Feature Extraction
sentence-transformers
Safetensors
embedding_gemma2
sentence-similarity
embeddings
text-embeddings
multimodal
image-text-retrieval
audio-embeddings
semantic-search
retrieval
quantized
4-bit precision
int4
mixed-precision
low-memory
gemma
embeddinggemma
multimodal-embeddings
cross-modal-retrieval
clip-alternative
quantization
post-training-quantization
cpu
edge
8-bit precision
Commit History
2 days ago
-
Caption retrieval on all 5,000 images: loss 0.23%, upper bound 0.41%
Verified signatureb017f48 -
Rename, SEO tags, add inference optimization section
Verified signature6119f13 -
Add peak memory measured on the second release
Verified signature9e82fdf
3 days ago
-
second release: sensitive layers at int8 (486 MiB), lower rounding error and fewer top-1 changes
Verified signature2e41616 -
correct top-1 change rates to the measured packed-model values
Verified signatureb501c62 -
add measured GPU speedups: fp16 autocast and 140-token mode
Verified signature85d305e -
add second memory run and loader A/B
Verified signature42113ac -
rewrite model card: results first, limits, memory, breakdown
Verified signaturec709ba7 -
add measured process memory and speed
Verified signature4dd6b93 -
int4 g16 text, vision and audio package
Verified signature798b305 -
initial commit
Verified signatured3be51c