Quant FAQ

Do I need to learn PyPScript to use Quant Mode?

No. Quant Mode is Python-based.

Use it when you want to define train() and predict() directly rather than author a .ppc strategy file.

Can I use multiple files in the Research workspace?

Not as a general Python project yet.

The current supported workspace is intentionally constrained to:

• strategy.py
• quant.config.json
• README.md

Does quant.config.json really matter?

Yes.

It controls:

• pair
• timeframe
• parameters
• artifact target
• training requirements
• inference compatibility declarations

What data is available for training?

PyP provides market data inside the quant workflow.

The exact training environment should be treated as structured platform data, not as arbitrary user-managed file ingestion.

Can I use my own dataset?

Custom dataset upload is not the current default path.

Quant Mode today is designed around PyP-provided market data and a controlled training pipeline.

What should I deploy from?

Deploy from a validated release, not directly from a raw job.

The correct progression is:

• job
• artifact
• release
• simulation
• deployment or marketplace listing

Is Quant Mode only for simple models?

No, but the runtime path must stay honest about what it can execute.

Some artifact targets fit lighter runtime paths better than others. Model preservation, runtime cost, and deployment fit should all influence the target you choose.

Is my source code exposed to subscribers?

No.

Subscribers receive the strategy's signal output and managed delivery behavior, not the underlying source code.