Skip to content

Instantly share code, notes, and snippets.

View maaquib's full-sized avatar

Aaqib maaquib

View GitHub Profile
@maaquib
maaquib / perf.txt
Created September 5, 2025 15:41
PrefilPerfComparision
Benchmark Configuration:
Batch sizes: [1]
Sequence lengths: [32, 64, 128, 256, 512, 1024, 1536, 2048, 4096, 8192, 16384]
Number of heads: [16, 32, 64, 128]
Head dimensions: [64, 128]
Causal: True
Data type: torch.bfloat16
Warmup iterations: 10
Benchmark iterations: 100
@maaquib
maaquib / benchmark_flash_comparison.py
Created September 5, 2025 15:40
benchmark_flash_comparison
#!/usr/bin/env python3
"""
Benchmark comparing FlashInfer vs Flash Attention on B200 GPU.
Tests various sequence lengths and head dimensions.
"""
import argparse
import sys
from collections.abc import Sequence
from dataclasses import dataclass
> uname -m && cat /etc/*release
x86_64
DISTRIB_ID=Ubuntu
DISTRIB_RELEASE=18.04
DISTRIB_CODENAME=bionic
DISTRIB_DESCRIPTION="Ubuntu 18.04.5 LTS"
NAME="Ubuntu"
VERSION="18.04.5 LTS (Bionic Beaver)"
ID=ubuntu
ID_LIKE=debian
> uname -m && cat /etc/*release
x86_64
DISTRIB_ID=Ubuntu
DISTRIB_RELEASE=18.04
DISTRIB_CODENAME=bionic
DISTRIB_DESCRIPTION="Ubuntu 18.04.5 LTS"
NAME="Ubuntu"
VERSION="18.04.5 LTS (Bionic Beaver)"
ID=ubuntu
ID_LIKE=debian
@maaquib
maaquib / index.html
Last active August 19, 2016 20:58
D3JS Bar Graph (Above and Below x-axis with animation)
<html><head>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
<script type="text/javascript" src="http://d3js.org/d3.v3.min.js"></script>
<style type="text/css">
.bar.positive {
fill: steelblue;
}
.bar.negative {
fill: brown;