Spaces:
Sleeping
Sleeping
File size: 4,003 Bytes
4022660 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 | # app.py
import os, pathlib, pandas as pd, gradio as gr
from agents import run_pipeline
from files_process import prepare_input_arg, load_input_text # load_input_text not used here but handy
def _scores_to_df(result_json: dict) -> pd.DataFrame:
rows = result_json.get("scores", []) or []
df = pd.DataFrame(rows)
# Drop justification from table
if "justification" in df.columns:
df = df.drop(columns=["justification"])
cols = ["agent", "clinical_completeness", "ai_rigor",
"trial_framing", "privacy_regulatory", "clarity_structure"]
for c in cols:
if c not in df.columns:
df[c] = None
return df[cols]
def _winner_md_block(result_json: dict) -> str:
winner = result_json.get("winner", "") or "N/A"
just = ""
for r in result_json.get("scores", []):
if r.get("agent") == winner:
just = r.get("justification", "") or ""
break
if just:
return f"### 🏆 Winner: **{winner}**\n\n> *{just}*"
return f"### 🏆 Winner: **{winner}**"
def run_ui(text_in, file_in, oai_model, gem_model, ds_model):
try:
input_arg = prepare_input_arg(text_in, file_in)
result_json = run_pipeline(
input_arg,
oai_model=oai_model,
gem_model=gem_model,
ds_model=ds_model,
)
# Read agent drafts saved by run_pipeline
p1, p2, p3 = [pathlib.Path(f"agent{i}.md") for i in range(1, 4)]
agent1_md = p1.read_text(encoding="utf-8") if p1.exists() else "*agent1.md not found*"
agent2_md = p2.read_text(encoding="utf-8") if p2.exists() else "*agent2.md not found*"
agent3_md = p3.read_text(encoding="utf-8") if p3.exists() else "*agent3.md not found*"
scores_df = _scores_to_df(result_json)
winner_md = _winner_md_block(result_json)
return (
agent1_md, agent2_md, agent3_md,
scores_df, winner_md,
str(p1) if p1.exists() else None,
str(p2) if p2.exists() else None,
str(p3) if p3.exists() else None
)
except Exception as e:
# Keep output shapes consistent
return f"**Error:** {e}", "", "", pd.DataFrame(), "", None, None, None
with gr.Blocks(title="Healthcare–AI Case Studies (3 Agents + Manager)") as demo:
gr.Markdown("# Healthcare–AI Case Studies\nProvide text or upload a .txt/.docx/.pdf, then click **Run**.")
with gr.Accordion("Models (optional)", open=False):
m1 = gr.Textbox(value="gpt-4o-mini", label="Agent 1 (OpenAI)")
m2 = gr.Textbox(value="gpt-4.1-nano", label="Agent 2 (style-2)")
m3 = gr.Textbox(value="gpt-4.1-mini", label="Agent 3 (style-3)")
gr.Markdown("### Manager Scores")
scores_df = gr.Dataframe(label="Scores (justification hidden)")
winner_md = gr.Markdown(label="Winner & rationale")
gr.Markdown("### Download agent drafts")
with gr.Row():
dl1 = gr.DownloadButton(label="Download agent1.md")
dl2 = gr.DownloadButton(label="Download agent2.md")
dl3 = gr.DownloadButton(label="Download agent3.md")
gr.Markdown("### Input")
with gr.Row():
txt = gr.Textbox(lines=10, label="Paste source text (optional)")
fil = gr.File(label="Upload .txt / .docx / .pdf (optional)", file_count="single",
file_types=["text", ".docx", ".pdf"])
run_btn = gr.Button("Run")
gr.Markdown("### Agent Drafts (expand to view)")
with gr.Accordion("Agent outputs", open=False):
with gr.Row():
a1_md = gr.Markdown(label="Agent 1 draft")
a2_md = gr.Markdown(label="Agent 2 draft")
a3_md = gr.Markdown(label="Agent 3 draft")
run_btn.click(
fn=run_ui,
inputs=[txt, fil, m1, m2, m3],
outputs=[a1_md, a2_md, a3_md, scores_df, winner_md, dl1, dl2, dl3],
)
# On Spaces, it's enough to expose `demo`; running locally calls launch().
if __name__ == "__main__":
demo.launch() |