Text Retrieval
Transformers
Safetensors
bidirectional_pplx_qwen3
feature-extraction
sentence-embeddings
contextual-embeddings
custom_code
Instructions to use seslami-pplx/pplx-embed-context-v1.2-4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use seslami-pplx/pplx-embed-context-v1.2-4B with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("seslami-pplx/pplx-embed-context-v1.2-4B", trust_remote_code=True, dtype="auto") - Notebooks
- Google Colab
- Kaggle
Upload SLERP-merged checkpoint (alpha=0.5) from two adversarial-FT runs at step-1500
43188fa verified | { | |
| "architectures": [ | |
| "PPLXQwen3ContextualModel" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "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, | |
| "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 | |
| } | |