MiniMax-M2.5.w4a16 / recipe.yaml
ChibuUkachi's picture
Upload folder using huggingface_hub
83e3699 verified
default_stage:
default_modifiers:
GPTQModifier:
targets: [Linear]
ignore: ['re:.*lm_head', 're:visual.*', 're:model.visual.*', 're:.*mlp.gate$', 're:.*embed_tokens$',
're:.*shared_expert_gate$', 're:.*linear_attn.*']
scheme: W4A16
weight_observer: mse
bypass_divisibility_checks: false
block_size: 128
dampening_frac: 0.01
actorder: static
offload_hessians: false