EXAONE-4.5-33B-DASHQ-INT4-g64 / dashq_config.json
jkim96's picture
Update evaluation results
e5047d7 verified
Raw
History Blame
163 kB
{
"base_model": "LGAI-EXAONE/EXAONE-4.5-33B",
"format": "dashq-packed-linear",
"format_version": 1,
"method": "dashq",
"model_class": "image_text_to_text",
"params": {
"bits": 4,
"group_size": 64,
"low_memory_optimization": false,
"moe_hessian_scope": "shared",
"n_samples": 128,
"scale_zero_dtype": "float16",
"symmetric": false,
"use_error_compensation": true,
"use_optimal_shrinkage": true,
"use_weighted_quantization": true
},
"quantized_modules": {
"model.language_model.layers.0.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.0.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.0.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.0.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.0.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.0.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.0.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.1.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.1.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.1.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.1.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.1.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.1.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.1.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.10.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.10.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.10.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.10.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.10.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.10.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.10.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.11.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.11.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.11.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.11.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.11.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.11.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.11.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.12.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.12.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.12.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.12.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.12.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.12.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.12.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.13.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.13.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.13.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.13.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.13.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.13.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.13.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.14.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.14.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.14.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.14.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.14.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.14.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.14.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.15.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.15.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.15.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.15.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.15.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.15.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.15.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.16.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.16.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.16.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.16.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.16.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.16.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.16.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.17.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.17.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.17.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.17.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.17.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.17.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.17.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.18.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.18.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.18.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.18.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.18.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.18.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.18.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.19.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.19.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.19.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.19.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.19.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.19.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.19.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.2.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.2.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.2.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.2.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.2.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.2.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.2.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.20.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.20.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.20.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.20.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.20.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.20.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.20.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.21.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.21.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.21.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.21.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.21.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.21.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.21.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.22.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.22.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.22.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.22.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.22.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.22.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.22.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.23.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.23.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.23.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.23.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.23.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.23.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.23.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.24.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.24.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.24.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.24.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.24.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.24.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.24.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.25.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.25.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.25.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.25.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.25.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.25.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.25.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.26.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.26.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.26.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.26.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.26.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.26.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.26.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.27.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.27.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.27.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.27.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.27.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.27.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.27.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.28.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.28.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.28.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.28.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.28.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.28.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.28.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.29.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.29.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.29.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.29.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.29.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.29.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.29.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.3.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.3.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.3.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.3.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.3.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.3.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.3.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.30.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.30.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.30.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.30.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.30.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.30.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.30.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.31.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.31.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.31.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.31.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.31.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.31.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.31.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.32.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.32.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.32.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.32.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.32.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.32.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.32.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.33.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.33.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.33.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.33.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.33.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.33.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.33.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.34.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.34.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.34.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.34.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.34.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.34.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.34.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.35.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.35.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.35.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.35.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.35.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.35.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.35.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.36.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.36.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.36.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.36.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.36.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.36.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.36.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.37.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.37.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.37.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.37.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.37.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.37.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.37.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.38.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.38.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.38.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.38.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.38.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.38.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.38.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.39.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.39.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.39.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.39.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.39.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.39.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.39.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.4.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.4.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.4.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.4.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.4.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.4.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.4.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.40.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.40.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.40.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.40.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.40.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.40.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.40.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.41.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.41.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.41.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.41.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.41.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.41.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.41.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.42.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.42.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.42.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.42.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.42.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.42.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.42.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.43.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.43.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.43.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.43.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.43.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.43.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.43.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.44.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.44.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.44.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.44.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.44.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.44.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.44.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.45.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.45.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.45.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.45.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.45.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.45.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.45.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.46.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.46.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.46.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.46.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.46.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.46.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.46.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.47.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.47.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.47.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.47.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.47.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.47.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.47.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.48.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.48.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.48.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.48.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.48.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.48.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.48.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.49.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.49.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.49.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.49.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.49.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.49.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.49.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.5.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.5.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.5.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.5.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.5.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.5.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.5.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.50.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.50.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.50.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.50.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.50.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.50.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.50.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.51.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.51.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.51.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.51.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.51.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.51.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.51.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.52.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.52.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.52.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.52.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.52.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.52.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.52.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.53.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.53.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.53.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.53.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.53.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.53.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.53.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.54.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.54.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.54.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.54.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.54.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.54.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.54.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.55.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.55.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.55.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.55.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.55.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.55.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.55.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.56.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.56.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.56.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.56.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.56.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.56.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.56.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.57.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.57.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.57.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.57.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.57.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.57.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.57.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.58.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.58.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.58.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.58.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.58.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.58.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.58.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.59.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.59.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.59.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.59.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.59.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.59.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.59.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.6.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.6.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.6.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.6.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.6.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.6.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.6.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.60.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.60.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.60.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.60.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.60.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.60.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.60.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.61.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.61.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.61.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.61.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.61.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.61.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.61.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.62.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.62.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.62.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.62.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.62.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.62.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.62.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.63.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.63.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.63.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.63.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.63.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.63.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.63.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.7.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.7.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.7.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.7.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.7.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.7.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.7.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.8.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.8.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.8.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.8.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.8.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.8.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.8.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.9.mlp.down_proj": {
"group_size": 64,
"in_features": 27392,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 27392,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.9.mlp.gate_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.9.mlp.up_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 2191360,
"out_features": 27392,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.9.self_attn.k_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.9.self_attn.o_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.9.self_attn.q_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 409600,
"out_features": 5120,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
},
"model.language_model.layers.9.self_attn.v_proj": {
"group_size": 64,
"in_features": 5120,
"linear_dtype": "bfloat16",
"nbits": 4,
"num_groups": 81920,
"out_features": 1024,
"packing": "int4_packed_u32",
"quant_in_features": 5120,
"runtime_backend": "torch",
"scale_zero_dtype": "float16"
}
},
"results": {
"Method": "dashq",
"Model": "LGAI-EXAONE/EXAONE-4.5-33B",
"ModelSizeGB": 23.131899224,
"OriginalSizeGB": 68.700331096,
"PPL": 8.442422866821289,
"Params": "{'bits': 4, 'group_size': 64, 'scale_zero_dtype': 'float16', 'n_samples': 128, 'moe_hessian_scope': 'shared', 'use_error_compensation': True, 'use_optimal_shrinkage': True, 'use_weighted_quantization': True, 'symmetric': False, 'low_memory_optimization': False}",
"QuantTime": 1540.4014463424683,
"arc_challenge": 57.67918088737202,
"arc_easy": 85.01683501683502,
"commonsense_qa": 78.05077805077805,
"gsm8k_cot": 75.3601213040182,
"hellaswag": 79.01812387970524,
"lambada_openai": null,
"mmlu": 78.15126050420169,
"openbookqa": null,
"piqa": 80.68552774755169,
"truthfulqa_mc2": 56.40180916893911,
"winogrande": 73.32280978689818,
"zeroshot_avg": 72.88215207686848
}
}