devflow / analysis_outputs /T8 /task1_kv_cache.txt
bhsinghgrid's picture
Update app/inference + ablation task outputs
27f26fd verified
TASK 1 — KV CACHE BENCHMARK
========================================
has_generate_cached=True
memory_profile=N/A (CPU/MPS)
src_len standard(s) cached(s) speedup encoder%
16 0.449 0.313 1.43x 36.2%
32 0.391 0.290 1.35x 34.9%
64 0.641 0.465 1.38x 35.6%
Saved graphs:
- task1_time_comparison.png
- task1_speedup.png
- task1_encoder_cost.png