requirements

crypto/brian/readme.md

@@ -5,6 +5,15 @@ run: >conda activate gpt-gpu
     python .\index.py
 
 
+conda create --name my_trading_env python=3.9
+conda activate my_trading_env
+pip install -r requirements.txt
+
+
+# --------------------------------------------------------
+
+
+
 Usage:
 Run the script with a command-line argument — for example:
 • python index-deep-new.py --mode train
@@ -52,4 +61,10 @@ existing (running  but unfinished ) code:
 implement these suggestions into our code and add arguments for easy switching of modes:
 - train (only): pool latest data and use it for backtesting with RL to learn to detect peaks/valleys 
 - live: load best checkpoint and latest HLOCv data to actively generate trade signals, but calculate and back propagate errors when closing positions. optimize for profit in the reward function
-- inference: optimize model loading for inference only - load historical data and periodically append new live data and generate siganls but without active RL
+- inference: optimize model loading for inference only - load historical data and periodically append new live data and generate siganls but without active RL
+
+
+
+
+
+does this code implement retrospective RL (when we detect with live data the valley/top sequence we should do a parameter optimisation (back propagation or other modern technique) with the best move we should have taken in retrospective). this will allow us to learn and become optimized over time. we should also incorporate MoE architecture so we will have one expert specialized at fastly tuning to the current data stream/market conditions.