Commits · 55fb97f7bd61273fe8464a72866a72eaa88b5759 · OpenDAS / vllm_cscc

26 Dec, 2024 1 commit

[Model] [Quantization] Support deepseek_v3 w8a8 fp8 block-wise quantization (#11523) · 2072924d

Michael Goin authored Dec 26, 2024


Signed-off-by: mgoin <michael@neuralmagic.com>
Signed-off-by: simon-mo <simon.mo@hey.com>
Signed-off-by: simon-mo <xmo@berkeley.edu>
Co-authored-by: simon-mo <simon.mo@hey.com>
Co-authored-by: simon-mo <xmo@berkeley.edu>
Co-authored-by: HandH1998 <1335248067@qq.com>

2072924d

23 Dec, 2024 1 commit
- [Misc] Add assertion and helpful message for marlin24 compressed models (#11388) · b866cdbd
  Dipika Sikka authored Dec 23, 2024
  
  b866cdbd
21 Dec, 2024 1 commit
- [Bugfix] update should_ignore_layer (#11354) · 51ff216d
  George authored Dec 21, 2024
```
Signed-off-by: George Ohashi <george@neuralmagic.com>
```
  51ff216d
19 Dec, 2024 2 commits
- [Bugfix] Fix broken CPU compressed-tensors test (#11338) · 276738ce
  Isotr0py authored Dec 20, 2024
```
Signed-off-by: Isotr0py <2037008807@qq.com>
```
  276738ce
- [Bugfix][Build/CI] Fix sparse CUTLASS compilation on CUDA [12.0, 12.2) (#11311) · 5a9da2e6
  Tyler Michael Smith authored Dec 18, 2024
```
Signed-off-by: Tyler Michael Smith <tyler@neuralmagic.com>
```
  5a9da2e6
18 Dec, 2024 1 commit

[Kernel]: Cutlass 2:4 Sparsity + FP8/Int8 Quant Support (#10995) · 60508ffd

Dipika Sikka authored Dec 18, 2024


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>

60508ffd

15 Dec, 2024 1 commit
- [[Misc]Upgrade bitsandbytes to the latest version 0.45.0 (#11201) · 15859f23
  Jee Jee Li authored Dec 15, 2024
  
  15859f23
27 Nov, 2024 2 commits
- [Bugfix] Fix GGUF inference with FP16 unquantized checkpoint (#10675) · b98c62ba
  Isotr0py authored Nov 28, 2024
```
Signed-off-by: Isotr0py <2037008807@qq.com>
```
  b98c62ba
- [bugfix] fix the default value of llm_int8_threshold in BitsAndBytesConfig (#10657) · cfb3bf25
  yansh97 authored Nov 27, 2024
  
  cfb3bf25
26 Nov, 2024 1 commit
- [Bugfix] Check bnb_4bit_quant_storage for bitsandbytes (#10642) · 7576cd38
  Michael Goin authored Nov 26, 2024
  
  7576cd38
21 Nov, 2024 1 commit
- [torch.compile] limit inductor threads and lazy import quant (#10482) · 388ee3de
  youkaichao authored Nov 20, 2024
```
Signed-off-by: youkaichao <youkaichao@gmail.com>
```
  388ee3de
19 Nov, 2024 1 commit
- [Model][Quantization] HQQ support through Marlin kernel expansion (#9766) · b00b33d7
  ElizaWszola authored Nov 19, 2024
```
Signed-off-by: ElizaWszola <eliza@neuralmagic.com>
```
  b00b33d7
18 Nov, 2024 3 commits
- [Kernel] Initial Machete W4A8 support + Refactors (#9855) · 96d999fb
  Lucas Wilkinson authored Nov 18, 2024
```
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
```
  96d999fb
- [Kernel] Explicitly specify other value in tl.load calls (#9014) · c2170a5b
  Angus Wang authored Nov 18, 2024
```
Signed-off-by: Angus Wang <wangjadehao@gmail.com>
```
  c2170a5b
- [Hardware][XPU] AWQ/GPTQ support for xpu backend (#10107) · 6b2d25ef
  Yan Ma authored Nov 19, 2024
```
Signed-off-by: yan ma <yan.ma@intel.com>
```
  6b2d25ef
15 Nov, 2024 1 commit

[bugfix] Fix static asymmetric quantization case (#10334) · bf2ddc66

Luka Govedič authored Nov 14, 2024


Signed-off-by: Daniël de Kok <me@danieldk.eu>
Signed-off-by: luka <luka@neuralmagic.com>
Co-authored-by: Daniël de Kok <me@danieldk.eu>

bf2ddc66

09 Nov, 2024 1 commit

[Kernel][Triton] Add Triton implementation for scaled_mm_triton to support fp8... · 127c0748

rasmith authored Nov 08, 2024


[Kernel][Triton] Add Triton implementation for scaled_mm_triton to support fp8 and int8 SmoothQuant, symmetric case (#9857)
Signed-off-by: Randall Smith <Randall.Smith@amd.com>

127c0748

07 Nov, 2024 2 commits
- [Hardware][CPU] Update torch 2.5 (#9911) · a4b3e0c1
  Li, Jiang authored Nov 07, 2024
```
Signed-off-by: jiang1.li <jiang1.li@intel.com>
```
  a4b3e0c1
- [Bugfix] Fix FP8 torch._scaled_mm fallback for torch>2.5 with CUDA<12.4 (#10095) · 4ab32566
  Michael Goin authored Nov 06, 2024
```
Signed-off-by: mgoin <michael@neuralmagic.com>
```
  4ab32566
06 Nov, 2024 1 commit
- Remove ScaledActivation for AWQ (#10057) · 399c7986
  Michael Goin authored Nov 06, 2024
```
Signed-off-by: mgoin <michael@neuralmagic.com>
```
  399c7986
05 Nov, 2024 1 commit
- [Misc] Modify BNB parameter name (#9997) · b9c64c0c
  Jee Jee Li authored Nov 06, 2024
```
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
```
  b9c64c0c
04 Nov, 2024 1 commit
- [Bugfix] Respect modules_to_not_convert within awq_marlin (#9895) · 8f0a9ca8
  Michael Goin authored Nov 04, 2024
```
Signed-off-by: mgoin <michael@neuralmagic.com>
```
  8f0a9ca8
01 Nov, 2024 2 commits
- [Bugfix/Core] Flashinfer k_scale and v_scale (#9861) · 598b6d7b
  Pavani Majety authored Nov 01, 2024
  
  598b6d7b
- [Bugfix] Fix layer skip logic with bitsandbytes (#9887) · 37a4947d
  Michael Goin authored Nov 01, 2024
```
Signed-off-by: mgoin <michael@neuralmagic.com>
```
  37a4947d
30 Oct, 2024 1 commit
- [Bugfix][VLM] Make apply_fp8_linear work with >2D input (#9812) · 226688bd
  Michael Goin authored Oct 29, 2024
  
  226688bd
29 Oct, 2024 1 commit
- [Model] Add BNB quantization support for Mllama (#9720) · 09500f7d
  Isotr0py authored Oct 29, 2024
  
  09500f7d
28 Oct, 2024 2 commits
- [torch.compile] support moe models (#9632) · 32176fee
  youkaichao authored Oct 27, 2024
```
Signed-off-by: youkaichao <youkaichao@gmail.com>
```
  32176fee
- [Hardware][ROCM] using current_platform.is_rocm (#9642) · 4e2d95e3
  wangshuai09 authored Oct 28, 2024
```
Signed-off-by: wangshuai09 <391746016@qq.com>
```
  4e2d95e3
25 Oct, 2024 1 commit
- [Bugfix] Fix compressed_tensors_moe bad config.strategy (#9677) · ca0d9222
  Michael Goin authored Oct 25, 2024
  
  ca0d9222
23 Oct, 2024 1 commit
- [VLM] Enable overriding whether post layernorm is used in vision encoder + fix quant args (#9217) · c18e1a34
  Cyrus Leung authored Oct 23, 2024
```
Co-authored-by: Isotr0py <2037008807@qq.com>
```
  c18e1a34
18 Oct, 2024 1 commit
- [BugFix] Stop silent failures on compressed-tensors parsing (#9381) · 48138a84
  Dipika Sikka authored Oct 17, 2024
  
  48138a84
17 Oct, 2024 1 commit
- [Kernel] Add Exllama as a backend for compressed-tensors (#9395) · e312e52b
  Lucas Wilkinson authored Oct 17, 2024
  
  e312e52b
15 Oct, 2024 1 commit
- [Misc] Directly use compressed-tensors for checkpoint definitions (#8909) · 22f8a695
  Michael Goin authored Oct 15, 2024
```
Co-authored-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  22f8a695
09 Oct, 2024 2 commits
- [Hardware][CPU] Support AWQ for CPU backend (#7515) · ca77dd7a
  Li, Jiang authored Oct 10, 2024
  
  ca77dd7a
- support bitsandbytes quantization with more models (#9148) · 2f4117c3
  chenqianfzh authored Oct 08, 2024
  
  2f4117c3
07 Oct, 2024 1 commit
- [Core] Refactor GGUF parameters packing and forwarding (#8859) · f19da648
  Isotr0py authored Oct 07, 2024
  
  f19da648
04 Oct, 2024 1 commit
- [Kernel] Zero point support in fused MarlinMoE kernel + AWQ Fused MoE (#8973) · 05d68643
  ElizaWszola authored Oct 04, 2024
```
Co-authored-by: Dipika <dipikasikka1@gmail.com>
Co-authored-by: Dipika Sikka <ds3822@columbia.edu>
```
  05d68643
28 Sep, 2024 1 commit
- [CI/Build] Update models tests & examples (#8874) · e1a3f5e8
  Cyrus Leung authored Sep 29, 2024
```
Co-authored-by: Roger Wang <ywang@roblox.com>
```
  e1a3f5e8
27 Sep, 2024 2 commits
- [Bugfix] fix for deepseek w4a16 (#8906) · c5d55356
  Lucas Wilkinson authored Sep 27, 2024
```
Co-authored-by: mgoin <michael@neuralmagic.com>
```
  c5d55356
- [Kernel] AQ AZP 4/4: Integrate asymmetric quantization to linear method (#7271) · 172d1cd2
  Luka Govedič authored Sep 27, 2024
  
  172d1cd2