Menu

1 / 8

Running Captum on Cloud TPUs with TorchTPU: a work log

DEV Community: pytorch·Joinal Ahmed·14 days ago
#PmHagnHJ
#devtopytorch#devto#dev#captum#steps
Reading 0:00
15s threshold

This is the log of a few days spent running Captum's attribution methods on Cloud TPUs through TorchTPU, in the order the work happened, with what I learned at each step. Code, scripts and raw findings are in github.com/joinalahmed/captum-on-torchtpu . Measurements and upstream status describe the work through 2026-11-01.

What I ended up with

Captum runs unmodified on TorchTPU for the small-model configurations tested: 27 deterministic methods agree with CPU to about 1e-7, and 11 stochastic methods pass execution checks. The tested LLM wrappers also run on Qwen3-0.6B. Broader coverage exposed missing MaxPool support for three DeepLift configurations; compiled LayerIG needed two patches.

Matmul precision matters for numerical agreement. With torch.set_float32_matmul_precision("highest") , small-model errors fell from about 1e-3 to 1e-7. On Qwen3-0.6B, the 5-token, 32-step test reached 8.5e-6; the 42-token, 64-step example was around 7e-5 to 9e-5.…

Continue reading — create a free account

Join HashtagPLUS to read full articles, follow hashtags, vote, and join the conversation.