quant_config_by_layer.json

{"model.layers.0.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.0.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.0.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.0.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.0.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.0.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.0.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.1.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.1.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.1.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.1.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.1.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.1.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.1.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.2.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.2.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.2.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.2.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.2.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.2.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.2.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.3.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.3.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.3.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.3.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.3.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.3.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.3.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.4.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.4.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.4.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.4.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.4.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.4.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.4.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.5.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.5.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.5.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.5.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.5.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.5.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.5.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.6.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.6.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.6.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.6.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.6.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.6.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.6.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.7.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.7.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.7.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.7.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.7.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.7.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.7.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.8.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.8.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.8.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.8.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.8.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.8.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.8.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.9.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.9.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.9.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.9.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.9.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.9.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.9.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.10.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.10.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.10.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.10.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.10.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.10.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.10.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.11.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.11.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.11.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.11.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.11.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.11.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.11.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.12.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.12.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.12.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.12.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.12.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.12.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.12.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.13.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.13.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.13.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.13.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.13.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.13.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.13.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.14.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.14.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.14.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.14.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.14.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.14.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.14.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "model.layers.15.self_attn.q_proj": {"wbits": 8, "groupsize": 128}, "model.layers.15.self_attn.k_proj": {"wbits": 8, "groupsize": 128}, "model.layers.15.self_attn.v_proj": {"wbits": 8, "groupsize": 128}, "model.layers.15.self_attn.o_proj": {"wbits": 8, "groupsize": 128}, "model.layers.15.mlp.gate_proj": {"wbits": 8, "groupsize": 128}, "model.layers.15.mlp.up_proj": {"wbits": 8, "groupsize": 128}, "model.layers.15.mlp.down_proj": {"wbits": 8, "groupsize": 128}, "quant_method": "gptq"}