Commit Graph
41 Commits
Author SHA1 Message Date
ZihaoMuandLZ-QWQ 9cacd0896c fix(moe): publish rocm cpu expert inputs before notification (#764)
* fix(moe): prevent stale cpu expert inputs in rocm graphs

* fix(moe): publish rocm cpu inputs with graph events

* fix(moe): qualify rocm kernel io with ordered copy fallback

* fix(moe): enable locked hybrid staging on rocm

---------

Co-authored-by: LZ-QWQ <lz1830837889@163.com>
2026-11-02 21:29:42 -08:00
Xiaoze Fan 3d23551c5b perf(cpu-moe): speed up the nvfp4 W4A8 kernels (#851)
The VNNI dots were compute bound per core; with fewer instructions per weight they reach the DRAM read bandwidth.
The W4A8 kernel now follows FREETOKEN_CPU_MOE_ISA.
2026-11-02 15:05:17 -08:00
Xiaoze Fan dacac5f5fa feat(spec): add MTP drafts for the Qwen3.5 family and Qwen3.8-Flash-Next (#844)
* feat(spec): add the MTP proposer for the qwen3_5 family

* test(models): read the qwen3_5 MTP head of bf16 and block-fp8 checkpoints

* feat(spec): verify qwen4_exp targets on qsa_sparse graphs with PLE verify handlers

* feat(spec): add the qwen4_exp MTP draft from the checkpoint or a Strata pack

* docs(cli): document the speculative decoding flags

* fix(spec): refuse MTP draft sources the draft cannot read before the target loads

* docs(qwen4_exp): describe the verify path in the PLE docstrings

* fix(spec): refuse a missing relative MTP draft path before the target loads

* fix(moe): read MTP head experts on the host and require both stacked tensors

* refactor(spec): tell the final prefill chunk from prompt_len instead of the scheduler's ChunkedReq

* refactor(spec): let each MTP draft family check its own weight source

* fix(qwen4_exp): size the disk PLE eager staging for decode and verify rows

* perf(kernel): read the KV-index gather's chunk count from the grid, not a constexpr

* refactor(moe): let each family name its MTP expert tensors

* refactor(spec): force the MTP draft's MoE resident in the proposer and check it after load

* fix(spec): refuse MTP with dummy weights

* refactor(spec): move HiddenRelay under spec/mtp, its only user

* refactor(spec): drop MTPDraft.config, fold its embed hooks into embed_tokens

* refactor(models): note FeatureSink's graph binding, drop its runtime_checkable

* fix(models): read quantized MTP fc projections under the quant config
2026-11-02 11:37:09 -08:00
Mukkesh Ganesh ea2916d31c feat(gguf): serve gpt-oss gguf checkpoints on gpt_oss (#838)
* feat(gguf): serve gpt-oss gguf checkpoints on gpt_oss

* fix(gguf): refuse missing gpt-oss bias tensors, honor swa pattern, scale nvfp4 before bias

* fix(gguf): honor gpt-oss yarn truncate, gate the bf16 float cast, scale nvfp4 before the down bias
2026-11-02 08:58:07 -08:00
Xiaoze Fan e328cf6381 refactor: add engine and scheduler seams for multi-row decode steps (#833)
* refactor: add engine and scheduler seams for multi-row decode steps

* refactor: rename the graph steps argument to rows_per_req

* refactor: drop VerifyInfo and require rows_per_req on verify batches

* refactor: fix stale seam comments and read decode_rows_per_req directly

* refactor: reword two GraphRunner comments for multi-row graphs

* fix(engine): scale the GPU-decode MoE pool limits by decode_rows_per_req

* refactor: drop the unread Context.moe_offload_cache field

* test: trim the multi-row seam tests
2026-10-27 16:03:01 -07:00
Chris Qian 0bf7dbbf23 fix(models): discover safetensors shards from the index (#819)
* fix(models): find index-only checkpoints that keep shards in a subdirectory

* fix(models): make the checkpoint index authoritative for shard discovery

The index is read directly so safetensors_weight_map stays the only
header-based reader, and _host_ram_fits_parallel sizes the shards the
reader will actually open, so the low-RAM serial fallback fires again
for subdirectory layouts.

* fix(models): fall back to root shards when the index is unusable

Share one index parser between iter_weight_files and safetensors_weight_map
instead of a third copy; warn and glob the root when the index is broken,
absent or empty; reject index keys that escape the model folder; let the hub
download fetch the index itself; tell the truth in the diagnose docstring.

* revert(models): back out the defensive index fallback, keep the no-index test

Per review: a broken index is a broken checkpoint, so the loader keeps the plain one rule -- use the index if there is one, otherwise glob the root. The no-index fallback test stays as the only coverage of the root glob and the consolidated.safetensors rule.
2026-10-24 19:10:52 -07:00
Xiaoze Fan 161ffa5a61 perf(moe): scan the hybrid eviction pool in chunks and add fetch rounding modes (#778) 2026-10-21 11:38:13 -07:00
LZ_QAQ cc4aa085b6 fix(moe): enable prefill hit-d2d on rocm (#771)
* fix(kernel): order the batch-memcpy probe against the current stream

Adapt the stream dependency and busy-stream regression from #414
(https://github.com/FlashML-org/FreeToken/pull/414). Warm the JIT
before queuing the backlog so compilation cannot mask the race.

* fix(moe): enable prefill hit-d2d on rocm
2026-10-20 23:15:17 -07:00
Xiaoze Fan c5b5b7243a feat(moe): fetch locked-layer experts through pinned staging (#742)
* feat(moe): fetch locked-layer experts through pinned staging and probe the pin limit

Under hybrid, layers whose banks the pin quota could not take keep GPU slots of their own pools and fetch part of their misses through pinned staging that CPU threads fill (--moe-locked-fetch, --moe-locked-max-fetch). --moe-cpu-layers auto without FREETOKEN_PIN_BUDGET_GB now loads the banks unpinned, pins the staging, then pins layers middle-out until CUDA refuses one and OS-locks the rest.

* fix(moe): resize the locked fetch on cache rebuild and gather staging by device address

* feat(moe): log decode moe stats per layer group with --moe-collect-stats

* perf(moe): default the locked fetch to max(2, min(4, logical cpus / 4)) copy threads

* feat(moe): take the locked fetch share from the ft bench bw profile by default

* fix(moe): fuse the locked fetch into publish and gather kernels and address review findings

* refactor(moe): gather the locked staging with fast_index_copy_multi

* refactor(moe): move the locked fetch setup out of the engine and address review findings

* fix(moe): fail the boot when the locked staging cannot be pinned and keep locked layers out of the hybrid share log
2026-10-20 14:59:05 -07:00
Xiaoze Fan 0db5023b60 fix(quant)!: keep nvfp4 global scales in fp32 (#739)
NVFP4 weight global scales (amax / 2688, 2e-5 to 4e-4 in the served checkpoints) sit at fp16's
normal minimum; stored as fp16, up to 97% of a checkpoint's expert globals were subnormal, with
up to 1.7e-3 relative error per row.

BREAKING CHANGE: an FTW checkpoint that stores NVFP4 global scales as fp16 is refused at load;
reconvert it with `ft checkpoint`.
2026-10-19 12:33:43 -07:00
Xiaoze Fan e4f69d9e23 perf(moe): prefill nvfp4 moe experts on nvfp4 activations on sm_120 (#736) 2026-10-18 20:10:51 -07:00
Xiaoze Fan 8f49e0af98 feat(moe): bench hybrid fetch fractions per moe layer kind (#734)
* feat(moe): bench hybrid fetch fractions per moe layer kind

The ft bench bw profile (version 6) holds only layer_kinds rows, keyed by CPU format, options, expert shape, top_k and CPU worker threads; older profiles are ignored. ft bench bw --model also takes checkpoint paths and benches each of their layer kinds. Hybrid decode reads each layer's own row for its fetch fraction, else keeps a fixed cap of 1. --moe-strategy auto no longer upgrades to hybrid from the profile.

* fix(bench): fill gguf mxfp4 scales, scope act_block to ds_fp4 and skip cpu-less rows
2026-10-17 23:25:12 -07:00
Xiaoze Fanandandy-yang-1 ac6a15de75 perf(moe): speed up the nvfp4 moe prefill gemm on rtx gpus (#733)
* perf(moe): speed up the nvfp4 moe prefill gemm on rtx gpus

* perf(moe): keep the byte e2m1 decode at every tile size on h100

---------

Co-authored-by: andy-yang-1 <andy_yang@berkeley.edu>
2026-10-17 17:50:59 -07:00
Xiaoze Fan 31c6ef4f77 feat(gguf): apply the weight scales of nvfp4 gguf tensors (#732) 2026-10-17 13:46:50 -07:00
Xiaoze Fan c8559fdff6 feat(cpu-moe): serve GGUF experts on the cpu and hybrid backends through ggml-cpu (#704)
* feat(cpu-moe): serve GGUF experts on the cpu and hybrid backends through ggml-cpu

* feat(cpu-moe): run float GGUF expert rows on the cpu through ggml-cpu's vec_dot

* feat(cpu-moe): build ggml-cpu for each llama.cpp x86 variant and run the best one the CPU supports
2026-10-16 22:28:34 -07:00
Xiaoze Fan 03c2c9c35e perf(kernel): rewrite the triton nvfp4 moe prefill gemm (#723)
* perf(kernel): rewrite the triton nvfp4 moe prefill gemm

* perf(moe): tune nvfp4 prefill tiles per gpu class

GeForce runs fp32-accumulate tensor MMA at half rate, so RTX parts want smaller BLOCK_SIZE_M than H100 and their own BLOCK_N / BLOCK_KB / stages.
Tables come from tile sweeps on RTX PRO 6000, RTX 5090, RTX 4090 and RTX 3080 Ti; H100-class GPUs and AMD keep the previous tuning.
2026-10-16 22:03:59 -07:00
Xiaoze Fan 4663b2b89e feat(gguf): serve MiniMax-M2 and MiniMax-M3 GGUF checkpoints (#725) 2026-10-16 19:01:30 -07:00
Xiaoze Fanandandy-yang-1 e8c6bf1123 feat(moe): serve experts of several formats through per-format slot pools (#700)
* feat(moe): serve experts of several formats through per-format slot pools

* refactor(cpu-moe): pass each layer's cpu format with its options

* fix(moe): read only each format's layers when loading experts of several formats

---------

Co-authored-by: andy-yang-1 <andy_yang@berkeley.edu>
2026-10-16 15:13:59 -07:00
Xiaoze Fanandzihaomu e5d3bd3faf feat(gguf): serve GGUF checkpoints on vendored ggml-cuda kernels (#690)
* build(kernel): vendor ggml-cuda from llama.cpp 8c2134a8 behind a tvm-ffi binding

* build(deps): vendor the gguf-py reader

* feat(quant): add the GGUF quant kind and an embedding quant method

* feat(gguf): serve Flash-Next and Gemma4 GGUF on the GGUF quant kind and drop the old Q4_0 path

* fix(kernel): build ggml-cuda for plain sm_120 by turning off the native fp4 MMA

* perf(gguf): write ggml results in the activation dtype and keep MoE intermediates in bf16

* build(kernel): compress the ggml fatbin to keep the multi-arch kernel cache small

* fix(gguf): read F32, F16 and BF16 PLE tables without ggml

* fix(moe): pad the offload slot cache for ggml MMQ tail reads

* fix(gguf): support hip builds and runtime devices

* fix(gguf): copy same-type dequantize and check for missing ggml converters

* fix(gguf): load quantized qwen4exp ple_key and ple_value

* feat(gguf): run float parts of a fused GGUF linear as bf16 segments

* fix(gguf): load quantized qwen4exp hyper-connection and gdn ba weights

* fix(moe): zero the prefill staging area after the last prefill layer

---------

Co-authored-by: zihaomu <zihaomu@outlook.com>
2026-10-15 16:33:16 -07:00
LZ_QAQ 0bca88f70c fix(rocm): reuse staged moe alignment on Triton 3.8 (#691)
* fix(rocm): reuse staged moe alignment below triton 3.9

* fix(rocm): restrict moe alignment workaround to Triton 3.8
2026-10-15 02:21:32 -07:00
Xiaoze Fan b416f85c40 feat(moe): split the offload expert cache into slot pools (#689)
* feat(moe): split the offload expert cache into slot pools

* perf(moe): stop recompiling the hybrid ensure kernel for each slot pool size

* fix(engine): turn off prefill overlap when --moe-cache-mem leaves under two layers of slots

* fix(engine): load the --moe-cache-split profile before the expert banks
2026-10-14 19:44:55 -07:00
Xiaoze Fan f02fbb16a2 perf(moe): split hybrid decode routes inside the ensure kernel (#685)
* perf(moe): split hybrid decode routes inside the ensure kernel

* refactor(moe): always split routes in the hybrid ensure kernel

* test(moe): drop the fetch split test left without its kernel check
2026-10-13 20:30:34 -07:00
Xiaoze Fanandzihaomu 2010271266 refactor(cpu-moe): split the executor into weight formats and gpu sync backends (#670)
* refactor(cpu-moe): split the executor into weight formats and gpu sync backends

- move each expert weight format into kernels/<fmt>.cpp behind the MoeFormat
  interface in format.h; the executor keeps the threads, the work split and
  the pass 2 route sum
- pass each format its banks and integer options by name; MoeParams keeps only
  the shared shape and activation fields
- key the formats by name in one table that weight_formats() exports, replacing
  the integer format ids and the fp8_block_isa_name hook
- replace the per-(layer, batch size) flag slots, pinned IO buffers and task
  table with one doorbell word MoeTask{layer, num_tokens} and one shared pinned
  IO block of max_tokens rows
- move the GPU/CPU handshake behind GpuSync with a CUDA and a HIP backend picked
  by setup.py; HIP waits on signal memory and adds explicit batch-memop graph
  nodes where the runtime does not record the native memops under capture
- raise on graph capture when the handshake is off and host functions are not
  replay-safe
- allocate format scratch on 64-byte boundaries
- size max_tokens by the explicit cuda_graph_bs as well

Co-authored-by: zihaomu <zihaomu@outlook.com>

* perf(cpu-moe): skip the worker pool for tasks without routed experts

* perf(cpu-moe): publish and combine the cpu handshake with kernels

* fix(cpu-moe): clean up failed hip capability probes

---------

Co-authored-by: zihaomu <zihaomu@outlook.com>
2026-10-13 20:12:58 -07:00
95ac86f215 feat(cpu-moe): serve block-fp8 experts on the cpu and hybrid backends (#662)
* feat(cpu-moe): block-FP8 expert GEMV (W8A16) for the cpu/hybrid backends

Block-fp8 checkpoints (Qwen3.5/3.6-FP8, DeepSeek-V3-style, GLM) had no CPU expert
path at all:

    NotImplementedError: --moe-backend cpu/hybrid ... supports
    ['bf16', 'ds_fp4', 'mxfp4_triton', 'nvfp4', 'q4_0'] formats,
    but this checkpoint's experts are 'fp8_block'

so every expert miss had to stream over PCIe at ~25 GB/s while the CPU sat idle.

The bank is fp8-e4m3 row-major [rows, K] with one bf16 scale per 128x128 block --
K contiguous per output row, unlike mxfp4's transposed bank, so it maps straight
onto a bf16 dot product. e4m3 -> bf16 is *exact* (3 mantissa bits into 7, and
bf16's exponent range covers e4m3's whole 2^-9..448 span), so the AVX-512 path
widens weights in-register and feeds `_mm512_dpbf16_ps` against the bf16
activations: identical products to the reference's fp32 math, differing only in
summation order -- the same latitude the bf16 dot already takes.

The normal range rebases the exponent with a shift and a constant +120; exp == 0
is subnormal (m * 2^-9) and does not obey that, so those lanes blend in from a
small table. AVX2 does the same in fp32; scalar goes through the existing 256-entry
e4m3 LUT.

`ft bench bw` gains an fp8_block CPU measurement (it could already build the banks).
Measured on main with this commit alone, `--dtype fp8_block --reps 5`, 2x Xeon Gold
6526Y:

                        CPU MoE      PCIe    ratio   backend picked
    main             not measurable  25.1     --     offload
    main + this      53.7 GB/s       25.1    2.14x   hybrid

Before this commit the format reports no CPU number, so the backend picker has
nothing to weigh and selects offload. The ratio is what decides the backend, so
adding the kernel is what lets a block-fp8 deployment use the CPU at all.

End to end on Qwen/Qwen3.6-35B-A3B-FP8, median of 3 generations at matched GPU
expert residency:

    residency   offload   hybrid    delta
    10%           43.94    50.94    +15.9%
    25%           82.65    55.12    -33%

The win is real but conditional: hybrid pays off once the miss rate is high enough
that PCIe is the bottleneck, and costs throughput when enough experts are resident
that the CPU becomes the limiter. See the PR discussion -- the ratio makes
`--moe-backend auto` select hybrid for every block-fp8 checkpoint, which is right
for the oversubscribed case this project targets and wrong for a model that nearly
fits, and the bench models bandwidth but not miss rate.

Correctness is checked against the Triton block-fp8 decode kernel on identical
banks and routing (bs 1/2/5), plus an exactness check on the e4m3 -> bf16 widening
over all 256 codes. A weight-format ABI probe (`max_weight_format_id`) makes a
stale prebuilt `.so` fail loudly instead of falling through to the wrong dequant
branch and returning silently wrong numbers.

Assisted-by: Claude

* fix(cpu-moe): honor padded fp8 scale strides

The scale banks' trailing dimension is padded so per-expert rows land on
a 16-byte boundary for the fused copy (kernel/aot_models.py:
fp8_block_scale_pad), so the row stride is not ceil(K/128). Recomputing
it from the unpadded block count reads scales at the wrong offset.

Qwen3.8 is the case in the tree: models/qwen3_5_moe/weight.py records
down_scale as 20x5 bf16 = 200 B, and 200 % 16 != 0 pads cols 5 -> 6.
The benchbw workload (H=2048, I=512) and DSV4-Flash (H=7168, I=2048)
both happen to give rows*cols divisible by 8, so neither trips it.

Read the stride off the tensor and pass it through to the kernel, and
let the shape check assert the unpadded extent is present rather than an
exact shape. The fixture now allocates padded banks, so the existing GPU
cross-check covers this too; the GPU kernel was already stride-aware.

* test(cpu-moe): add float64 parity reference and an avx512f fp8 tier

The existing GPU-vs-CPU check compares two fp32 accumulations at a 2e-2
tolerance, which cannot see accumulator precision loss. Add a float64
reference over mixed-sign, full-range e4m3 weights and zero-mean
activations, and force each ISA tier explicitly so the sweep is real.

Also add an avx512f tier between avx512bf16 and avx2, so CPUs with
AVX-512 but no BF16 extension (Skylake-SP through Ice Lake-SP) stop
falling back to 8-wide fp32.

Measured on Xeon Platinum 8368 (Ice Lake-SP, no avx512_bf16), three
repetitions each: +19% single-thread cache-resident (3.31 -> 3.94 GB/s,
disjoint intervals), but only +4% under `ft bench bw --dtype fp8` at 75
threads (60.4 -> 62.7 GB/s, overlapping intervals). At that thread count
the kernel is memory-bound, so the project's own benchmark cannot
resolve the tier; the single-thread gain is the honest figure.

* test(cpu-moe): cover fp8 dispatch and cpu-only regressions

The new tier shipped without test coverage: deleting the avx512f branch
from select_fp8dot left every test green, because isa_name() reports
select_dot() (the bf16 chain) rather than the fp8 dispatch. Injecting a
ragged-tail bug into dot_fp8_block_avx2 was invisible for the same
reason -- no test ran numerics under a forced tier.

Force each available fp8 tier through the float64 parity GEMV, expose
the selected fp8 tier so the assertion can see it, and make the padded
bank test run a GEMV instead of only constructing an executor.

Also move the module-level CUDA guard onto the tests that actually need
a GPU. A CPU MoE kernel's most relevant hosts are the ones without a
usable GPU, and all ten tests were skipping there.

* review: force each ISA tier in the fp8_block test, and stop the docstring overclaiming

The comparison test only varied batch size, so all three cases ran
whatever tier this machine auto-selects and the scalar / AVX2 /
AVX-512-BF16 fp8_block kernels were never compared against the GPU at
all.

Parametrise FREETOKEN_CPU_MOE_ISA over the four tiers. Setting the
variable alone would not be enough: pick_isa() caps *downward*, so asking
for a tier the CPU or build lacks silently returns a lower one and the
test would pass four times over the same kernel. The executor latches its
choice into isa_name() at construction, so read it back and skip when it
does not match what was asked for. On a machine with all four, that is 4
x 3 cases against a 8-expert, 256-hidden problem.

test_subnormal_and_sign_decode_exactly said it validated that "every e4m3
byte must widen to the value the reference LUT gives", but it never calls
the CPU kernel -- it checks that bf16 round-trips e4m3 losslessly and
that subnormals are finite and small. That is the premise the widening
rests on, which is worth pinning, but it is not the widening. Renamed to
test_bf16_carries_every_e4m3_byte and the docstring now says what the
body asserts.

Not run here: this file is skipif(not torch.cuda.is_available()), and
this machine has no GPU and no torch. Compiles clean and ruff reports the
same two pre-existing E741 findings before and after.

Assisted-by: Claude

* fix(moe): let fp8_block experts decode on the cpu executor

* fix(cpu-moe): build the bf16 fp8 tier only when the compiler has avx512bf16

* fix(bench): version the bw profile and ignore older ones

* perf(cpu-moe): read fp8 weights as fp16 bits in the cpu dot kernels

* refactor(cpu-moe): write the fp8_block bank checks like the other formats

* refactor(cpu-moe): name the fp8 cpu symbols fp8_block

* docs(server): list fp8_block in the --moe-cpu-layers help

---------

Co-authored-by: Gabriel A. Devenyi <gdevenyi@gmail.com>
Co-authored-by: ChenyuHe <hechenyu@zju.edu.cn>
2026-10-11 22:12:41 -07:00
Xiaoze Fan 76449847b6 perf(kernel): avoid layout conversions in nvfp4 marlin decode (#631) 2026-10-07 22:42:44 -07:00
Xiaoze Fanandandy-yang-1 127c85665b feat(deepseek-v41): support text and vision serving (#603)
* test(deepseek-v41): add independent text and vision reference

* feat(quant): support block-32 and unrounded inference outputs

* feat(attention): add csa2 packed cache and sparse backend

* feat(deepseek-v41): serve text and images through shared runtime

* perf(mhc): tune batch-stable reduction tiles

* perf(moe): autotune fp4 decode launch geometry

* perf(moe): admit recurring misses against actual lru victims

* revert: perf(moe): admit recurring misses against actual lru victims

This reverts commit 1652d200f0ebb825379cf44c28baf52971e3a461.

* perf(mhc): derive the reduction tile from the width, not a timing

* perf(moe): replace the ds_fp4 decode autotune with a fixed two-warp launch

* fix(deepseek-v41): fence engram staging against in-flight lookups

* fix(moe): return the cpu routed sum in fp32 and round it in python

* fix(deepseek-v41): keep the vision tower and aligner unquantized

* fix(scheduler): admit bounded-replay hits by the history their resume reads

* refactor(deepseek-v41): name the attention family after the model, not csa2

* refactor(kvcache): import the dsv41 pool only for dsv41 models

* refactor(deepseek-v41): name the replay switch --swa-decoder-replay

* fix(server): keep reasoning_effort a string on the wire

* refactor(models): keep the prefill replay hooks off BaseLLMModel

* refactor(deepseek-v41): drop the fp32 routed sum and batch-invariant prefill

* fix(moe): weight ds_fp4 routes in the down epilogue again

* docs(models): anchor the v4.1 replay paragraph on the tech report

* refactor(deepseek-v41): run the decoder source's narrowed prefill on Block

* refactor(kvcache): let the dsv41 pool own the history a resume keeps live

* refactor(deepseek-v41): resume bounded-replay hits by capping the prefix match

* refactor(deepseek-v41): describe the attention group as a dsv4 variant and derive its geometry from the args

* refactor(deepseek-v41): keep the engram table locator in the model package

* refactor(deepseek-v41): drop the fp32 lm head output

- return the head's bf16 logits, as DeepSeek-V4 does
- revert ParallelLMHead(out_dtype=) and LinearKernel.apply_out_dtype, which only this call used
- remove test_head_preserves_fp32_logit_margin
- cast test_vision's prefill logits to fp32 before comparing, as the harness does

* refactor(quant): carry the w4a8 activation block on MoEConfig

- drop ActDesc and QuantScheme.act; the scheme only describes stored tensors
- Fp8BlockConfig sets MoEConfig.act_block from weight_block_size for its fp4 experts via a layer_config hook
- the ds_fp4 GPU kernel and CPU executor read the block from the layer config and require it

* refactor(deepseek-v41): size the kv pool from the args, not a geometry copy

- delete kvcache/dsv41_geometry.py; the pool and cost model read dsv41_args by attribute, as the DSV4 pool reads dsv4_args
- source / ratio aliasing reads args.roles; the duplicated source and ratio validation is gone
- the cost model owns the tier row formats, row bytes and the private window layers
- the pool keeps the format and window/ratio checks, and exposes the formats and is_private_window to the attention backend

* refactor(kvcache): move the dsv4 pool into kvcache/dsv4

* refactor(kvcache): move the dsv41 pool into kvcache/dsv4

* chore(benchmarks): drop the dsv41 selection and fp8 linear benchmarks

* test(deepseek-v41): drop tests that repeat other coverage

* test(deepseek-v41): load the tiny model's resident experts as banks

* fix(server): parse deepseek-v4.1's space-led dsml tool calls

---------

Co-authored-by: andy-yang-1 <andy_yang@berkeley.edu>
2026-10-07 00:15:11 -07:00
Xiaoze Fanandshubham3-ucb 1c183e74ae perf(moe): spin-wait the cpu moe pool between layers (#619)
Co-authored-by: shubham3-ucb <229030647+shubham3-ucb@users.noreply.github.com>
2026-10-06 21:38:10 -07:00
SuperHotDogCat c1a504a50b perf(moe): Improved fast_index_copy_multi_jit for NVLink-C2C hosts (#610)
* perf(moe): default fused gather to 32 blocks/bank on NVLink-C2C hosts

* build(aot): prebuild fast_index_copy_multi (1024, 32) variant

* perf(moe): make fused H2D gather grid configurable via env var

* docs: document FREETOKEN_H2D_BLOCKS_PER_BANK for NVLink-C2C hosts

* fix(moe): auto-detect H2D gather grid per device in the copy plan

* chore(docs): revised docs for NVLinkC2C
2026-10-06 19:54:38 -07:00
Xiaoze Fan 189c8ad0c3 feat(moe): serve NVFP4 and MXFP4 experts fused through the expert banks (#601)
* feat(moe): serve NVFP4 and MXFP4 experts fused through the expert banks

Resident experts skip the state dict and load through the offload expert-bank loader (HF pieces or FTW banks) into per-layer GPU banks, so fused serves bf16, fp8_block, NVFP4 and MXFP4 alike. Unified-memory GPUs auto-select fused for these formats and ignore offload-only flags with a warning.

BREAKING CHANGE: ft checkpoint always writes MoE experts as banks; an FTW converted with --moe-backend fused or triton (dense experts) no longer loads and must be reconverted.

* docs(moe): name only GB10 / DGX Spark for unified memory
2026-10-04 13:51:55 -07:00
Xiaoze Fanandandy-yang-1 cee23bf44a fix(runtime): correct overlap token limits and cache safety (#546)
* fix(scheduler): count delivered tokens under overlap

* fix(moe): initialize slots used by inactive hybrid routes

* fix(kvcache): report usable window capacity floors

* refactor(moe): inline the zeroed slot cache allocation

* test: drop the new regression tests

---------

Co-authored-by: andy-yang-1 <andy_yang@berkeley.edu>
2026-09-23 22:22:48 -07:00
Xiaoze Fan 477c8601b2 refactor(quant): config, scheme and method layers for quantization (#418)
* refactor(quant): config, scheme and method layers for checkpoint quantization

QuantConfig parses the checkpoint's quantization_config into a QuantScheme per
projection; every Linear*, ParallelLMHead and MoE layer takes quant_config and a
prefix, declares its tensors through a QuantMethod and selects its kernel at
construction. MoE kernels own the expert bank layout, pack and apply, so the
offload cache, the FTW converter and the CPU executor follow one layout.

- --quant-backend layer[.kind]=name selects kernels (--nvfp4-backend forwards to it)
- --moe-strategy replaces --moe-backend; the old spelling stays as an alias
- --moe-cpu-layers auto for Windows/WSL pin budgets; --moe-cache-auto respects the kernel's slot limit
- removes the per-family quantized layer classes, bank providers, format env switches and GLM runtime fp8
2026-09-08 23:37:23 -07:00
Xiaoze FanandShuo Yang a2538a428b feat(glm5_next): support GLM-5.3-Flash (#332)
Hybrid 45-layer model (34 KDA + 11 DSA) with NVFP4 MoE (288 experts):

- KDA linear-attention kernels and op, hybrid-linear pool dispatch
  generalized to any paged family
- NoPE-MLA attention with a kpool DSA indexer backend: 64-token pages,
  1/ratio shadow index slab, per-request tail rings, CUDA-graph-safe
  decode, prefix-cache snapshot contract
- mHC (x4 residual streams) layer with a fused split-K triton kernel;
  torch reference kept as the semantic pin
- clamped-SwiGLU activation across the MoE stack
- weight loading for NVFP4 exports in the multimodal-wrapper layout:
  ModelOpt and compressed-tensors (RedHatAI) tensor kinds;

Co-authored-by: Shuo Yang <73746844+andy-yang-1@users.noreply.github.com>
2026-09-01 15:18:15 -07:00
Xiaoze Fan e05cff83a0 perf(moe): route fused_topk through the in-repo triton router (#319) 2026-08-31 20:21:11 -07:00
Xiaoze Fan 3a20a79038 fix(kernel): unbreak the nightly kernel-cache wheel build (#310)
* fix(kernel): move fp8_block_scale_pad into aot_models to unbreak the kernel-cache build

* fix(kernel): exclude bank rows fast_index_copy cannot compile from aot specs

* fix(moe): fail fast when fused copy is off and a bank row cannot fall back
2026-08-30 22:43:50 -07:00
Xiaoze Fan bd8f3d519a feat(qwen4_exp): support Qwen3.8-Flash-Next (#257)
Serve Qwen3.8-Flash-Next (HF model_type qwen4_exp) text-only: 36 GDN +
12 QSA compressed-sparse attention layers on 4 hyper-connection residual
streams, a PLE n-gram embedding layer backed by a 47.7 GiB pinned-host
table with UVA gather, and 512 NVFP4 / block-fp8 routed experts (top-10)
plus a gated shared expert.

- attention: qsa_sparse backend (AttnType.QSA) over QSAKVCache -- paged
  GQA K/V, a 1/ratio compressed index-key slab shadowing the KV pages,
  and a per-request pending ring sized from index_ratio
- kvcache: declarative slot-sibling states (ModelConfig.slot_states) on
  LinearStatePool carry the PLE conv history and n-gram context through
  the hybrid-radix snapshot/COW lifecycle
- scheduler: hybrid prefill chunks align to the page size so snapshots
  land on donatable boundaries
- kernels: triton kernels adapted from vLLM/SGLang (hc, qsa, ple gather,
  moe router / shared gate) plus an original radix block top-k; int64
  row addressing throughout
- moe: non-power-of-2 top-k router, deep-K marlin decode config, one
  fp8 scale-bank padding rule shared with the AOT row table
- engine: the PLE table load reserves its pinned bytes from the pin
  budget before the expert banks plan their residency
2026-08-28 15:33:15 -07:00
Xiaoze Fan 2757bb5f91 feat(server)!: add --gpu to choose the GPU on multi-GPU machines (#117)
* feat(server)!: add --gpu to choose the GPU on multi-GPU machines

--gpu takes a GPU UUID (as nvidia-smi -L prints) or an nvidia-smi index.

It is applied as CUDA_VISIBLE_DEVICES in the parent before the workers spawn, so the engine still binds cuda:<rank>.

/v1/stats reports the engine's GPU under "gpus".

ft bench bw takes the same --gpu and writes one profile per GPU (benchbw/<gpu-uuid>.json); the legacy benchbw.json is still read by GPU name.

The daemon's /bench/profile returns the running serve's GPU profile.

BREAKING CHANGE: ft checkpoint --device is removed; use --gpu.

* refactor!: resolve --gpu via NVML and bind by UUID, not CUDA_VISIBLE_DEVICES

BREAKING CHANGE: ft bench bw --device is removed; use --gpu.

* refactor: name the id namespaces and drop the multi-device e4m3 scan

Splits the published id into _assigned_physical (UUID) and
_assigned_visible (CUDA ordinal). One process runs on one GPU, so
e4m3_native() judges that card instead of scanning every visible device.
2026-08-24 16:44:01 -07:00
Xiaoze Fan 831d38a66b fix(moe): report the residency banks actually settle at 2026-08-23 23:06:56 +00:00
Xiaoze Fan eebb3f53f3 feat(moe): estimate expert-bank bytes from the model config 2026-08-23 23:06:56 +00:00
Xiaoze Fan c41833b717 feat(moe): per-layer host-bank residency -- split lock-CPU/pin-GPU layers, auto selection under capped pin quota 2026-08-23 21:02:59 +00:00
Shuo Yang 9dc3ff4ca5 Support minimax m3 (#1)
* feat(minimax_m3): serve MiniMax-M3-NVFP4 single-GPU (block-sparse GQA + MXFP8 dense + swigluoai NVFP4 experts)
2026-08-13 16:22:41 -07:00
Xiaoze FanandShuo Yang 3af9d90ee5 feat: initial open-source release
Co-authored-by: Shuo Yang <andy_yang@berkeley.edu>
2026-08-11 22:53:25 +00:00