Downloads · 30 days
3
12% of all-time downloads
embeddinggemma-300m-distilled-depth20pct-15-768dim-best
Pieces/embeddinggemma-300m-distilled-depth20pct-15-768dim-best
embeddinggemma-300m-distilled-depth20pct-15-768dim-best is a machine learning model from Pieces. Use it for the machine learning task on the model card, and read the license before you ship it in a product.
This is a distilled/compressed version of google/embeddinggemma-300m.
All-time downloads
25
Public
Parameters
47.2M
246 MB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors208 MB · 84%
At a glance
- Model type
- gemma3_text
- Access
- Public
- Created
- Dec 22, 2025
- Updated
- Dec 22, 2025
- SHA
- de402da2
Model card
From the Hugging Face model README
Distilled Backbone: embeddinggemma-300m-distilled-768dim
This is a distilled/compressed version of google/embeddinggemma-300m.
Compression Details
- Base model: google/embeddinggemma-300m
- Width reduction factor: None
- Target hidden size: None
- Final embedding dimension: 768
- Had projection layer: False
- Projection incorporated: False
Usage
To use this as a pretrained backbone in validation:
python -m playground.validate_from_checkpoint \
--pretrained_backbone_model "Pieces/embeddinggemma-300m-distilled-depth20pct-15-768dim-best" \
--backbone_embedding_dim 768 \
--backbone_pooling_mode "mean" \
--num_samples 100 \
--dataset_split "val" \
--num_general_tags_high 100 \
--num_general_tags_low 100
Important: The path must point to the sentence_transformer subdirectory (not the parent directory).
The exported model has been configured to output embeddings of dimension 768 directly, incorporating any projection layers from the distillation process.
- Type
- gemma3_text
- Created
- Dec 22, 2025
- Updated
- Dec 22, 2025