Spaces:
Runtime error
Runtime error
clean up the figure, add data caching, add headers
Browse files
app.py
CHANGED
@@ -10,6 +10,11 @@ import numpy as np
|
|
10 |
import plotly.figure_factory as ff
|
11 |
import plotly.express as px
|
12 |
|
|
|
|
|
|
|
|
|
|
|
13 |
tokenizer_names_to_test = [
|
14 |
"openai/gpt4",
|
15 |
"xlm-roberta-base", # old style
|
@@ -24,27 +29,30 @@ tokenizer_names_to_test = [
|
|
24 |
]
|
25 |
|
26 |
with st.sidebar:
|
|
|
|
|
|
|
|
|
|
|
27 |
with st.spinner('Loading dataset...'):
|
28 |
-
val_data =
|
29 |
st.success(f'Data loaded: {len(val_data)}')
|
30 |
|
31 |
languages = st.multiselect(
|
32 |
'Select languages',
|
33 |
options=sorted(val_data.lang.unique()),
|
34 |
default=['English', 'Spanish' ,'Chinese'],
|
35 |
-
max_selections=
|
36 |
)
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
st.write('You selected:', tokenizer_name)
|
42 |
|
43 |
# with st.spinner('Loading tokenizer...'):
|
44 |
# tokenizer = AutoTokenizer.from_pretrained(tokenizer_name)
|
45 |
# st.success(f'Tokenizer loaded: {tokenizer_name}')
|
46 |
|
47 |
-
# # TODO - preload the tokenized versions ... much easier!
|
48 |
# # TODO - add the metadata data as well??? later on maybe
|
49 |
# with st.spinner('Calculating tokenization for data...'):
|
50 |
# if tokenizer_name not in val_data.columns:
|
@@ -55,18 +63,27 @@ with st.container():
|
|
55 |
if tokenizer_name in val_data.columns:
|
56 |
subset_df = val_data[val_data.lang.isin(languages)]
|
57 |
subset_data = [val_data[val_data.lang==_lang][tokenizer_name] for _lang in languages]
|
|
|
|
|
|
|
|
|
58 |
|
59 |
-
|
60 |
-
|
|
|
|
|
|
|
|
|
|
|
61 |
st.plotly_chart(fig, use_container_width=True)
|
62 |
|
|
|
|
|
|
|
|
|
63 |
|
64 |
-
|
65 |
-
|
66 |
-
|
67 |
-
|
68 |
-
# curve_type='normal', # override default 'kde'
|
69 |
-
# colors=colors)
|
70 |
-
|
71 |
-
|
72 |
|
|
|
10 |
import plotly.figure_factory as ff
|
11 |
import plotly.express as px
|
12 |
|
13 |
+
@st.cache_data
|
14 |
+
def load_data():
|
15 |
+
return pd.read_csv('MassiveDatasetValidationData.csv')
|
16 |
+
|
17 |
+
# TODO allow new tokenizers from HF
|
18 |
tokenizer_names_to_test = [
|
19 |
"openai/gpt4",
|
20 |
"xlm-roberta-base", # old style
|
|
|
29 |
]
|
30 |
|
31 |
with st.sidebar:
|
32 |
+
st.subheader('Model')
|
33 |
+
# TODO multi-select tokenizers
|
34 |
+
tokenizer_name = st.sidebar.selectbox('Select tokenizer', options=tokenizer_names_to_test)
|
35 |
+
|
36 |
+
st.subheader('Data')
|
37 |
with st.spinner('Loading dataset...'):
|
38 |
+
val_data = load_data()
|
39 |
st.success(f'Data loaded: {len(val_data)}')
|
40 |
|
41 |
languages = st.multiselect(
|
42 |
'Select languages',
|
43 |
options=sorted(val_data.lang.unique()),
|
44 |
default=['English', 'Spanish' ,'Chinese'],
|
45 |
+
max_selections=6
|
46 |
)
|
47 |
+
|
48 |
+
st.subheader('Figure')
|
49 |
+
show_hist = st.checkbox('Show histogram', value=False)
|
50 |
+
# dist_marginal = st.radio('Select distribution', options=['box', 'violin', 'rug'], horizontal=True)
|
|
|
51 |
|
52 |
# with st.spinner('Loading tokenizer...'):
|
53 |
# tokenizer = AutoTokenizer.from_pretrained(tokenizer_name)
|
54 |
# st.success(f'Tokenizer loaded: {tokenizer_name}')
|
55 |
|
|
|
56 |
# # TODO - add the metadata data as well??? later on maybe
|
57 |
# with st.spinner('Calculating tokenization for data...'):
|
58 |
# if tokenizer_name not in val_data.columns:
|
|
|
63 |
if tokenizer_name in val_data.columns:
|
64 |
subset_df = val_data[val_data.lang.isin(languages)]
|
65 |
subset_data = [val_data[val_data.lang==_lang][tokenizer_name] for _lang in languages]
|
66 |
+
|
67 |
+
st.header('Tokenization in different languages')
|
68 |
+
st.divider()
|
69 |
+
fig = ff.create_distplot(subset_data, group_labels=languages, show_hist=show_hist)
|
70 |
|
71 |
+
fig.update_layout(
|
72 |
+
title=dict(text=tokenizer_name, font=dict(size=25), automargin=True, yref='paper', ),
|
73 |
+
# title=tokenizer_name,
|
74 |
+
xaxis_title="Number of Tokens",
|
75 |
+
yaxis_title="Density",
|
76 |
+
# title_font_family='"Source Sans Pro", sans-serif'
|
77 |
+
)
|
78 |
st.plotly_chart(fig, use_container_width=True)
|
79 |
|
80 |
+
st.subheader('Median Token Length')
|
81 |
+
metric_cols = st.columns(len(languages))
|
82 |
+
for i, _lang in enumerate(languages):
|
83 |
+
metric_cols[i].metric(_lang, int(np.median(subset_df[subset_df.lang==_lang][tokenizer_name])))
|
84 |
|
85 |
+
if tokenizer_name not in ['openai/gpt4']:
|
86 |
+
url = f'https://huggingface.co/{tokenizer_name}'
|
87 |
+
link = f'[Find on the HuggingFace hub]({url})'
|
88 |
+
st.markdown(link, unsafe_allow_html=True)
|
|
|
|
|
|
|
|
|
89 |
|