From d8eeb1a736fac122134172800db29b7c9a3f4da3 Mon Sep 17 00:00:00 2001
From: Travis Fischer <fisch0920@gmail.com>
Date: Sun, 19 Feb 2023 03:48:06 -0600
Subject: [PATCH] feat: switch from gpt-3-encoder to gpt3-tokenizer

---
 package.json       |  2 +-
 pnpm-lock.yaml     | 15 +++++++++++----
 src/chatgpt-api.ts |  4 ++--
 src/tokenizer.ts   | 12 ++++++++++++
 4 files changed, 26 insertions(+), 7 deletions(-)
 create mode 100644 src/tokenizer.ts

diff --git a/package.json b/package.json
index e04a532..6db3d40 100644
--- a/package.json
+++ b/package.json
@@ -36,7 +36,7 @@
   },
   "dependencies": {
     "eventsource-parser": "^0.0.5",
-    "gpt-3-encoder": "^1.1.4",
+    "gpt3-tokenizer": "^1.1.5",
     "keyv": "^4.5.2",
     "p-timeout": "^6.0.0",
     "quick-lru": "^6.1.1",
diff --git a/pnpm-lock.yaml b/pnpm-lock.yaml
index 70c1242..8a65a24 100644
--- a/pnpm-lock.yaml
+++ b/pnpm-lock.yaml
@@ -8,7 +8,7 @@ specifiers:
   del-cli: ^5.0.0
   dotenv-safe: ^8.2.0
   eventsource-parser: ^0.0.5
-  gpt-3-encoder: ^1.1.4
+  gpt3-tokenizer: ^1.1.5
   husky: ^8.0.2
   keyv: ^4.5.2
   lint-staged: ^13.0.3
@@ -26,7 +26,7 @@ specifiers:
 
 dependencies:
   eventsource-parser: 0.0.5
-  gpt-3-encoder: 1.1.4
+  gpt3-tokenizer: 1.1.5
   keyv: 4.5.2
   p-timeout: 6.1.0
   quick-lru: 6.1.1
@@ -506,6 +506,10 @@ packages:
       picomatch: 2.3.1
     dev: true
 
+  /array-keyed-map/2.1.3:
+    resolution: {integrity: sha512-JIUwuFakO+jHjxyp4YgSiKXSZeC0U+R1jR94bXWBcVlFRBycqXlb+kH9JHxBGcxnVuSqx5bnn0Qz9xtSeKOjiA==}
+    dev: false
+
   /array-union/2.1.0:
     resolution: {integrity: sha512-HGyxoOTYUyCM6stUe6EJgnd4EoewAI7zMdfqO+kGjnlZmBDz/cR5pf8r/cR4Wq60sL/p0IkcjUEEPwS3GFrIyw==}
     engines: {node: '>=8'}
@@ -1376,8 +1380,11 @@ packages:
       get-intrinsic: 1.2.0
     dev: true
 
-  /gpt-3-encoder/1.1.4:
-    resolution: {integrity: sha512-fSQRePV+HUAhCn7+7HL7lNIXNm6eaFWFbNLOOGtmSJ0qJycyQvj60OvRlH7mee8xAMjBDNRdMXlMwjAbMTDjkg==}
+  /gpt3-tokenizer/1.1.5:
+    resolution: {integrity: sha512-O9iCL8MqGR0Oe9wTh0YftzIbysypNQmS5a5JG3cB3M4LMYjlAVvNnf8LUzVY9MrI7tj+YLY356uHtO2lLX2HpA==}
+    engines: {node: '>=12'}
+    dependencies:
+      array-keyed-map: 2.1.3
     dev: false
 
   /graceful-fs/4.2.10:
diff --git a/src/chatgpt-api.ts b/src/chatgpt-api.ts
index 1bd32cf..f13fd30 100644
--- a/src/chatgpt-api.ts
+++ b/src/chatgpt-api.ts
@@ -1,9 +1,9 @@
-import { encode as gptEncode } from 'gpt-3-encoder'
 import Keyv from 'keyv'
 import pTimeout from 'p-timeout'
 import QuickLRU from 'quick-lru'
 import { v4 as uuidv4 } from 'uuid'
 
+import * as tokenizer from './tokenizer'
 import * as types from './types'
 import { fetch as globalFetch } from './fetch'
 import { fetchSSE } from './fetch-sse'
@@ -438,7 +438,7 @@ Current date: ${currentDate}${this._sepToken}\n\n`
       text = text.replace(/<\|im_sep\|>/g, '<|endoftext|>')
     }
 
-    return gptEncode(text).length
+    return tokenizer.encode(text).length
   }
 
   protected get _isChatGPTModel() {
diff --git a/src/tokenizer.ts b/src/tokenizer.ts
new file mode 100644
index 0000000..81209ed
--- /dev/null
+++ b/src/tokenizer.ts
@@ -0,0 +1,12 @@
+import GPT3TokenizerImport from 'gpt3-tokenizer'
+
+const GPT3Tokenizer: typeof GPT3TokenizerImport =
+  typeof GPT3TokenizerImport === 'function'
+    ? GPT3TokenizerImport
+    : (GPT3TokenizerImport as any).default
+
+export const tokenizer = new GPT3Tokenizer({ type: 'gpt3' })
+
+export function encode(input: string): number[] {
+  return tokenizer.encode(input).bpe
+}