Dipika Sikka
60508ffda9
[Kernel]: Cutlass 2:4 Sparsity + FP8/Int8 Quant Support ( #10995 )
...
Co-authored-by: Faraz Shahsavan <faraz.shahsavan@gmail.com>
Co-authored-by: ilmarkov <markovilya197@gmail.com>
Co-authored-by: Rahul Tuli <rahul@neuralmagic.com>
Co-authored-by: rshaw@neuralmagic.com <rshaw@neuralmagic.com>
2024-12-18 09:57:16 -05:00
Luka Govedič
30870b4f66
[torch.compile] Dynamic fp8 + rms_norm fusion ( #10906 )
...
Signed-off-by: luka <luka@neuralmagic.com>
Co-authored-by: Varun Sundar Rabindranath <varun@neuralmagic.com>
2024-12-13 03:19:23 +00:00
kliuae
7c25fe45a6
[AMD] Add support for GGUF quantization on ROCm ( #10254 )
2024-11-22 21:14:49 -08:00
Luka Govedič
4f93dfe952
[torch.compile] Fuse RMSNorm with quant ( #9138 )
...
Signed-off-by: luka <luka@neuralmagic.com>
Co-authored-by: youkaichao <youkaichao@126.com>
2024-11-08 21:20:08 +00:00
Hanzhi Zhou
6192e9b8fe
[Core][Distributed] Refactor ipc buffer init in CustomAllreduce ( #10030 )
...
Signed-off-by: Hanzhi Zhou <hanzhi713@gmail.com>
2024-11-06 23:50:47 -08:00
youkaichao
8549c82660
[core] cudagraph output with tensor weak reference ( #9724 )
...
Signed-off-by: youkaichao <youkaichao@gmail.com>
2024-10-27 00:19:28 -07:00
Charlie Fu
59449095ab
[Performance][Kernel] Fused_moe Performance Improvement ( #9384 )
...
Signed-off-by: charlifu <charlifu@amd.com>
2024-10-24 15:37:52 -07:00
Jee Jee Li
295a061fb3
[Kernel] add kernel for FATReLU ( #9610 )
...
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
2024-10-24 16:18:27 +08:00
Mor Zusman
fb60ae9b91
[Kernel][Model] Improve continuous batching for Jamba and Mamba ( #9189 )
2024-10-16 12:12:43 -04:00
Lucas Wilkinson
aeb37c2a72
[CI/Build] Per file CUDA Archs (improve wheel size and dev build times) ( #8845 )
2024-10-03 22:55:25 -04:00
Mor Zusman
f13a07b1f8
[Kernel][Model] Varlen prefill + Prefill chunking support for mamba kernels and Jamba model ( #8533 )
2024-09-29 17:35:58 -04:00
Lucas Wilkinson
86e9c8df29
[Kernel] (2/N) Machete - Integrate into CompressedTensorsWNA16 and GPTQMarlin ( #7701 )
...
Co-authored-by: mgoin <michael@neuralmagic.com>
Co-authored-by: Divakar Verma <137818590+divakar-amd@users.noreply.github.com>
Co-authored-by: Tyler Michael Smith <tyler@neuralmagic.com>
2024-09-23 13:46:26 -04:00
Tyler Michael Smith
8110e44529
[Kernel] Change interface to Mamba causal_conv1d_update for continuous batching ( #8012 )
2024-09-17 23:44:27 +00:00
youkaichao
99aa4eddaf
[torch.compile] register allreduce operations as custom ops ( #8526 )
2024-09-16 22:57:57 -07:00
Luka Govedič
5d73ae49d6
[Kernel] AQ AZP 3/4: Asymmetric quantization kernels ( #7270 )
2024-09-16 11:52:40 -07:00
William Lin
a6c0f3658d
[multi-step] add flashinfer backend ( #7928 )
2024-09-12 11:16:22 -07:00
bnellnm
73202dbe77
[Kernel][Misc] register ops to prevent graph breaks ( #6917 )
...
Co-authored-by: Sage Moore <sage@neuralmagic.com>
2024-09-11 12:52:19 -07:00
Dipika Sikka
23f322297f
[Misc] Remove SqueezeLLM ( #8220 )
2024-09-06 16:29:03 -06:00
Mor Zusman
fdd9daafa3
[Kernel/Model] Migrate mamba_ssm and causal_conv1d kernels to vLLM ( #7651 )
2024-08-28 15:06:52 -07:00
Lucas Wilkinson
5288c06aa0
[Kernel] (1/N) Machete - Hopper Optimized Mixed Precision Linear Kernel ( #7174 )
2024-08-20 07:09:33 -06:00
bnellnm
37fd47e780
[Kernel] fix types used in aqlm and ggml kernels to support dynamo ( #7596 )
2024-08-16 14:00:11 -07:00
Luka Govedič
8d59dbb000
[Kernel] Add per-tensor and per-token AZP epilogues ( #5941 )
...
Co-authored-by: Tyler Michael Smith <tyler@neuralmagic.com>
2024-08-06 18:17:08 +00:00
Isotr0py
360bd67cf0
[Core] Support loading GGUF model ( #5191 )
...
Co-authored-by: Michael Goin <michael@neuralmagic.com>
2024-08-05 17:54:23 -06:00
Lucas Wilkinson
a8d604ca2a
[Misc] Disambiguate quantized types via a new ScalarType ( #6396 )
2024-08-02 13:51:58 -07:00
HandH1998
6512937de1
Support W4A8 quantization for vllm ( #5218 )
2024-07-31 07:55:21 -06:00
Alexander Matveev
75acdaa4b6
[Kernel] Increase precision of GPTQ/AWQ Marlin kernel ( #6795 )
2024-07-27 17:52:33 -04:00
Alexander Matveev
396d92d5e0
[Kernel][Core] Add AWQ support to the Marlin kernel ( #6612 )
2024-07-21 19:41:42 -04:00
Varun Sundar Rabindranath
2e26564259
[ Kernel ] FP8 Dynamic Per Token Quant - Add scale_ub ( #6593 )
...
Co-authored-by: Varun Sundar Rabindranth <varun@neuralmagic.com>
2024-07-19 18:15:26 -07:00
Varun Sundar Rabindranath
b5241e41d9
[ Kernel ] FP8 Dynamic-Per-Token Quant Kernel ( #6511 )
...
Co-authored-by: Varun Sundar Rabindranath <varun@neuralmagic.com>
2024-07-18 01:38:35 +00:00
Alexander Matveev
e76466dde2
[Core] draft_model_runner: Implement prepare_inputs on GPU for advance_step ( #6338 )
2024-07-17 14:30:28 -07:00
Michael Goin
978aed5300
[Kernel][Attention] Separate Attention.kv_scale into k_scale and v_scale ( #6081 )
2024-07-16 15:31:32 -07:00
Michael Goin
47f0954af0
[Kernel] Expand FP8 support to Ampere GPUs using FP8 Marlin ( #5975 )
2024-07-03 17:38:00 +00:00
Chip Kerchner
38a1674abb
Support CPU inference with VSX PowerPC ISA ( #5652 )
2024-06-26 21:53:04 +00:00
Luka Govedič
5bfd1bbc98
[Kernel] Adding bias epilogue support for cutlass_scaled_mm ( #5560 )
...
Co-authored-by: Chih-Chieh-Yang <7364402+cyang49@users.noreply.github.com>
Co-authored-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
2024-06-26 15:16:00 +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
Tyler Michael Smith
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
bnellnm
5467ac3196
[Kernel][Misc] Use TORCH_LIBRARY instead of PYBIND11_MODULE for custom ops ( #5047 )
2024-06-09 16:23:30 -04:00
Dipika Sikka
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
cbb2f59cc8
[Kernel] Pass a device pointer into the quantize kernel for the scales ( #5159 )
2024-06-03 09:52:30 -07:00
Eric Xihui Lin
8e192ff967
[Kernel][Backend][Model] Blocksparse flash attention kernel and Phi-3-Small model ( #4799 )
...
Co-authored-by: beagleski <yunanzhang@microsoft.com>
Co-authored-by: bapatra <bapatra@microsoft.com>
Co-authored-by: Barun Patra <codedecde@users.noreply.github.com>
Co-authored-by: Michael Goin <michael@neuralmagic.com>
2024-05-24 22:00:52 -07:00
Dipika Sikka
a1242324c9
[Kernel] Initial Activation Quantization Support ( #4525 )
...
Co-authored-by: Varun Sundar Rabindranath <varunsundar08@gmail.com>
Co-authored-by: Varun Sundar Rabindranath <varun@neuralmagic.com>
2024-05-23 21:29:18 +00:00
Michael Goin
5f6d10c14c
[CI/Build] Enforce style for C++ and CUDA code with clang-format ( #4722 )
2024-05-22 07:18:41 +00:00
Tyler Michael Smith
2060e93659
[Kernel] Add w8a8 CUTLASS kernels ( #4749 )
2024-05-16 18:32:50 -04:00
Alexander Matveev
6979ade384
Add GPTQ Marlin 2:4 sparse structured support ( #4790 )
...
Co-authored-by: Robert Shaw <rshaw@neuralmagic.com>
2024-05-16 12:56:15 -04:00
SangBin Cho
3521ba4f25
[Core][Model runner refactoring 1/N] Refactor attn metadata term ( #4518 )
2024-05-03 10:20:12 -07:00
alexm-nm
7038e8b803
[Kernel] Support running GPTQ 8-bit models in Marlin ( #4533 )
2024-05-02 12:56:22 -04:00
Robert Shaw
73c8d677e5
[Kernel] Marlin Expansion: Support AutoGPTQ Models with Marlin ( #3922 )
...
Co-authored-by: alexm <alexm@neuralmagic.com>
Co-authored-by: mgoin <michael@neuralmagic.com>
2024-04-29 09:35:34 -07:00
Philipp Moritz
12628d3c78
[Kernel] Optimize FP8 support for MoE kernel / Mixtral via static scales ( #4343 )
...
Co-authored-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
2024-04-27 04:49:59 +00:00
Philipp Moritz
eace8bf0b9
[Kernel] FP8 support for MoE kernel / Mixtral ( #4244 )
...
This PR is the first step towards fixing https://github.com/vllm-project/vllm/pull/3208
It implements dynamic per-tensor scaling (see https://github.com/vllm-project/vllm/pull/4118 ), so users do not need to compute activation scales on a calibration dataset and they also don't need to convert their model checkpoints. It is enough to specify the `quantization="fp8"` argument. You can try out the PR like this:
```python
from vllm import LLM, SamplingParams
prompts = [
"Hello, my name is",
"The president of the United States is",
"The capital of France is",
"The future of AI is",
]
sampling_params = SamplingParams(temperature=0.8, top_p=0.95)
llm = LLM(model="mistralai/Mixtral-8x7B-Instruct-v0.1", tensor_parallel_size=2, quantization="fp8")
outputs = llm.generate(prompts, sampling_params)
# Print the outputs.
for output in outputs:
prompt = output.prompt
generated_text = output.outputs[0].text
print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")
```
**Performance**: For this PR, the focus is on making the code clean (while still trying to get reasonable performance), there is a bunch of optimizations that we will submit as a follow up PR that significantly improve the performance (similar to the numbers in https://github.com/vllm-project/vllm/pull/3954 ). With this PR, the results are as follows:
<img width="725" alt="Screenshot 2024-04-21 at 1 31 50 PM" src="https://github.com/vllm-project/vllm/assets/113316/d8fe1118-07a0-4d4e-8530-37a77d465a03 ">
**Accuracy**: The accuracy with this PR on MMLU on `mistralai/Mixtral-8x7B-v0.1` is as follows:
```
| Groups |Version|Filter|n-shot|Metric|Value | |Stderr|
|------------------|-------|------|-----:|------|-----:|---|-----:|
|mmlu |N/A |none | 0|acc |0.7018|± |0.0036|
| - humanities |N/A |none | 5|acc |0.6472|± |0.0065|
| - other |N/A |none | 5|acc |0.7673|± |0.0072|
| - social_sciences|N/A |none | 5|acc |0.8099|± |0.0070|
| - stem |N/A |none | 5|acc |0.6131|± |0.0083|
```
this compares favorably with the fp16 results which are
```
| Groups |Version|Filter|n-shot|Metric|Value | |Stderr|
|------------------|-------|------|-----:|------|-----:|---|-----:|
|mmlu |N/A |none | 0|acc |0.7020|± |0.1313|
| - humanities |N/A |none | 5|acc |0.6425|± |0.1349|
| - other |N/A |none | 5|acc |0.7744|± |0.1038|
| - social_sciences|N/A |none | 5|acc |0.8131|± |0.0695|
| - stem |N/A |none | 5|acc |0.6108|± |0.1383|
```
Happy hacking!
2024-04-24 01:18:23 +00:00
