Pull Request Labeler · Workflow runs · shalinib-ibm/llama.cpp · GitHub

Actions

All workflows
Workflows
- Pull Request Labeler Pull Request Labeler
- Build Actions Cache Build Actions Cache
- Build on Linux using cross-compiler Build on Linux using cross-compiler
- Build on RISCV Linux Machine by Cloud-V Build on RISCV Linux Machine by Cloud-V
- Build relocatable cmake package Build relocatable cmake package
- Check Pre-Tokenizer Hashes Check Pre-Tokenizer Hashes
- CI CI
- CI (AMD) CI (AMD)
- Close inactive issues Close inactive issues
- Copilot Setup Steps Copilot Setup Steps
- EditorConfig Checker EditorConfig Checker
Management
- Caches

Pull Request Labeler

Actions

Loading...
Loading

30 workflow runs

30 workflow runs

Q4/Q8 Tiled Gemm Optimization. Pull Request Labeler #30: Pull request #22 synchronize by shalinib-ibm

12s

12s

Q4/Q8 Tiled Gemm Optimization. Pull Request Labeler #29: Pull request #22 opened by shalinib-ibm

21s

21s

Q4 Tiled Gemm Optimization Pull Request Labeler #28: Pull request #21 opened by shalinib-ibm

20s

20s

Prepack all k blocks of Matrices A and B Pull Request Labeler #27: Pull request #20 opened by shalinib-ibm

5m 59s

5m 59s

Pre Compute scale factor once and reuse inside kenrel. Pull Request Labeler #26: Pull request #19 opened by shalinib-ibm

11s

11s

Implement a 16x8 kernel for Q4 GEMM Pull Request Labeler #25: Pull request #18 opened by shalinib-ibm

12s

12s

Q4 tiled Gemm Implementation Pull Request Labeler #24: Pull request #17 opened by shalinib-ibm

11s

11s

Q4 gemm trail1 Pull Request Labeler #23: Pull request #16 synchronize by shalinib-ibm

16s

16s

Q4 gemm trail1 Pull Request Labeler #22: Pull request #16 opened by shalinib-ibm

11s

11s

PowerPC: Sgemm Optimization Pull Request Labeler #21: Pull request #15 synchronize by shalinib-ibm

11s

11s

PowerPC: Sgemm Optimization Pull Request Labeler #20: Pull request #15 synchronize by shalinib-ibm

16s

16s

PowerPC: Sgemm Optimization Pull Request Labeler #19: Pull request #15 synchronize by shalinib-ibm

13s

13s

PowerPC: Sgemm Optimization Pull Request Labeler #18: Pull request #15 synchronize by shalinib-ibm

22s

22s

PowerPC: Sgemm Optimization Pull Request Labeler #17: Pull request #15 synchronize by shalinib-ibm

16s

16s

PowerPC: Sgemm Optimization Pull Request Labeler #16: Pull request #15 opened by shalinib-ibm

12s

12s

POC:Avoid PackTranspose TinyBLAS_PPC in MMA kernel Pull Request Labeler #15: Pull request #14 opened by shalinib-ibm

13s

13s

PowerPC: Replace static_assert with runtime assertion Pull Request Labeler #14: Pull request #13 opened by shalinib-ibm

6m 11s

6m 11s

Exp: Perf Benefit with xvi8ger4pp signed version Pull Request Labeler #13: Pull request #12 opened by shalinib-ibm

12s

12s

Move packing from inside MMA kernels Pull Request Labeler #12: Pull request #11 opened by shalinib-ibm

13s

13s

Refactor: Move matrix packing outside GEMM kernels Pull Request Labeler #11: Pull request #10 opened by shalinib-ibm

11s

11s

Avoid PackTranspose calls Fp32 tiny blas kernels Pull Request Labeler #10: Pull request #9 opened by shalinib-ibm

14s

14s

Optimize ggml_vec_dot_q8_q8 for ppc Pull Request Labeler #9: Pull request #8 opened by shalinib-ibm

12s

12s

Manual unrolling of loop for vec_dot_q8_q8 Pull Request Labeler #8: Pull request #7 opened by shalinib-ibm

14m 4s

14m 4s

ggml_vec_dot_q8_q8 mma optimize Pull Request Labeler #7: Pull request #6 opened by shalinib-ibm

14s

14s

Bug fix Pull Request Labeler #6: Pull request #5 opened by shalinib-ibm

19s

19s