Feature Extraction
MLX
Safetensors
qwen3
embeddings
sentence-similarity
quantization
omlx
oq4e
4-bit precision
Instructions to use TiGa-RCE/Qwen3-Embedding-8B-MLX-oQ4e with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use TiGa-RCE/Qwen3-Embedding-8B-MLX-oQ4e with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir Qwen3-Embedding-8B-MLX-oQ4e TiGa-RCE/Qwen3-Embedding-8B-MLX-oQ4e
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
File size: 509 Bytes
9db6783 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 | {
"pair_count": 24,
"top1": 1.0,
"recall_at_5": 1.0,
"mrr": 1.0,
"mean_aligned_embedding_cosine_vs_bf16": 0.9876885414123535,
"minimum_aligned_embedding_cosine_vs_bf16": 0.9831511974334717,
"score_rmse_vs_bf16": 0.012937835417687893,
"queries_with_rank_change": 0,
"gate_passed": false,
"gate_criteria": {
"top1_delta_min": 0.0,
"recall_at_5_delta_min": 0.0,
"mrr_delta_min": -0.01,
"minimum_aligned_embedding_cosine_min": 0.99,
"queries_with_rank_change_max": 2
}
}
|