Abdullah Zaki
commited on
Commit
·
574b1b5
1
Parent(s):
7aa3e0f
Add plotly t
Browse files
app.py
CHANGED
@@ -8,14 +8,9 @@ from supabase import create_client, Client
|
|
8 |
import os
|
9 |
import plotly.express as px
|
10 |
|
11 |
-
# Initialize Supabase client with API key from environment variables
|
12 |
-
SUPABASE_URL = os.getenv("SUPABASE_URL")
|
13 |
-
SUPABASE_KEY = os.getenv("SUPABASE_KEY")
|
14 |
-
if not SUPABASE_URL or not SUPABASE_KEY:
|
15 |
-
raise ValueError("SUPABASE_URL and SUPABASE_KEY must be set as environment variables.")
|
16 |
-
supabase: Client = create_client(SUPABASE_URL, SUPABASE_KEY)
|
17 |
-
|
18 |
# Initialize Chronos-T5-Large for forecasting
|
|
|
|
|
19 |
chronos_pipeline = ChronosPipeline.from_pretrained(
|
20 |
"amazon/chronos-t5-large",
|
21 |
device_map="cuda" if torch.cuda.is_available() else "cpu",
|
@@ -23,6 +18,7 @@ chronos_pipeline = ChronosPipeline.from_pretrained(
|
|
23 |
)
|
24 |
|
25 |
# Initialize Prophet-Qwen3-4B-SFT for Arabic reports
|
|
|
26 |
qwen_tokenizer = AutoTokenizer.from_pretrained("radm/prophet-qwen3-4b-sft")
|
27 |
qwen_model = AutoModelForCausalLM.from_pretrained(
|
28 |
"radm/prophet-qwen3-4b-sft",
|
@@ -30,41 +26,98 @@ qwen_model = AutoModelForCausalLM.from_pretrained(
|
|
30 |
torch_dtype=torch.bfloat16
|
31 |
)
|
32 |
|
33 |
-
def fetch_supabase_data(table_name: str = "sentiment_data") -> pd.DataFrame:
|
34 |
-
"""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
35 |
try:
|
36 |
-
|
|
|
|
|
|
|
|
|
37 |
if response.data:
|
38 |
df = pd.DataFrame(response.data)
|
|
|
39 |
df['date'] = pd.to_datetime(df['date'])
|
|
|
|
|
40 |
return df
|
41 |
else:
|
42 |
-
raise ValueError("No data found in Supabase table.")
|
43 |
except Exception as e:
|
44 |
raise Exception(f"Error fetching Supabase data: {str(e)}")
|
45 |
|
46 |
-
def forecast_and_report(
|
47 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
48 |
try:
|
49 |
-
# Load data
|
|
|
50 |
if data_source == "Supabase":
|
51 |
-
df = fetch_supabase_data(table_name)
|
52 |
-
else:
|
53 |
-
if
|
54 |
-
return {"error": "Please upload a CSV file."}, None, None
|
55 |
-
df = pd.read_csv(csv_file)
|
|
|
56 |
if "sentiment" not in df.columns or "date" not in df.columns:
|
57 |
-
return {"error": "CSV must contain 'date' and 'sentiment' columns."}, None, None
|
58 |
df['date'] = pd.to_datetime(df['date'])
|
|
|
|
|
|
|
|
|
|
|
59 |
|
60 |
-
# Prepare time series
|
|
|
61 |
context = torch.tensor(df["sentiment"].values, dtype=torch.float32)
|
62 |
|
63 |
-
# Run forecast
|
64 |
forecast = chronos_pipeline.predict(context, prediction_length)
|
|
|
65 |
low, median, high = np.quantile(forecast[0].numpy(), [0.1, 0.5, 0.9], axis=0)
|
66 |
|
67 |
-
# Format forecast results
|
|
|
68 |
forecast_dates = pd.date_range(start=df["date"].iloc[-1] + pd.Timedelta(days=1), periods=prediction_length, freq="D")
|
69 |
forecast_df = pd.DataFrame({
|
70 |
"date": forecast_dates,
|
@@ -73,14 +126,15 @@ def forecast_and_report(data_source: str, csv_file=None, prediction_length: int
|
|
73 |
"high": high
|
74 |
})
|
75 |
|
76 |
-
# Create forecast plot
|
77 |
-
|
78 |
-
fig = px.line(
|
79 |
fig.update_traces(line=dict(color="blue"), selector=dict(name="median"))
|
80 |
fig.update_traces(line=dict(color="red", dash="dash"), selector=dict(name="low"))
|
81 |
fig.update_traces(line=dict(color="green", dash="dash"), selector=dict(name="high"))
|
82 |
|
83 |
-
# Generate Arabic report
|
|
|
84 |
prompt = (
|
85 |
"اكتب تقريراً رسمياً بالعربية يلخص توقعات المشاعر للأيام الثلاثين القادمة بناءً على البيانات التالية:\n"
|
86 |
f"- متوسط التوقعات: {median[:5].tolist()} (أول 5 أيام)...\n"
|
@@ -88,37 +142,59 @@ def forecast_and_report(data_source: str, csv_file=None, prediction_length: int
|
|
88 |
f"- الحد الأعلى (90%): {high[:5].tolist()}...\n"
|
89 |
"التقرير يجب أن يكون موجزاً (200-300 كلمة)، يشرح الاتجاهات، ويستخدم لغة رسمية."
|
90 |
)
|
|
|
91 |
inputs = qwen_tokenizer(prompt, return_tensors="pt").to(qwen_model.device)
|
|
|
92 |
outputs = qwen_model.generate(
|
93 |
inputs["input_ids"],
|
94 |
-
max_new_tokens=500,
|
95 |
-
do_sample=True,
|
96 |
-
temperature=0.7,
|
97 |
-
top_p=0.9
|
98 |
)
|
|
|
99 |
report = qwen_tokenizer.decode(outputs[0], skip_special_tokens=True)
|
100 |
|
101 |
-
return forecast_df.to_dict(), fig, report
|
102 |
-
|
103 |
except Exception as e:
|
104 |
-
|
|
|
105 |
|
106 |
-
# Gradio interface
|
107 |
with gr.Blocks() as demo:
|
108 |
gr.Markdown("# Sentiment Forecasting and Arabic Reporting")
|
109 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
110 |
csv_file = gr.File(label="Upload CSV (if CSV selected)")
|
111 |
table_name = gr.Textbox(label="Supabase Table Name", value="sentiment_data")
|
112 |
prediction_length = gr.Slider(1, 60, value=30, step=1, label="Prediction Length (days)")
|
|
|
113 |
submit = gr.Button("Run Forecast and Generate Report")
|
|
|
|
|
114 |
output = gr.JSON(label="Forecast Results")
|
115 |
plot = gr.Plot(label="Forecast Plot")
|
116 |
-
report = gr.Textbox(label="Arabic Report", lines=10)
|
|
|
117 |
|
|
|
118 |
submit.click(
|
119 |
fn=forecast_and_report,
|
120 |
-
inputs=[
|
121 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
122 |
)
|
123 |
|
124 |
-
|
|
|
|
8 |
import os
|
9 |
import plotly.express as px
|
10 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
# Initialize Chronos-T5-Large for forecasting
|
12 |
+
# These models are loaded once at the start of the Gradio app for efficiency.
|
13 |
+
# The device_map automatically handles CPU/GPU allocation.
|
14 |
chronos_pipeline = ChronosPipeline.from_pretrained(
|
15 |
"amazon/chronos-t5-large",
|
16 |
device_map="cuda" if torch.cuda.is_available() else "cpu",
|
|
|
18 |
)
|
19 |
|
20 |
# Initialize Prophet-Qwen3-4B-SFT for Arabic reports
|
21 |
+
# These models are also loaded once at the start.
|
22 |
qwen_tokenizer = AutoTokenizer.from_pretrained("radm/prophet-qwen3-4b-sft")
|
23 |
qwen_model = AutoModelForCausalLM.from_pretrained(
|
24 |
"radm/prophet-qwen3-4b-sft",
|
|
|
26 |
torch_dtype=torch.bfloat16
|
27 |
)
|
28 |
|
29 |
+
def fetch_supabase_data(supabase_url: str, supabase_key: str, table_name: str = "sentiment_data") -> pd.DataFrame:
|
30 |
+
"""
|
31 |
+
Fetches time series data from Supabase using the provided URL and API key.
|
32 |
+
|
33 |
+
Args:
|
34 |
+
supabase_url (str): The URL of your Supabase project.
|
35 |
+
supabase_key (str): Your Supabase API key (anon key).
|
36 |
+
table_name (str): The name of the table to fetch data from.
|
37 |
+
|
38 |
+
Returns:
|
39 |
+
pd.DataFrame: A DataFrame containing 'date' and 'sentiment' columns.
|
40 |
+
|
41 |
+
Raises:
|
42 |
+
Exception: If there's an error connecting to Supabase or no data is found.
|
43 |
+
"""
|
44 |
+
if not supabase_url or not supabase_key:
|
45 |
+
raise ValueError("Supabase URL and Key must be provided to fetch data from Supabase.")
|
46 |
+
|
47 |
try:
|
48 |
+
# Create a new Supabase client instance for each call, using the provided URL and key.
|
49 |
+
# This allows the user to input different keys/URLs without restarting the app.
|
50 |
+
supabase_client: Client = create_client(supabase_url, supabase_key)
|
51 |
+
response = supabase_client.table(table_name).select("date, sentiment").order("date", desc=False).execute()
|
52 |
+
|
53 |
if response.data:
|
54 |
df = pd.DataFrame(response.data)
|
55 |
+
# Ensure 'date' column is in datetime format
|
56 |
df['date'] = pd.to_datetime(df['date'])
|
57 |
+
# Ensure 'sentiment' column is numeric for forecasting
|
58 |
+
df['sentiment'] = pd.to_numeric(df['sentiment'])
|
59 |
return df
|
60 |
else:
|
61 |
+
raise ValueError(f"No data found in Supabase table '{table_name}'.")
|
62 |
except Exception as e:
|
63 |
raise Exception(f"Error fetching Supabase data: {str(e)}")
|
64 |
|
65 |
+
def forecast_and_report(
|
66 |
+
data_source: str,
|
67 |
+
supabase_url: str, # New input for Supabase URL
|
68 |
+
supabase_key: str, # New input for Supabase Key
|
69 |
+
csv_file=None,
|
70 |
+
prediction_length: int = 30,
|
71 |
+
table_name: str = "sentiment_data"
|
72 |
+
):
|
73 |
+
"""
|
74 |
+
Runs forecasting with Chronos-T5-Large and generates an Arabic report with Qwen3-4B-SFT.
|
75 |
+
|
76 |
+
Args:
|
77 |
+
data_source (str): Specifies whether to use "Supabase" or "CSV Upload".
|
78 |
+
supabase_url (str): The Supabase project URL (used if data_source is "Supabase").
|
79 |
+
supabase_key (str): The Supabase API key (used if data_source is "Supabase").
|
80 |
+
csv_file: The uploaded CSV file (used if data_source is "CSV Upload").
|
81 |
+
prediction_length (int): The number of days to forecast.
|
82 |
+
table_name (str): The name of the Supabase table.
|
83 |
+
|
84 |
+
Returns:
|
85 |
+
tuple: A tuple containing:
|
86 |
+
- dict: Forecast results as a dictionary.
|
87 |
+
- plotly.graph_objects.Figure: A Plotly figure of the forecast.
|
88 |
+
- str: The generated Arabic report.
|
89 |
+
- str: An error message if an error occurs.
|
90 |
+
"""
|
91 |
try:
|
92 |
+
# Load data based on selected source
|
93 |
+
df = pd.DataFrame() # Initialize df to avoid UnboundLocalError
|
94 |
if data_source == "Supabase":
|
95 |
+
df = fetch_supabase_data(supabase_url, supabase_key, table_name)
|
96 |
+
else: # data_source == "CSV Upload"
|
97 |
+
if csv_file is None:
|
98 |
+
return {"error": "Please upload a CSV file when 'CSV Upload' is selected."}, None, None, "Error: CSV file not provided."
|
99 |
+
df = pd.read_csv(csv_file.name) # Access the file path
|
100 |
+
# Basic validation for required columns in CSV
|
101 |
if "sentiment" not in df.columns or "date" not in df.columns:
|
102 |
+
return {"error": "CSV must contain 'date' and 'sentiment' columns."}, None, None, "Error: Missing 'date' or 'sentiment' columns in CSV."
|
103 |
df['date'] = pd.to_datetime(df['date'])
|
104 |
+
df['sentiment'] = pd.to_numeric(df['sentiment'])
|
105 |
+
|
106 |
+
# Ensure there's data to process
|
107 |
+
if df.empty:
|
108 |
+
return {"error": "No data available for forecasting or reporting."}, None, None, "Error: No data available."
|
109 |
|
110 |
+
# Prepare time series data for Chronos
|
111 |
+
# Ensure sentiment is float32 for the model
|
112 |
context = torch.tensor(df["sentiment"].values, dtype=torch.float32)
|
113 |
|
114 |
+
# Run forecast using Chronos-T5-Large pipeline
|
115 |
forecast = chronos_pipeline.predict(context, prediction_length)
|
116 |
+
# Calculate quantiles for low, median, and high predictions
|
117 |
low, median, high = np.quantile(forecast[0].numpy(), [0.1, 0.5, 0.9], axis=0)
|
118 |
|
119 |
+
# Format forecast results into a DataFrame
|
120 |
+
# Generate future dates starting from the day after the last historical date
|
121 |
forecast_dates = pd.date_range(start=df["date"].iloc[-1] + pd.Timedelta(days=1), periods=prediction_length, freq="D")
|
122 |
forecast_df = pd.DataFrame({
|
123 |
"date": forecast_dates,
|
|
|
126 |
"high": high
|
127 |
})
|
128 |
|
129 |
+
# Create forecast plot using Plotly
|
130 |
+
# Combine historical data for plotting if desired, but here we plot only forecast
|
131 |
+
fig = px.line(forecast_df, x="date", y=["median", "low", "high"], title="Sentiment Forecast")
|
132 |
fig.update_traces(line=dict(color="blue"), selector=dict(name="median"))
|
133 |
fig.update_traces(line=dict(color="red", dash="dash"), selector=dict(name="low"))
|
134 |
fig.update_traces(line=dict(color="green", dash="dash"), selector=dict(name="high"))
|
135 |
|
136 |
+
# Generate Arabic report using Prophet-Qwen3-4B-SFT
|
137 |
+
# Construct the prompt with relevant forecast snippets
|
138 |
prompt = (
|
139 |
"اكتب تقريراً رسمياً بالعربية يلخص توقعات المشاعر للأيام الثلاثين القادمة بناءً على البيانات التالية:\n"
|
140 |
f"- متوسط التوقعات: {median[:5].tolist()} (أول 5 أيام)...\n"
|
|
|
142 |
f"- الحد الأعلى (90%): {high[:5].tolist()}...\n"
|
143 |
"التقرير يجب أن يكون موجزاً (200-300 كلمة)، يشرح الاتجاهات، ويستخدم لغة رسمية."
|
144 |
)
|
145 |
+
# Tokenize the prompt and move to the model's device (CPU/GPU)
|
146 |
inputs = qwen_tokenizer(prompt, return_tensors="pt").to(qwen_model.device)
|
147 |
+
# Generate the report text
|
148 |
outputs = qwen_model.generate(
|
149 |
inputs["input_ids"],
|
150 |
+
max_new_tokens=500, # Max length for the generated report
|
151 |
+
do_sample=True, # Enable sampling for more diverse text
|
152 |
+
temperature=0.7, # Control randomness (lower for less random)
|
153 |
+
top_p=0.9 # Nucleus sampling parameter
|
154 |
)
|
155 |
+
# Decode the generated tokens back to text, skipping special tokens
|
156 |
report = qwen_tokenizer.decode(outputs[0], skip_special_tokens=True)
|
157 |
|
158 |
+
return forecast_df.to_dict(), fig, report, "Success" # Return success message
|
|
|
159 |
except Exception as e:
|
160 |
+
# Catch any exceptions and return an error message
|
161 |
+
return {}, None, None, f"An error occurred: {str(e)}"
|
162 |
|
163 |
+
# Gradio interface definition
|
164 |
with gr.Blocks() as demo:
|
165 |
gr.Markdown("# Sentiment Forecasting and Arabic Reporting")
|
166 |
+
|
167 |
+
# Input components for Supabase credentials and data source selection
|
168 |
+
with gr.Row():
|
169 |
+
data_source = gr.Radio(["Supabase", "CSV Upload"], label="Data Source", value="Supabase")
|
170 |
+
supabase_url = gr.Textbox(label="Supabase URL", placeholder="e.g., https://your-project-ref.supabase.co", interactive=True)
|
171 |
+
supabase_key = gr.Textbox(label="Supabase Key", placeholder="Your Supabase anon key", type="password", interactive=True)
|
172 |
+
|
173 |
csv_file = gr.File(label="Upload CSV (if CSV selected)")
|
174 |
table_name = gr.Textbox(label="Supabase Table Name", value="sentiment_data")
|
175 |
prediction_length = gr.Slider(1, 60, value=30, step=1, label="Prediction Length (days)")
|
176 |
+
|
177 |
submit = gr.Button("Run Forecast and Generate Report")
|
178 |
+
|
179 |
+
# Output components for results
|
180 |
output = gr.JSON(label="Forecast Results")
|
181 |
plot = gr.Plot(label="Forecast Plot")
|
182 |
+
report = gr.Textbox(label="Arabic Report", lines=10, rtl=True, show_copy_button=True) # Added rtl=True for Arabic display
|
183 |
+
status_message = gr.Textbox(label="Status", interactive=False) # For displaying success/error messages
|
184 |
|
185 |
+
# Define the click event handler for the submit button
|
186 |
submit.click(
|
187 |
fn=forecast_and_report,
|
188 |
+
inputs=[
|
189 |
+
data_source,
|
190 |
+
supabase_url,
|
191 |
+
supabase_key,
|
192 |
+
csv_file,
|
193 |
+
prediction_length,
|
194 |
+
table_name
|
195 |
+
],
|
196 |
+
outputs=[output, plot, report, status_message]
|
197 |
)
|
198 |
|
199 |
+
# Launch the Gradio application
|
200 |
+
demo.launch()
|