Hi,
The grammar that is passed to the OCaml enumeration backend has the probability for all production rules as 0.0.
I have a few questions related to that:
- Do these probabilities change as a part of the enumeration process (so that programs generated later have more chance of solving the tasks)?
- Which part of the code changes these probabilities using the I/O examples?
- Where can I find the final PCFG after training has been completed?
Hi,
The grammar that is passed to the OCaml enumeration backend has the probability for all production rules as 0.0.
I have a few questions related to that: