Update README.md

README.md

@@ -24,10 +24,10 @@ It is _**the largest open-source vision/vision-language foundation model (14B)**
 ## Model Details
 - **Model Type:** vision foundation model, feature backbone
 - **Model Stats:**
-  - Params (M): 5903
+  - Params (M): 5540 (the last 3 blocks are discarded)
   - Image size: 448 x 448
 - **Pretrain Dataset:** LAION-en, LAION-COCO, COYO, CC12M, CC3M, SBU, Wukong, LAION-multi
-
+- **Note!!:** InternViT-6B originally had 48 blocks, and we found that using the output after the fourth-to-last block worked best for VLLM. For ease of use and to save GPU memory, we simply discarded the last 3 blocks. Now the model has only 45 blocks and the number of parameters has been reduced from 5.9B to 5.5B. Please set `mm_vision_select_layer=-1` when using this model.
 ## Model Usage (Image Embeddings)
 
 ```python