Hugging Face's logo

nm-testing
/
Qwen2-VL-2B-Instruct-Sparse-0.6

Safetensors
qwen2_vl
compressed-tensors
Model card Files
xet
Community
Qwen2-VL-2B-Instruct-Sparse-0.6
File size: 166 Bytes 
9ba370a
 
 
 
 
 
 
1
2
3
4
5
6
7
 
DEFAULT_stage:
  DEFAULT_modifiers:
    SparseGPTModifier:
      sparsity: 0.6
      sequential_targets: [Qwen2VLDecoderLayer]
      ignore: [lm_head, 're:visual.*']