diff --git "a/quality_report.json" "b/quality_report.json" --- "a/quality_report.json" +++ "b/quality_report.json" @@ -2,8 +2,8 @@ "source": "Agnes-AI/Agnes-3.0-Flash", "source_revision": "24f712ce59379b54c4a141d2708c35daf5ff613b", "target": "ProCreations/Agnes-3.0-Flash-NVFP4", - "status": "build_failed", - "code_revision": "fc2119a0164ef0e766f4889754d3d16440f3478a", + "status": "building", + "code_revision": "aec16a201750078614bed85f35fccc463ec9edbf", "modelopt_revision": "5cae3940402f1ced98069a666b0bec72ec8b33b5", "quantization": { "method": "plain NVFP4 max", @@ -14,10679 +14,7 @@ "activation_global_scale": 1.0, "block_size": 16, "activation_block_scales": "dynamic E4M3", - "kv_cache_quantized": false, - "configuration": { - "quant_cfg": [ - { - "quantizer_name": "*", - "enable": false - }, - { - "quantizer_name": "model.language_model.layers.0.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.0.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.0.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.0.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.0.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.0.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.1.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.1.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.1.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.1.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.1.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.1.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.2.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.2.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.2.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.2.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.2.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.2.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.3.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.3.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.3.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.3.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.3.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.3.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.3.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.3.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.3.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.3.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.3.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.3.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.3.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.3.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.4.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.4.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.4.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.4.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.4.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.4.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.5.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.5.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.5.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.5.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.5.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.5.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.6.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.6.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.6.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.6.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.6.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.6.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.7.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.7.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.7.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.7.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.7.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.7.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.7.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.7.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.7.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.7.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.7.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.7.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.7.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.7.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.8.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.8.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.8.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.8.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.8.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.8.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.9.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.9.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.9.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.9.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.9.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.9.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.10.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.10.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.10.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.10.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.10.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.10.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.11.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.11.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.11.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.11.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.11.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.11.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.11.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.11.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.11.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.11.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.11.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.11.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.11.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.11.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.12.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.12.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.12.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.12.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.12.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.12.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.13.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.13.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.13.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.13.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.13.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.13.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.14.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.14.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.14.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.14.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.14.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.14.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.15.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.15.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.15.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.15.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.15.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.15.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.15.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.15.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.15.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.15.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.15.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.15.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.15.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.15.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.16.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.16.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.16.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.16.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.16.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.16.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.17.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.17.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.17.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.17.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.17.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.17.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.18.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.18.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.18.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.18.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.18.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.18.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.19.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.19.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.19.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.19.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.19.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.19.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.19.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.19.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.19.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.19.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.19.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.19.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.19.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.19.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.20.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.20.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.20.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.20.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.20.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.20.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.21.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.21.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.21.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.21.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.21.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.21.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.22.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.22.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.22.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.22.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.22.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.22.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.23.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.23.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.23.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.23.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.23.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.23.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.23.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.23.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.23.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.23.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.23.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.23.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.23.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.23.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.24.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.24.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.24.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.24.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.24.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.24.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.25.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.25.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.25.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.25.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.25.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.25.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.26.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.26.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.26.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.26.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.26.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.26.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.27.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.27.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.27.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.27.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.27.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.27.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.27.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.27.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.27.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.27.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.27.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.27.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.27.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.27.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.28.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.28.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.28.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.28.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.28.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.28.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.29.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.29.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.29.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.29.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.29.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.29.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.30.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.30.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.30.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.30.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.30.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.30.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.31.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.31.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.31.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.31.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.31.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.31.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.31.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.31.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.31.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.31.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.31.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.31.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.31.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.31.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.32.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.32.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.32.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.32.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.32.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.32.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.33.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.33.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.33.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.33.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.33.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.33.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.34.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.34.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.34.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.34.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.34.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.34.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.35.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.35.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.35.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.35.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.35.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.35.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.35.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.35.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.35.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.35.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.35.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.35.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.35.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.35.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.36.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.36.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.36.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.36.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.36.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.36.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.37.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.37.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.37.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.37.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.37.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.37.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.38.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.38.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.38.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.38.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.38.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.38.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.39.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.39.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.39.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.39.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.39.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.39.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.39.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.39.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.39.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.39.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.39.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.39.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.39.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.39.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.40.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.40.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.40.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.40.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.40.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.40.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.41.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.41.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.41.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.41.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.41.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.41.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.42.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.42.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.42.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.42.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.42.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.42.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.43.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.43.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.43.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.43.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.43.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.43.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.43.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.43.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.43.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.43.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.43.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.43.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.43.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.43.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.44.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.44.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.44.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.44.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.44.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.44.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.45.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.45.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.45.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.45.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.45.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.45.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.46.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.46.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.46.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.46.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.46.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.46.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.47.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.47.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.47.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.47.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.47.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.47.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.47.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.47.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.47.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.47.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.47.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.47.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.47.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.47.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.48.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.48.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.48.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.48.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.48.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.48.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.49.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.49.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.49.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.49.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.49.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.49.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.50.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.50.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.50.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.50.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.50.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.50.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.51.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.51.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.51.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.51.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.51.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.51.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.51.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.51.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.51.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.51.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.51.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.51.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.51.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.51.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.52.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.52.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.52.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.52.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.52.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.52.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.53.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.53.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.53.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.53.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.53.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.53.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.54.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.54.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.54.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.54.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.54.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.54.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.55.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.55.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.55.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.55.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.55.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.55.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.55.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.55.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.55.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.55.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.55.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.55.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.55.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.55.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.56.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.56.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.56.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.56.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.56.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.56.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.57.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.57.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.57.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.57.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.57.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.57.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.58.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.58.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.58.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.58.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.58.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.58.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.59.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.59.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.59.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.59.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.59.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.59.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.59.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.59.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.59.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.59.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.59.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.59.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.59.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.59.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.60.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.60.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.60.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.60.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.60.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.60.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.61.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.61.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.61.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.61.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.61.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.61.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.62.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.62.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.62.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.62.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.62.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.62.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.63.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.63.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.63.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.63.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.63.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.63.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.63.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.63.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.63.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.63.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.63.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.63.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.63.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.63.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.64.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.64.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.64.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.64.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.64.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.64.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.65.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.65.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.65.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.65.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.65.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.65.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.66.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.66.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.66.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.66.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.66.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.66.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.67.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.67.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.67.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.67.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.67.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.67.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.67.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.67.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.67.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.67.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.67.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.67.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.67.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.67.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.68.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.68.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.68.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.68.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.68.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.68.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.69.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.69.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.69.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.69.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.69.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.69.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.70.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.70.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.70.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.70.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.70.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.70.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.71.global_attn.q_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.71.global_attn.q_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.71.global_attn.k_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.71.global_attn.k_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.71.global_attn.v_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.71.global_attn.v_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.71.global_attn.o_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.71.global_attn.o_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.71.mlp.gate_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.71.mlp.gate_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.71.mlp.up_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.71.mlp.up_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - }, - { - "quantizer_name": "model.language_model.layers.71.mlp.down_proj.weight_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - } - } - }, - { - "quantizer_name": "model.language_model.layers.71.mlp.down_proj.input_quantizer", - "cfg": { - "num_bits": [ - 2, - 1 - ], - "effective_bits": 4.5, - "block_sizes": { - "-1": 16, - "type": "dynamic", - "scale_bits": [ - 4, - 3 - ] - }, - "constant_amax": 2688.0 - } - } - ], - "algorithm": { - "method": "max", - "layerwise": { - "enable": false - }, - "skip_forward_without_activation_calib": true - } - }, - "weight_statistics_seconds": 4.781414550961927 + "kv_cache_quantized": false }, "folding": { "main_width": 17408, @@ -10708,10 +36,5 @@ }, "hardware": [ "NVIDIA H200" - ], - "export_graph_inspection": { - "dummy_forward_calls": 1, - "input_tokens": 2, - "quantizers_disabled": true - } + ] }