File size: 6,081 Bytes
75264b6 9a0f06e 75264b6 9a0f06e 75264b6 |
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 |
<!DOCTYPE html>
<html>
<head>
<meta charset="UTF-8"/>
<meta name="viewport" content="width=device-width, initial-scale=1.0"/>
<script src="https://cdn.tailwindcss.com"></script>
<!-- polyfill for firefox + import maps -->
<script src="https://unpkg.com/[email protected]/dist/es-module-shims.js"></script>
<script type="importmap">
{
"imports": {
"@huggingface/inference": "https://cdn.jsdelivr.net/npm/@huggingface/[email protected]/+esm"
}
}
</script>
</head>
<body>
<form class="w-[90%] mx-auto pt-8" onsubmit="launch(); return false;">
<h1 class="text-3xl font-bold">
<span
class="bg-clip-text text-transparent bg-gradient-to-r from-pink-500 to-violet-500"
>
Document & visual question answering demo with
<a href="https://github.com/huggingface/huggingface.js">
<kbd>@huggingface/inference</kbd>
</a>
</span>
</h1>
<p class="mt-8">
First, input your token if you have one! Otherwise, you may encounter
rate limiting. You can create a token for free at
<a
target="_blank"
href="https://huggingface.co/settings/tokens"
class="underline text-blue-500"
>hf.co/settings/tokens</a
>
</p>
<input
type="text"
id="token"
class="rounded border-2 border-blue-500 shadow-md px-3 py-2 w-96 mt-6"
placeholder="token (optional)"
/>
<p class="mt-8">
Pick the model type and the model you want to run. Check out models for
<a
href="https://huggingface.co/tasks/document-question-answering"
class="underline text-blue-500"
target="_blank"
>
document</a
> and
<a
href="https://huggingface.co/tasks/visual-question-answering"
class="underline text-blue-500"
target="_blank"
>image</a> question answering.
</p>
<div class="space-x-2 flex text-sm mt-8">
<label>
<input class="sr-only peer" name="type" type="radio" value="document" onclick="update_model(this.value)" checked />
<div class="px-3 py-3 rounded-lg shadow-md flex items-center justify-center text-slate-700 bg-gradient-to-r peer-checked:font-semibold peer-checked:from-pink-500 peer-checked:to-violet-500 peer-checked:text-white">
Document
</div>
</label>
<label>
<input class="sr-only peer" name="type" type="radio" value="image" onclick="update_model(this.value)" />
<div class="px-3 py-3 rounded-lg shadow-md flex items-center justify-center text-slate-700 bg-gradient-to-r peer-checked:font-semibold peer-checked:from-pink-500 peer-checked:to-violet-500 peer-checked:text-white">
Image
</div>
</label>
</div>
<input
id="model"
class="rounded border-2 border-blue-500 shadow-md px-3 py-2 w-96 mt-6"
value="impira/layoutlm-document-qa"
required
/>
<p class="mt-8">The input image</p>
<input type="file" required accept="image/*"
class="rounded border-blue-500 shadow-md px-3 py-2 w-96 mt-6 block"
rows="5"
id="image"
/>
<p class="mt-8">The question</p>
<input
type="text"
id="question"
class="rounded border-2 border-blue-500 shadow-md px-3 py-2 w-96 mt-6"
required
/>
<button
id="submit"
class="my-8 bg-green-500 rounded py-3 px-5 text-white shadow-md disabled:bg-slate-300"
>
Run
</button>
<p class="text-gray-400 text-sm">Output logs</p>
<div id="logs" class="bg-gray-100 rounded p-3 mb-8 text-sm">
Output will be here
</div>
<p>Check out the <a class="underline text-blue-500"
href="https://huggingface.co/spaces/huggingfacejs/doc-vis-qa/blob/main/index.html"
target="_blank">source code</a></p>
</form>
<script type="module">
import {HfInference} from "@huggingface/inference";
const default_models = {
"document": "impira/layoutlm-document-qa",
"image": "dandelin/vilt-b32-finetuned-vqa",
};
let running = false;
async function launch() {
if (running) {
return;
}
running = true;
try {
const hf = new HfInference(
document.getElementById("token").value.trim() || undefined
);
const model = document.getElementById("model").value.trim();
const model_type = document.querySelector("[name=type]:checked").value;
const image = document.getElementById("image").files[0];
const question = document.getElementById("question").value.trim();
document.getElementById("logs").textContent = "";
const method = model_type === "document" ? hf.documentQuestionAnswering : hf.visualQuestionAnswering;
const result = await method({model, inputs: { //{answer, score, start, end}
image, question
}});
document.getElementById("logs").textContent = JSON.stringify(result, null, 2); //answer + ": " + score + start + end;
} catch (err) {
alert("Error: " + err.message);
} finally {
running = false;
}
}
window.launch = launch;
window.update_model = (model_type) => {
const model_input = document.getElementById("model");
const cur_model = model_input.value.trim();
let new_model = "";
if (
model_type === "document" && cur_model === default_models["image"]
|| model_type === "image" && cur_model === default_models["document"]
|| cur_model === ""
) {
new_model = default_models[model_type];
}
model_input.value = new_model;
};
</script>
</body>
</html> |