Feature Extraction
MLX
Safetensors
multilingual
embedding_gemma2
mlx-vlm
embedding
sentence-similarity
multimodal
image-feature-extraction
audio-feature-extraction
video-feature-extraction
8-bit precision
Instructions to use mlx-community/embeddinggemma-2-mxfp8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use mlx-community/embeddinggemma-2-mxfp8 with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] hf download mlx-community/embeddinggemma-2-mxfp8 --local-dir embeddinggemma-2-mxfp8
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Download validation.json from mlx-community/embeddinggemma-2-mxfp8: direct link, hf CLI and curl.
- Browser
- Download file 2.43 kB
-
https://huggingface.co/mlx-community/embeddinggemma-2-mxfp8/resolve/main/validation.json
- Command line
-
hf download hf://mlx-community/embeddinggemma-2-mxfp8/validation.json
-
curl -L -o validation.json https://huggingface.co/mlx-community/embeddinggemma-2-mxfp8/resolve/main/validation.json
2.43 kB
| { | |
| "source_model": "google/embeddinggemma-2", | |
| "source_revision": "914f7f89142e33e77833254d9c9b90c3cef7303b", | |
| "mlx_vlm_revision": "3d87e88402f307efbf68e568971aa887ee7d9ed0", | |
| "variant": "mxfp8", | |
| "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": 1596, | |
| "tensor_dtypes": [ | |
| "BF16", | |
| "U32", | |
| "U8" | |
| ], | |
| "weight_bytes": 1225720015, | |
| "cases": { | |
| "audio": { | |
| "shape": [ | |
| 1, | |
| 768 | |
| ], | |
| "finite": true, | |
| "minimum_cosine_vs_torch_fp32": 0.9983353865128464, | |
| "maximum_absolute_error_vs_torch_fp32": 0.006557946804956866, | |
| "minimum_cosine_vs_source_bf16": 0.9981795014663992, | |
| "minimum_truncated_cosine_vs_torch_fp32": 0.9984341378931347 | |
| }, | |
| "image": { | |
| "shape": [ | |
| 1, | |
| 768 | |
| ], | |
| "finite": true, | |
| "minimum_cosine_vs_torch_fp32": 0.9994873791960924, | |
| "maximum_absolute_error_vs_torch_fp32": 0.003609755933630241, | |
| "minimum_cosine_vs_source_bf16": 0.9995206654001738, | |
| "minimum_truncated_cosine_vs_torch_fp32": 0.9995135458227378 | |
| }, | |
| "text": { | |
| "shape": [ | |
| 6, | |
| 768 | |
| ], | |
| "finite": true, | |
| "minimum_cosine_vs_torch_fp32": 0.9983049609155468, | |
| "maximum_absolute_error_vs_torch_fp32": 0.006092039627625164, | |
| "minimum_cosine_vs_source_bf16": 0.9982079929326719, | |
| "minimum_truncated_cosine_vs_torch_fp32": 0.998331601337103, | |
| "retrieval_scores_venus_mars": [ | |
| 0.6707783937454224, | |
| 0.8356978297233582 | |
| ] | |
| }, | |
| "text_image": { | |
| "shape": [ | |
| 1, | |
| 768 | |
| ], | |
| "finite": true, | |
| "minimum_cosine_vs_torch_fp32": 0.9992635574576849, | |
| "maximum_absolute_error_vs_torch_fp32": 0.0048226723343564, | |
| "minimum_cosine_vs_source_bf16": 0.9992596731314828, | |
| "minimum_truncated_cosine_vs_torch_fp32": 0.9993020285279928 | |
| }, | |
| "video": { | |
| "shape": [ | |
| 1, | |
| 768 | |
| ], | |
| "finite": true, | |
| "minimum_cosine_vs_torch_fp32": 0.9992312775532464, | |
| "maximum_absolute_error_vs_torch_fp32": 0.007177200518493687, | |
| "minimum_cosine_vs_source_bf16": 0.9992132664387819, | |
| "minimum_truncated_cosine_vs_torch_fp32": 0.9992793313591022 | |
| } | |
| } | |
| } | |