dnth commited on
Commit
d3bb4ae
1 Parent(s): 51ee882

Upload caption_map.ipynb

Browse files
Files changed (1) hide show
  1. caption_map.ipynb +154 -0
caption_map.ipynb ADDED
@@ -0,0 +1,154 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cells": [
3
+ {
4
+ "cell_type": "code",
5
+ "execution_count": 1,
6
+ "metadata": {},
7
+ "outputs": [
8
+ {
9
+ "data": {
10
+ "application/vnd.jupyter.widget-view+json": {
11
+ "model_id": "0dd1298de3c84ea3ab8ed31b2a0b2888",
12
+ "version_major": 2,
13
+ "version_minor": 0
14
+ },
15
+ "text/plain": [
16
+ "Loading checkpoint shards: 0%| | 0/2 [00:00<?, ?it/s]"
17
+ ]
18
+ },
19
+ "metadata": {},
20
+ "output_type": "display_data"
21
+ }
22
+ ],
23
+ "source": [
24
+ "import torch\n",
25
+ "from multiprocessing import set_start_method\n",
26
+ "from transformers import Blip2Processor, Blip2ForConditionalGeneration\n",
27
+ "from datasets import load_dataset\n",
28
+ "\n",
29
+ "# Load BLIP-2 model and processor\n",
30
+ "processor = Blip2Processor.from_pretrained(\"Salesforce/blip2-opt-2.7b\")\n",
31
+ "model = Blip2ForConditionalGeneration.from_pretrained(\"Salesforce/blip2-opt-2.7b\", torch_dtype=torch.float16)"
32
+ ]
33
+ },
34
+ {
35
+ "cell_type": "code",
36
+ "execution_count": 11,
37
+ "metadata": {},
38
+ "outputs": [],
39
+ "source": [
40
+ "def gpu_computation(batch, rank):\n",
41
+ " device = f\"cuda:{(rank or 0) % torch.cuda.device_count()}\"\n",
42
+ " model.to(device)\n",
43
+ " inputs = processor(images=batch[\"image\"], return_tensors=\"pt\").to(device, torch.float16)\n",
44
+ "\n",
45
+ " with torch.no_grad():\n",
46
+ " generated_ids = model.generate(**inputs, max_length=51)\n",
47
+ " \n",
48
+ " batch[\"caption\"] = processor.batch_decode(generated_ids, skip_special_tokens=True)\n",
49
+ " return batch"
50
+ ]
51
+ },
52
+ {
53
+ "cell_type": "code",
54
+ "execution_count": 12,
55
+ "metadata": {},
56
+ "outputs": [
57
+ {
58
+ "data": {
59
+ "application/vnd.jupyter.widget-view+json": {
60
+ "model_id": "61fe62d696904a7c894bd2c6f082b426",
61
+ "version_major": 2,
62
+ "version_minor": 0
63
+ },
64
+ "text/plain": [
65
+ "Map: 0%| | 0/10 [00:00<?, ? examples/s]"
66
+ ]
67
+ },
68
+ "metadata": {},
69
+ "output_type": "display_data"
70
+ }
71
+ ],
72
+ "source": [
73
+ "import multiprocessing\n",
74
+ "\n",
75
+ "\n",
76
+ "if __name__ == \"__main__\":\n",
77
+ " # Check if start method is already set\n",
78
+ " try:\n",
79
+ " multiprocessing.get_start_method()\n",
80
+ " except RuntimeError:\n",
81
+ " multiprocessing.set_start_method(\"spawn\")\n",
82
+ "\n",
83
+ " # Load your dataset\n",
84
+ " dataset = load_dataset(\"visual-layer/oxford-iiit-pet-vl-enriched\", split=\"train\")\n",
85
+ " dataset = dataset.select(range(10))\n",
86
+ "\n",
87
+ " updated_dataset = dataset.map(\n",
88
+ " gpu_computation,\n",
89
+ " batched=True,\n",
90
+ " batch_size=4, # Adjust based on your GPU memory\n",
91
+ " with_rank=True,\n",
92
+ " num_proc=torch.cuda.device_count(), # one process per GPU\n",
93
+ " )"
94
+ ]
95
+ },
96
+ {
97
+ "cell_type": "code",
98
+ "execution_count": 13,
99
+ "metadata": {},
100
+ "outputs": [
101
+ {
102
+ "data": {
103
+ "text/plain": [
104
+ "['a cat walking on grass\\n',\n",
105
+ " 'a white dog playing with a ball\\n',\n",
106
+ " 'a dog sitting in the grass\\n',\n",
107
+ " 'a dog laying in the grass\\n',\n",
108
+ " 'a dog standing in the snow\\n',\n",
109
+ " 'a dog laying in the grass\\n',\n",
110
+ " 'a dog laying on a brick sidewalk\\n',\n",
111
+ " 'a man holding a black dog\\n',\n",
112
+ " 'a large dog standing in the grass\\n',\n",
113
+ " 'a pug dog with its tongue out standing on a tiled floor\\n']"
114
+ ]
115
+ },
116
+ "execution_count": 13,
117
+ "metadata": {},
118
+ "output_type": "execute_result"
119
+ }
120
+ ],
121
+ "source": [
122
+ "updated_dataset['caption']"
123
+ ]
124
+ },
125
+ {
126
+ "cell_type": "code",
127
+ "execution_count": null,
128
+ "metadata": {},
129
+ "outputs": [],
130
+ "source": []
131
+ }
132
+ ],
133
+ "metadata": {
134
+ "kernelspec": {
135
+ "display_name": "Python 3",
136
+ "language": "python",
137
+ "name": "python3"
138
+ },
139
+ "language_info": {
140
+ "codemirror_mode": {
141
+ "name": "ipython",
142
+ "version": 3
143
+ },
144
+ "file_extension": ".py",
145
+ "mimetype": "text/x-python",
146
+ "name": "python",
147
+ "nbconvert_exporter": "python",
148
+ "pygments_lexer": "ipython3",
149
+ "version": "3.10.12"
150
+ }
151
+ },
152
+ "nbformat": 4,
153
+ "nbformat_minor": 2
154
+ }