daquexian
99ded1e1c4
[Doc] Remove comments incorrectly copied from another project ( #6286 )
2024-07-10 17:05:26 -07:00
Baoyuan Qi and Robert Shaw
d3a245138a
[Bugfix]fix and needs_scalar_to_array logic check ( #6238 )
...
Co-authored-by: Robert Shaw <114415538+robertgshaw2-neuralmagic@users.noreply.github.com >
2024-07-09 23:43:24 +00:00
Avshalom Manevich
f7a8fa39d8
[Kernel] reloading fused_moe config on the last chunk ( #6210 )
2024-07-08 08:00:38 -07:00
Robert Shaw and Robert Shaw
abfe705a02
[ Misc ] Support Fp8 via llm-compressor ( #6110 )
...
Co-authored-by: Robert Shaw <rshaw@neuralmagic>
2024-07-07 20:42:11 +00:00
Robert Shaw
62963d129e
[ Misc ] Clean Up CompressedTensorsW8A8 ( #6113 )
2024-07-03 22:50:08 +00:00
Michael Goin
47f0954af0
[Kernel] Expand FP8 support to Ampere GPUs using FP8 Marlin ( #5975 )
2024-07-03 17:38:00 +00:00
youkaichao
482045ee77
[hardware][misc] introduce platform abstraction ( #6080 )
2024-07-02 20:12:22 -07:00
ee93f4f92a
[CORE] Quantized lm-head Framework ( #4442 )
...
Co-authored-by: Robert Shaw <rshaw@neuralmagic.com >
Co-authored-by: ZX <zx@lbx.dev >
2024-07-02 22:25:17 +00:00
7c008c51a9
[ Misc ] Refactor MoE to isolate Fp8 From Mixtral ( #5970 )
...
Co-authored-by: Robert Shaw <rshaw@neuralmagic>
Co-authored-by: Michael Goin <michael@neuralmagic.com >
2024-07-02 21:54:35 +00:00
Avshalom Manevich
12a59959ed
[Bugfix] adding chunking mechanism to fused_moe to handle large inputs ( #6029 )
2024-07-01 21:08:29 +00:00
sroy745
80ca1e6a3a
[Speculative Decoding 2/2 ] Integrate typical acceptance sampler into Spec Decode Worker ( #5348 )
2024-07-01 00:33:05 -07:00
youkaichao
614aa51203
[misc][cuda] use nvml to avoid accidentally cuda initialization ( #6007 )
2024-06-30 20:07:34 -07:00
Robert Shaw and Robert Shaw
af9ad46fca
[ Misc ] Refactor w8a8 to use process_weights_after_load (Simplify Weight Loading) ( #5940 )
...
Co-authored-by: Robert Shaw <rshaw@neuralmagic>
2024-06-30 23:06:27 +00:00
Dipika Sikka
7836fdcc11
[Misc] Fix get_min_capability ( #5971 )
2024-06-30 20:15:16 +00:00
8dbfcd35bf
[ CI/Build ] Added E2E Test For Compressed Tensors ( #5839 )
...
Co-authored-by: Michael Goin <michael@neuralmagic.com >
Co-authored-by: Robert Shaw <rshaw@neuralmagic>
2024-06-29 21:12:58 +08:00
Cody Yu
f7dac83d95
[Kernel] Raise an exception in MoE kernel if the batch size is larger then 65k ( #5939 )
2024-06-29 21:04:20 +08:00
wangding zeng and Philipp Moritz
be0b3af9e0
Support Deepseek-V2 ( #4650 )
...
Co-authored-by: Philipp Moritz <pcmoritz@gmail.com >
2024-06-28 13:24:57 -07:00
Robert Shaw and Robert Shaw
2cd402e169
[ Bugfix ] Enabling Loading Models With Fused QKV/MLP on Disk with FP8 ( #5921 )
...
Co-authored-by: Robert Shaw <rshaw@neuralmagic>
2024-06-28 18:43:49 +00:00
Robert Shaw and Robert Shaw
b185230744
[ Misc ] Remove fp8_shard_indexer from Col/Row Parallel Linear (Simplify Weight Loading) ( #5928 )
...
Co-authored-by: Robert Shaw <rshaw@neuralmagic>
2024-06-28 13:49:57 -04:00
Ilya Lavrenov
57f09a419c
[Hardware][Intel] OpenVINO vLLM backend ( #5379 )
2024-06-28 13:50:16 +00:00
Tyler Michael Smith
5932634409
Unmark fused_moe config json file as executable ( #5960 )
2024-06-28 06:36:12 -07:00
Divakar Verma
c3dde367f1
[Kernel][ROCm][AMD] fused_moe Triton configs v2 for mi300X ( #5932 )
2024-06-27 13:41:08 -07:00
Woosuk Kwon
79c92c7c8a
[Model] Add Gemma 2 ( #5908 )
2024-06-27 13:33:56 -07:00
Woosuk Kwon
6806998bf9
[Bugfix] Fix embedding to support 2D inputs ( #5829 )
2024-06-26 00:15:22 -07:00
Dipika Sikka
dd248f7675
[Misc] Update w4a16 compressed-tensors support to include w8a16 ( #5794 )
2024-06-25 19:23:35 +00:00
Roger Wang
bd620b01fb
[Kernel][CPU] Add Quick gelu to CPU ( #5717 )
2024-06-21 06:39:40 +00:00
Tyler Michael Smith
3f3b6b2150
[Bugfix] Fix the CUDA version check for FP8 support in the CUTLASS kernels ( #5715 )
2024-06-20 18:36:10 +00:00
Roger Wang
ad137cd111
[Model] Port over CLIPVisionModel for VLMs ( #5591 )
2024-06-20 11:52:09 +00:00
Dipika Sikka
4a30d7e3cc
[Misc] Add per channel support for static activation quantization; update w8a8 schemes to share base classes ( #5650 )
2024-06-19 18:06:44 -04:00
Shukant Pal
59a1eb59c9
[Bugfix] Fix Phi-3 Long RoPE scaling implementation ( #5628 )
2024-06-19 01:46:38 +00:00
Dipika Sikka
95db455e7f
[Misc] Add channel-wise quantization support for w8a8 dynamic per token activation quantization ( #5542 )
2024-06-18 12:45:05 -04:00
sroy745
fa9e385229
[Speculative Decoding 1/2 ] Add typical acceptance sampling as one of the sampling techniques in the verifier ( #5131 )
2024-06-17 21:29:09 -05:00
728c4c8a06
[Hardware][Intel GPU] Add Intel GPU(XPU) inference backend ( #3814 )
...
Co-authored-by: Jiang Li <jiang1.li@intel.com >
Co-authored-by: Abhilash Majumder <abhilash.majumder@intel.com >
Co-authored-by: Abhilash Majumder <30946547+abhilash1910@users.noreply.github.com >
2024-06-17 11:01:25 -07:00
Dipika Sikka
890d8d960b
[Kernel] compressed-tensors marlin 24 support ( #5435 )
2024-06-17 12:32:48 -04:00
Amit Garg
9333fb8eb9
[Model] Rename Phi3 rope scaling type ( #5595 )
2024-06-17 12:04:14 -04:00
Cyrus Leung
0e9164b40a
[mypy] Enable type checking for test directory ( #5017 )
2024-06-15 04:45:31 +00:00
15985680e2
[ Misc ] Rs/compressed tensors cleanup ( #5432 )
...
Co-authored-by: mgoin <michael@neuralmagic.com >
Co-authored-by: Dipika Sikka <dipikasikka1@gmail.com >
2024-06-14 10:01:46 -07:00
Tyler Michael Smith
703475f6c2
[Kernel] Fix CUTLASS 3.x custom broadcast load epilogue ( #5516 )
2024-06-14 09:30:15 -07:00
Tyler Michael Smith
e38042d4af
[Kernel] Disable CUTLASS kernels for fp8 ( #5505 )
2024-06-13 13:38:05 -07:00
85657b5607
[Kernel] Factor out epilogues from cutlass kernels ( #5391 )
...
Co-authored-by: Michael Goin <michael@neuralmagic.com >
Co-authored-by: youkaichao <youkaichao@gmail.com >
Co-authored-by: zifeitong <zifei.tong@parasail.io >
Co-authored-by: Robert Shaw <114415538+robertgshaw2-neuralmagic@users.noreply.github.com >
2024-06-13 11:22:19 -07:00
wenyujin333
bd43973522
[Kernel] Tune Qwen2MoE kernel configurations with tp2,4 ( #5497 )
...
Tune Qwen2-57B-A14B configs based on #4921
Throughput Performance
command: python benchmarks/benchmark_throughput.py --model=Qwen/Qwen2-57B-A14B-Instruct --input-len 1000 --output-len 50 -tp 2
A100 GPU
benchmark no config w/ PR
tp=2 10.53 requests/s, 11058.17 tokens/s 12.47 requests/s, 13088.57 tokens/s
tp=4 17.77 requests/s, 18662.95 tokens/s 20.20 requests/s, 21212.32 tokens/s
2024-06-13 09:01:10 -07:00
Dipika Sikka and Robert Shaw
c2637a613b
[Kernel] w4a16 support for compressed-tensors ( #5385 )
...
Co-authored-by: Robert Shaw <114415538+robertgshaw2-neuralmagic@users.noreply.github.com >
2024-06-13 10:19:56 -04:00
Woosuk Kwon
1a8bfd92d5
[Hardware] Initial TPU integration ( #5292 )
2024-06-12 11:53:03 -07:00
Nick Hill
a008629807
[Misc] Various simplifications and typing fixes ( #5368 )
2024-06-11 10:29:02 +08:00
Dipika Sikka and Michael Goin
5884c2b454
[Misc] Update to comply with the new compressed-tensors config ( #5350 )
...
Co-authored-by: Michael Goin <michael@neuralmagic.com >
2024-06-10 03:49:46 +00:00
bnellnm
5467ac3196
[Kernel][Misc] Use TORCH_LIBRARY instead of PYBIND11_MODULE for custom ops ( #5047 )
2024-06-09 16:23:30 -04:00
Michael Goin
c09dade2a2
[Misc][Breaking] Change FP8 checkpoint format from act_scale -> input_scale ( #5353 )
2024-06-08 13:54:05 -04:00
Cheng Li
e69ded7d1c
[Bug Fix] Fix the support check for FP8 CUTLASS ( #5352 )
...
Bug description:
With torch 2.4.0.dev20240603+cu121,
cutlass_fp8_supported outputs False, and the (capability, version) before the comparison is (90, 11111111112)
This PR fixes the support check for FP8 CUTLASS ( cutlass_fp8_supported) which was introduced in https://github.com/vllm-project/vllm/pull/5183 .
2024-06-08 00:42:05 +00:00
ca3ea51bde
[Kernel] Dynamic Per-Token Activation Quantization ( #5037 )
...
Co-authored-by: Varun Sundar Rabindranath <varunsundar08@gmail.com >
Co-authored-by: Varun Sundar Rabindranath <varun@neuralmagic.com >
2024-06-07 09:36:26 -07:00
Tyler Michael Smith
8d75fe48ca
[Kernel] Switch fp8 layers to use the CUTLASS kernels ( #5183 )
...
Switching from torch._scaled_mm to vLLM's cutlass fp8 kernels when supported as we are seeing 5-15% improvement in e2e performance on neuralmagic/Meta-Llama-3-8B-Instruct-FP8
see https://docs.google.com/spreadsheets/d/1GiAnmzyGHgZ6zL_LDSTm35Bdrt4A8AaFEurDlISYYA4/ for some quick e2e benchmarks and #5144 for comparisons across different GEMM sizes.
2024-06-07 08:42:35 +00:00
daquexian