Text, images, speech and environmental sounds in one embedding space. Vela Omni Mini supports multimodal search, routing and clustering with normalized vectors that can be compared directly. Scores are 0–100; higher is better. The comparison uses the same held-out evaluation examples, complete retrieval pools and 128-token text cap for both models. Bold marks improvement over the original large model. These known test pools are reused across releases. The common protocol uses labeled TRAIN prototypes for text classification and all matching positives for retrieval; it is separate from official MTEB classification. All 14 metrics, Macro-F1, exact counts and uncertainty. The primary metric…
Open weights
apache-2.0
1.4B parameters
pytorch
This is the model card of a transformers model that has been pushed on the Hub. Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. Use the code below to get started with the model. Carbon emissions can be estimated using the Machine Learning Impact calculator presented in Lacoste et al. (2019).
Open weights
1.9B parameters
transformers
Dense retriever for patent prior-art search (abstract → abstract), fine-tuned from on patent-family pairs (same invention filed in different jurisdictions / languages). Trained for Korean ↔ English cross-lingual and monolingual patent retrieval. Do not use the base model's asymmetric convention (query: / passage: ). It raises no error but silently degrades retrieval quality, because training was symmetric. corpus (other family members masked; one relevant document per query). crosssame: translation pairs (KR↔EN), crosssimilar: cross-lingual same-family non-translation pairs, monosimilar: monolingual same-family pairs. Scores on the 50k-document subset; full-corpus (209,671 documents)…
Open weights
other
1.1B parameters
262,144 tokens
sentence-transformers
Model Type: Feature backbone; Params: 1136M (giant); Image size: 224 x 224 x 3; Patch size: 14 x 14 x 3; Repository: github.com:marrlab/DinoBloom; Original Weights: Zenodo.
Open weights
apache-2.0
1.1B parameters
timm
Native MLX conversion of ewin-reg/WeMM-Embedding-2B-Quantized — the hybrid FP8-protected attention + group-16 INT4 checkpoint — that preserves its mixed-precision allocation instead of flattening everything to one precision. The source checkpoint is curvature-guided: it deliberately keeps the precision-sensitive path at ~8-bit (FP8 E4M3) and compresses only the quantization-robust masses to INT4. Uniform 4-bit quantization destroys that — the source README itself documents softmax attention collapse (DKL >= 7.50) when Q/K go to 4 bits. This build mirrors the Two deviations from the source, both improvements or forced by MLX: - 8-bit group-64 is finer than the source's per-tensor-scalar FP8…
Open weights
other
2.2B parameters
262,144 tokens
mlx
EmbeddingGemma 2 is an open multimodal embedding model built by Google DeepMind which maps text (incl. code), images, video, and audio inputs—and combinations thereof—into a single, unified 768-dimensional vector space. The model has 740M total parameters, combining a 270M parameter text model with modular vision (170M) and audio (300M) encoders. Designed to run on consumer hardware such as mobile devices and laptops, EmbeddingGemma 2 delivers low-latency semantic representations for on-device applications, like search, retrieval-augmented generation (RAG), classification, and clustering. EmbeddingGemma 2 builds upon the architectural and capability advancements of Gemma 4, offering several…
Open weights
apache-2.0
744M parameters
262,144 tokens
transformers
