Conversation
- posture_weight 0.2 → 2.0: make posture matter vs alive bonus - energy_penalty_weight 0.0005 → 0.0: remove distraction for balance stage - smoothness_weight 0.0 → 0.1: encourage smooth corrective movements - learning_rate 3e-4 → 1e-4: prevent sharp policy degradation - ent_coef 0.05 → 0.1: maintain exploration, avoid freeze-and-fall - min_avg_episode_length 400 → 300: more achievable curriculum gate https://claude.ai/code/session_019KsnCT9nvDg5cpTMquZTzQ
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This pull request updates the reinforcement learning configuration for the Velociraptor's stage 1 balance task. The main focus is on tuning reward weights and learning parameters to improve the agent's ability to stand and balance. The most important changes are grouped below:
Reward and environment parameter adjustments:
posture_weightfrom 0.2 to 2.0 to emphasize maintaining proper posture.energy_penalty_weightto 0.0 (from 0.0005) to remove the penalty for energy use.smoothness_weightof 0.1 (was 0.0) to encourage smoother movements.Learning algorithm parameter tuning:
learning_ratefrom 3e-4 to 1e-4 and increasedent_coeffrom 0.05 to 0.1 to promote more exploration.Curriculum criteria update:
min_avg_episode_lengthfrom 400 to 300 to allow curriculum progression with shorter average episodes.