ESM & typescript (#22)

* move to ES

Author: duck4i

.gitignore

@@ -2,4 +2,5 @@ node_modules/
 build/
 example/
 model.gguf
-.vscode/
+.vscode/
+dist

README.md

@@ -46,7 +46,7 @@ sudo apt-get install -y build-essential cmake g++
 
 ```javascript
 
-const { RunInference } = require('@duck4i/llama');
+import { RunInference } = from "@duck4i/llama";
 
 const system_prompt = "The following is a conversation with an AI assistant. The assistant is helpful, creative, clever, and very friendly.";
 const user_prompt = "What is life expectancy of a duck?";
@@ -60,7 +60,7 @@ console.log("Answer", inference);
 It is likely you will want async functions for better memory management with multiple prompts, which is done like this:
 
 ```javascript
-const { LoadModelAsync, CreateContextAsync, RunInferenceAsync, ReleaseContextAsync, ReleaseModelAsync } = require('@duck4i/llama');
+import { LoadModelAsync, CreateContextAsync, RunInferenceAsync, ReleaseContextAsync, ReleaseModelAsync } = from "@duck4i/llama";
 
 const system_prompt = "The following is a conversation with an AI assistant. The assistant is helpful, creative, clever, and very friendly.";
 const prompts = [
@@ -118,7 +118,7 @@ You can control log levels coming from llamacpp like this:
 
 ```javascript
 
-const { SetLogLevel } = require('@duck4i/llama');
+import { SetLogLevel } = from '@duck4i/llama';
 
 // 0 - none, 1 - debug, 2 - info, 3 - warn, 4 - error
 SetLogLevel(1);

__tests__/basic.js renamed to __tests__/basic.ts

@@ -1,8 +1,9 @@
-const { execSync } = require('child_process');
-const fs = require('fs');
+import { execSync } from 'child_process';
+import { existsSync } from 'fs';
+import assert from 'assert';
 
-const { ChatManager, Role } = require('../chatManager');
-const {
+import { ChatManager, Role } from '../src/chat';
+import {
     RunInference,
     LoadModelAsync,
     CreateContextAsync,
@@ -11,83 +12,81 @@ const {
     ReleaseModelAsync,
     SetLogLevel,
     GetModelToken,
-} = require("bindings")("npm-llama");
+} from '../src/index';
+
 
 const model = "model.gguf";
 const modelUrl = "https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct-GGUF/resolve/main/qwen2.5-0.5b-instruct-fp16.gguf?download=true";
-const system_prompt = "The following is a conversation with an AI assistant. The assistant is helpful, creative, clever, and very friendly.";
-
+const systemPrompt = "The following is a conversation with an AI assistant. The assistant is helpful, creative, clever, and very friendly.";
 
-describe('Node LLaMA Test Suite', () => {
+describe("Llama tests - basic", () => {
 
     beforeAll(() => {
-
-        if (!fs.existsSync(model)) {
+        // Setup - Download model if needed
+        if (!existsSync(model)) {
             execSync(`npx llama-download -p ${model} -u ${modelUrl}`, { stdio: 'inherit' });
         } else {
             console.log("Model already downloaded");
         }
-    });
+    })
 
-    test('log level works', () => {
+    test('log level works', async () => {
         SetLogLevel(1); // debug logs
-        expect(true).toBeTruthy();
+        assert.ok(true);
     });
 
-    test('direct inference works', () => {
-        const inference = RunInference(model, "How old can ducks get?", system_prompt);
+    test('direct inference works', async () => {
+        const inference: string = RunInference(model, "How old can ducks get?", systemPrompt);
         console.log("Result", inference);
-        expect(inference.includes('10 years')).toBeTruthy();
+        assert.ok(inference.includes('10 years'));
     });
 
     test('async inference works', async () => {
-
-        const prompts = [
+        const prompts: string[] = [
             "How old can ducks get?",
             "Why are ducks so cool?",
             "Is there a limit on number of ducks I can own?"
-        ]
+        ];
 
         const modelHandle = await LoadModelAsync(model);
         const ctx = await CreateContextAsync(modelHandle);
         console.log("Model loaded", model);
 
         for (const prompt of prompts) {
-            const inference = await RunInferenceAsync(modelHandle, ctx, prompt, system_prompt, 64);
+            const inference: string = await RunInferenceAsync(modelHandle, ctx, prompt, systemPrompt, 64);
             console.log("Reply:", inference);
-            expect(inference.length > 0).toBeTruthy();
+            assert.ok(inference.length > 0);
         }
 
         await ReleaseContextAsync(ctx);
         await ReleaseModelAsync(modelHandle);
     });
 
     test('custom inference works', async () => {
-
         const user = "How old can ducks get?";
-        const prompt = `"!#<|im_start|>system ${system_prompt}<|im_end|><|im_start|>user ${user}<|im_end|><|im_start|>assistant"`;
+        const prompt = `"!#<|im_start|>system ${systemPrompt}<|im_end|><|im_start|>user ${user}<|im_end|><|im_start|>assistant"`;
 
         const modelHandle = await LoadModelAsync(model);
         const context = await CreateContextAsync(modelHandle);
-        const result = await RunInferenceAsync(modelHandle, context, prompt);
+        const result: string = await RunInferenceAsync(modelHandle, context, prompt);
         await ReleaseContextAsync(context);
         await ReleaseModelAsync(modelHandle);
 
         console.log("Result", result);
-        expect(result.length > 1).toBeTruthy();
+        assert.ok(result.length > 1);
     });
 
-    test('tokens work', async () => {
 
+    test('tokens work', async () => {
         const modelHandle = await LoadModelAsync(model);
         const ctx = await CreateContextAsync(modelHandle);
 
-        const eos = GetModelToken(modelHandle, "EOS");
-        const bos = GetModelToken(modelHandle, "BOS");
-        const eot = GetModelToken(modelHandle, "EOT");
-        const sep = GetModelToken(modelHandle, "SEP");
-        const cls = GetModelToken(modelHandle, "CLS");
-        const nl = GetModelToken(modelHandle, "NL");
+        const eos: string = GetModelToken(modelHandle, "EOS");
+        const bos: string = GetModelToken(modelHandle, "BOS");
+        const eot: string = GetModelToken(modelHandle, "EOT");
+        const sep: string = GetModelToken(modelHandle, "SEP");
+        const cls: string = GetModelToken(modelHandle, "CLS");
+        const nl: string = GetModelToken(modelHandle, "NL");
 
         console.log("EOS", eos);
         console.log("BOS", bos);
@@ -99,17 +98,17 @@ describe('Node LLaMA Test Suite', () => {
         await ReleaseContextAsync(ctx);
         await ReleaseModelAsync(modelHandle);
 
-        expect(eos.length > 1).toBeTruthy();
-        expect(bos.length > 1).toBeTruthy();
-    })
+        assert.ok(eos.length > 1);
+        assert.ok(bos.length > 1);
+    });
 
     test('chat works', async () => {
         SetLogLevel(4); // warn
 
         const modelHandle = await LoadModelAsync(model);
         const ctx = await CreateContextAsync(modelHandle);
 
-        const chat = new ChatManager(system_prompt);
+        const chat = new ChatManager(systemPrompt);
 
         let reply = "";
         let prompt = chat.getNextPrompt("Hello, my name is Duck!");
@@ -128,8 +127,6 @@ describe('Node LLaMA Test Suite', () => {
         await ReleaseContextAsync(ctx);
         await ReleaseModelAsync(modelHandle);
 
-        expect(reply.includes("Duck")).toBeTruthy();
+        assert.ok(reply.includes("Duck"));
     });
-
-    
 });

downloadModel.js

@@ -1,3 +1,9 @@
-module.exports = {
-    testTimeout: 480000, // 480 seconds (8m)
+/** @type {import('ts-jest').JestConfigWithTsJest} **/
+export default {
+  testEnvironment: "node",
+  transform: {
+    "^.+.tsx?$": ["ts-jest",{}],
+  },
+  extensionsToTreatAsEsm: ['.ts'],
+  testTimeout: 480000 // 480 seconds (8m)
 };