File size: 1,778 Bytes
8a8bbbf
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ed82cc8
8a8bbbf
d2f66eb
 
ed82cc8
d2f66eb
8a8bbbf
d2f66eb
 
 
 
ed82cc8
d2f66eb
ed82cc8
d2f66eb
 
 
 
ed82cc8
d2f66eb
ed82cc8
d2f66eb
 
8a8bbbf
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
name: StockSense-AI
version: "3.0.0"
description: >
  Enterprise-grade AI trading engine for Indian markets (NSE/BSE). 
  Blends Deep Learning (LSTM), technical analysis, sentiment analysis, 
  and LLM reasoning to provide high-precision stock signals.

tags:
  - openenv
  - finance
  - trading
  - nse-bse
  - stock-prediction
  - ai-vit

# Task definitions for the OpenEnv grader
tasks:
  - id: "prediction_task"
    name: "prediction"
    description: "Full market direction prediction (bullish/bearish/neutral) and a confidence score."
    difficulty: "medium"
    max_steps: 1
    score_range: [0.01, 0.99]
    grader: "backend.graders:grade_prediction"
  - id: "sentiment_task"
    name: "sentiment"
    description: "Analyze recent news and social media sentiment for a specific ticker."
    difficulty: "easy"
    max_steps: 1
    score_range: [0.01, 0.99]
    grader: "backend.graders:grade_sentiment"
  - id: "technical_task"
    name: "technical"
    description: "Calculate and interpret technical indicators (RSI, MACD, etc.) for a ticker."
    difficulty: "easy"
    max_steps: 1
    score_range: [0.01, 0.99]
    grader: "backend.graders:grade_technical"

# High-level definition of the environment's interaction space
observation_space:
  type: StockMarketState
  fields:
    ticker: "The stock ticker symbol (e.g., RELIANCE.NS)"
    timestamp: "The current date and time"
    price_data: "Historical OHLCV data with technical indicators"
    sentiment_data: "Recent news headlines and sentiment scores"

action_space:
  type: TradingSignal
  fields:
    direction: "Predicted market direction: bullish, bearish, or neutral"
    score: "Task confidence score strictly in the (0, 1) interval"
    reasoning: "Brief explanation of the factor-based justification"