Add PPO kernels#5
Open
haixuanTao wants to merge 2 commits into
Open
Conversation
Adds vortx::linalg::{Activation (tanh + tanh_backward), Adam} and their shaders,
the GPU building blocks for MLP training (used by nexus RL demos / zealot-rl).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
New `ppo` op (host + shader) producing the per-sample OUTPUT gradients that feed the generic GEMM/elu_backward backward backbone: - gpu_ppo_actor_grad: clipped-surrogate actor gradient (logp over the action dims, ratio = exp(logp - logp_old), clip mask) -> g_mean plus the state-independent log_std gradient contribution. - gpu_ppo_value_grad: clipped value-loss gradient. Both are an exact port of zealot-rl's minibatch_step. Every per-sample tensor is row-major [rows x M] (M = minibatch columns); one thread handles one sample column and loops over the (small) action dim. No Shape uniform -- dims ride in PpoActorParams/PpoValueParams. Exports Ppo, PpoActorParams, PpoValueParams (host) and GpuPpoActorGrad, GpuPpoValueGrad (shaders). Verified vs CPU minibatch_step (~1e-7, ~25% of samples on the clip branch). Note: the one-line glamx Cargo.toml dependency is shared with the ELU and GEMM-vec4 branches; whichever lands first, the others need a trivial rebase of that line. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Member
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Add PPO gradient kernel for RL training