Spaces:
Runtime error
Runtime error
Upload folder using huggingface_hub
Browse files- README.md +4 -11
- app.py +121 -0
- requirements.txt +2 -0
README.md
CHANGED
|
@@ -1,12 +1,5 @@
|
|
| 1 |
-
|
| 2 |
-
title: Kakoverse Browser
|
| 3 |
-
emoji: 🚀
|
| 4 |
-
colorFrom: red
|
| 5 |
-
colorTo: yellow
|
| 6 |
-
sdk: gradio
|
| 7 |
-
sdk_version: 5.49.1
|
| 8 |
-
app_file: app.py
|
| 9 |
-
pinned: false
|
| 10 |
-
---
|
| 11 |
|
| 12 |
-
|
|
|
|
|
|
|
|
|
| 1 |
+
# KakoVerse Conversation Browser
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2 |
|
| 3 |
+
Load a KakoVerse conversation dataset from the Hugging Face Hub and inspect each dialogue in a chat-style UI. Double-click any row to open the full conversation with seeker/supporter bubbles and care levels.
|
| 4 |
+
|
| 5 |
+
The default dataset is `Reza2kn/kakoverse-conversations-v0`.
|
app.py
ADDED
|
@@ -0,0 +1,121 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import json
|
| 2 |
+
from functools import lru_cache
|
| 3 |
+
from typing import Any, Dict, List, Tuple
|
| 4 |
+
|
| 5 |
+
import gradio as gr
|
| 6 |
+
from datasets import Dataset, load_dataset
|
| 7 |
+
|
| 8 |
+
DEFAULT_DATASET_ID = "reza2kn/kakoverse-conversations-v0"
|
| 9 |
+
|
| 10 |
+
|
| 11 |
+
@lru_cache(maxsize=4)
|
| 12 |
+
def _load_dataset(repo_id: str) -> Dataset:
|
| 13 |
+
return load_dataset(repo_id)["train"]
|
| 14 |
+
|
| 15 |
+
|
| 16 |
+
def load_conversations(repo_id: str) -> Tuple[List[Dict[str, Any]], List[List[Any]]]:
|
| 17 |
+
dataset = _load_dataset(repo_id)
|
| 18 |
+
table_rows: List[List[Any]] = []
|
| 19 |
+
records: List[Dict[str, Any]] = []
|
| 20 |
+
|
| 21 |
+
for row in dataset:
|
| 22 |
+
records.append(row)
|
| 23 |
+
table_rows.append(
|
| 24 |
+
[
|
| 25 |
+
row["persona_id"],
|
| 26 |
+
row.get("crisis_category"),
|
| 27 |
+
row.get("crisis_summary"),
|
| 28 |
+
row.get("turns_recorded"),
|
| 29 |
+
row.get("desired_turns"),
|
| 30 |
+
]
|
| 31 |
+
)
|
| 32 |
+
return records, table_rows
|
| 33 |
+
|
| 34 |
+
|
| 35 |
+
def conversation_to_html(conversation: Dict[str, Any]) -> str:
|
| 36 |
+
turns = conversation.get("turns") or []
|
| 37 |
+
persona_id = conversation.get("persona_id")
|
| 38 |
+
crisis_category = conversation.get("crisis_category")
|
| 39 |
+
crisis_summary = conversation.get("crisis_summary")
|
| 40 |
+
|
| 41 |
+
head = f"""
|
| 42 |
+
<style>
|
| 43 |
+
.thread {{ font-family: Inter, sans-serif; max-width: 760px; margin: 1rem auto; }}
|
| 44 |
+
.meta {{ margin-bottom: 1rem; color: #444; }}
|
| 45 |
+
.bubble {{ border-radius: 14px; padding: 12px 16px; margin: 6px 0; display: inline-block; max-width: 90%; }}
|
| 46 |
+
.seeker {{ background-color: #1f2937; color: white; float: left; clear: both; }}
|
| 47 |
+
.supporter {{ background-color: #e5f3ff; color: #0f172a; float: right; clear: both; }}
|
| 48 |
+
.cold {{ background-color: #f5f5f5; color: #1f2937; }}
|
| 49 |
+
.care-pill {{ font-size: 12px; margin-left: 6px; padding: 2px 6px; border-radius: 999px; background: rgba(15, 23, 42, 0.1); }}
|
| 50 |
+
.turn {{ overflow: hidden; padding-bottom: 12px; }}
|
| 51 |
+
</style>
|
| 52 |
+
<div class="thread">
|
| 53 |
+
<div class="meta"><strong>Persona:</strong> {persona_id} | <strong>Crisis:</strong> {crisis_category} — {crisis_summary}</div>
|
| 54 |
+
"""
|
| 55 |
+
|
| 56 |
+
body_parts = [head]
|
| 57 |
+
for turn in turns:
|
| 58 |
+
seeker_msg = turn.get("seeker_message", "")
|
| 59 |
+
body_parts.append(f'<div class="turn"><div class="bubble seeker">{seeker_msg}</div></div>')
|
| 60 |
+
for variant in turn.get("supporter_responses", []):
|
| 61 |
+
care_level = variant.get("care_level")
|
| 62 |
+
label = variant.get("style")
|
| 63 |
+
classes = "bubble supporter"
|
| 64 |
+
if label == "cold":
|
| 65 |
+
classes += " cold"
|
| 66 |
+
message_html = variant.get("message", "")
|
| 67 |
+
body_parts.append(
|
| 68 |
+
f'<div class="bubble {classes}">{message_html}'
|
| 69 |
+
f'<span class="care-pill">{label} · care={care_level}</span></div>'
|
| 70 |
+
)
|
| 71 |
+
body_parts.append("</div>")
|
| 72 |
+
return "".join(body_parts)
|
| 73 |
+
|
| 74 |
+
|
| 75 |
+
def on_load(repo_id: str):
|
| 76 |
+
records, table_rows = load_conversations(repo_id)
|
| 77 |
+
table_update = gr.update(value=table_rows, visible=True)
|
| 78 |
+
status_update = gr.update(value=f"Loaded {len(records)} conversations from {repo_id}", visible=True)
|
| 79 |
+
return records, table_update, status_update
|
| 80 |
+
|
| 81 |
+
|
| 82 |
+
def on_select(evt: gr.SelectData, records: List[Dict[str, Any]]) -> str:
|
| 83 |
+
row_index = evt.index[0]
|
| 84 |
+
if row_index < 0 or row_index >= len(records):
|
| 85 |
+
return "<div>Select a valid conversation row.</div>"
|
| 86 |
+
record = records[row_index]
|
| 87 |
+
return conversation_to_html(record)
|
| 88 |
+
|
| 89 |
+
|
| 90 |
+
with gr.Blocks(css=".footer {display:none !important;}") as demo:
|
| 91 |
+
gr.Markdown("# KakoVerse Conversations Browser")
|
| 92 |
+
gr.Markdown(
|
| 93 |
+
"Load a conversation dataset from the Hugging Face Hub, then double-click any row to view it in a chat-style layout."
|
| 94 |
+
)
|
| 95 |
+
|
| 96 |
+
dataset_state = gr.State([])
|
| 97 |
+
with gr.Row():
|
| 98 |
+
dataset_id = gr.Textbox(label="Dataset repo", value=DEFAULT_DATASET_ID, scale=4)
|
| 99 |
+
browse_btn = gr.Button("Browse conversations", scale=1)
|
| 100 |
+
status = gr.Markdown("", visible=False)
|
| 101 |
+
|
| 102 |
+
table = gr.Dataframe(
|
| 103 |
+
headers=["persona_id", "crisis_category", "crisis_summary", "turns_recorded", "desired_turns"],
|
| 104 |
+
datatype=["str", "str", "str", "int", "int"],
|
| 105 |
+
row_count=0,
|
| 106 |
+
col_count=5,
|
| 107 |
+
interactive=False,
|
| 108 |
+
visible=False,
|
| 109 |
+
)
|
| 110 |
+
chat_panel = gr.HTML()
|
| 111 |
+
|
| 112 |
+
browse_btn.click(
|
| 113 |
+
on_load,
|
| 114 |
+
inputs=[dataset_id],
|
| 115 |
+
outputs=[dataset_state, table, status],
|
| 116 |
+
)
|
| 117 |
+
table.select(on_select, inputs=[dataset_state], outputs=chat_panel)
|
| 118 |
+
|
| 119 |
+
|
| 120 |
+
if __name__ == "__main__":
|
| 121 |
+
demo.launch()
|
requirements.txt
ADDED
|
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
gradio>=4.44.0
|
| 2 |
+
datasets>=2.20.0
|