Upload app.py
Browse files
app.py
ADDED
@@ -0,0 +1,275 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import json
|
2 |
+
import gradio as gr
|
3 |
+
import pandas as pd
|
4 |
+
import plotly.express as px
|
5 |
+
|
6 |
+
PIPELINE_TAGS = [
|
7 |
+
'text-generation',
|
8 |
+
'text-to-image',
|
9 |
+
'text-classification',
|
10 |
+
'text2text-generation',
|
11 |
+
'audio-to-audio',
|
12 |
+
'feature-extraction',
|
13 |
+
'image-classification',
|
14 |
+
'translation',
|
15 |
+
'reinforcement-learning',
|
16 |
+
'fill-mask',
|
17 |
+
'text-to-speech',
|
18 |
+
'automatic-speech-recognition',
|
19 |
+
'image-text-to-text',
|
20 |
+
'token-classification',
|
21 |
+
'sentence-similarity',
|
22 |
+
'question-answering',
|
23 |
+
'image-feature-extraction',
|
24 |
+
'summarization',
|
25 |
+
'zero-shot-image-classification',
|
26 |
+
'object-detection',
|
27 |
+
'image-segmentation',
|
28 |
+
'image-to-image',
|
29 |
+
'image-to-text',
|
30 |
+
'audio-classification',
|
31 |
+
'visual-question-answering',
|
32 |
+
'text-to-video',
|
33 |
+
'zero-shot-classification',
|
34 |
+
'depth-estimation',
|
35 |
+
'text-ranking',
|
36 |
+
'image-to-video',
|
37 |
+
'multiple-choice',
|
38 |
+
'unconditional-image-generation',
|
39 |
+
'video-classification',
|
40 |
+
'text-to-audio',
|
41 |
+
'time-series-forecasting',
|
42 |
+
'any-to-any',
|
43 |
+
'video-text-to-text',
|
44 |
+
'table-question-answering',
|
45 |
+
]
|
46 |
+
|
47 |
+
def is_audio_speech(repo_dct):
|
48 |
+
res = (repo_dct.get("pipeline_tag", None) and "audio" in repo_dct.get("pipeline_tag", "").lower()) or \
|
49 |
+
(repo_dct.get("pipeline_tag", None) and "speech" in repo_dct.get("pipeline_tag", "").lower()) or \
|
50 |
+
(repo_dct.get("tags", None) and any("audio" in tag.lower() for tag in repo_dct.get("tags", []))) or \
|
51 |
+
(repo_dct.get("tags", None) and any("speech" in tag.lower() for tag in repo_dct.get("tags", [])))
|
52 |
+
return res
|
53 |
+
|
54 |
+
def is_music(repo_dct):
|
55 |
+
res = (repo_dct.get("tags", None) and any("music" in tag.lower() for tag in repo_dct.get("tags", [])))
|
56 |
+
return res
|
57 |
+
|
58 |
+
def is_robotics(repo_dct):
|
59 |
+
res = (repo_dct.get("tags", None) and any("robot" in tag.lower() for tag in repo_dct.get("tags", [])))
|
60 |
+
return res
|
61 |
+
|
62 |
+
def is_biomed(repo_dct):
|
63 |
+
res = (repo_dct.get("tags", None) and any("bio" in tag.lower() for tag in repo_dct.get("tags", []))) or \
|
64 |
+
(repo_dct.get("tags", None) and any("medic" in tag.lower() for tag in repo_dct.get("tags", [])))
|
65 |
+
return res
|
66 |
+
|
67 |
+
def is_timeseries(repo_dct):
|
68 |
+
res = (repo_dct.get("tags", None) and any("series" in tag.lower() for tag in repo_dct.get("tags", [])))
|
69 |
+
return res
|
70 |
+
|
71 |
+
def is_science(repo_dct):
|
72 |
+
res = (repo_dct.get("tags", None) and any("science" in tag.lower() and not "bigscience" in tag for tag in repo_dct.get("tags", [])))
|
73 |
+
return res
|
74 |
+
|
75 |
+
def is_video(repo_dct):
|
76 |
+
res = (repo_dct.get("tags", None) and any("video" in tag.lower() for tag in repo_dct.get("tags", [])))
|
77 |
+
return res
|
78 |
+
|
79 |
+
def is_image(repo_dct):
|
80 |
+
res = (repo_dct.get("tags", None) and any("image" in tag.lower() for tag in repo_dct.get("tags", [])))
|
81 |
+
return res
|
82 |
+
|
83 |
+
def is_text(repo_dct):
|
84 |
+
res = (repo_dct.get("tags", None) and any("text" in tag.lower() for tag in repo_dct.get("tags", [])))
|
85 |
+
return res
|
86 |
+
|
87 |
+
TAG_FILTER_FUNCS = {
|
88 |
+
"Audio & Speech": is_audio_speech,
|
89 |
+
"Time series": is_timeseries,
|
90 |
+
"Robotics": is_robotics,
|
91 |
+
"Music": is_music,
|
92 |
+
"Video": is_video,
|
93 |
+
"Images": is_image,
|
94 |
+
"Text": is_text,
|
95 |
+
"Biomedical": is_biomed,
|
96 |
+
"Sciences": is_science,
|
97 |
+
}
|
98 |
+
|
99 |
+
def make_org_stats(repo_type, count_by, org_stats, top_k=20, filter_func=None):
|
100 |
+
assert count_by in ["likes", "downloads", "downloads_all"]
|
101 |
+
assert repo_type in ["all", "datasets", "models"]
|
102 |
+
repos = ["datasets", "models"] if repo_type == "all" else [repo_type]
|
103 |
+
if filter_func is None:
|
104 |
+
filter_func = lambda x: True
|
105 |
+
sorted_stats = sorted(
|
106 |
+
[(
|
107 |
+
author,
|
108 |
+
sum(dct[count_by] for dct in author_dct[repo] if filter_func(dct))
|
109 |
+
) for repo in repos for author, author_dct in org_stats.items()],
|
110 |
+
key=lambda x:x[1],
|
111 |
+
reverse=True,
|
112 |
+
)
|
113 |
+
res = sorted_stats[:top_k] + [("Others...", sum(st for auth, st in sorted_stats[top_k:]))]
|
114 |
+
total_st = sum(st for o, st in res)
|
115 |
+
res_plot_df = []
|
116 |
+
for org, st in res:
|
117 |
+
if org == "Others...":
|
118 |
+
res_plot_df += [("Others...", "other", st * 100 / total_st)]
|
119 |
+
else:
|
120 |
+
for repo in repos:
|
121 |
+
for dct in org_stats[org][repo]:
|
122 |
+
if filter_func(dct):
|
123 |
+
res_plot_df += [(org, dct["id"], dct[count_by] * 100 / total_st)]
|
124 |
+
return ([(o, 100 * st / total_st) for o, st in res if st > 0], res_plot_df)
|
125 |
+
|
126 |
+
def make_figure(count_by, repo_type, org_stats, tag_filter=None, pipeline_filter=None):
|
127 |
+
assert count_by in ["downloads", "likes", "downloads_all"]
|
128 |
+
assert repo_type in ["all", "models", "datasets"]
|
129 |
+
assert tag_filter is None or pipeline_filter is None
|
130 |
+
filter_func = None
|
131 |
+
if tag_filter:
|
132 |
+
filter_func = TAG_FILTER_FUNCS[tag_filter]
|
133 |
+
if pipeline_filter:
|
134 |
+
filter_func = lambda dct: dct.get("pipeline_tag", None) and dct.get("pipeline_tag", "") == pipeline_filter
|
135 |
+
_, res_plot_df = make_org_stats(repo_type, count_by, org_stats, top_k=25, filter_func=filter_func)
|
136 |
+
df = pd.DataFrame(
|
137 |
+
dict(
|
138 |
+
organizations=[o for o, _, _ in res_plot_df],
|
139 |
+
repo=[r for _, r, _ in res_plot_df],
|
140 |
+
stats=[s for _, _, s in res_plot_df],
|
141 |
+
)
|
142 |
+
)
|
143 |
+
df[repo_type] = repo_type # in order to have a single root node
|
144 |
+
fig = px.treemap(df, path=[repo_type, 'organizations', 'repo'], values='stats')
|
145 |
+
fig.update_layout(
|
146 |
+
treemapcolorway = ["pink" for _ in range(len(res_plot_df))],
|
147 |
+
margin = dict(t=50, l=25, r=25, b=25)
|
148 |
+
)
|
149 |
+
return fig
|
150 |
+
|
151 |
+
|
152 |
+
with gr.Blocks() as demo:
|
153 |
+
org_stats_data = gr.State(value=None) # To store loaded data
|
154 |
+
|
155 |
+
with gr.Row():
|
156 |
+
gr.Markdown("""
|
157 |
+
## Hugging Face Organization Stats
|
158 |
+
|
159 |
+
This app shows how different organizations are contributing to different aspects of the open AI ecosystem.
|
160 |
+
Use the dropdowns on the left to select repository types, metrics, and optionally tags representing topics or modalities of interest.
|
161 |
+
""")
|
162 |
+
with gr.Row():
|
163 |
+
with gr.Column(scale=1):
|
164 |
+
repo_type_dropdown = gr.Dropdown(
|
165 |
+
label="Repository Type",
|
166 |
+
choices=["all", "models", "datasets"],
|
167 |
+
value="all"
|
168 |
+
)
|
169 |
+
count_by_dropdown = gr.Dropdown(
|
170 |
+
label="Metric",
|
171 |
+
choices=["downloads", "likes", "downloads_all"],
|
172 |
+
value="downloads"
|
173 |
+
)
|
174 |
+
|
175 |
+
filter_choice_radio = gr.Radio(
|
176 |
+
label="Filter by",
|
177 |
+
choices=["None", "Tag Filter", "Pipeline Filter"],
|
178 |
+
value="None"
|
179 |
+
)
|
180 |
+
|
181 |
+
tag_filter_dropdown = gr.Dropdown(
|
182 |
+
label="Select Tag",
|
183 |
+
choices=list(TAG_FILTER_FUNCS.keys()),
|
184 |
+
value=None,
|
185 |
+
visible=False
|
186 |
+
)
|
187 |
+
pipeline_filter_dropdown = gr.Dropdown(
|
188 |
+
label="Select Pipeline Tag",
|
189 |
+
choices=PIPELINE_TAGS,
|
190 |
+
value=None,
|
191 |
+
visible=False
|
192 |
+
)
|
193 |
+
|
194 |
+
generate_plot_button = gr.Button("Generate Plot")
|
195 |
+
|
196 |
+
with gr.Column(scale=3):
|
197 |
+
plot_output = gr.Plot()
|
198 |
+
|
199 |
+
def generate_plot_on_click(repo_type, count_by, filter_choice, tag_filter, pipeline_filter, data):
|
200 |
+
# Print the current state of the input variables
|
201 |
+
print(f"Generating plot with the following inputs:")
|
202 |
+
print(f" Repository Type: {repo_type}")
|
203 |
+
print(f" Metric (Count By): {count_by}")
|
204 |
+
print(f" Filter Choice: {filter_choice}")
|
205 |
+
if filter_choice == "Tag Filter":
|
206 |
+
print(f" Tag Filter: {tag_filter}")
|
207 |
+
elif filter_choice == "Pipeline Filter":
|
208 |
+
print(f" Pipeline Filter: {pipeline_filter}")
|
209 |
+
|
210 |
+
if data is None:
|
211 |
+
print("Error: Data not loaded yet.")
|
212 |
+
return None
|
213 |
+
|
214 |
+
selected_tag_filter = None
|
215 |
+
selected_pipeline_filter = None
|
216 |
+
|
217 |
+
if filter_choice == "Tag Filter":
|
218 |
+
selected_tag_filter = tag_filter
|
219 |
+
elif filter_choice == "Pipeline Filter":
|
220 |
+
selected_pipeline_filter = pipeline_filter
|
221 |
+
|
222 |
+
fig = make_figure(
|
223 |
+
count_by=count_by,
|
224 |
+
repo_type=repo_type,
|
225 |
+
org_stats=data,
|
226 |
+
tag_filter=selected_tag_filter,
|
227 |
+
pipeline_filter=selected_pipeline_filter
|
228 |
+
)
|
229 |
+
return fig
|
230 |
+
|
231 |
+
def update_filter_visibility(filter_choice):
|
232 |
+
if filter_choice == "Tag Filter":
|
233 |
+
return gr.update(visible=True), gr.update(visible=False)
|
234 |
+
elif filter_choice == "Pipeline Filter":
|
235 |
+
return gr.update(visible=False), gr.update(visible=True)
|
236 |
+
else: # "None"
|
237 |
+
return gr.update(visible=False), gr.update(visible=False)
|
238 |
+
|
239 |
+
filter_choice_radio.change(
|
240 |
+
fn=update_filter_visibility,
|
241 |
+
inputs=[filter_choice_radio],
|
242 |
+
outputs=[tag_filter_dropdown, pipeline_filter_dropdown]
|
243 |
+
)
|
244 |
+
|
245 |
+
# Load data once at startup
|
246 |
+
def load_org_data():
|
247 |
+
print("Loading organization statistics data...")
|
248 |
+
loaded_org_stats = json.load(open("org_to_artifacts_2l_stats.json"))
|
249 |
+
print("Data loaded successfully.")
|
250 |
+
return loaded_org_stats
|
251 |
+
|
252 |
+
demo.load(
|
253 |
+
fn=load_org_data,
|
254 |
+
inputs=[], # No inputs needed to just load data
|
255 |
+
outputs=[org_stats_data] # Only output to the state
|
256 |
+
)
|
257 |
+
|
258 |
+
# Button click event to generate plot
|
259 |
+
generate_plot_button.click(
|
260 |
+
fn=generate_plot_on_click,
|
261 |
+
inputs=[
|
262 |
+
repo_type_dropdown,
|
263 |
+
count_by_dropdown,
|
264 |
+
filter_choice_radio,
|
265 |
+
tag_filter_dropdown,
|
266 |
+
pipeline_filter_dropdown,
|
267 |
+
org_stats_data
|
268 |
+
],
|
269 |
+
outputs=[plot_output]
|
270 |
+
)
|
271 |
+
|
272 |
+
|
273 |
+
if __name__ == "__main__":
|
274 |
+
# org_stats = json.load(open("org_to_artifacts_2l_stats.json")) # Data loading handled by demo.load
|
275 |
+
demo.launch()
|