Upload folder using huggingface_hub

README.md

@@ -1,10 +1,10 @@
 ---
 library_name: transformers
 tags:
-- biology
-- protein-structure
-- esmfold2
-- multimodal-protein-model
+  - biology
+  - protein-structure
+  - esmfold2
+  - multimodal-protein-model
 ---
 
 # FastPLMs ESMFold2
@@ -20,12 +20,12 @@ from transformers import AutoModel
 model = AutoModel.from_pretrained(
     "Synthyra/ESMFold2-Fast",
     trust_remote_code=True,
-    dtype=torch.bfloat16,
-    device_map="cuda",
-).eval()
+    dtype=torch.float32,
+).eval().cuda()
 ```
 
 Use `Synthyra/ESMFold2` for the full model and `Synthyra/ESMFold2-Fast` for the faster release variant.
+The folding trunk runs in fp32; the 6B ESMC backbone is loaded in bf16 by default via `esmc_precision="bf16"`.
 
 ## Fold One Protein