Merge pull request #3 from samestrin/1.0.1

1.0.1

Author: samestrin

.eslintrc.json

@@ -0,0 +1,11 @@
+{
+  "env": {
+    "node": true,
+    "commonjs": true,
+    "es2021": true
+  },
+  "extends": "eslint:recommended",
+  "parserOptions": {
+    "ecmaVersion": 12
+  }
+}

.github/workflows/main.yml

@@ -5,11 +5,11 @@ on:
     branches:
       - main
     paths:
-      - "package.json"
+      - 'package.json'
   workflow_dispatch:
   schedule:
     # Run the workflow every week (adjust as needed)
-    - cron: "0 0 * * 0"
+    - cron: '0 0 * * 0'
 
 jobs:
   update-packages:
@@ -21,7 +21,7 @@ jobs:
       - name: Setup Node.js
         uses: actions/setup-node@v3
         with:
-          node-version: "20"
+          node-version: '20'
 
       - name: Update NPM packages
         run: |

.prettierrc

@@ -0,0 +1,4 @@
+{
+  "singleQuote": true,
+  "trailingComma": "all"
+}

README.md

@@ -20,6 +20,12 @@ The LLM Interface project is a versatile and comprehensive wrapper designed to i
 
 ## Updates
 
+**v1.0.01**
+
+- **LLMInterfaceSendMessage**: Send a message to any LLM provider without creating a new instance of the `llm-interface`.
+- **Model Aliases**: Simplified model selection, `default`, `small`, and `large` model aliases now available.
+- **Major Refactor**: Improved comments, test cases, centralized LLM provider definitions.
+
 **v1.0.00**
 
 - **Initial 1.0 Release**
@@ -28,14 +34,6 @@ The LLM Interface project is a versatile and comprehensive wrapper designed to i
 
 - **Simple Prompt Handler**: Added support for simplified prompting.
 
-**v0.0.10**
-
-- **Hugging Face**: Added support for new LLM provider Hugging Face (_over 150,000 publicly accessible machine learning models_)
-- **Perplexity**: Added support for new LLM provider Perplexity
-- **AI21**: Add support for new LLM provider AI21 Studio
-- **JSON Output Improvements**: The `json_object` mode now guarantees the return a valid JSON object or null.
-- **Graceful Retries**: Retry LLM queries upon failure with progressive delays.
-
 ## Dependencies
 
 The project relies on several npm packages and APIs. Here are the primary dependencies:
@@ -64,13 +62,13 @@ npm install llm-interface
 Import `llm-interface` using:
 
 ```javascript
-const LLMInterface = require("llm-interface");
+const LLMInterface = require('llm-interface');
 ```
 
 or
 
 ```javascript
-import LLMInterface from "llm-interface";
+import LLMInterface from 'llm-interface';
 ```
 
 then call the handler you want to use:
@@ -79,10 +77,10 @@ then call the handler you want to use:
 const openai = new LLMInterface.openai(process.env.OPENAI_API_KEY);
 
 const message = {
-  model: "gpt-3.5-turbo",
+  model: 'gpt-3.5-turbo',
   messages: [
-    { role: "system", content: "You are a helpful assistant." },
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'system', content: 'You are a helpful assistant.' },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 
@@ -102,7 +100,7 @@ or if you want to keep things _simple_ you can use:
 const openai = new LLMInterface.openai(process.env.OPENAI_API_KEY);
 
 openai
-  .sendMessage("Explain the importance of low latency LLMs.")
+  .sendMessage('Explain the importance of low latency LLMs.')
   .then((response) => {
     console.log(response);
   })

docs/API.md

@@ -28,7 +28,7 @@ Different LLMs may have their own message object rules. For example, both Anthro
 
 ```javascript
 openai
-  .sendMessage(message, { max_tokens: 150, response_format: "json_object" })
+  .sendMessage(message, { max_tokens: 150, response_format: 'json_object' })
   .then((response) => {
     console.log(response);
   })
@@ -241,7 +241,7 @@ perplexity
 
 - **Parameters:**
   - `message`: An object containing the model and messages to send.
-  - `options`: An optional object containing `model`. This method currently has no token limitation.
+  - `options`: An optional object containing `max_tokens` and `model`.
   - `interfaceOptions`: An optional object specifying `cacheTimeoutSeconds` and `retryAttempts`.
 - **Returns:** A promise that resolves to the response text.
 - **Example:**

docs/USAGE.md

@@ -32,13 +32,13 @@ How to use `llm-interface` in your project.
 First, require the LLMInterface from the `llm-interface` package:
 
 ```javascript
-const LLMInterface = require("llm-interface");
+const LLMInterface = require('llm-interface');
 ```
 
 or import it:
 
 ```javascript
-import LLMInterface from "llm-interface";
+import LLMInterface from 'llm-interface';
 ```
 
 ## Basic Usage Examples
@@ -55,10 +55,10 @@ The OpenAI interface allows you to send messages to the OpenAI API.
 const openai = new LLMInterface.openai(process.env.OPENAI_API_KEY);
 
 const message = {
-  model: "gpt-3.5-turbo",
+  model: 'gpt-3.5-turbo',
   messages: [
-    { role: "system", content: "You are a helpful assistant." },
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'system', content: 'You are a helpful assistant.' },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 
@@ -169,10 +169,10 @@ The Gemini interface allows you to send messages to the Google Gemini API.
 const gemini = new LLMInterface.gemini(process.env.GEMINI_API_KEY);
 
 const message = {
-  model: "gemini-1.5-flash",
+  model: 'gemini-1.5-flash',
   messages: [
-    { role: "system", content: "You are a helpful assistant." },
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'system', content: 'You are a helpful assistant.' },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 
@@ -196,10 +196,10 @@ The Goose AI interface allows you to send messages to the Goose AI API.
 const goose = new LLMInterface.goose(process.env.GROQ_API_KEY);
 
 const message = {
-  model: "gpt-neo-20b",
+  model: 'gpt-neo-20b',
   messages: [
-    { role: "system", content: "You are a helpful assistant." },
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'system', content: 'You are a helpful assistant.' },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 
@@ -223,10 +223,10 @@ The Groq interface allows you to send messages to the Groq API.
 const groq = new LLMInterface.groq(process.env.GROQ_API_KEY);
 
 const message = {
-  model: "llama3-8b-8192",
+  model: 'llama3-8b-8192',
   messages: [
-    { role: "system", content: "You are a helpful assistant." },
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'system', content: 'You are a helpful assistant.' },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 
@@ -250,15 +250,15 @@ The HuggingFace interface allows you to send messages to the HuggingFace API.
 const huggingface = new LLMInterface.huggingface(process.env.ANTHROPIC_API_KEY);
 
 const message = {
-  model: "claude-3-opus-20240229",
+  model: 'claude-3-opus-20240229',
   messages: [
     {
-      role: "user",
+      role: 'user',
       content:
-        "You are a helpful assistant. Say OK if you understand and stop.",
+        'You are a helpful assistant. Say OK if you understand and stop.',
     },
-    { role: "system", content: "OK" },
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'system', content: 'OK' },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 
@@ -282,10 +282,10 @@ The Mistral AI interface allows you to send messages to the Mistral AI API.
 const mistral = new LLMInterface.mistral(process.env.GROQ_API_KEY);
 
 const message = {
-  model: "llama3-8b-8192",
+  model: 'llama3-8b-8192',
   messages: [
-    { role: "system", content: "You are a helpful assistant." },
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'system', content: 'You are a helpful assistant.' },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 
@@ -309,15 +309,15 @@ The Perplexity interface allows you to send messages to the Perplexity API.
 const perplexity = new LLMInterface.perplexity(process.env.ANTHROPIC_API_KEY);
 
 const message = {
-  model: "claude-3-opus-20240229",
+  model: 'claude-3-opus-20240229',
   messages: [
     {
-      role: "user",
+      role: 'user',
       content:
-        "You are a helpful assistant. Say OK if you understand and stop.",
+        'You are a helpful assistant. Say OK if you understand and stop.',
     },
-    { role: "system", content: "OK" },
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'system', content: 'OK' },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 
@@ -341,22 +341,22 @@ The Reka AI interface allows you to send messages to the Reka AI REST API.
 const reka = new LLMInterface.reka(process.env.REKA_API_KEY);
 
 const message = {
-  model: "reka-core",
+  model: 'reka-core',
   messages: [
     {
-      role: "user",
+      role: 'user',
       content:
-        "You are a helpful assistant. Say OK if you understand and stop.",
+        'You are a helpful assistant. Say OK if you understand and stop.',
     },
-    { role: "system", content: "OK" },
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'system', content: 'OK' },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 
 reka
   .sendMessage(message, {})
-  .then((response) => console.log("Response:", response))
-  .catch((error) => console.error("Error:", error));
+  .then((response) => console.log('Response:', response))
+  .catch((error) => console.error('Error:', error));
 ```
 
 ### LLaMA.cpp Interface
@@ -369,9 +369,9 @@ The LLaMA.cpp interface allows you to send messages to the LLaMA.cpp API; this i
 const llamacpp = new LLMInterface.llamacpp(process.env.LLAMACPP_URL);
 
 const message = {
-  model: "some-llamacpp-model",
+  model: 'some-llamacpp-model',
   messages: [
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 
@@ -398,7 +398,7 @@ This simplified example uses a string based prompt with the default OpenAI model
 ```javascript
 const openai = new LLMInterface.openai(process.env.OPENAI_API_KEY);
 
-const message = "Explain the importance of low latency LLMs.";
+const message = 'Explain the importance of low latency LLMs.';
 
 openai
   .sendMessage(message)
@@ -424,22 +424,22 @@ Some interfaces allows you request the response back in JSON, currently **OpenAI
 const openai = new LLMInterface.openai(process.env.OPENAI_API_KEY);
 
 const message = {
-  model: "gpt-3.5-turbo",
+  model: 'gpt-3.5-turbo',
   messages: [
     {
-      role: "system",
-      content: "You are a helpful assistant.",
+      role: 'system',
+      content: 'You are a helpful assistant.',
     },
     {
-      role: "user",
+      role: 'user',
       content:
-        "Explain the importance of low latency LLMs. Return the results as a JSON object. Follow this format: [{reason, reasonDescription}].",
+        'Explain the importance of low latency LLMs. Return the results as a JSON object. Follow this format: [{reason, reasonDescription}].',
     },
   ],
 };
 
 openai
-  .sendMessage(message, { max_tokens: 150, response_format: "json_object" })
+  .sendMessage(message, { max_tokens: 150, response_format: 'json_object' })
   .then((response) => {
     console.log(response);
   })
@@ -458,10 +458,10 @@ To reduce operational costs and improve performance you can optionally specify a
 const openai = new LLMInterface.openai(process.env.OPENAI_API_KEY);
 
 const message = {
-  model: "gpt-3.5-turbo",
+  model: 'gpt-3.5-turbo',
   messages: [
-    { role: "system", content: "You are a helpful assistant." },
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'system', content: 'You are a helpful assistant.' },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 
@@ -485,10 +485,10 @@ You can gracefully retry your requests. In this example we use OpenAI and up to
 const openai = new LLMInterface.openai(process.env.OPENAI_API_KEY);
 
 const message = {
-  model: "gpt-3.5-turbo",
+  model: 'gpt-3.5-turbo',
   messages: [
-    { role: "system", content: "You are a helpful assistant." },
-    { role: "user", content: "Explain the importance of low latency LLMs." },
+    { role: 'system', content: 'You are a helpful assistant.' },
+    { role: 'user', content: 'Explain the importance of low latency LLMs.' },
   ],
 };
 

eslint.config.mjs

@@ -0,0 +1,8 @@
+import globals from 'globals';
+import pluginJs from '@eslint/js';
+
+export default [
+  { files: ['**/*.js'], languageOptions: { sourceType: 'commonjs' } },
+  { languageOptions: { globals: globals.browser } },
+  pluginJs.configs.recommended,
+];