Commit History
-
Fix word_embedding_dimension in 1_Pooling/config.json
Verified signature1_Pooling/config.json declares word_embedding_dimension: 1024, but this model's actual hidden_size (per config.json) is 640. This mismatch causes sentence-transformers' MSELoss (which sizes its projection layer via model.get_embedding_dimension(), trusting this file) to build a Linear(1024, ...) layer instead of Linear(640, ...), raising a shape mismatch at runtime when real 640-dim embeddings are passed in. It also affects any other pooling-dimension-derived Matryoshka dimension defaults. This PR corrects word_embedding_dimension to 640 to match config.json's hidden_size; all other pooling settings are unchanged.4002318
-
Update README.md
Verified signature6022959
-
Truncate layer_types to match num_hidden_layers (#1)
- Truncate layer_types to match num_hidden_layers (a6f9c87cd90598b9580dcc6bb1b582e8fb6984d1) Co-authored-by: Alexander Miller <alex-miller-gov@users.noreply.huggingface.co>4013fb0
-
Update README.md
Verified signatureebbe08e
-
Update README.md
Verified signature7684401
-
Update README.md
Verified signatureb63009a
-
Update README.md
Verified signature9eb3500 -
Upload folder using huggingface_hub
Verified signature4d145cb -
initial commit
Verified signature86a59a1