Expose APIs from RL4LM to Consume 2 Models, 1 Scenario and 1 Morality Score to align the schema generator LM - Githubissues

Pinaka-ai / SchemaBottleneck

0 stars 0 forks source link

Expose APIs from RL4LM to Consume 2 Models, 1 Scenario and 1 Morality Score to align the schema generator LM #8

Open logisticloon opened 5 months ago

logisticloon commented 5 months ago

The purpose of this subtask is to ensure Alignment of LM_SchemaGenerator

The two models are LM_SchemaGenerator and LM_Task.

The API should expect path of the two models (perhaps checkpoints and Another API callback function for LM_Task) and a scenario as input parameters.

Ensure a fork from RL4LM and not RL4F.

Subtasks:

[ ] Extend the metric function to incorporate MSE ( Reuse your last implementation )
[ ] PPO training setup for LM_SchemaGenerator.