Commits · fa7ca75dbade521b3fe05f42dac327c0ce504e52 · jerrrrry / infinilm

26 Nov, 2025 2 commits
- Merge pull request #84 from pengcheng888/issue/83 · fa7ca75d
  PanZezhong1725 authored Nov 26, 2025
```
issue/83 - 添加AutoLlama类，支持创建不同backend的模型
```
  fa7ca75d
- issue/83 - 添加AutoLlama类，支持创建不同backend的模型 · e4114c03
  pengcheng888 authored Nov 26, 2025
  
  e4114c03
22 Nov, 2025 7 commits
- Merge pull request #82 from InfiniTensor/issue/80 · 5d182420
  PanZezhong1725 authored Nov 22, 2025
```
修复attention prefill计时方式，重构目录
```
  5d182420
- issue/80 增加对musa同步支持 · 02676be8
  PanZezhong authored Nov 22, 2025
  
  02676be8
- issue/80 修复attention prefill计时方式，重构目录 · d7d0889d
  PanZezhong authored Nov 22, 2025
  
  d7d0889d
- Merge pull request #81 from pengcheng888/issue/80 · 469f2884
  pengcheng888 authored Nov 22, 2025
```
issue/80 模型文件夹名字改为为model_path,增加llama模型的moore, iluvatar平台的参数
```
  469f2884
- issue/80 模型文件夹名字改为为model_path,增加moore平台的参数 · faaa9ec3
  pengcheng888 authored Nov 22, 2025
  
  faaa9ec3
- Merge pull request #79 from pengcheng888/issue/78 · 7d222d83
  pengcheng888 authored Nov 22, 2025
```
issue/78 - 在test文件夹目录,添加qwen3的moe和attention的耗时测试脚本
```
  7d222d83
- issue/78 - 在test文件夹目录,添加qwen3的moe和attention的耗时测试脚本 · d80bd14a
  pengcheng888 authored Nov 22, 2025
  
  d80bd14a
21 Nov, 2025 2 commits
- Merge pull request #77 from pengcheng888/main-python · 9e727b14
  pengcheng888 authored Nov 21, 2025
```
issue/76 - 添加python的llama模型实现
```
  9e727b14
- 调整repeat_kv的代码 · c53104e2
  pengcheng888 authored Nov 21, 2025
  
  c53104e2
20 Nov, 2025 1 commit
- issue/76 - 添加python的llama模型实现 · 3ddffe8d
  pengcheng888 authored Nov 20, 2025
  
  3ddffe8d
06 Nov, 2025 1 commit
- issue/71: support vllm benchmark text completion in InfiniLM · 4fd9d490
  spike-zhu authored Nov 06, 2025
  
  4fd9d490
29 Oct, 2025 1 commit
- Merge pull request #65 from InfiniTensor/issue/64 · 1c710c19
  PanZezhong1725 authored Oct 29, 2025
```
issue/64 - jiuge.py verbose output
```
  1c710c19
28 Oct, 2025 1 commit
- issue/64 - jiuge.py verbose output · 7a412203
  wooway777 authored Oct 28, 2025
  
  7a412203
22 Oct, 2025 1 commit
- Issue/60 -main 修复输出token乱码并适配了qwen3模型 · 80909bee
  pengcheng888 authored Oct 22, 2025
  
  80909bee
10 Oct, 2025 2 commits
- 增加issue模板 · 753a4f60
  PanZezhong authored Oct 10, 2025
  
  753a4f60
- feat: add ceval test script · 515b9245
  PanZezhong authored Oct 10, 2025
  
  515b9245
29 Sep, 2025 2 commits
- Merge pull request #46 from gongchensu/feature/support_hygon · ddea3d19
  PanZezhong1725 authored Sep 29, 2025
```
Compatible with Hygon DCU.
```
  ddea3d19
- Compatible with Hygon DCU. · 2a51ff9e
  zhuyue authored Sep 12, 2025
  
  2a51ff9e
24 Sep, 2025 1 commit
- Merge pull request #55 from InfiniTensor/issue/53 · d39a5e4d
  PanZezhong1725 authored Sep 24, 2025
```
issue/53: update AWQ-dequantize op name to match infinicore op dequantizeAWQ
```
  d39a5e4d
23 Sep, 2025 1 commit
- fix: update AWQ-dequantize op name to match infinicore op dequantizeAWQ · 33f5256a
  zhushuang authored Sep 23, 2025
  
  33f5256a
19 Sep, 2025 1 commit
- Merge pull request #51 from InfiniTensor/issue/50 · 664589d0
  PanZezhong1725 authored Sep 19, 2025
```
issue50: update infiniopDequantize args to macth IniniCore-infiniopDequantize
```
  664589d0
18 Sep, 2025 2 commits
- fix: update InfiniLM-infiniopDequantize args to math IniniCore-infiniopDequantize · 73671c25
  zhushuang authored Sep 18, 2025
  
  73671c25
- issue/48: p800支持 · 6fa3a12a
  zhangyue authored Sep 18, 2025
  
  6fa3a12a
16 Sep, 2025 2 commits
- fix: use new rope (#47) · 1ab1e668
  PanZezhong1725 authored Sep 16, 2025
  
  1ab1e668
- fix: use new rope · 36890e40
  PanZezhong1725 authored Sep 16, 2025
  
  36890e40
08 Sep, 2025 1 commit
- rename register weight · a179dcc3
  PanZezhong1725 authored Sep 08, 2025
  
  a179dcc3
04 Sep, 2025 1 commit
- Merge pull request #41 from InfiniTensor/model_scripts_modualization · 3998658b
  thatPepe authored Sep 04, 2025
```
Model scripts modualization
```
  3998658b
03 Sep, 2025 3 commits
- feat: specify distribution policy when register · 8bd0f91c
  PanZezhong1725 authored Sep 03, 2025
  
  8bd0f91c
- fix jiuge_awq qk_buf · 11c6e423
  PanZezhong1725 authored Sep 03, 2025
  
  11c6e423
- issue/39 - Used macro on op cache (#40) · 5330d5fa
  PanZezhong1725 authored Sep 03, 2025
  
  5330d5fa
02 Sep, 2025 3 commits
- issue/39 - Used macro on op cache · c219718e
  wooway777 authored Sep 02, 2025
  
  c219718e
- [T2-3-1]blkmjsian · 22804eaa
  blkmjsian authored Sep 02, 2025
```
- deepseek
- jiuge 4B awq 
```
  22804eaa
- Merge pull request #38 from InfiniTensor/issue/37 · 5c6000ec
  PanZezhong1725 authored Sep 02, 2025
```
issue/37 - fixed an inappropriate qk buffer slicing
```
  5c6000ec
01 Sep, 2025 1 commit
- issue/37 - fixed an inappropriate qk buffer slicing · 61ae257c
  wooway777 authored Sep 01, 2025
  
  61ae257c
21 Aug, 2025 1 commit
- Merge pull request #30 from InfiniTensor/issue/29 · e753b372
  PanZezhong1725 authored Aug 21, 2025
```
issue/29 - fixing tensor length mismatches across requests
```
  e753b372
20 Aug, 2025 3 commits
- issue/29 - further fixed tensor dimensions on qk_gemm and qk_softmax · c3f8b79b
  wooway777 authored Aug 20, 2025
  
  c3f8b79b
- issue/29 - fixing tensor length mismatches across requests · 9038f7ab
  wooway777 authored Aug 20, 2025
  
  9038f7ab
- Merge pull request #28 from InfiniTensor/issue/27 · c41055b5
  PanZezhong1725 authored Aug 20, 2025
```
issue/27 - fixed slicing for req offset
```
  c41055b5