Public/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-09-20 01:31:31 +02:00
Code Issues Packages Projects Releases Wiki Activity

cuda : enable i16 and i32 for DUP (#28897)

Browse Source 
* cuda : enable i16 and i32 for DUP

* docs : update ops table for DUP on CUDA
This commit is contained in:
Aman Karki
2026-09-15 09:12:21 +05:30
committed by GitHub
parent 69eb250670
commit 4c9233c034
4 changed files with 14 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/ops.md
+1 -1
View File
@@ -44,7 +44,7 @@ Legend:
| DSV4_HC_COMB | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ✅ | ❌ | ❌ | ❌ | ❌ |
| DSV4_HC_POST | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ✅ | ❌ | ❌ | ❌ | ❌ |
| DSV4_HC_PRE | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ❌ | ✅ | ❌ | ❌ | ❌ | ❌ |
| DUP | ❌ | ✅ | ✅ | 🟡 | ❌ | ❌ | 🟡 | 🟡 | ✅ | ✅ | ❌ | ❌ | ❌ |
| DUP | ❌ | ✅ | ✅ | ✅ | ❌ | ❌ | 🟡 | 🟡 | ✅ | ✅ | ❌ | ❌ | ❌ |
| ELU | ❌ | ✅ | ✅ | 🟡 | 🟡 | ❌ | ✅ | ❌ | ✅ | ✅ | ✅ | ❌ | ❌ |
| EXP | ❌ | ✅ | ✅ | 🟡 | 🟡 | 🟡 | ✅ | ❌ | ✅ | ✅ | ✅ | ❌ | ❌ |
| EXPM1 | ❌ | ❌ | ✅ | 🟡 | 🟡 | ❌ | ✅ | ❌ | ✅ | ✅ | ✅ | ❌ | ❌ |
docs/ops/CUDA.csv
+4 -4
View File
@@ -5000,14 +5000,14 @@
"CUDA0","REPEAT_BACK","type=f32,ne=[8,6,4,2],nr=[1,1,1,2],v=1","support","1","yes","CUDA"
"CUDA0","DUP","type=f32,ne=[10,10,20,1]","support","1","yes","CUDA"
"CUDA0","DUP","type=f16,ne=[10,10,20,1]","support","1","yes","CUDA"
"CUDA0","DUP","type=i32,ne=[10,10,20,1]","support","0","no","CUDA"
"CUDA0","DUP","type=i16,ne=[10,10,20,1]","support","0","no","CUDA"
"CUDA0","DUP","type=i32,ne=[10,10,20,1]","support","1","yes","CUDA"
"CUDA0","DUP","type=i16,ne=[10,10,20,1]","support","1","yes","CUDA"
"CUDA0","DUP","type=f32,ne=[10,10,5,1],permute=[0,2,1,3]","support","1","yes","CUDA"
"CUDA0","DUP","type=f16,ne=[10,10,5,1],permute=[0,2,1,3]","support","1","yes","CUDA"
"CUDA0","DUP","type=f32,ne=[10,10,5,1],permute=[1,0,2,3]","support","1","yes","CUDA"
"CUDA0","DUP","type=f16,ne=[10,10,5,1],permute=[1,0,2,3]","support","1","yes","CUDA"
"CUDA0","DUP","type=i16,ne=[10,8,3,1],permute=[0,2,1,3]","support","0","no","CUDA"
"CUDA0","DUP","type=i16,ne=[10,8,3,1],permute=[1,2,0,3]","support","0","no","CUDA"
"CUDA0","DUP","type=i16,ne=[10,8,3,1],permute=[0,2,1,3]","support","1","yes","CUDA"
"CUDA0","DUP","type=i16,ne=[10,8,3,1],permute=[1,2,0,3]","support","1","yes","CUDA"
"CUDA0","SET","type_src=f32,type_dst=f32,ne=[6,5,4,3],dim=1","support","1","yes","CUDA"
"CUDA0","SET","type_src=f32,type_dst=f32,ne=[6,5,4,3],dim=2","support","1","yes","CUDA"
"CUDA0","SET","type_src=f32,type_dst=f32,ne=[6,5,4,3],dim=3","support","1","yes","CUDA"
Can't render this file because it is too large.