=self.api_key)
+        except Exception as e:
+            raise Exception(f"Failed to initialize Cohere client: {str(e)}")
+
+    def predict(self, prompt: Any) -> str:
+        try:
+            response = self.client.chat(
+                model="command-r-plus-08-2024",
+                messages=[{"role": "user", "content": prompt}]
+            )
+            return response.message.content[0].text
+        except Exception as e:
+            raise Exception(f"Failed to generate prediction: {str(e)}")
+
+    @staticmethod
+    def load_from_kwargs(self, kwargs: Dict):
+        model_config = ModelConfig(**kwargs)
+        self.config = model_config
+        self.load_llm()
+
+if __name__ == "__main__":
+    import os
+
+    # set the API key in an environment variable
+    os.environ['COHERE_API_KEY'] = " "
+
+    # Create an instance of CohereModel
+    llm = CohereModel()
+
+    # Make a prediction
+    prompt = "Write a Linkedin post on generative AI using emojis and symbols?"
+    response = llm.predict(prompt)
+
+    print(f"Response: {response}")