Conversation
… NF4 table lookup
Introduce EjkernelRuntimeError for consistent unsupported-feature
reporting across all kernel backends. The kernel registry now wraps
implementations with a validation layer that detects unsupported
parameters via AST inspection and raises structured errors before
dispatch.
CUDA quantized matmul gains cuBLASLt and CUTLASS GEMM backends
(selectable via EJKERNEL_QMM_CUDA_GEMM env var), group-size-specialized
dequantization kernels for sizes 8-1024, and 4-elements-per-thread
throughput optimization. The Triton NF4 kernel replaces its polynomial
approximation with an exact codebook table lookup.
CUDA flash attention is simplified by removing the CUTLASS autotuning
infrastructure in favor of direct dispatch with paged-KV cache support.
The executor now auto-prefers native CUDA implementations on NVIDIA GPUs
when available.
Additional changes:
- Expand affine quantization group_size support to {8..512}
- Add benchmark platform ignore-list (EJKERNEL_BENCH_IGNORE_PLATFORMS)
- Update tests to expect EjkernelRuntimeError for unsupported features
- Improve docstrings across callib, kernels, ops, types, and utilities
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Testing