| Title: | Interface for Large Language Model APIs in R |
|---|---|
| Description: | A unified interface to interact with various Large Language Model (LLM) APIs such as 'OpenAI' (see <https://platform.openai.com/docs/quickstart> for details), 'Anthropic' (see <https://docs.anthropic.com/en/api/getting-started> for details), 'Groq' (see <https://console.groq.com/docs/api-reference> for details), and 'Together AI' (see <https://docs.together.ai/docs/quickstart> for details). Allows users to configure API parameters, send messages, and retrieve responses seamlessly within R. |
| Authors: | Ali Sanaei [aut, cre] |
| Maintainer: | Ali Sanaei <[email protected]> |
| License: | MIT + file LICENSE |
| Version: | 0.1.1 |
| Built: | 2024-10-22 07:20:56 UTC |
| Source: | CRAN |
Sends a message to the specified LLM API and retrieves the response.
call_llm(config, messages, verbose = FALSE, json = FALSE)call_llm(config, messages, verbose = FALSE, json = FALSE)
config |
An 'llm_config' object created by 'llm_config()'. |
messages |
A list of message objects to send to the API. |
verbose |
Logical. If 'TRUE', prints the full API response. |
json |
Logical. If 'TRUE', returns the raw JSON response as an attribute. |
The generated text response with additional attributes based on parameters.
## Not run: # ----- Groq Example ----- groq_config <- llm_config( provider = "groq", model = "llama-3.1-8b-instant", api_key = Sys.getenv("GROQ_KEY"), temperature = 0.7, max_tokens = 500 ) # Define the message with a system prompt message <- list( list(role = "system", content = "You ONLY fill in the blank. Do NOT answer in full sentences."), list(role = "user", content = "What's the capital of France? ----") ) # Call the LLM response <- call_llm(groq_config, message) # Display the response cat("Groq Response:", response, "\n") # Extract and print the full API response full.response <- attr(response, which = 'full_response') print(full.response) # ----- OpenAI Example with More Parameters ----- # Create a configuration with more parameters comprehensive_openai_config <- llm_config( provider = "openai", model = "gpt-4o-mini", api_key = Sys.getenv("OPENAI_KEY"), temperature = 1, # Controls the randomness of the output max_tokens = 750, # Maximum number of tokens to generate top_p = 1, # Nucleus sampling parameter frequency_penalty = 0.5, # Penalizes new tokens based on their frequency presence_penalty = 0.3 # Penalizes new tokens based on their presence ) # Define a more complex message comprehensive_message <- list( list(role = "system", content = "You are an expert data scientist."), list(role = "user", content = "When will you ever use OLS?") ) # Call the LLM with all parameters comprehensive_response <- call_llm( config = comprehensive_openai_config, messages = comprehensive_message, json = TRUE # Retrieve the raw JSON response as an attribute ) # Display the generated text response cat("Comprehensive OpenAI Response:", comprehensive_response, "\n") # Access and print the raw JSON response raw_json_response <- attr(comprehensive_response, "raw_json") print(raw_json_response) ## End(Not run)## Not run: # ----- Groq Example ----- groq_config <- llm_config( provider = "groq", model = "llama-3.1-8b-instant", api_key = Sys.getenv("GROQ_KEY"), temperature = 0.7, max_tokens = 500 ) # Define the message with a system prompt message <- list( list(role = "system", content = "You ONLY fill in the blank. Do NOT answer in full sentences."), list(role = "user", content = "What's the capital of France? ----") ) # Call the LLM response <- call_llm(groq_config, message) # Display the response cat("Groq Response:", response, "\n") # Extract and print the full API response full.response <- attr(response, which = 'full_response') print(full.response) # ----- OpenAI Example with More Parameters ----- # Create a configuration with more parameters comprehensive_openai_config <- llm_config( provider = "openai", model = "gpt-4o-mini", api_key = Sys.getenv("OPENAI_KEY"), temperature = 1, # Controls the randomness of the output max_tokens = 750, # Maximum number of tokens to generate top_p = 1, # Nucleus sampling parameter frequency_penalty = 0.5, # Penalizes new tokens based on their frequency presence_penalty = 0.3 # Penalizes new tokens based on their presence ) # Define a more complex message comprehensive_message <- list( list(role = "system", content = "You are an expert data scientist."), list(role = "user", content = "When will you ever use OLS?") ) # Call the LLM with all parameters comprehensive_response <- call_llm( config = comprehensive_openai_config, messages = comprehensive_message, json = TRUE # Retrieve the raw JSON response as an attribute ) # Display the generated text response cat("Comprehensive OpenAI Response:", comprehensive_response, "\n") # Access and print the raw JSON response raw_json_response <- attr(comprehensive_response, "raw_json") print(raw_json_response) ## End(Not run)
Creates a configuration object for interacting with a specified LLM API provider.
llm_config(provider, model, api_key, ...)llm_config(provider, model, api_key, ...)
provider |
A string specifying the API provider. Supported providers include: "openai" for OpenAI, "anthropic" for Anthropic, "groq" for Groq, "together" for Together AI; This configuration is extendable; to add support for additional providers, define a new S3 method for 'call_llm' corresponding to the provider. |
model |
The model name to use. This depends on the provider. For example: OpenAI: "gpt-4o-mini", Anthropic: "claude-3-opus-20240229", Groq: "llama-3.1-8b-instant", Together AI: "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo". |
api_key |
Your API key for the provider. |
... |
Additional model-specific parameters (e.g., 'temperature', 'max_tokens'). |
An object of class 'llm_config' containing API and model parameters.
## Not run: # Obtain API keys from: # OpenAI: https://platform.openai.com/api-keys # Groq: https://console.groq.com/keys # Anthropic: https://console.anthropic.com/settings/keys # Together AI: https://api.together.ai/settings/api-keys # ----- OpenAI Example ----- openai_config <- llm_config( provider = "openai", model = "gpt-4o-mini", api_key = OPENAI_KEY, temperature = 0.7, max_tokens = 500 ) # ----- Anthropic Example ----- anthropic_config <- llm_config( provider = "anthropic", model = "claude-3-opus-20240229", api_key = ANTHROPIC_API_KEY, max_tokens = 500 ) # ----- Groq Example ----- groq_config <- llm_config( provider = "groq", model = "llama-3.1-8b-instant", api_key = GROQ_API_KEY, temperature = 0.3, max_tokens = 1000 ) # ----- Together AI Example ----- together_config <- llm_config( provider = "together", model = "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo", api_key = TOGETHER_KEY, temperature = 0.5, max_tokens = 1000 ) # This configuration is extendable by defining new `call_llm` methods for additional providers. ## End(Not run)## Not run: # Obtain API keys from: # OpenAI: https://platform.openai.com/api-keys # Groq: https://console.groq.com/keys # Anthropic: https://console.anthropic.com/settings/keys # Together AI: https://api.together.ai/settings/api-keys # ----- OpenAI Example ----- openai_config <- llm_config( provider = "openai", model = "gpt-4o-mini", api_key = OPENAI_KEY, temperature = 0.7, max_tokens = 500 ) # ----- Anthropic Example ----- anthropic_config <- llm_config( provider = "anthropic", model = "claude-3-opus-20240229", api_key = ANTHROPIC_API_KEY, max_tokens = 500 ) # ----- Groq Example ----- groq_config <- llm_config( provider = "groq", model = "llama-3.1-8b-instant", api_key = GROQ_API_KEY, temperature = 0.3, max_tokens = 1000 ) # ----- Together AI Example ----- together_config <- llm_config( provider = "together", model = "meta-llama/Meta-Llama-3.1-8B-Instruct-Turbo", api_key = TOGETHER_KEY, temperature = 0.5, max_tokens = 1000 ) # This configuration is extendable by defining new `call_llm` methods for additional providers. ## End(Not run)