Spaces:
Running
Running
chore: Reorder arguments in run_model_task function
Browse files- code_efficiency_calculator.py +1 -1
- leaderboard.py +31 -9
- requirements.txt +3 -1
code_efficiency_calculator.py
CHANGED
|
@@ -618,7 +618,7 @@ def fetch_completion(dataset,model):
|
|
| 618 |
return dataset
|
| 619 |
|
| 620 |
|
| 621 |
-
def run_model_task(model
|
| 622 |
|
| 623 |
if "/" in model:
|
| 624 |
model = model.split("/")[1]
|
|
|
|
| 618 |
return dataset
|
| 619 |
|
| 620 |
|
| 621 |
+
def run_model_task(task,model):
|
| 622 |
|
| 623 |
if "/" in model:
|
| 624 |
model = model.split("/")[1]
|
leaderboard.py
CHANGED
|
@@ -1,8 +1,11 @@
|
|
| 1 |
# Original code by https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard
|
| 2 |
# Modified by EffiBench
|
| 3 |
|
|
|
|
|
|
|
| 4 |
import pandas as pd
|
| 5 |
import gradio as gr
|
|
|
|
| 6 |
|
| 7 |
|
| 8 |
def make_default_md_1():
|
|
@@ -13,15 +16,11 @@ def make_default_md_1():
|
|
| 13 |
<a href='https://github.com/huangd1999/EffiBench' style='color: {link_color}; text-decoration: none;'>GitHub</a> |
|
| 14 |
<a href='https://github.com/huangd1999/EffiBench/tree/main/data' style='color: {link_color}; text-decoration: none;'>Dataset</a>
|
| 15 |
"""
|
| 16 |
-
|
| 17 |
return leaderboard_md
|
| 18 |
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
"""
|
| 23 |
-
|
| 24 |
-
return leaderboard_md
|
| 25 |
|
| 26 |
leaderboard_md = """
|
| 27 |
Three benchmarks are displayed: **EffiBench**, **HumanEval** and **MBPP**.
|
|
@@ -50,9 +49,26 @@ Please cite the following paper if you find our leaderboard or dataset helpful.
|
|
| 50 |
}
|
| 51 |
"""
|
| 52 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 53 |
def build_leaderboard_tab(leaderboard_table_file):
|
| 54 |
gr.Markdown(make_default_md_1(), elem_id="leaderboard_markdown")
|
| 55 |
-
gr.Markdown(
|
| 56 |
|
| 57 |
df = pd.read_csv(leaderboard_table_file)
|
| 58 |
def filter_leaderboard(dataset, timeout):
|
|
@@ -79,7 +95,13 @@ def build_leaderboard_tab(leaderboard_table_file):
|
|
| 79 |
dataset_dropdown.change(fn=update_leaderboard, inputs=[dataset_dropdown, timeout_dropdown], outputs=leaderboard)
|
| 80 |
timeout_dropdown.change(fn=update_leaderboard, inputs=[dataset_dropdown, timeout_dropdown], outputs=leaderboard)
|
| 81 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 82 |
with gr.Accordion("Citation", open=True):
|
| 83 |
gr.Markdown(citation_md, elem_id="leaderboard_markdown")
|
| 84 |
gr.Markdown(acknowledgment_md, elem_id="ack_markdown")
|
| 85 |
-
|
|
|
|
| 1 |
# Original code by https://huggingface.co/spaces/lmsys/chatbot-arena-leaderboard
|
| 2 |
# Modified by EffiBench
|
| 3 |
|
| 4 |
+
import json
|
| 5 |
+
from pathlib import Path
|
| 6 |
import pandas as pd
|
| 7 |
import gradio as gr
|
| 8 |
+
from calculate_memory_usage import report_results
|
| 9 |
|
| 10 |
|
| 11 |
def make_default_md_1():
|
|
|
|
| 16 |
<a href='https://github.com/huangd1999/EffiBench' style='color: {link_color}; text-decoration: none;'>GitHub</a> |
|
| 17 |
<a href='https://github.com/huangd1999/EffiBench/tree/main/data' style='color: {link_color}; text-decoration: none;'>Dataset</a>
|
| 18 |
"""
|
|
|
|
| 19 |
return leaderboard_md
|
| 20 |
|
| 21 |
+
add_model_md = f"""
|
| 22 |
+
🤗 [filing a request](https://github.com/huangd1999/EffiBench/issues/new?assignees=&labels=model+eval&projects=&template=model_eval_request.yml&title=%F0%9F%92%A1+%5BREQUEST%5D+-+%3CMODEL_NAME%3E) to add your models on our leaderboard!**
|
| 23 |
+
"""
|
|
|
|
|
|
|
|
|
|
| 24 |
|
| 25 |
leaderboard_md = """
|
| 26 |
Three benchmarks are displayed: **EffiBench**, **HumanEval** and **MBPP**.
|
|
|
|
| 49 |
}
|
| 50 |
"""
|
| 51 |
|
| 52 |
+
def process_uploaded_file(file):
|
| 53 |
+
if file is None:
|
| 54 |
+
return "No file uploaded."
|
| 55 |
+
try:
|
| 56 |
+
file = Path(file)
|
| 57 |
+
json_data = json.loads(file.read_text())
|
| 58 |
+
except Exception as e:
|
| 59 |
+
return f"Error processing the file: {str(e)}"
|
| 60 |
+
|
| 61 |
+
try:
|
| 62 |
+
task, model = file.stem.split("_")
|
| 63 |
+
except Exception as e:
|
| 64 |
+
return f"Error parsing the task and model name from the file name: {str(e)}! Should be in the format of <task>_<model>.json"
|
| 65 |
+
|
| 66 |
+
|
| 67 |
+
|
| 68 |
+
|
| 69 |
def build_leaderboard_tab(leaderboard_table_file):
|
| 70 |
gr.Markdown(make_default_md_1(), elem_id="leaderboard_markdown")
|
| 71 |
+
gr.Markdown(add_model_md, elem_id="leaderboard_markdown")
|
| 72 |
|
| 73 |
df = pd.read_csv(leaderboard_table_file)
|
| 74 |
def filter_leaderboard(dataset, timeout):
|
|
|
|
| 95 |
dataset_dropdown.change(fn=update_leaderboard, inputs=[dataset_dropdown, timeout_dropdown], outputs=leaderboard)
|
| 96 |
timeout_dropdown.change(fn=update_leaderboard, inputs=[dataset_dropdown, timeout_dropdown], outputs=leaderboard)
|
| 97 |
|
| 98 |
+
with gr.Tab("XXX"):
|
| 99 |
+
file_upload = gr.File(label="Upload JSON File")
|
| 100 |
+
upload_button = gr.Button("Process File")
|
| 101 |
+
output_text = gr.Textbox(label="Output")
|
| 102 |
+
|
| 103 |
+
upload_button.click(process_uploaded_file, inputs=file_upload, outputs=output_text)
|
| 104 |
+
|
| 105 |
with gr.Accordion("Citation", open=True):
|
| 106 |
gr.Markdown(citation_md, elem_id="leaderboard_markdown")
|
| 107 |
gr.Markdown(acknowledgment_md, elem_id="ack_markdown")
|
|
|
requirements.txt
CHANGED
|
@@ -1 +1,3 @@
|
|
| 1 |
-
plotly
|
|
|
|
|
|
|
|
|
| 1 |
+
plotly
|
| 2 |
+
line_profiler
|
| 3 |
+
memory_profiler
|