fengzi

icon indicating a website link https://github.com/fengzi258 icon indicating an email [email protected]

icon location 武汉 icon location A newbie

Results 1 comments of fengzi

When using Accelerate for data parallel inference, using different numbers of GPUs results in different results

> > It's probably because of #1308. So the fewshot samples used for a particular `doc_id` will vary depending on whether DP is used and the number of ranks. Best...