c

practice.ipynb

@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": 2,
    "id": "259a585b",
    "metadata": {},
    "outputs": [],
@@ -30,7 +30,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 3,
    "id": "fe7bf6eb",
    "metadata": {},
    "outputs": [
@@ -58,7 +58,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 4,
    "id": "4bddedb8",
    "metadata": {},
    "outputs": [
@@ -95,18 +95,70 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 5,
    "id": "b33dcb00",
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "65\n",
+      "\n",
+      " !$&',-.3:;?ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz\n"
+     ]
+    }
+   ],
    "source": [
-    "set(text)"
+    "chars = sorted(list(set(text)))\n",
+    "vocab_size = len(chars)\n",
+    "print(vocab_size\n",
+    ")\n",
+    "print(\"\".join(chars\n",
+    "))\n",
+    "\n"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "7c84058a",
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "'!'"
+      ]
+     },
+     "execution_count": 11,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "stoi = {char:i for i, char in enumerate(chars)}\n",
+    "itos = {i:char for i, char in enumerate(chars)}\n",
+    "# print (itos)\n",
+    "\n",
+    "stoi['a'] # 0\n",
+    "itos[2]\n",
+    "\n",
+    "encode = lambda s: \n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "4f43742c",
+   "metadata": {},
+   "outputs": [],
+   "source": []
   }
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": ".venv",
+   "display_name": ".venv (3.12.10)",
    "language": "python",
    "name": "python3"
   },