RL 457 SUPERVISION v4.0

TRAINING ACTIVITIES CONSOLE

Log Filters

All Messages

Dataset

processed progress started error

Other Messages

0 logs

MODEL CONFIGURATION

Auto-Force Config on ConnectSends local config automatically on every page load / reconnect

Snapshot Interval (ms)

Sequence Length

Expiry (seconds)

Training

Learning Rate

Learning Rate Decay

Batch Size

Epochs

Patience (Early Stop)

Reduce LR Patience

Reduce LR Factor

Warmup Threshold

Network Architecture

Dense Units (comma separated)

Activation

Optimizer

Batch Normalization

Regularization

Weight Decay

Dropout Rate

Gradient Clip Norm

Signals & Market

Volatility Gate Multiplier

Feature Selection

TIME WINDOWS (ms)

SHARED DATASET

Location:

/storage/shared/ALL_IMORTANTS/RL_457_PROJECT/

Checking for dataSet.csv/json...

TRAINING CONTROL

Ready

Epoch Progress0 / 0

Batch:

0 / 0

BATCH STATISTICS

GRADIENT: UNKNOWN

Current Loss

—

trend: —

Val Loss

—

Train Acc

—

Val Acc

—

Batch

—

Speed

—

Elapsed

—

Remaining

—

Grad Stability

—

lower = stable

Grad Trend

—

+rising / −falling

Grad Mean

—

Grad Health

—

Loss Std Dev

—

batch variance

Avg Batch Time

—

Loss Trend

—

Batches Left

—

Learn Momentum

—

loss improvement rate

Signal / Noise

—

higher = cleaner

Loss Ratio

—

batch vs avg

Acc Ratio

—

batch vs avg

Grad Ratio

—

vs rolling grad

Win Rate

—

rolling accuracy

Avg Loss

—

rolling window

Avg Acc

—

rolling window

BATCH HISTORY (last 100)

Loss

Accuracy

FEATURE IMPORTANCE — LIVE PER BATCH

All Features

—

Strong

—

Noisy

—

Overfit Risk

—

Top Shown

20

Waiting for batch feature importance data — fires every few batches during training...

EPOCH ANALYTICS

0

Live Readiness

Score / 100

Awaiting data

0

Learning Quality

Score / 100

Awaiting data

PhaseInitializing

Training has not started yet.

Projected Live Accuracy

—

Epoch

0

of 200

Train Loss

—

Val Loss

—

Best Val Loss

—

checkpoint

Train Acc

—

Val Acc

—

Learning Rate

—

Patience Left

—

epochs

Loss Curves (Train vs Val)

Accuracy Curves (Train vs Val)

Epoch Diagnostics

Overfit Risk

—

Underfit Risk

—

Train Trend

—

Val Trend

—

Gen Gap

—

Acc Gap

—

Convergence

—

LR Health

—

Grad Health

—

Grad Norm

—

Weight Health

—

Weight L2

—

Learn Momentum

—

Signal/Noise

—

Noise Sensitivity

—

Noise CV

—

Concept Drift

—

Calib Status

—

Calib Gap

—

Avg Gap

—

Module Health

OTC Detector

—

Trend Analyst

—

Behavior Detector

—

MTF Integrator

—

Confirmation Engine

—

Sim Engine

—

Regime Classifier

—

Generalization Gap

Live Projection

Projected Acc

—

Est. Live Drop

—

Live Confidence

—

Overfit Risk

—

Trend

—

Peak Val Acc

—

Peak Epoch

—

Best Gen Gap

—

Memory & LR (last 10 batches)

Memory (MB used)

Learning Rate

LEARNING INSIGHTS

RECOMMENDATIONS & ANOMALIES

MODULE PERFORMANCE

Epoch Ratios

Loss Ratio

—

Acc Ratio

—

Grad Ratio

—

Win Rate

—

Overfit Summary

—

INFERENCE RESULTS

HOLDs suppressed: 0

0

Total Signals (BUY + SELL)

0

BUY

0

SELL

—

Win Rate

Action Distribution (BUY vs SELL)

Confidence Distribution (10 buckets)

Time	Entry Price	Probability	Confidence	Action	Outcome	Price Δ	P&L
No inference data yet. Start playback.

LIVE METRICS & SYSTEM HEALTH

Model Versions

Edit Model Configuration

Edit Time Windows

Configuration Conflict

Console Auto-Clear Timer

BATCH STATISTICS

Current Loss

Val Loss

Train Acc

Val Acc

Batch

Speed

Elapsed

Remaining

Grad Stability

Grad Trend

Grad Mean

Grad Health

Loss Std Dev

Avg Batch Time

Loss Trend

Batches Left

Learn Momentum

Signal / Noise

Loss Ratio

Acc Ratio

Grad Ratio

Win Rate

Avg Loss

Avg Acc

BATCH HISTORY (last 100)

Feature Chart