{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# test our tokenizer on a simple sentence\n",
    "tokens = tokenizer.encode('ciao, come va?')  # 'hi, how are you?'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "Encoding(num_tokens=512, attributes=[ids, type_ids, tokens, offsets, attention_mask, special_tokens_mask, overflowing])"
      ]
     },
     "metadata": {},
     "execution_count": 6
    }
   ],
   "source": [
    "tokens  # this is our encodings object, with several tensors including ids and attention_mask"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "['[CLS]', 'ciao', ',', 'Ġcome', 'Ġva', '?', '[SEP]', '[PAD]', '[PAD]', '[PAD]']"
      ]
     },
     "metadata": {},
     "execution_count": 7
    }
   ],
   "source": [
    "tokens.tokens[:10]  # we can view the tokens here (eg output of merges.txt)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "output_type": "execute_result",
     "data": {
      "text/plain": [
       "[1, 16834, 16, 488, 611, 35, 2, 0, 0, 0]"
      ]
     },
     "metadata": {},
     "execution_count": 8
    }
   ],
   "source": [
    "tokens.ids[:10]  # and here are the token ids (output of vocab.json)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "ML",
   "language": "python",
   "name": "ml"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}