Feature Extraction
MLX
Safetensors
bidirectional_pplx_qwen3
apple-silicon
sentence-similarity
contextual-embeddings
perplexity
qwen3
custom_code
Instructions to use agentmish/pplx-embed-context-v1-4b-mlx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use agentmish/pplx-embed-context-v1-4b-mlx with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir pplx-embed-context-v1-4b-mlx agentmish/pplx-embed-context-v1-4b-mlx
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
| { | |
| "architectures": [ | |
| "PPLXQwen3Model" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "attn_implementation": "sdpa", | |
| "auto_map": { | |
| "AutoConfig": "configuration.PPLXQwen3Config", | |
| "AutoModel": "modeling.PPLXQwen3ContextualModel" | |
| }, | |
| "bos_token_id": 151643, | |
| "dtype": "float32", | |
| "eos_token_id": 151643, | |
| "head_dim": 128, | |
| "hidden_act": "silu", | |
| "hidden_size": 2560, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 9728, | |
| "layer_types": [ | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention", | |
| "full_attention" | |
| ], | |
| "max_position_embeddings": 32768, | |
| "max_window_layers": 36, | |
| "mlx_contextual_embedding": { | |
| "source_repo": "perplexity-ai/pplx-embed-context-v1-4b", | |
| "source_revision": "0cb9b89b219a9b8ac95aa31aa0b67f1d5801c500", | |
| "converter": "pplx-mlx-convert", | |
| "dtype": "bfloat16" | |
| }, | |
| "model_type": "bidirectional_pplx_qwen3", | |
| "num_attention_heads": 32, | |
| "num_hidden_layers": 36, | |
| "num_key_value_heads": 8, | |
| "rms_norm_eps": 1e-06, | |
| "rope_parameters": { | |
| "rope_theta": 1000000, | |
| "rope_type": "default" | |
| }, | |
| "rope_theta": 1000000, | |
| "sliding_window": null, | |
| "tie_word_embeddings": true, | |
| "transformers_version": "5.0.0.dev0", | |
| "use_bidirectional_attention": true, | |
| "use_cache": false, | |
| "use_sliding_window": false, | |
| "vocab_size": 151936 | |
| } |