Devesh Sarda

The issue is because of the following line of code https://github.com/sthalles/SimCLR/blob/1848fc934ad844ae630e6c452300433fe99acfd9/simclr.py#L28 which assumes that the number of features you have is a multiple of the batch size but this is...

I ran into the same issue on a private dataset - Fixed the issue by upgrading to `duckdb-1.1.0`

I wrote a simple script to do this: ``` import os import shutil # Prefix to only include certain files FILE_NAME_PREFIX = "wiki_" # The path to the directory containing...

I wasn't aware this was possible. Do you mind sharing an example before/after json?

Oh, I see - I thought they were input files specifying certain benchmark configurations. I will run this for before and after the PR and share the results here.

I have also implemented `contains_block_parallel` similar to `contains_warp_parallel`. I ran the benchmark on the previous version and current version and have attached the associated json files: - Before (Commit `3e658cf53c07340e885ee49afbdb6daa6bc147ba`):...

Hey, here are some initial numbers: ``` API | row_width | num_rows | hit_rate | Ref Time | Cmp Time | Diff | %Diff find | 1024 | 50000 |...

@davidwendt I am happy to explore in adding in the KMP/chunking but it is still surprising that even for strings of length 32k, the block based method is not showing...

@GregoryKimball Thanks for that suggestion. I will look into how the workload is generated to adjust this. We are currently in a midterm week here in Wisconsin so I might...