Welcome to My Tech Blog

Optimizing VLM Training on One GPU: A Five-Layer Recipe How I got SiQ-VL from 14,713 to 100,923 real tokens per second on a single Blackwell GPU, and the four places that surprised me along the way. TL;DR I trained a small vision-language model (SigLIP-2 vision tower + Qwen2.5 LLM, projector-aligned) on one NVIDIA RTX PRO 6000 Blackwell, ran a 48-configuration sweep across two model sizes and both training stages, and ended up with a recipe that compounds five optimization layers. ...

Why Variable Sequence Length Breaks DDP Throughput How to reproduce, measure, and fix token skew in transformer training with length bucketing and token-budget batching. TL;DR In transformer training, DDP can look balanced by sample count while being badly imbalanced by actual work. I built a small one-machine lab that uses a tiny transformer-like model with variable sequence lengths and four distributed ranks. The headline result was simple: uniform 128-token batches: 250,959 tokens/s variable lengths with fixed sample count: 122,006 tokens/s variable lengths with length bucketing: 208,668 tokens/s variable lengths with token-budget batching: 193,289 tokens/s The bad case was not a kernel problem. It was a batching problem: ...

Learning PyTorch DDP Performance Tuning on a One-GPU Machine How to build real intuition for DistributedDataParallel scaling, stragglers, communication, and synchronization even when you only have one GPU. TL;DR Most DDP performance problems are easier to understand than they first look. In this post I built a small single-machine lab that uses CPU gloo processes to reproduce the part of DDP reasoning that matters most: the slowest rank often sets the pace small per-rank work hurts scaling communication can dominate step time rank-0-only host work becomes everyone’s problem once you synchronize The important numbers from the lab were: ...

Profiling a PyTorch Training Job End to End How to decide whether a training job is blocked on data, PyTorch overhead, or a hot CUDA kernel, using torch.profiler, Nsight Systems, and Nsight Compute in the right order. TL;DR When a PyTorch training job feels slow, the most expensive mistake is starting at the wrong layer. In this case study, I built a small synthetic training lab and used it to force three common bottlenecks: ...

Most “Scalable” Multimodal Pipelines Don’t Survive Foundation-Model Scale A lot of multimodal pipelines claim to scale. In practice, they often depend on at least one of the following: global shuffles (groupBy/join/repartition), materializing massive intermediate datasets, centralized coordination that becomes a bottleneck, or brittle recovery logic (rerun-the-world on failure). That works for demos. It breaks at foundation-model scale. This series is about a different design point: A streaming-first multimodal pipeline that scales linearly with data and hardware — with no global shuffle, and resumable at partition granularity. ...