Feature Extraction
MLX
Safetensors
multilingual
embedding_gemma2
mlx-vlm
embedding
sentence-similarity
multimodal
image-feature-extraction
audio-feature-extraction
video-feature-extraction
4-bit precision
Instructions to use mlx-community/embeddinggemma-2-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use mlx-community/embeddinggemma-2-4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] hf download mlx-community/embeddinggemma-2-4bit --local-dir embeddinggemma-2-4bit
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Download validation.json from mlx-community/embeddinggemma-2-4bit: direct link, hf CLI and curl.
- Browser
- Download file 2.41 kB
-
https://huggingface.co/mlx-community/embeddinggemma-2-4bit/resolve/main/validation.json
- Command line
-
hf download hf://mlx-community/embeddinggemma-2-4bit/validation.json
-
curl -L -o validation.json https://huggingface.co/mlx-community/embeddinggemma-2-4bit/resolve/main/validation.json
2.41 kB
| { | |
| "source_model": "google/embeddinggemma-2", | |
| "source_revision": "914f7f89142e33e77833254d9c9b90c3cef7303b", | |
| "mlx_vlm_revision": "3d87e88402f307efbf68e568971aa887ee7d9ed0", | |
| "variant": "4bit", | |
| "hardware": "Apple M3 Ultra", | |
| "mlx_version": "0.32.3", | |
| "transformers_version": "5.18.0.dev0", | |
| "validation_scope": "Six multilingual text inputs and synthetic image, audio, two-frame video, and text+image; numerical smoke checks, not a retrieval benchmark.", | |
| "tensor_count": 1816, | |
| "tensor_dtypes": [ | |
| "BF16", | |
| "U32" | |
| ], | |
| "weight_bytes": 1098381958, | |
| "cases": { | |
| "audio": { | |
| "shape": [ | |
| 1, | |
| 768 | |
| ], | |
| "finite": true, | |
| "minimum_cosine_vs_torch_fp32": 0.9744351326382886, | |
| "maximum_absolute_error_vs_torch_fp32": 0.03714944629709238, | |
| "minimum_cosine_vs_source_bf16": 0.9745487588365271, | |
| "minimum_truncated_cosine_vs_torch_fp32": 0.9755004802420051 | |
| }, | |
| "image": { | |
| "shape": [ | |
| 1, | |
| 768 | |
| ], | |
| "finite": true, | |
| "minimum_cosine_vs_torch_fp32": 0.9903859935386443, | |
| "maximum_absolute_error_vs_torch_fp32": 0.016958792220801044, | |
| "minimum_cosine_vs_source_bf16": 0.9904275597536485, | |
| "minimum_truncated_cosine_vs_torch_fp32": 0.9908756584718292 | |
| }, | |
| "text": { | |
| "shape": [ | |
| 6, | |
| 768 | |
| ], | |
| "finite": true, | |
| "minimum_cosine_vs_torch_fp32": 0.9808765992496479, | |
| "maximum_absolute_error_vs_torch_fp32": 0.022356300154592085, | |
| "minimum_cosine_vs_source_bf16": 0.9812440662582766, | |
| "minimum_truncated_cosine_vs_torch_fp32": 0.9822185364113407, | |
| "retrieval_scores_venus_mars": [ | |
| 0.675256609916687, | |
| 0.8269824981689453 | |
| ] | |
| }, | |
| "text_image": { | |
| "shape": [ | |
| 1, | |
| 768 | |
| ], | |
| "finite": true, | |
| "minimum_cosine_vs_torch_fp32": 0.9864358647643854, | |
| "maximum_absolute_error_vs_torch_fp32": 0.019337336771483207, | |
| "minimum_cosine_vs_source_bf16": 0.9864374751342606, | |
| "minimum_truncated_cosine_vs_torch_fp32": 0.986872206893313 | |
| }, | |
| "video": { | |
| "shape": [ | |
| 1, | |
| 768 | |
| ], | |
| "finite": true, | |
| "minimum_cosine_vs_torch_fp32": 0.986145886202646, | |
| "maximum_absolute_error_vs_torch_fp32": 0.021159542036788805, | |
| "minimum_cosine_vs_source_bf16": 0.9861580254276772, | |
| "minimum_truncated_cosine_vs_torch_fp32": 0.9863870540059549 | |
| } | |
| } | |
| } | |