Update README.md

README.md

@@ -8,7 +8,7 @@ pipeline_tag: text-ranking
 
 ## Highlights
 
-Our reranker is on the Pareto frontier across 5 key areas:
+Our reranker is on the cost/performance Pareto frontier across 5 key areas:
 - Instruction following (including capability to rank more recent information higher)
 - Question answering
 - Multilinguality
@@ -26,14 +26,14 @@ For more details on these and other benchmarks, please refer to our [blogpost](h
 - Model Type: Text Reranking
 - Supported Languages: 100+
 - Number of Paramaters: 1B
-- Context Length: base model supports 32K but we tested up to 8K
+- Context Length: up to 32K
 - Blogpost: https://contextual.ai/blog/rerank-v2
 
 ## Quickstart
 
 ### vLLM usage
 
-Requires vllm>=0.10.0 for FP4 or vllm>=0.8.5 for FP16.
+Requires vllm==0.10.0 for NVFP4 or vllm>=0.8.5 for BF16.
 
 ```python
 import os
@@ -102,7 +102,7 @@ def infer_w_vllm(model_path: str, query: str, instruction: str, documents: list[
 
 ### Transformers Usage
 
-Requires transformers>=4.51.0 for FP16. Not supported for FP4.
+Requires transformers>=4.51.0 for BF16. Not supported for NVFP4.
 
 ```python
 import torch
@@ -174,7 +174,7 @@ If you use this model, please cite:
 
 ## License
 
-Creative Commons Attribution Non Commercial Share Alike 4.0
+Creative Commons Attribution Non Commercial Share Alike 4.0 (cc-by-nc-sa-4.0)
 
 ## Contact