Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -2,60 +2,65 @@ import gradio as gr
|
|
| 2 |
from bertopic import BERTopic
|
| 3 |
from sentence_transformers import SentenceTransformer
|
| 4 |
|
|
|
|
| 5 |
def run_from_textfile(file):
|
| 6 |
if file is None:
|
| 7 |
return "Please upload a .txt file.", "", None
|
| 8 |
|
| 9 |
-
#
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
docs = [line.strip() for line in text.split("\n") if line.strip()]
|
| 12 |
|
| 13 |
if len(docs) < 3:
|
| 14 |
return "Need at least 3 documents (one per line).", "", None
|
| 15 |
|
| 16 |
-
# Embedding
|
| 17 |
embedder = SentenceTransformer("all-MiniLM-L6-v2")
|
| 18 |
|
| 19 |
-
# Topic
|
| 20 |
topic_model = BERTopic(embedding_model=embedder)
|
| 21 |
topics, probs = topic_model.fit_transform(docs)
|
| 22 |
|
| 23 |
-
# Topic
|
| 24 |
topic_info = topic_model.get_topic_info().to_string()
|
| 25 |
|
| 26 |
-
#
|
| 27 |
assignments = "\n".join([f"Doc {i+1}: Topic {topics[i]}" for i in range(len(docs))])
|
| 28 |
|
| 29 |
-
# Visualization
|
| 30 |
fig = topic_model.visualize_barchart(top_n_topics=10)
|
| 31 |
|
| 32 |
return topic_info, assignments, fig
|
| 33 |
|
| 34 |
|
|
|
|
| 35 |
with gr.Blocks() as demo:
|
| 36 |
gr.Markdown("# 🧠 Topic Modeling from TXT File (BERTopic)")
|
| 37 |
gr.Markdown(
|
| 38 |
-
"Upload a plain text (.txt) file. Each line should contain one LLM response.\n
|
| 39 |
-
"
|
| 40 |
-
"```\n"
|
| 41 |
-
"Response 1...\n"
|
| 42 |
-
"Response 2...\n"
|
| 43 |
-
"Response 3...\n"
|
| 44 |
-
"```\n"
|
| 45 |
)
|
| 46 |
|
| 47 |
file_input = gr.File(label="Upload .txt file")
|
| 48 |
|
| 49 |
-
|
| 50 |
|
| 51 |
-
topic_output = gr.Textbox(label="Topic Overview", lines=
|
| 52 |
-
assignment_output = gr.Textbox(label="Document → Topic Assignments", lines=
|
| 53 |
fig_output = gr.Plot(label="Topic Visualization")
|
| 54 |
|
| 55 |
-
|
| 56 |
fn=run_from_textfile,
|
| 57 |
inputs=file_input,
|
| 58 |
outputs=[topic_output, assignment_output, fig_output]
|
| 59 |
)
|
| 60 |
|
|
|
|
| 61 |
demo.launch()
|
|
|
|
| 2 |
from bertopic import BERTopic
|
| 3 |
from sentence_transformers import SentenceTransformer
|
| 4 |
|
| 5 |
+
|
| 6 |
def run_from_textfile(file):
|
| 7 |
if file is None:
|
| 8 |
return "Please upload a .txt file.", "", None
|
| 9 |
|
| 10 |
+
# ---- Handle file input for both HuggingFace and local environments ----
|
| 11 |
+
try:
|
| 12 |
+
# HuggingFace Spaces: file is NamedString and supports .decode()
|
| 13 |
+
text = file.decode("utf-8")
|
| 14 |
+
except:
|
| 15 |
+
# Local Gradio: file is a TemporaryFile-like object
|
| 16 |
+
text = file.read().decode("utf-8")
|
| 17 |
+
|
| 18 |
+
# Split the text into documents (one per line)
|
| 19 |
docs = [line.strip() for line in text.split("\n") if line.strip()]
|
| 20 |
|
| 21 |
if len(docs) < 3:
|
| 22 |
return "Need at least 3 documents (one per line).", "", None
|
| 23 |
|
| 24 |
+
# ---- Embedding Model ----
|
| 25 |
embedder = SentenceTransformer("all-MiniLM-L6-v2")
|
| 26 |
|
| 27 |
+
# ---- Topic Modeling ----
|
| 28 |
topic_model = BERTopic(embedding_model=embedder)
|
| 29 |
topics, probs = topic_model.fit_transform(docs)
|
| 30 |
|
| 31 |
+
# ---- Topic Summary ----
|
| 32 |
topic_info = topic_model.get_topic_info().to_string()
|
| 33 |
|
| 34 |
+
# ---- Document → Topic Assignments ----
|
| 35 |
assignments = "\n".join([f"Doc {i+1}: Topic {topics[i]}" for i in range(len(docs))])
|
| 36 |
|
| 37 |
+
# ---- Visualization ----
|
| 38 |
fig = topic_model.visualize_barchart(top_n_topics=10)
|
| 39 |
|
| 40 |
return topic_info, assignments, fig
|
| 41 |
|
| 42 |
|
| 43 |
+
# ---- Gradio Interface ----
|
| 44 |
with gr.Blocks() as demo:
|
| 45 |
gr.Markdown("# 🧠 Topic Modeling from TXT File (BERTopic)")
|
| 46 |
gr.Markdown(
|
| 47 |
+
"Upload a plain text (.txt) file. Each line should contain **one LLM response**.\n"
|
| 48 |
+
"\nExample format:\n```\nResponse 1...\nResponse 2...\nResponse 3...\n```"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
)
|
| 50 |
|
| 51 |
file_input = gr.File(label="Upload .txt file")
|
| 52 |
|
| 53 |
+
run_button = gr.Button("Run Topic Modeling")
|
| 54 |
|
| 55 |
+
topic_output = gr.Textbox(label="Topic Overview", lines=12)
|
| 56 |
+
assignment_output = gr.Textbox(label="Document → Topic Assignments", lines=12)
|
| 57 |
fig_output = gr.Plot(label="Topic Visualization")
|
| 58 |
|
| 59 |
+
run_button.click(
|
| 60 |
fn=run_from_textfile,
|
| 61 |
inputs=file_input,
|
| 62 |
outputs=[topic_output, assignment_output, fig_output]
|
| 63 |
)
|
| 64 |
|
| 65 |
+
# Launch app
|
| 66 |
demo.launch()
|