Commits · 31bc990677e82f9396b4ea3050a1fdaafcb1cce3 · OpenDAS / ktransformers

25 Feb, 2025 23 commits
- Merge pull request #667 from Azure-Tang/update-readme · 31bc9906
  Azure authored Feb 26, 2025
```
[update] Update doc.
```
  31bc9906
- update doc · bb6920ed
  Azure authored Feb 25, 2025
  
  bb6920ed
- Merge pull request #665 from KMSorSMS/v0.2.2rc1 · 9c71bcb0
  ZiWei Yuan authored Feb 25, 2025
```
⚡ release v0.2.2rc1
```
  9c71bcb0
- ⚡ release v0.2.2rc1 · ddf33393
  liam authored Feb 25, 2025
  
  ddf33393
- Merge pull request #663 from kvcache-ai/develop-0.2.2 · 8333a4d8
  Azure authored Feb 25, 2025
```
[release]  Release 0.2.2rc.
```
  8333a4d8
- Merge pull request #662 from Azure-Tang/support-fp8 · c6e4e1c3
  Azure authored Feb 25, 2025
```
[update] Update readme.
```
  c6e4e1c3
- Merge branch 'develop-0.2.2' into support-fp8 · 91c16192
  Azure authored Feb 25, 2025
```
 Update README.md
```
  91c16192
- Update DeepseekR1_V3_tutorial.md · 13974eb2
  Atream authored Feb 25, 2025
  
  13974eb2
- Update DeepseekR1_V3_tutorial.md add long context · 03f8bc9f
  Atream authored Feb 25, 2025
  
  03f8bc9f
- add fp8 multi gpu yaml example · 2c0cce90
  Azure authored Feb 25, 2025
  
  2c0cce90
- Merge branch 'fix-update-flashinfer_wrapper_local_chat' into develop-0.2.2 · d9b2895b
  Atream authored Feb 25, 2025
  
  d9b2895b
- fix-update-flashinfer_wrapper_local_chat · 477ac28a
  Atream authored Feb 25, 2025
  
  477ac28a
- fix fp8 multi gpu; update FQA · 7e5962af
  Azure authored Feb 25, 2025
  
  7e5962af
- Merge pull request #659 from KMSorSMS/develop-0.2.2 · 89b55052
  ZiWei Yuan authored Feb 25, 2025
```
📝 add benchmark.md
```
  89b55052
- 📝 add benchmark.md · 1b5ac67f
  liam authored Feb 25, 2025
  
  1b5ac67f
- Merge pull request #658 from KMSorSMS/develop-0.2.2 · 1aa10e93
  ZiWei Yuan authored Feb 25, 2025
```
⚡ update git ignore add docker dev container
```
  1aa10e93
- ⚡ update git ignore add docker dev container · 0ca0b99f
  liam authored Feb 25, 2025
  
  0ca0b99f
- Merge branch 'main' into develop-0.2.2 · 5474be52
  Azure authored Feb 25, 2025
  
  5474be52
- update FAQ · 021822dd
  Azure authored Feb 25, 2025
  
  021822dd
- Merge pull request #657 from kvcache-ai/feat-absorb-for-long-prefill · b443c7df
  Atream authored Feb 25, 2025
```
Feat absorb for long prefill
```
  b443c7df
- support absorb for prefill long context · f4c198bd
  Atream authored Feb 25, 2025
  
  f4c198bd
- Merge pull request #643 from Azure-Tang/support-fp8 · 050b745a
  Azure authored Feb 25, 2025
```
[feat] Support fp8 linear kernel;
```
  050b745a
- Update doc · 36fbeee3
  Azure authored Feb 25, 2025
  
  36fbeee3
24 Feb, 2025 9 commits
- update fp8 kernel tutorial · 4dc5518e
  Azure authored Feb 24, 2025
  
  4dc5518e
- Merge pull request #608 from makllama/fix_musa_ext · 7b2a6690
  Atream authored Feb 24, 2025
```
musa: support bf16
```
  7b2a6690
- Merge pull request #645 from makllama/torch2.2 · 6f9ea689
  Atream authored Feb 24, 2025
```
Ensure backward compatibility with PyTorch 2.2
```
  6f9ea689
- Ensure backward compatibility with Torch 2.2 · f88c05a6
  Xiaodong Ye authored Feb 24, 2025
```
Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>
```
  f88c05a6
- Merge remote-tracking branch 'upstream/develop-0.2.2' into support-fp8 · ca7366d2
  Azure authored Feb 24, 2025
  
  ca7366d2
- Add data loader to read special weights for fp8; Add special weight process script · 581a524f
  Azure authored Feb 24, 2025
  
  581a524f
- Merge branch 'main' into feat-absorb-for-long-prefill · e9b1216a
  Atream authored Feb 24, 2025
  
  e9b1216a
- Merge pull request #638 from kvcache-ai/feat-moonlight · 4b5991e7
  Atream authored Feb 24, 2025
```
fix KExpertsMarlin on GPU with out CUDA Graph
```
  4b5991e7
- fix KExpertsMarlin on GPU with out CUDA Graph · f3276950
  Atream authored Feb 24, 2025
  
  f3276950
23 Feb, 2025 8 commits
- Merge pull request #621 from kvcache-ai/feat-moonlight · eb039b72
  Atream authored Feb 23, 2025
```
support moonlight, use ktransformers/optimize/optimize_rules/Moonlight-16B-A3B.yaml
```
  eb039b72
- update yaml · f5f6c6b9
  Atream authored Feb 23, 2025
  
  f5f6c6b9
- support Moonlight · e8e02e5c
  Atream authored Feb 23, 2025
  
  e8e02e5c
- tmp · 95d937c5
  DDong Jianwei authored Feb 23, 2025
  
  95d937c5
- remove causal mask · 006e8c6a
  Atream authored Feb 23, 2025
  
  006e8c6a
- Merge pull request #612 from kvcache-ai/fix-bf16-load · cdb6f896
  Atream authored Feb 23, 2025
```
fix bf16 load, TODO: refactor cpu dequant
```
  cdb6f896
- fix bf16 load, TODO: refactor cpu dequant · 036ae25a
  Atream authored Feb 23, 2025
  
  036ae25a
- musa: support bf16 · 18b1d183
  Xiaodong Ye authored Feb 23, 2025
```
Signed-off-by: Xiaodong Ye <xiaodong.ye@mthreads.com>
```
  18b1d183