dusterbloom: perf(models): fused MoE gate+up — 3→2 expert matmuls per layer#12
Open
dusterbloom wants to merge 14 commits into
Open
dusterbloom: perf(models): fused MoE gate+up — 3→2 expert matmuls per layer#12dusterbloom wants to merge 14 commits into
dusterbloom wants to merge 14 commits into
Conversation
dusterbloom
force-pushed
the
dusterbloom/perf-foundational-wins
branch
from
2e9e083 to
18bf33f
Compare
5 tasks
fix(deps): update rust crate toml to v1
…l-action-digest chore(deps): update taiki-e/install-action digest to cca35ed
…file chore(deps): update rust crate tokio to v1.52.2
…-lockfile chore(deps): update rust crate tower-http to v0.6.9
…anbanda#143) Adds AnyCache::trim_by to roll back KV layers for speculative decode while leaving hybrid Arrays state untouched.\n\nCI: https://github.com/panbanda/higgs/actions/runs/25312580791
…#148) * feat(qwen3_next): mixed-bit Qwen3.5 GDN BA loading fallback Adds a fallback path for loading Qwen3.5 models with mixed-bit GDN projection weights (some layers q4, some q8 — common in unsloth's dynamic-quant variants). The default fused-projection loader fuses `in_proj_a` + `in_proj_b` into a single matmul; mixed-bit weights have incompatible shapes and the fusion fails. Behaviour: 1. Detect via `is_mixed_bit_gdn_ba_fusion_error` — matches a `ModelError::ShapeMismatch` whose message contains both `in_proj_ba` and `requires separate GDN projections`. 2. On detection, retry the load with `args.use_separate_gdn_projections = true`, taking the `load_qwen3_5_moe_weights_direct` path. Forward dispatches go from 2 to 4 GDN ops per layer — slightly slower but correct. 3. Forced separate (via `args.use_separate_gdn_projections` config or `HIGGS_SEPARATE_GDN_PROJ` env var) skips the fused attempt entirely. Also adds: * `qwen3_5_quantization_config` — parses `{group_size, bits}` from the per-layer `quantization` map in `config.json`. * `qwen3_5_mixed_ba_quantization_layers` — scans for the layers where `in_proj_a` and `in_proj_b` differ in bits or group_size. * `can_concatenate_axis0` — guard used inside `load_qwen3_5_moe_weights_fused` to emit the diagnostic `ShapeMismatch` error rather than panicking on the concat. * `load_qwen3_5_model_with_gdn_fallback` — private helper called by both `load_qwen3_5_model` (dense) and `load_qwen3_5_moe_model` (MoE), unifying the fallback path. Adaptations from feat/magic-canvas → origin/main: * The dense `load_qwen3_5_model` previously only honoured the env var; now it honours `args.use_separate_gdn_projections` too, matching the MoE path. Strict improvement: the config flag is set only by the env var or by mixed-bit detection. * No `unwrap()`, no `as` casts (use `i32::try_from`); match arms enumerate variants. No file-level allows added. Verification on origin/main (rustc 1.95.0): * `cargo check -p higgs-models` — clean * `cargo clippy --all-targets --all-features -- -D warnings` — clean * `cargo fmt --check` — clean * `cargo test -p higgs-models --lib` — 333/333 pass (3 new) Source: feat/magic-canvas commit `061e500c`. Direct cherry-pick had 5 conflict regions because origin/main has evolved the load functions independently; this is a manual surgical port that preserves origin/main's structure while adding the fallback behaviour. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com> * fix(qwen3_next): preserve explicit GDN projection config --------- Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com> Co-authored-by: Jonathan Reyes <me@jonathanreyes.com>
SwitchMlpWeights::forward_gather_fused() lazy-concatenates gate+up weights on first call, then dispatches a single gather_qmm instead of two separate calls. FfnBlock::forward() now routes through the fused path instead of forward_gather_global_sort(). Measured on 35B-A3B-3bit M4 base: - S=1 decode: 27ms → 17ms (−37%) - S=16 verify: 253ms → 112ms (−56%) - MoE/layer at K=1: 0.47ms (down from ~0.68ms) Co-Authored-By: Claude Sonnet 4.6 (1M context) <noreply@anthropic.com>
Reflow `let fw/fs/fb = ops::concatenate_axis(..)` from broken-line indentation back onto single lines so `cargo fmt --all -- --check` passes in CI.
Two errors flagged by `-D clippy::doc-markdown` and `-D clippy::cast-sign-loss`/`-D clippy::as-conversions`: - Backtick `MoE` and `gather_qmm` in the `fused_gate_up` doc comment. - Replace `top_k as u32` with the same `u32::try_from(top_k).map_err(...)` pattern already used by `forward_gather_global_sort`.
panbanda
force-pushed
the
dusterbloom/perf-foundational-wins
branch
from
009280b to
371c2e3
Compare
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
3 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.
Summary
SwitchMlpWeights::forward_gather_fused()lazy-concatenates gate+up weights on first call, then dispatches a singlegather_qmminstead of two separate calls.FfnBlock::forward()now routes through the fused path instead offorward_gather_global_sort().Numbers (35B-A3B-3bit on M4 base)
Test plan
cargo check -p higgs-models— cleancargo test -p higgs-models --lib qwen3_next::— 71 passed, 0 failed (includes newtest_moe_gate_up_fusion_parity)cargo run -p higgs-bench --release --bin bench_decode -- --model <qwen3_5_moe-key>— A/B against main on a Qwen3.5-MoE checkpointcargo clippy -p higgs-models— cleanNotes
crates/higgs-models/src/qwen3_next.rs(+175 / −9)