Token Classification
Transformers
Safetensors
English
kompress_v2
text-compression
modernbert
lora
kompress
Instructions to use chopratejas/kompress-v2-base with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use chopratejas/kompress-v2-base with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("token-classification", model="chopratejas/kompress-v2-base")# Load model directly from transformers import HeadroomCompressorV2 model = HeadroomCompressorV2.from_pretrained("chopratejas/kompress-v2-base", dtype="auto") - Notebooks
- Google Colab
- Kaggle
| { | |
| "model_type": "kompress_v2", | |
| "base_model_name": "answerdotai/ModernBERT-base", | |
| "num_labels": 2, | |
| "use_lora": true, | |
| "lora_r": 16, | |
| "lora_alpha": 32, | |
| "lora_dropout": 0.05, | |
| "lora_target_modules": [ | |
| "Wqkv", | |
| "Wi", | |
| "Wo" | |
| ], | |
| "span_hidden": 256, | |
| "span_kernels": [ | |
| 5, | |
| 3 | |
| ], | |
| "span_loss_weight": 0.3, | |
| "head_dropout": 0.1, | |
| "max_length": 8192, | |
| "architectures": [ | |
| "HeadroomCompressorV2" | |
| ] | |
| } |