UPSTREAM PR #21985: Feat/modelscope support #1354
Open
LOCI Review / Performance Per Binary #1354
succeeded
Apr 17, 2026
Performance varied across binaries, overall acceptable
1 binary improved · 9 binaries unchanged · 5 binaries stable ~ within threshold · 0 binaries degraded ~ beyond threshold
| Binary | Δ % Response | Δ % Throughput | Performance (based on response time) |
|---|---|---|---|
build.bin.libggml-base.so |
0 | 0 | unchanged |
build.bin.libggml-cpu.so |
0.39 | 0 | stable |
build.bin.libggml.so |
0 | 0.02 | unchanged |
build.bin.libllama.so |
0.07 | -0.01 | stable |
build.bin.libmtmd.so |
0 | 0 | unchanged |
build.bin.llama-bench |
3.15 | 0.54 | stable |
build.bin.llama-cvector-generator |
0.43 | 0.2 | stable |
build.bin.llama-gemma3-cli |
0 | 0.38 | unchanged |
build.bin.llama-gguf-split |
0 | 0.04 | unchanged |
build.bin.llama-llava-cli |
0 | 0.38 | unchanged |
build.bin.llama-minicpmv-cli |
0 | 0.38 | unchanged |
build.bin.llama-quantize |
-0.06 | -0.13 | improved |
build.bin.llama-qwen2vl-cli |
0 | 0.38 | unchanged |
build.bin.llama-tokenize |
0 | -0.01 | unchanged |
build.bin.llama-tts |
0.07 | 0.27 | stable |
Performance threshold: 30%
Default configuration used.
Note: Performance status is evaluated only from Δ% Response. Throughput is displayed for reference.
Explore the complete analysis inside the Loci Inspector.
Open the Pull Request linked to this check-run.
Loading