Upload README.md with huggingface_hub

README.md

@@ -12,6 +12,7 @@ tags:
 
 # MetaEncoder v1
 
+![](assets/benchmarks.png)
 
 ## Usage
 
@@ -21,7 +22,7 @@ the model in bf16.
 ```python
 from modeling_metaencoder import MetaEncoder
 
-model = MetaEncoder.from_pretrained("<path to this repo>")
+model = MetaEncoder.from_pretrained("facebook/meta-encoder-v1")
 ```
 
 `MetaEncoder` has three calls. Callers pass content and task descriptions; the wrapper writes every prompt itself, in
@@ -98,8 +99,7 @@ similarities). Questions are answered independently; for a question that only ap
 decide the first and ignore the second as needed.
 
 `decide_batch(states, questions, images=None, video=None)` answers the same questions for many states (one
-`images` / `video` entry per state). Option embeddings depend only on the question, so they are computed once and
-cached: a fixed question set costs one forward pass per state and question.
+`images` / `video` entry per state).
 
 Probabilities are `softmax(score / 0.02)`, the contrastive training temperature. To fit the temperature to your
 data, decide a labelled set and call `model.calibrate(answers, labels)` with the per-question answers