Open Peterande opened 1 week ago
作者您好,在RAR sample的过程中,我发现将sampled = torch.multinomial(probs, num_samples=1)改成torch.argmax(probs, dim=-1),生成的结果完全崩坏、没有意义。这是什么原因导致的呢?ARLoss中算token是否正确不就是用的torch.argmax(shift_logits, dim=1)吗?
sampled = torch.multinomial(probs, num_samples=1)
torch.argmax(probs, dim=-1)
torch.argmax(shift_logits, dim=1)
Hi,
Thanks for sharing the findings. Unfortunately I do not have experience on using argmax (greedy search) for image generation, so I may not be able to share any insights on this
作者您好,在RAR sample的过程中,我发现将
sampled = torch.multinomial(probs, num_samples=1)
改成torch.argmax(probs, dim=-1)
,生成的结果完全崩坏、没有意义。这是什么原因导致的呢?ARLoss中算token是否正确不就是用的torch.argmax(shift_logits, dim=1)
吗?