transformers

5f0fd118 - Optimize Qwen2VL vision model by precomputing cos/sin embeds before ViT blocks (#35837)

Optimize Qwen2VL vision model by precomputing cos/sin embeds before ViT blocks (#35837) * Optimize Qwen2VL vision model by precomputing cos/sin embeds before ViT blocks * Make rotary_pos_emb optional & fix type * Adapt pre-computed cos/sin to Qwen2.5VL * More concise