Scott Roy

Results 32 comments of Scott Roy

Build PERF for HQQ vs. Affine

Thanks! The unit test looks good to me. I'll let @jainapurva review the benchmark script, as I'm less familiar with that part of the codebase and if there are coding...

The `embedding_4bit` implementation has a strong assumption about weight_quant_min and weight_quant_max

I think @manuelcandales added the 4-bit embedding op, but I agree we should clean them up. Let me create a EE/BE task for this.