Spaces:
Sleeping
Sleeping
Added updated intent
Browse files- Dockerfile +4 -10
- main.py +16 -16
- voice/intent_classifier.py +22 -22
Dockerfile
CHANGED
@@ -1,4 +1,3 @@
|
|
1 |
-
# Use official Python image
|
2 |
FROM python:3.10-slim
|
3 |
|
4 |
# Install system dependencies
|
@@ -6,14 +5,12 @@ RUN apt-get update && \
|
|
6 |
apt-get install -y wget ffmpeg unzip curl gcc && \
|
7 |
rm -rf /var/lib/apt/lists/*
|
8 |
|
9 |
-
|
10 |
-
WORKDIR /app
|
11 |
|
12 |
-
|
13 |
-
ENV HF_HOME=/app/cache
|
14 |
|
15 |
-
|
16 |
-
|
17 |
|
18 |
# Copy your code into the container
|
19 |
COPY . .
|
@@ -22,9 +19,6 @@ COPY . .
|
|
22 |
RUN pip install --upgrade pip && pip install -r requirements.txt
|
23 |
RUN pip install --no-cache-dir torch==2.2.0+cpu torchvision==0.17.0+cpu torchaudio==2.2.0+cpu -f https://download.pytorch.org/whl/torch_stable.html
|
24 |
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
# Install spaCy model
|
29 |
RUN python -m spacy download en_core_web_lg
|
30 |
# RUN python -m spacy download en_core_web_sm
|
|
|
|
|
1 |
FROM python:3.10-slim
|
2 |
|
3 |
# Install system dependencies
|
|
|
5 |
apt-get install -y wget ffmpeg unzip curl gcc && \
|
6 |
rm -rf /var/lib/apt/lists/*
|
7 |
|
8 |
+
WORKDIR /app
|
|
|
9 |
|
10 |
+
ENV HF_HOME=/app/cache
|
|
|
11 |
|
12 |
+
# Create cache directory and set permissions
|
13 |
+
RUN mkdir -p /app/cache /app/db && chmod -R 777 /app/cache /app/db
|
14 |
|
15 |
# Copy your code into the container
|
16 |
COPY . .
|
|
|
19 |
RUN pip install --upgrade pip && pip install -r requirements.txt
|
20 |
RUN pip install --no-cache-dir torch==2.2.0+cpu torchvision==0.17.0+cpu torchaudio==2.2.0+cpu -f https://download.pytorch.org/whl/torch_stable.html
|
21 |
|
|
|
|
|
|
|
22 |
# Install spaCy model
|
23 |
RUN python -m spacy download en_core_web_lg
|
24 |
# RUN python -m spacy download en_core_web_sm
|
main.py
CHANGED
@@ -55,22 +55,22 @@ async def process_query(vosk_model_path, audio_data=None, query_text=None, use_r
|
|
55 |
|
56 |
if intent:
|
57 |
intent_to_module = {
|
58 |
-
"
|
59 |
-
"
|
60 |
-
"
|
61 |
-
"
|
62 |
-
"
|
63 |
-
"
|
64 |
-
"
|
65 |
-
"
|
66 |
-
"
|
67 |
-
"
|
68 |
-
"
|
69 |
-
"
|
70 |
-
"
|
71 |
-
"
|
72 |
-
"
|
73 |
-
"
|
74 |
|
75 |
}
|
76 |
|
|
|
55 |
|
56 |
if intent:
|
57 |
intent_to_module = {
|
58 |
+
"net_income": ("modules.get_net_income", "GetNetIncome"),
|
59 |
+
"revenue": ("modules.get_revenue", "GetRevenue"),
|
60 |
+
"stock_price": ("modules.get_stock_price", "GetStockPrice"),
|
61 |
+
"profit_margin": ("modules.get_profit_margin", "GetProfitMargin"),
|
62 |
+
"company_info": ("modules.get_company_profile", "GetCompanyProfile"),
|
63 |
+
"market_capitalization": ("modules.get_market_cap", "GetMarketCap"),
|
64 |
+
"historical_stock_price": ("modules.get_historical_stock_price", "GetHistoricalStockPrice"),
|
65 |
+
"dividend_info": ("modules.get_dividend_info", "GetDividendInfo"),
|
66 |
+
"balance_sheet": ("modules.get_balance_sheet", "GetBalanceSheet"),
|
67 |
+
"cash_flow": ("modules.get_cash_flow", "GetCashFlow"),
|
68 |
+
"financial_ratios": ("modules.get_financial_ratios", "GetFinancialRatios"),
|
69 |
+
"earnings_per_share": ("modules.get_earnings_per_share", "GetEarningsPerShare"),
|
70 |
+
"interest_rate": ("modules.get_interest", "GetInterest"),
|
71 |
+
"income_tax": ("modules.get_income_tax", "GetIncomeTax"),
|
72 |
+
"cost_info": ("modules.get_cost_info", "GetCostInfo"),
|
73 |
+
"research_info": ("modules.get_research_info", "GetResearchInfo")
|
74 |
|
75 |
}
|
76 |
|
voice/intent_classifier.py
CHANGED
@@ -12,22 +12,22 @@ class IntentClassifier:
|
|
12 |
self.nlp = spacy.load("en_core_web_lg") # "en_core_web_sm"
|
13 |
self.classifier = pipeline("zero-shot-classification", model="facebook/bart-large-mnli", from_pt=True)
|
14 |
self.intents = [
|
15 |
-
"
|
16 |
-
"
|
17 |
-
"
|
18 |
-
"
|
19 |
-
"
|
20 |
-
"
|
21 |
-
"
|
22 |
-
"
|
23 |
-
"
|
24 |
-
"
|
25 |
-
"
|
26 |
-
"
|
27 |
-
"
|
28 |
-
"
|
29 |
-
"
|
30 |
-
"
|
31 |
]
|
32 |
|
33 |
# Mapping of company names to ticker symbols (case-insensitive)
|
@@ -168,7 +168,7 @@ class IntentClassifier:
|
|
168 |
|
169 |
# Step 3: Extract metric using keyword matching with synonyms
|
170 |
text_lower = text.lower()
|
171 |
-
if any(keyword in text_lower for keyword in ["
|
172 |
entities["metric"] = "netIncome"
|
173 |
elif "revenue" in text_lower:
|
174 |
entities["metric"] = "revenue"
|
@@ -176,19 +176,19 @@ class IntentClassifier:
|
|
176 |
entities["metric"] = "netProfitMargin"
|
177 |
elif any(keyword in text_lower for keyword in ["market cap", "market capitalization", "market"]):
|
178 |
entities["metric"] = "mktCap"
|
179 |
-
elif any(keyword in text_lower for keyword in ["payout ratio", "dividend payout"]):
|
180 |
entities["metric"] = "payoutRatio"
|
181 |
elif any(keyword in text_lower for keyword in ["current ratio", "liquidity ratio"]):
|
182 |
entities["metric"] = "currentRatio"
|
183 |
-
elif any(keyword in text_lower for keyword in ["eps", "earnings per share", "earnings"]):
|
184 |
entities["metric"] = "eps"
|
185 |
elif any(keyword in text_lower for keyword in ["stock", "stock price", "current price", "valuation", "price"]):
|
186 |
entities["metric"] = "price"
|
187 |
-
elif any(keyword in text_lower for keyword in ["company info", "about company", "who is"]):
|
188 |
entities["metric"] = "ceo"
|
189 |
-
elif any(keyword in text_lower for keyword in ["balance sheet", "sheet", "assets"]):
|
190 |
entities["metric"] = "Assets&Liabilities"
|
191 |
-
elif any(keyword in text_lower for keyword in ["historical"
|
192 |
entities["metric"] = "historical"
|
193 |
elif any(keyword in text_lower for keyword in ["cash", "flow", "cash flow"]):
|
194 |
entities["metric"] = "cashFlowFromOperatingActivities"
|
|
|
12 |
self.nlp = spacy.load("en_core_web_lg") # "en_core_web_sm"
|
13 |
self.classifier = pipeline("zero-shot-classification", model="facebook/bart-large-mnli", from_pt=True)
|
14 |
self.intents = [
|
15 |
+
"net_income",
|
16 |
+
"revenue",
|
17 |
+
"stock_price",
|
18 |
+
"profit_margin",
|
19 |
+
"company_info",
|
20 |
+
"market_capitalization",
|
21 |
+
"historical_stock_price",
|
22 |
+
"dividend_info",
|
23 |
+
"balance_sheet",
|
24 |
+
"cash_flow",
|
25 |
+
"financial_ratios",
|
26 |
+
"earnings_per_share",
|
27 |
+
"interest_rate",
|
28 |
+
"research_info",
|
29 |
+
"cost_info",
|
30 |
+
"income_tax"
|
31 |
]
|
32 |
|
33 |
# Mapping of company names to ticker symbols (case-insensitive)
|
|
|
168 |
|
169 |
# Step 3: Extract metric using keyword matching with synonyms
|
170 |
text_lower = text.lower()
|
171 |
+
if any(keyword in text_lower for keyword in ["income"]):
|
172 |
entities["metric"] = "netIncome"
|
173 |
elif "revenue" in text_lower:
|
174 |
entities["metric"] = "revenue"
|
|
|
176 |
entities["metric"] = "netProfitMargin"
|
177 |
elif any(keyword in text_lower for keyword in ["market cap", "market capitalization", "market"]):
|
178 |
entities["metric"] = "mktCap"
|
179 |
+
elif any(keyword in text_lower for keyword in ["payout ratio", "payout", "dividend" "dividend payout"]):
|
180 |
entities["metric"] = "payoutRatio"
|
181 |
elif any(keyword in text_lower for keyword in ["current ratio", "liquidity ratio"]):
|
182 |
entities["metric"] = "currentRatio"
|
183 |
+
elif any(keyword in text_lower for keyword in ["eps", "earning", "earnings per share", "earnings"]):
|
184 |
entities["metric"] = "eps"
|
185 |
elif any(keyword in text_lower for keyword in ["stock", "stock price", "current price", "valuation", "price"]):
|
186 |
entities["metric"] = "price"
|
187 |
+
elif any(keyword in text_lower for keyword in ["company info", "company", "about company", "who is"]):
|
188 |
entities["metric"] = "ceo"
|
189 |
+
elif any(keyword in text_lower for keyword in ["balance sheet", "balance", "sheet", "assets"]):
|
190 |
entities["metric"] = "Assets&Liabilities"
|
191 |
+
elif any(keyword in text_lower for keyword in ["historical"]):
|
192 |
entities["metric"] = "historical"
|
193 |
elif any(keyword in text_lower for keyword in ["cash", "flow", "cash flow"]):
|
194 |
entities["metric"] = "cashFlowFromOperatingActivities"
|