Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Switch to differentiable head, larger input sequence (#489)
* Switch to differentiable head, larger input sequence * change calibration ds; shorten quantization code * generate notebook outputs * Updated plots * minor * set CPU freq. governor to 'performance'; vertical throughput arrow * bf16 and fp32 baselines * IPEX+TS+AMX Working * results with "powersave" setting * results with governor="perf" * remove ipex-fp32 results * clear fp32 output * restore `bge-small (transformers)` output * move `generate_random_sequences` out of `PerformanceBenchmark` * Fix typo at the end * Final touches to throughput plot --------- Co-authored-by: Tom Aarsen <[email protected]>
- Loading branch information