Xiang An

The largest public face training set.

1

Hello, a new distributed training algorithm for face recognition and the largest public face training set glint360k release is here https://github.com/deepinsight/insightface/tree/master/recognition/partial_fc

如题

MegatronLM 微信讨论组

1

老哥，方便交流可以进这个微信讨论组交流吗？ 

Fix the bug in multi-machine training.

2

When using multiple machines, the rank exceeds 8, which is more than the maximum number of GPUs on the current node. The correct approach should not require passing in the...

A New Powerful Visual Tower for LLaVA!

3

We adopted the official [LLaVA-NeXT](https://github.com/LLaVA-VL/LLaVA-NeXT) and the official training dataset [LLaVA-NeXT-Data](https://huggingface.co/datasets/lmms-lab/LLaVA-NeXT-Data) for evaluating the foundational visual models. The language model is Qwen2.5-7B. | Vision Tower | RoPE2D | ChartQA |...