File size: 6,312 Bytes
6fe7180 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 |
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"id": "initial_id",
"metadata": {
"collapsed": true,
"ExecuteTime": {
"end_time": "2024-01-12T16:31:17.690349522Z",
"start_time": "2024-01-12T16:31:15.472874479Z"
}
},
"outputs": [],
"source": [
"import torch"
]
},
{
"cell_type": "code",
"execution_count": 2,
"outputs": [],
"source": [
"%reload_ext autoreload\n",
"%autoreload 2"
],
"metadata": {
"collapsed": false,
"ExecuteTime": {
"end_time": "2024-01-12T16:31:17.717430741Z",
"start_time": "2024-01-12T16:31:17.695066680Z"
}
},
"id": "ecefdad828c7daa3"
},
{
"cell_type": "code",
"execution_count": 3,
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Some weights of BartForConditionalGeneration were not initialized from the model checkpoint at facebook/bart-large-cnn and are newly initialized: ['model.shared.weight']\n",
"You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n"
]
}
],
"source": [
"\n",
"from transformers import AutoTokenizer, BartForConditionalGeneration\n",
"\n",
"tokenizer = AutoTokenizer.from_pretrained(\"facebook/bart-large-cnn\")\n",
"model = BartForConditionalGeneration.from_pretrained(\"facebook/bart-large-cnn\")\n"
],
"metadata": {
"collapsed": false,
"ExecuteTime": {
"end_time": "2024-01-12T16:31:26.058437142Z",
"start_time": "2024-01-12T16:31:17.720106168Z"
}
},
"id": "8c32b182fbcac2b6"
},
{
"cell_type": "code",
"execution_count": 4,
"outputs": [],
"source": [
"from rsasumm.beam_search import RSAContextualDecoding"
],
"metadata": {
"collapsed": false,
"ExecuteTime": {
"end_time": "2024-01-12T16:31:26.097766981Z",
"start_time": "2024-01-12T16:31:26.056626187Z"
}
},
"id": "cb33d902fe736c25"
},
{
"cell_type": "code",
"execution_count": 5,
"outputs": [],
"source": [
"\n",
"\n",
"texts = ['The paper gives really interesting insights on the topic of transfer learning. It is well presented and the experiment are extensive. I believe the authors missed Jane and al 2021. In addition, I think, there is a mistake in the math.',\n",
" 'The paper gives really interesting insights on the topic of transfer learning. It is well presented and the experiment are extensive. However, some parts remain really unclear and I would like to see a more detailed explanation of the proposed method.',\n",
" 'The paper gives really interesting insights on the topic of transfer learning. It is not well presented and lack experiments. In addition, some parts remain really unclear and I would like to see a more detailed explanation of the proposed method.'\n",
" ]\n",
"\n",
"# texts = [texts[2], texts[1], texts[0]]\n",
"\n"
],
"metadata": {
"collapsed": false,
"ExecuteTime": {
"end_time": "2024-01-12T16:31:26.127922110Z",
"start_time": "2024-01-12T16:31:26.098805312Z"
}
},
"id": "436ef1482c361159"
},
{
"cell_type": "code",
"execution_count": 6,
"outputs": [],
"source": [
"source_texts = tokenizer(texts, return_tensors=\"pt\", padding=True)\n",
"\n",
"rsa = RSAContextualDecoding(model, tokenizer, 'cpu')\n",
"\n"
],
"metadata": {
"collapsed": false,
"ExecuteTime": {
"end_time": "2024-01-12T16:31:26.169520864Z",
"start_time": "2024-01-12T16:31:26.125283164Z"
}
},
"id": "84b9943cac6cd7b2"
},
{
"cell_type": "code",
"execution_count": 7,
"outputs": [],
"source": [
"output = rsa.generate(target_id=1, source_texts_ids=source_texts.input_ids, source_text_attention_mask=source_texts.attention_mask, max_length=50, top_p=0.95, do_sample=True, rationality=8.0, temperature=1.0, process_logits_before_rsa=True)"
],
"metadata": {
"collapsed": false,
"ExecuteTime": {
"end_time": "2024-01-12T16:32:14.857034731Z",
"start_time": "2024-01-12T16:31:26.164578792Z"
}
},
"id": "620e54a63dd2099c"
},
{
"cell_type": "code",
"execution_count": 8,
"outputs": [
{
"data": {
"text/plain": "['Some parts of the paper remain unclear. I would like to see a more detailed explanation of the proposed method.',\n 'Some parts of the paper remain unclear. I would like to see a more detailed explanation of the proposed method.',\n 'Some parts of the paper remain unclear. I would like to see a more detailed explanation of the proposed method.',\n 'Some parts of the paper remain unclear. I would like to see a more detailed explanation of the proposed method.',\n 'Some parts of the paper remain unclear. I would like to see a more detailed explanation of the proposed method.',\n 'Some parts of the paper remain unclear. I would like to see a more detailed explanation of the proposed method.',\n 'Some parts of the paper remain unclear. I would like to see a more detailed explanation of the proposed method.',\n 'Some parts of the paper remain unclear. I would like to see a more detailed explanation of the proposed method.']"
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"\n",
"tokenizer.batch_decode(output[0], skip_special_tokens=True)\n",
"\n"
],
"metadata": {
"collapsed": false,
"ExecuteTime": {
"end_time": "2024-01-12T16:32:14.858531480Z",
"start_time": "2024-01-12T16:32:14.856763396Z"
}
},
"id": "fb3a5a9a8f9990ee"
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 3",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 2
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython2",
"version": "2.7.6"
}
},
"nbformat": 4,
"nbformat_minor": 5
}
|