Spaces:
Runtime error
Runtime error
Victoria Slocum
commited on
Commit
·
2d2033e
1
Parent(s):
df8e4de
fix: scrollbar, feat: download depen
Browse files- app.py +77 -22
- scrollbar.css +16 -7
app.py
CHANGED
|
@@ -1,9 +1,13 @@
|
|
|
|
|
| 1 |
import spacy
|
| 2 |
from spacy import displacy
|
| 3 |
import random
|
| 4 |
from spacy.tokens import Span
|
| 5 |
import gradio as gr
|
| 6 |
import pandas as pd
|
|
|
|
|
|
|
|
|
|
| 7 |
|
| 8 |
DEFAULT_MODEL = "en_core_web"
|
| 9 |
DEFAULT_TEXT = "Apple is looking at buying U.K. startup for $1 billion."
|
|
@@ -16,6 +20,7 @@ texts = {"en": DEFAULT_TEXT, "ca": "Apple està buscant comprar una startup del
|
|
| 16 |
"ja": "アップルがイギリスの新興企業を10億ドルで購入を検討", "ko": "애플이 영국의 스타트업을 10억 달러에 인수하는 것을 알아보고 있다.", "lt": "Jaunikis pirmąją vestuvinę naktį iškeitė į areštinės gultą", "nb": "Apple vurderer å kjøpe britisk oppstartfirma for en milliard dollar.", "nl": "Apple overweegt om voor 1 miljard een U.K. startup te kopen",
|
| 17 |
"pl": "Poczuł przyjemną woń mocnej kawy.", "pt": "Apple está querendo comprar uma startup do Reino Unido por 100 milhões de dólares", "ro": "Apple plănuiește să cumpere o companie britanică pentru un miliard de dolari", "ru": "Apple рассматривает возможность покупки стартапа из Соединённого Королевства за $1 млрд", "sv": "Apple överväger att köpa brittisk startup för 1 miljard dollar.", "zh": "作为语言而言,为世界使用人数最多的语言,目前世界有五分之一人口做为母语。"}
|
| 18 |
|
|
|
|
| 19 |
|
| 20 |
def get_all_models():
|
| 21 |
with open("requirements.txt") as f:
|
|
@@ -31,22 +36,44 @@ def get_all_models():
|
|
| 31 |
|
| 32 |
models = get_all_models()
|
| 33 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 34 |
|
| 35 |
def dependency(text, col_punct, col_phrase, compact, bg, font, model):
|
| 36 |
nlp = spacy.load(model + "_sm")
|
| 37 |
doc = nlp(text)
|
| 38 |
options = {"compact": compact, "collapse_phrases": col_phrase,
|
| 39 |
"collapse_punct": col_punct, "bg": bg, "color": font}
|
| 40 |
-
|
| 41 |
-
|
|
|
|
| 42 |
|
| 43 |
|
| 44 |
def entity(text, ents, model):
|
| 45 |
nlp = spacy.load(model + "_sm")
|
| 46 |
doc = nlp(text)
|
| 47 |
options = {"ents": ents}
|
| 48 |
-
|
| 49 |
-
|
|
|
|
| 50 |
|
| 51 |
|
| 52 |
def token(text, attributes, model):
|
|
@@ -128,8 +155,9 @@ def span(text, span1, span2, label1, label2, model):
|
|
| 128 |
Span(doc, idx2_1, idx2_2, label2),
|
| 129 |
]
|
| 130 |
|
| 131 |
-
|
| 132 |
-
|
|
|
|
| 133 |
|
| 134 |
|
| 135 |
def get_text(model):
|
|
@@ -196,8 +224,19 @@ with demo:
|
|
| 196 |
label="Text Color", value="black")
|
| 197 |
|
| 198 |
depen_output = gr.HTML(value=dependency(
|
| 199 |
-
DEFAULT_TEXT, True, True, False, DEFAULT_COLOR, "black", DEFAULT_MODEL))
|
| 200 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 201 |
gr.Markdown(" ")
|
| 202 |
with gr.Box():
|
| 203 |
with gr.Column():
|
|
@@ -205,11 +244,20 @@ with demo:
|
|
| 205 |
"## [Entity Recognizer](https://spacy.io/usage/visualizers#ent)")
|
| 206 |
gr.Markdown(
|
| 207 |
"The entity visualizer highlights named entities and their labels in a text.")
|
| 208 |
-
|
| 209 |
DEFAULT_ENTS, value=DEFAULT_ENTS)
|
| 210 |
-
|
| 211 |
DEFAULT_TEXT, DEFAULT_ENTS, DEFAULT_MODEL))
|
| 212 |
-
ent_button = gr.Button("
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 213 |
with gr.Box():
|
| 214 |
with gr.Column():
|
| 215 |
gr.Markdown(
|
|
@@ -224,7 +272,7 @@ with demo:
|
|
| 224 |
gr.Markdown("")
|
| 225 |
tok_output = gr.Dataframe(headers=DEFAULT_TOK_ATTR, value=default_token(
|
| 226 |
DEFAULT_TEXT, DEFAULT_TOK_ATTR, DEFAULT_MODEL), overflow_row_behaviour="paginate")
|
| 227 |
-
tok_button = gr.Button("
|
| 228 |
with gr.Box():
|
| 229 |
with gr.Column():
|
| 230 |
gr.Markdown(
|
|
@@ -244,7 +292,7 @@ with demo:
|
|
| 244 |
with gr.Column():
|
| 245 |
gr.Markdown("")
|
| 246 |
sim_random_button = gr.Button("Generate random words")
|
| 247 |
-
sim_button = gr.Button("
|
| 248 |
with gr.Box():
|
| 249 |
with gr.Column():
|
| 250 |
gr.Markdown(
|
|
@@ -276,31 +324,38 @@ with demo:
|
|
| 276 |
gr.Markdown("")
|
| 277 |
span_output = gr.HTML(value=span(
|
| 278 |
DEFAULT_TEXT, "U.K. startup", "U.K.", "ORG", "GPE", DEFAULT_MODEL))
|
| 279 |
-
gr.
|
| 280 |
-
gr.
|
| 281 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 282 |
|
| 283 |
model_input.change(get_text, inputs=[model_input], outputs=text_input)
|
| 284 |
button.click(dependency, inputs=[
|
| 285 |
-
text_input, col_punct, col_phrase, compact, bg, text, model_input], outputs=depen_output)
|
| 286 |
button.click(
|
| 287 |
-
entity, inputs=[text_input,
|
| 288 |
button.click(
|
| 289 |
token, inputs=[text_input, tok_input, model_input], outputs=tok_output)
|
| 290 |
button.click(vectors, inputs=[sim_text1,
|
| 291 |
sim_text2, model_input], outputs=sim_output)
|
| 292 |
button.click(
|
| 293 |
-
span, inputs=[text_input, span1, span2, label1, label2, model_input], outputs=span_output)
|
| 294 |
dep_button.click(dependency, inputs=[
|
| 295 |
-
text_input, col_punct, col_phrase, compact, bg, text, model_input], outputs=depen_output)
|
| 296 |
ent_button.click(
|
| 297 |
-
entity, inputs=[text_input,
|
| 298 |
tok_button.click(
|
| 299 |
token, inputs=[text_input, tok_input, model_input], outputs=[tok_output])
|
| 300 |
sim_button.click(vectors, inputs=[
|
| 301 |
sim_text1, sim_text2, model_input], outputs=sim_output)
|
| 302 |
span_button.click(
|
| 303 |
-
span, inputs=[text_input, span1, span2, label1, label2, model_input], outputs=span_output)
|
| 304 |
sim_random_button.click(random_vectors, inputs=[text_input, model_input], outputs=[
|
| 305 |
sim_output, sim_text1, sim_text2])
|
| 306 |
demo.launch()
|
|
|
|
| 1 |
+
from turtle import down
|
| 2 |
import spacy
|
| 3 |
from spacy import displacy
|
| 4 |
import random
|
| 5 |
from spacy.tokens import Span
|
| 6 |
import gradio as gr
|
| 7 |
import pandas as pd
|
| 8 |
+
import cairosvg
|
| 9 |
+
import base64
|
| 10 |
+
|
| 11 |
|
| 12 |
DEFAULT_MODEL = "en_core_web"
|
| 13 |
DEFAULT_TEXT = "Apple is looking at buying U.K. startup for $1 billion."
|
|
|
|
| 20 |
"ja": "アップルがイギリスの新興企業を10億ドルで購入を検討", "ko": "애플이 영국의 스타트업을 10억 달러에 인수하는 것을 알아보고 있다.", "lt": "Jaunikis pirmąją vestuvinę naktį iškeitė į areštinės gultą", "nb": "Apple vurderer å kjøpe britisk oppstartfirma for en milliard dollar.", "nl": "Apple overweegt om voor 1 miljard een U.K. startup te kopen",
|
| 21 |
"pl": "Poczuł przyjemną woń mocnej kawy.", "pt": "Apple está querendo comprar uma startup do Reino Unido por 100 milhões de dólares", "ro": "Apple plănuiește să cumpere o companie britanică pentru un miliard de dolari", "ru": "Apple рассматривает возможность покупки стартапа из Соединённого Королевства за $1 млрд", "sv": "Apple överväger att köpa brittisk startup för 1 miljard dollar.", "zh": "作为语言而言,为世界使用人数最多的语言,目前世界有五分之一人口做为母语。"}
|
| 22 |
|
| 23 |
+
button_css = "--tw-border-opacity: 1; border-color: rgb(75 85 99 / var(--tw-border-opacity)); --tw-gradient-from: #4b5563; --tw-gradient-stops: var(--tw-gradient-from), var(--tw-gradient-to, rgb(75 85 99 / 0));--tw-gradient-to: #374151; --tw-text-opacity: 1; color: rgb(255 255 255 / var(--tw-text-opacity)); background-color: rgb(55 65 81 / var(--tw-bg-opacity)); border-radius: 0.5rem; padding-top: 0.5rem; padding-bottom: 0.5rem; padding-left: 1rem; padding-right: 1rem; font-size: 1rem; line-height: 1.5rem; font-weight: 600; -webkit-appearance: button; --tw-shadow: 0 1px 2px 0 rgb(0 0 0 / 0.05); --tw-shadow-colored: 0 1px 2px 0 var(--tw-shadow-color); box-shadow: var(--tw-ring-offset-shadow, 0 0 #0000), var(--tw-ring-shadow, 0 0 #0000), var(--tw-shadow); border-width: 1px; align-items: center; justify-content: center; display: inline-flex;"
|
| 24 |
|
| 25 |
def get_all_models():
|
| 26 |
with open("requirements.txt") as f:
|
|
|
|
| 36 |
|
| 37 |
models = get_all_models()
|
| 38 |
|
| 39 |
+
def download_svg(svg):
|
| 40 |
+
encode = base64.b64encode(bytes(svg, 'utf-8'))
|
| 41 |
+
img = 'data:image/svg+xml;base64,' + str(encode)[2:-1]
|
| 42 |
+
html = f'<a download="displacy.svg" href="{img}" style="{button_css}">Download as SVG</a>'
|
| 43 |
+
return html
|
| 44 |
+
|
| 45 |
+
# def download_png(svg):
|
| 46 |
+
# encode = base64.b64encode(bytes(svg, 'utf-8'))
|
| 47 |
+
# svg_uri = 'data:image/svg+xml;base64,' + str(encode)[2:-1]
|
| 48 |
+
# output = cairosvg.svg2png(url=svg_uri)
|
| 49 |
+
# encoded = base64.b64encode(output)
|
| 50 |
+
# img = 'data:image/png;base64,' + str(encoded)[2:-1]
|
| 51 |
+
# html = f'<a download="displacy.png" href="{img}" style="{button_css}">Download as PNG</a>'
|
| 52 |
+
# return html
|
| 53 |
+
|
| 54 |
+
# def download(type, svg):
|
| 55 |
+
# if type == 'png':
|
| 56 |
+
# return download_png(svg)
|
| 57 |
+
# elif type == 'svg':
|
| 58 |
+
# return download_svg(svg)
|
| 59 |
|
| 60 |
def dependency(text, col_punct, col_phrase, compact, bg, font, model):
|
| 61 |
nlp = spacy.load(model + "_sm")
|
| 62 |
doc = nlp(text)
|
| 63 |
options = {"compact": compact, "collapse_phrases": col_phrase,
|
| 64 |
"collapse_punct": col_punct, "bg": bg, "color": font}
|
| 65 |
+
svg = displacy.render(doc, style="dep", options=options)
|
| 66 |
+
download = download_svg(svg)
|
| 67 |
+
return svg, download
|
| 68 |
|
| 69 |
|
| 70 |
def entity(text, ents, model):
|
| 71 |
nlp = spacy.load(model + "_sm")
|
| 72 |
doc = nlp(text)
|
| 73 |
options = {"ents": ents}
|
| 74 |
+
svg = displacy.render(doc, style="ent", options=options)
|
| 75 |
+
# download = download_svg('<svg xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink" xml:lang="en" id="97e9d3ac65344f2bb6e6ce517bd13b1e-0" class="displacy" width="1800" height="399.5" direction="ltr" style="max-width: none; height: 399.5px; color: black; font-family: Arial; direction: ltr">' + svg + "</svg>")
|
| 76 |
+
return svg
|
| 77 |
|
| 78 |
|
| 79 |
def token(text, attributes, model):
|
|
|
|
| 155 |
Span(doc, idx2_1, idx2_2, label2),
|
| 156 |
]
|
| 157 |
|
| 158 |
+
svg = displacy.render(doc, style="span")
|
| 159 |
+
# download = download_svg(svg)
|
| 160 |
+
return svg
|
| 161 |
|
| 162 |
|
| 163 |
def get_text(model):
|
|
|
|
| 224 |
label="Text Color", value="black")
|
| 225 |
|
| 226 |
depen_output = gr.HTML(value=dependency(
|
| 227 |
+
DEFAULT_TEXT, True, True, False, DEFAULT_COLOR, "black", DEFAULT_MODEL)[0])
|
| 228 |
+
with gr.Row():
|
| 229 |
+
with gr.Column():
|
| 230 |
+
dep_button = gr.Button("Regenerate Dependency Parser", variant="primary")
|
| 231 |
+
with gr.Column():
|
| 232 |
+
dep_download_button = gr.HTML(value=download_svg(depen_output.value))
|
| 233 |
+
with gr.Column():
|
| 234 |
+
gr.Markdown(" ")
|
| 235 |
+
with gr.Column():
|
| 236 |
+
gr.Markdown(" ")
|
| 237 |
+
|
| 238 |
+
|
| 239 |
+
|
| 240 |
gr.Markdown(" ")
|
| 241 |
with gr.Box():
|
| 242 |
with gr.Column():
|
|
|
|
| 244 |
"## [Entity Recognizer](https://spacy.io/usage/visualizers#ent)")
|
| 245 |
gr.Markdown(
|
| 246 |
"The entity visualizer highlights named entities and their labels in a text.")
|
| 247 |
+
ent_input = gr.CheckboxGroup(
|
| 248 |
DEFAULT_ENTS, value=DEFAULT_ENTS)
|
| 249 |
+
ent_output = gr.HTML(value=entity(
|
| 250 |
DEFAULT_TEXT, DEFAULT_ENTS, DEFAULT_MODEL))
|
| 251 |
+
ent_button = gr.Button("Regenerate Entity Recognizer", variant="primary")
|
| 252 |
+
# with gr.Row():
|
| 253 |
+
# with gr.Column():
|
| 254 |
+
# ent_button = gr.Button("Regenerate Entity Recognizer", variant="primary")
|
| 255 |
+
# with gr.Column():
|
| 256 |
+
# ent_download_button = gr.HTML(value=download_svg(ent_output.value))
|
| 257 |
+
# with gr.Column():
|
| 258 |
+
# gr.Markdown(" ")
|
| 259 |
+
# with gr.Column():
|
| 260 |
+
# gr.Markdown(" ")
|
| 261 |
with gr.Box():
|
| 262 |
with gr.Column():
|
| 263 |
gr.Markdown(
|
|
|
|
| 272 |
gr.Markdown("")
|
| 273 |
tok_output = gr.Dataframe(headers=DEFAULT_TOK_ATTR, value=default_token(
|
| 274 |
DEFAULT_TEXT, DEFAULT_TOK_ATTR, DEFAULT_MODEL), overflow_row_behaviour="paginate")
|
| 275 |
+
tok_button = gr.Button("Regenerate Token Properties", variant="primary")
|
| 276 |
with gr.Box():
|
| 277 |
with gr.Column():
|
| 278 |
gr.Markdown(
|
|
|
|
| 292 |
with gr.Column():
|
| 293 |
gr.Markdown("")
|
| 294 |
sim_random_button = gr.Button("Generate random words")
|
| 295 |
+
sim_button = gr.Button("Regenerate similarity", variant="primary")
|
| 296 |
with gr.Box():
|
| 297 |
with gr.Column():
|
| 298 |
gr.Markdown(
|
|
|
|
| 324 |
gr.Markdown("")
|
| 325 |
span_output = gr.HTML(value=span(
|
| 326 |
DEFAULT_TEXT, "U.K. startup", "U.K.", "ORG", "GPE", DEFAULT_MODEL))
|
| 327 |
+
span_button = gr.Button("Regenerate Spans", variant="primary")
|
| 328 |
+
# with gr.Row():
|
| 329 |
+
# with gr.Column():
|
| 330 |
+
# span_button = gr.Button("Regenerate Spans", variant="primary")
|
| 331 |
+
# with gr.Column():
|
| 332 |
+
# span_download_button = gr.HTML(value=download_svg(span_output.value))
|
| 333 |
+
# with gr.Column():
|
| 334 |
+
# gr.Markdown(" ")
|
| 335 |
+
# with gr.Column():
|
| 336 |
+
# gr.Markdown(" ")
|
| 337 |
|
| 338 |
model_input.change(get_text, inputs=[model_input], outputs=text_input)
|
| 339 |
button.click(dependency, inputs=[
|
| 340 |
+
text_input, col_punct, col_phrase, compact, bg, text, model_input], outputs=[depen_output, dep_download_button])
|
| 341 |
button.click(
|
| 342 |
+
entity, inputs=[text_input, ent_input, model_input], outputs=[ent_output])
|
| 343 |
button.click(
|
| 344 |
token, inputs=[text_input, tok_input, model_input], outputs=tok_output)
|
| 345 |
button.click(vectors, inputs=[sim_text1,
|
| 346 |
sim_text2, model_input], outputs=sim_output)
|
| 347 |
button.click(
|
| 348 |
+
span, inputs=[text_input, span1, span2, label1, label2, model_input], outputs=[span_output])
|
| 349 |
dep_button.click(dependency, inputs=[
|
| 350 |
+
text_input, col_punct, col_phrase, compact, bg, text, model_input], outputs=[depen_output, dep_download_button])
|
| 351 |
ent_button.click(
|
| 352 |
+
entity, inputs=[text_input, ent_input, model_input], outputs=[ent_output])
|
| 353 |
tok_button.click(
|
| 354 |
token, inputs=[text_input, tok_input, model_input], outputs=[tok_output])
|
| 355 |
sim_button.click(vectors, inputs=[
|
| 356 |
sim_text1, sim_text2, model_input], outputs=sim_output)
|
| 357 |
span_button.click(
|
| 358 |
+
span, inputs=[text_input, span1, span2, label1, label2, model_input], outputs=[span_output])
|
| 359 |
sim_random_button.click(random_vectors, inputs=[text_input, model_input], outputs=[
|
| 360 |
sim_output, sim_text1, sim_text2])
|
| 361 |
demo.launch()
|
scrollbar.css
CHANGED
|
@@ -1,22 +1,31 @@
|
|
| 1 |
-
.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
-webkit-appearance: none;
|
| 3 |
}
|
| 4 |
|
| 5 |
-
.
|
| 6 |
-
width:
|
| 7 |
}
|
| 8 |
|
| 9 |
-
.
|
| 10 |
height: 11px;
|
| 11 |
}
|
| 12 |
|
| 13 |
-
.
|
| 14 |
border-radius: 8px;
|
| 15 |
border: 2px solid white; /* should match background, can't be transparent */
|
| 16 |
background-color: rgba(0, 0, 0, .5);
|
| 17 |
}
|
| 18 |
|
| 19 |
-
.
|
| 20 |
background-color: #fff;
|
| 21 |
border-radius: 8px;
|
| 22 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
.output-html {
|
| 2 |
+
overflow-x: auto;
|
| 3 |
+
overflow-y: visible;
|
| 4 |
+
}
|
| 5 |
+
|
| 6 |
+
.output-html::-webkit-scrollbar {
|
| 7 |
-webkit-appearance: none;
|
| 8 |
}
|
| 9 |
|
| 10 |
+
.output-html::-webkit-scrollbar:vertical {
|
| 11 |
+
width: 0px;
|
| 12 |
}
|
| 13 |
|
| 14 |
+
.output-html::-webkit-scrollbar:horizontal {
|
| 15 |
height: 11px;
|
| 16 |
}
|
| 17 |
|
| 18 |
+
.output-html::-webkit-scrollbar-thumb {
|
| 19 |
border-radius: 8px;
|
| 20 |
border: 2px solid white; /* should match background, can't be transparent */
|
| 21 |
background-color: rgba(0, 0, 0, .5);
|
| 22 |
}
|
| 23 |
|
| 24 |
+
.output-html::-webkit-scrollbar-track {
|
| 25 |
background-color: #fff;
|
| 26 |
border-radius: 8px;
|
| 27 |
+
}
|
| 28 |
+
|
| 29 |
+
.spans {
|
| 30 |
+
min-height: 75px;
|
| 31 |
+
}
|