Spaces:
Runtime error
Runtime error
clean layout
Browse files- .gitignore +1 -0
- app.py +103 -87
- src/envs.py +1 -1
.gitignore
CHANGED
@@ -11,3 +11,4 @@ eval-results/
|
|
11 |
eval-queue-bk/
|
12 |
eval-results-bk/
|
13 |
logs/
|
|
|
|
11 |
eval-queue-bk/
|
12 |
eval-results-bk/
|
13 |
logs/
|
14 |
+
envs/
|
app.py
CHANGED
@@ -140,104 +140,120 @@ with demo:
|
|
140 |
with gr.Tabs(elem_classes="tab-buttons") as tabs:
|
141 |
with gr.TabItem("π
LLM Benchmark", elem_id="llm-benchmark-tab-table", id=0):
|
142 |
with gr.Row():
|
143 |
-
|
144 |
-
with
|
145 |
-
|
146 |
-
|
147 |
-
|
148 |
-
|
149 |
-
|
150 |
-
|
151 |
-
|
152 |
-
|
153 |
-
|
154 |
-
|
155 |
-
|
156 |
-
|
157 |
-
|
158 |
-
|
159 |
-
|
160 |
-
|
161 |
-
|
162 |
-
|
163 |
-
|
164 |
-
|
165 |
-
|
166 |
-
|
167 |
-
|
168 |
-
|
169 |
-
|
170 |
-
|
171 |
-
|
172 |
-
|
173 |
-
|
174 |
-
|
175 |
-
|
176 |
-
|
177 |
-
|
178 |
-
|
179 |
-
|
180 |
-
|
181 |
-
|
182 |
-
|
183 |
-
|
184 |
-
|
185 |
-
|
186 |
-
|
187 |
-
|
188 |
-
|
189 |
-
|
190 |
-
|
191 |
-
|
192 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
193 |
|
194 |
leaderboard_table = gr.components.Dataframe(
|
195 |
value=leaderboard_df[
|
196 |
[c.name for c in fields(AutoEvalColumn) if c.never_hidden]
|
197 |
-
+ shown_columns.value
|
198 |
],
|
199 |
-
headers=[c.name for c in fields(AutoEvalColumn) if c.never_hidden]
|
200 |
datatype=TYPES,
|
201 |
elem_id="leaderboard-table",
|
202 |
interactive=False,
|
203 |
visible=True,
|
204 |
)
|
205 |
|
206 |
-
# Dummy leaderboard for handling the case when the user uses backspace key
|
207 |
-
hidden_leaderboard_table_for_search = gr.components.Dataframe(
|
208 |
-
|
209 |
-
|
210 |
-
|
211 |
-
|
212 |
-
)
|
213 |
-
search_bar.submit(
|
214 |
-
|
215 |
-
|
216 |
-
|
217 |
-
|
218 |
-
|
219 |
-
|
220 |
-
|
221 |
-
|
222 |
-
|
223 |
-
|
224 |
-
|
225 |
-
)
|
226 |
-
for selector in [shown_columns, filter_columns_type, filter_columns_precision, filter_columns_size, deleted_models_visibility]:
|
227 |
-
|
228 |
-
|
229 |
-
|
230 |
-
|
231 |
-
|
232 |
-
|
233 |
-
|
234 |
-
|
235 |
-
|
236 |
-
|
237 |
-
|
238 |
-
|
239 |
-
|
240 |
-
|
241 |
|
242 |
with gr.TabItem("π About", elem_id="llm-benchmark-tab-table", id=2):
|
243 |
gr.Markdown(LLM_BENCHMARKS_TEXT, elem_classes="markdown-text")
|
|
|
140 |
with gr.Tabs(elem_classes="tab-buttons") as tabs:
|
141 |
with gr.TabItem("π
LLM Benchmark", elem_id="llm-benchmark-tab-table", id=0):
|
142 |
with gr.Row():
|
143 |
+
search_bar = gr.Textbox(
|
144 |
+
placeholder=" π Search for the model (separate multiple queries with `;`) and press ENTER...",
|
145 |
+
show_label=False,
|
146 |
+
elem_id="search-bar",
|
147 |
+
)
|
148 |
+
# with gr.Column():
|
149 |
+
# with gr.Row():
|
150 |
+
# search_bar = gr.Textbox(
|
151 |
+
# placeholder=" π Search for the model (separate multiple queries with `;`) and press ENTER...",
|
152 |
+
# show_label=False,
|
153 |
+
# elem_id="search-bar",
|
154 |
+
# )
|
155 |
+
# with gr.Row():
|
156 |
+
# shown_columns = gr.CheckboxGroup(
|
157 |
+
# choices=[
|
158 |
+
# c.name
|
159 |
+
# for c in fields(AutoEvalColumn)
|
160 |
+
# if not c.hidden and not c.never_hidden
|
161 |
+
# ],
|
162 |
+
# value=[
|
163 |
+
# c.name
|
164 |
+
# for c in fields(AutoEvalColumn)
|
165 |
+
# if c.displayed_by_default and not c.hidden and not c.never_hidden
|
166 |
+
# ],
|
167 |
+
# label="Select columns to show",
|
168 |
+
# elem_id="column-select",
|
169 |
+
# interactive=True,
|
170 |
+
# )
|
171 |
+
# with gr.Row():
|
172 |
+
# deleted_models_visibility = gr.Checkbox(
|
173 |
+
# value=False, label="Show gated/private/deleted models", interactive=True
|
174 |
+
# )
|
175 |
+
# with gr.Column(min_width=320):
|
176 |
+
# #with gr.Box(elem_id="box-filter"):
|
177 |
+
# filter_columns_type = gr.CheckboxGroup(
|
178 |
+
# label="Model types",
|
179 |
+
# choices=[t.to_str() for t in ModelType],
|
180 |
+
# value=[t.to_str() for t in ModelType],
|
181 |
+
# interactive=True,
|
182 |
+
# elem_id="filter-columns-type",
|
183 |
+
# )
|
184 |
+
# filter_columns_precision = gr.CheckboxGroup(
|
185 |
+
# label="Precision",
|
186 |
+
# choices=[i.value.name for i in Precision],
|
187 |
+
# value=[i.value.name for i in Precision],
|
188 |
+
# interactive=True,
|
189 |
+
# elem_id="filter-columns-precision",
|
190 |
+
# )
|
191 |
+
# filter_columns_size = gr.CheckboxGroup(
|
192 |
+
# label="Model sizes (in billions of parameters)",
|
193 |
+
# choices=list(NUMERIC_INTERVALS.keys()),
|
194 |
+
# value=list(NUMERIC_INTERVALS.keys()),
|
195 |
+
# interactive=True,
|
196 |
+
# elem_id="filter-columns-size",
|
197 |
+
# )
|
198 |
+
|
199 |
+
# leaderboard_table = gr.components.Dataframe(
|
200 |
+
# value=leaderboard_df[
|
201 |
+
# [c.name for c in fields(AutoEvalColumn) if c.never_hidden]
|
202 |
+
# + shown_columns.value
|
203 |
+
# ],
|
204 |
+
# headers=[c.name for c in fields(AutoEvalColumn) if c.never_hidden] + shown_columns.value,
|
205 |
+
# datatype=TYPES,
|
206 |
+
# elem_id="leaderboard-table",
|
207 |
+
# interactive=False,
|
208 |
+
# visible=True,
|
209 |
+
# )
|
210 |
|
211 |
leaderboard_table = gr.components.Dataframe(
|
212 |
value=leaderboard_df[
|
213 |
[c.name for c in fields(AutoEvalColumn) if c.never_hidden]
|
|
|
214 |
],
|
215 |
+
headers=[c.name for c in fields(AutoEvalColumn) if c.never_hidden],
|
216 |
datatype=TYPES,
|
217 |
elem_id="leaderboard-table",
|
218 |
interactive=False,
|
219 |
visible=True,
|
220 |
)
|
221 |
|
222 |
+
# # Dummy leaderboard for handling the case when the user uses backspace key
|
223 |
+
# hidden_leaderboard_table_for_search = gr.components.Dataframe(
|
224 |
+
# value=original_df[COLS],
|
225 |
+
# headers=COLS,
|
226 |
+
# datatype=TYPES,
|
227 |
+
# visible=False,
|
228 |
+
# )
|
229 |
+
# search_bar.submit(
|
230 |
+
# update_table,
|
231 |
+
# [
|
232 |
+
# hidden_leaderboard_table_for_search,
|
233 |
+
# shown_columns,
|
234 |
+
# filter_columns_type,
|
235 |
+
# filter_columns_precision,
|
236 |
+
# filter_columns_size,
|
237 |
+
# deleted_models_visibility,
|
238 |
+
# search_bar,
|
239 |
+
# ],
|
240 |
+
# leaderboard_table,
|
241 |
+
# )
|
242 |
+
# for selector in [shown_columns, filter_columns_type, filter_columns_precision, filter_columns_size, deleted_models_visibility]:
|
243 |
+
# selector.change(
|
244 |
+
# update_table,
|
245 |
+
# [
|
246 |
+
# hidden_leaderboard_table_for_search,
|
247 |
+
# shown_columns,
|
248 |
+
# filter_columns_type,
|
249 |
+
# filter_columns_precision,
|
250 |
+
# filter_columns_size,
|
251 |
+
# deleted_models_visibility,
|
252 |
+
# search_bar,
|
253 |
+
# ],
|
254 |
+
# leaderboard_table,
|
255 |
+
# queue=True,
|
256 |
+
# )
|
257 |
|
258 |
with gr.TabItem("π About", elem_id="llm-benchmark-tab-table", id=2):
|
259 |
gr.Markdown(LLM_BENCHMARKS_TEXT, elem_classes="markdown-text")
|
src/envs.py
CHANGED
@@ -6,7 +6,7 @@ from huggingface_hub import HfApi
|
|
6 |
# ----------------------------------
|
7 |
TOKEN = os.environ.get("TOKEN") # A read/write token for your org
|
8 |
|
9 |
-
OWNER = "
|
10 |
# ----------------------------------
|
11 |
|
12 |
REPO_ID = f"{OWNER}/leaderboard"
|
|
|
6 |
# ----------------------------------
|
7 |
TOKEN = os.environ.get("TOKEN") # A read/write token for your org
|
8 |
|
9 |
+
OWNER = "yale-nlp" # Change to your org - don't forget to create a results and request dataset, with the correct format!
|
10 |
# ----------------------------------
|
11 |
|
12 |
REPO_ID = f"{OWNER}/leaderboard"
|