Llama-3.2-1B-Instruct-GPTQ-8bit-128g / quantize_config.json
adriabama06's picture
Upload 8 files
ba1c107 verified
raw
history blame
203 Bytes
{"damp_percent": 0.01, "group_size": 128, "desc_act": false, "bits": 8, "sym": false, "allow_mix_bits": false, "true_sequential": false, "static_groups": false, "version": "GPTQ", "quant_method": "gptq"}