{ "format": "thrush-ternary-v2", "names": "hf", "packing": { "base": 3, "code_offset": 1, "elements_per_byte": 5, "row_bytes": "ceil(in/5)", "digit_order": "element i of a row is base-3 digit i%5 of byte i//5, least significant digit first; pad digits 0", "row_major": true }, "weight_rule": "w = scales[row, col // group_size] * (code - 1) with code in {0,1,2}", "quant": { "mode": "ternary", "group_size": 128 }, "source": { "teacher": "nvidia/parakeet-tdt-0.6b-v3" }, "vad_head": { "tensors": [ "vad_head.ctx.bias", "vad_head.ctx.weight", "vad_head.out.bias", "vad_head.out.weight", "vad_head.proj.bias", "vad_head.proj.weight" ], "d_model": 1024, "hidden": 128, "kernel": 5, "frame_seconds": 0.08, "dtype": "float16", "params": 213377 }, "quantized_modules": [ { "name": "encoder.layers.0.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5044150352478027 }, { "name": "encoder.layers.0.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5184123516082764 }, { "name": "encoder.layers.0.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.15679550170898438 }, { "name": "encoder.layers.0.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.46837711334228516 }, { "name": "encoder.layers.0.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.7815647125244141 }, { "name": "encoder.layers.0.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.7256069183349609 }, { "name": "encoder.layers.0.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.13124942779541016 }, { "name": "encoder.layers.0.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.49237823486328125 }, { "name": "encoder.layers.0.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.5364360809326172 }, { "name": "encoder.layers.0.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5409379005432129 }, { "name": "encoder.layers.0.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5457351207733154 }, { "name": "encoder.layers.1.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5231430530548096 }, { "name": "encoder.layers.1.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4283597469329834 }, { "name": "encoder.layers.1.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.34552860260009766 }, { "name": "encoder.layers.1.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5812540054321289 }, { "name": "encoder.layers.1.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.41185569763183594 }, { "name": "encoder.layers.1.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3538980484008789 }, { "name": "encoder.layers.1.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.13209915161132812 }, { "name": "encoder.layers.1.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4484829902648926 }, { "name": "encoder.layers.1.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.43465423583984375 }, { "name": "encoder.layers.1.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.42896461486816406 }, { "name": "encoder.layers.1.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5293712615966797 }, { "name": "encoder.layers.2.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4617030620574951 }, { "name": "encoder.layers.2.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.45351243019104004 }, { "name": "encoder.layers.2.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.32970237731933594 }, { "name": "encoder.layers.2.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5406455993652344 }, { "name": "encoder.layers.2.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.40648365020751953 }, { "name": "encoder.layers.2.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.38976478576660156 }, { "name": "encoder.layers.2.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.1539907455444336 }, { "name": "encoder.layers.2.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.45189952850341797 }, { "name": "encoder.layers.2.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.45950984954833984 }, { "name": "encoder.layers.2.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.43088412284851074 }, { "name": "encoder.layers.2.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5336544513702393 }, { "name": "encoder.layers.3.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4594447612762451 }, { "name": "encoder.layers.3.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4711878299713135 }, { "name": "encoder.layers.3.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.34746265411376953 }, { "name": "encoder.layers.3.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5459375381469727 }, { "name": "encoder.layers.3.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3894500732421875 }, { "name": "encoder.layers.3.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.37818431854248047 }, { "name": "encoder.layers.3.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.15802383422851562 }, { "name": "encoder.layers.3.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4486575126647949 }, { "name": "encoder.layers.3.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4589385986328125 }, { "name": "encoder.layers.3.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4405934810638428 }, { "name": "encoder.layers.3.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.540421724319458 }, { "name": "encoder.layers.4.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4638183116912842 }, { "name": "encoder.layers.4.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.48339200019836426 }, { "name": "encoder.layers.4.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3684682846069336 }, { "name": "encoder.layers.4.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5612239837646484 }, { "name": "encoder.layers.4.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.38426876068115234 }, { "name": "encoder.layers.4.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3797454833984375 }, { "name": "encoder.layers.4.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.14944171905517578 }, { "name": "encoder.layers.4.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.45036983489990234 }, { "name": "encoder.layers.4.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4643983840942383 }, { "name": "encoder.layers.4.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.447796106338501 }, { "name": "encoder.layers.4.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.547126054763794 }, { "name": "encoder.layers.5.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.47231388092041016 }, { "name": "encoder.layers.5.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.48183250427246094 }, { "name": "encoder.layers.5.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3488779067993164 }, { "name": "encoder.layers.5.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5331039428710938 }, { "name": "encoder.layers.5.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.40555763244628906 }, { "name": "encoder.layers.5.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3914680480957031 }, { "name": "encoder.layers.5.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.15366840362548828 }, { "name": "encoder.layers.5.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.45035457611083984 }, { "name": "encoder.layers.5.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4670143127441406 }, { "name": "encoder.layers.5.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.45580148696899414 }, { "name": "encoder.layers.5.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5230386257171631 }, { "name": "encoder.layers.6.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4766519069671631 }, { "name": "encoder.layers.6.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4740920066833496 }, { "name": "encoder.layers.6.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3270740509033203 }, { "name": "encoder.layers.6.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5052204132080078 }, { "name": "encoder.layers.6.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4151468276977539 }, { "name": "encoder.layers.6.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4025430679321289 }, { "name": "encoder.layers.6.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.15303707122802734 }, { "name": "encoder.layers.6.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.45003604888916016 }, { "name": "encoder.layers.6.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4495658874511719 }, { "name": "encoder.layers.6.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4526941776275635 }, { "name": "encoder.layers.6.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5148673057556152 }, { "name": "encoder.layers.7.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.48210573196411133 }, { "name": "encoder.layers.7.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.46149134635925293 }, { "name": "encoder.layers.7.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.33362388610839844 }, { "name": "encoder.layers.7.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5477542877197266 }, { "name": "encoder.layers.7.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.42502307891845703 }, { "name": "encoder.layers.7.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3920621871948242 }, { "name": "encoder.layers.7.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.1467151641845703 }, { "name": "encoder.layers.7.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4432258605957031 }, { "name": "encoder.layers.7.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4387655258178711 }, { "name": "encoder.layers.7.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4537210464477539 }, { "name": "encoder.layers.7.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5217123031616211 }, { "name": "encoder.layers.8.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.48615384101867676 }, { "name": "encoder.layers.8.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4634566307067871 }, { "name": "encoder.layers.8.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.34348583221435547 }, { "name": "encoder.layers.8.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5461740493774414 }, { "name": "encoder.layers.8.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3690071105957031 }, { "name": "encoder.layers.8.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.37335777282714844 }, { "name": "encoder.layers.8.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.14555931091308594 }, { "name": "encoder.layers.8.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4336113929748535 }, { "name": "encoder.layers.8.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.42722225189208984 }, { "name": "encoder.layers.8.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.46222662925720215 }, { "name": "encoder.layers.8.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5234966278076172 }, { "name": "encoder.layers.9.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.49318695068359375 }, { "name": "encoder.layers.9.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.471712589263916 }, { "name": "encoder.layers.9.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3633537292480469 }, { "name": "encoder.layers.9.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5625839233398438 }, { "name": "encoder.layers.9.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.37224864959716797 }, { "name": "encoder.layers.9.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3777170181274414 }, { "name": "encoder.layers.9.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.15102577209472656 }, { "name": "encoder.layers.9.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4328136444091797 }, { "name": "encoder.layers.9.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.41849327087402344 }, { "name": "encoder.layers.9.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.47850704193115234 }, { "name": "encoder.layers.9.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5117299556732178 }, { "name": "encoder.layers.10.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5326597690582275 }, { "name": "encoder.layers.10.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4868049621582031 }, { "name": "encoder.layers.10.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.38233089447021484 }, { "name": "encoder.layers.10.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.6121854782104492 }, { "name": "encoder.layers.10.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3678264617919922 }, { "name": "encoder.layers.10.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.36456298828125 }, { "name": "encoder.layers.10.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.14080429077148438 }, { "name": "encoder.layers.10.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4323916435241699 }, { "name": "encoder.layers.10.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.42212963104248047 }, { "name": "encoder.layers.10.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4855043888092041 }, { "name": "encoder.layers.10.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5134315490722656 }, { "name": "encoder.layers.11.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5217080116271973 }, { "name": "encoder.layers.11.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4963808059692383 }, { "name": "encoder.layers.11.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.35190773010253906 }, { "name": "encoder.layers.11.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5736579895019531 }, { "name": "encoder.layers.11.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3828563690185547 }, { "name": "encoder.layers.11.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3663148880004883 }, { "name": "encoder.layers.11.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.15211105346679688 }, { "name": "encoder.layers.11.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.43471431732177734 }, { "name": "encoder.layers.11.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.42731475830078125 }, { "name": "encoder.layers.11.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4918859004974365 }, { "name": "encoder.layers.11.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5454516410827637 }, { "name": "encoder.layers.12.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5295507907867432 }, { "name": "encoder.layers.12.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4964902400970459 }, { "name": "encoder.layers.12.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3329486846923828 }, { "name": "encoder.layers.12.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5321378707885742 }, { "name": "encoder.layers.12.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3993492126464844 }, { "name": "encoder.layers.12.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3653535842895508 }, { "name": "encoder.layers.12.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.16003036499023438 }, { "name": "encoder.layers.12.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.43686389923095703 }, { "name": "encoder.layers.12.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4244871139526367 }, { "name": "encoder.layers.12.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.47611141204833984 }, { "name": "encoder.layers.12.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5230574607849121 }, { "name": "encoder.layers.13.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.531564474105835 }, { "name": "encoder.layers.13.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5122246742248535 }, { "name": "encoder.layers.13.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.36064720153808594 }, { "name": "encoder.layers.13.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5534095764160156 }, { "name": "encoder.layers.13.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.37798595428466797 }, { "name": "encoder.layers.13.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.35416412353515625 }, { "name": "encoder.layers.13.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.14722442626953125 }, { "name": "encoder.layers.13.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.43766117095947266 }, { "name": "encoder.layers.13.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.41414833068847656 }, { "name": "encoder.layers.13.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4815671443939209 }, { "name": "encoder.layers.13.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4973268508911133 }, { "name": "encoder.layers.14.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5406713485717773 }, { "name": "encoder.layers.14.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.520122766494751 }, { "name": "encoder.layers.14.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.37984371185302734 }, { "name": "encoder.layers.14.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5693721771240234 }, { "name": "encoder.layers.14.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.37279415130615234 }, { "name": "encoder.layers.14.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.35245513916015625 }, { "name": "encoder.layers.14.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.14053821563720703 }, { "name": "encoder.layers.14.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4381241798400879 }, { "name": "encoder.layers.14.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.41631603240966797 }, { "name": "encoder.layers.14.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.48996448516845703 }, { "name": "encoder.layers.14.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5002038478851318 }, { "name": "encoder.layers.15.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5357084274291992 }, { "name": "encoder.layers.15.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5198726654052734 }, { "name": "encoder.layers.15.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3787193298339844 }, { "name": "encoder.layers.15.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.6187000274658203 }, { "name": "encoder.layers.15.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.37151432037353516 }, { "name": "encoder.layers.15.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3600912094116211 }, { "name": "encoder.layers.15.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.14568805694580078 }, { "name": "encoder.layers.15.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.43272924423217773 }, { "name": "encoder.layers.15.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.41810035705566406 }, { "name": "encoder.layers.15.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.49303531646728516 }, { "name": "encoder.layers.15.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.53395676612854 }, { "name": "encoder.layers.16.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5362865924835205 }, { "name": "encoder.layers.16.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5048935413360596 }, { "name": "encoder.layers.16.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3586082458496094 }, { "name": "encoder.layers.16.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5664215087890625 }, { "name": "encoder.layers.16.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3739347457885742 }, { "name": "encoder.layers.16.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.36450672149658203 }, { "name": "encoder.layers.16.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.1494426727294922 }, { "name": "encoder.layers.16.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.43798398971557617 }, { "name": "encoder.layers.16.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.3995380401611328 }, { "name": "encoder.layers.16.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.46408939361572266 }, { "name": "encoder.layers.16.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.47188329696655273 }, { "name": "encoder.layers.17.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4892551898956299 }, { "name": "encoder.layers.17.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4277915954589844 }, { "name": "encoder.layers.17.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.35779476165771484 }, { "name": "encoder.layers.17.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.6346902847290039 }, { "name": "encoder.layers.17.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3553304672241211 }, { "name": "encoder.layers.17.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3436117172241211 }, { "name": "encoder.layers.17.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.2155017852783203 }, { "name": "encoder.layers.17.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4304986000061035 }, { "name": "encoder.layers.17.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.39414215087890625 }, { "name": "encoder.layers.17.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4398188591003418 }, { "name": "encoder.layers.17.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4505887031555176 }, { "name": "encoder.layers.18.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4843440055847168 }, { "name": "encoder.layers.18.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.43118953704833984 }, { "name": "encoder.layers.18.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3386106491088867 }, { "name": "encoder.layers.18.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5647468566894531 }, { "name": "encoder.layers.18.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.37237071990966797 }, { "name": "encoder.layers.18.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3634309768676758 }, { "name": "encoder.layers.18.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.1491680145263672 }, { "name": "encoder.layers.18.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.43354082107543945 }, { "name": "encoder.layers.18.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4006175994873047 }, { "name": "encoder.layers.18.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4384031295776367 }, { "name": "encoder.layers.18.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4480724334716797 }, { "name": "encoder.layers.19.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.47231221199035645 }, { "name": "encoder.layers.19.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.40505099296569824 }, { "name": "encoder.layers.19.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3529081344604492 }, { "name": "encoder.layers.19.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5809326171875 }, { "name": "encoder.layers.19.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.36546993255615234 }, { "name": "encoder.layers.19.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.35294342041015625 }, { "name": "encoder.layers.19.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.17095470428466797 }, { "name": "encoder.layers.19.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4216432571411133 }, { "name": "encoder.layers.19.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.3793497085571289 }, { "name": "encoder.layers.19.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.42360615730285645 }, { "name": "encoder.layers.19.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.45482897758483887 }, { "name": "encoder.layers.20.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4815986156463623 }, { "name": "encoder.layers.20.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.36296963691711426 }, { "name": "encoder.layers.20.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.39092063903808594 }, { "name": "encoder.layers.20.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5931243896484375 }, { "name": "encoder.layers.20.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3127937316894531 }, { "name": "encoder.layers.20.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.30979347229003906 }, { "name": "encoder.layers.20.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.14920997619628906 }, { "name": "encoder.layers.20.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.4177713394165039 }, { "name": "encoder.layers.20.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.33641719818115234 }, { "name": "encoder.layers.20.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.43540430068969727 }, { "name": "encoder.layers.20.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.39700770378112793 }, { "name": "encoder.layers.21.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.48596787452697754 }, { "name": "encoder.layers.21.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3406181335449219 }, { "name": "encoder.layers.21.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.37795257568359375 }, { "name": "encoder.layers.21.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5576667785644531 }, { "name": "encoder.layers.21.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3066892623901367 }, { "name": "encoder.layers.21.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.2996206283569336 }, { "name": "encoder.layers.21.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.15436172485351562 }, { "name": "encoder.layers.21.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.40650415420532227 }, { "name": "encoder.layers.21.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.314422607421875 }, { "name": "encoder.layers.21.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.39052605628967285 }, { "name": "encoder.layers.21.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3677656650543213 }, { "name": "encoder.layers.22.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4427375793457031 }, { "name": "encoder.layers.22.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3055906295776367 }, { "name": "encoder.layers.22.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3857412338256836 }, { "name": "encoder.layers.22.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5995121002197266 }, { "name": "encoder.layers.22.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.2965202331542969 }, { "name": "encoder.layers.22.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.28084754943847656 }, { "name": "encoder.layers.22.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.14216327667236328 }, { "name": "encoder.layers.22.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.37828636169433594 }, { "name": "encoder.layers.22.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.2952146530151367 }, { "name": "encoder.layers.22.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4962489604949951 }, { "name": "encoder.layers.22.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.32378721237182617 }, { "name": "encoder.layers.23.feed_forward1.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.4867126941680908 }, { "name": "encoder.layers.23.feed_forward1.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.2900993824005127 }, { "name": "encoder.layers.23.self_attn.q_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3742027282714844 }, { "name": "encoder.layers.23.self_attn.k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.5684337615966797 }, { "name": "encoder.layers.23.self_attn.v_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.2869300842285156 }, { "name": "encoder.layers.23.self_attn.o_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.29657459259033203 }, { "name": "encoder.layers.23.self_attn.relative_k_proj", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.1371021270751953 }, { "name": "encoder.layers.23.conv.pointwise_conv1", "out_features": 2048, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.38350915908813477 }, { "name": "encoder.layers.23.conv.pointwise_conv2", "out_features": 1024, "in_features": 1024, "group_size": 128, "as_conv1d": true, "has_bias": false, "zero_fraction": 0.3024635314941406 }, { "name": "encoder.layers.23.feed_forward2.linear1", "out_features": 4096, "in_features": 1024, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.30291128158569336 }, { "name": "encoder.layers.23.feed_forward2.linear2", "out_features": 1024, "in_features": 4096, "group_size": 128, "as_conv1d": false, "has_bias": false, "zero_fraction": 0.3130035400390625 } ], "n_quantized_params": 603979776, "n_dense_params": 23290911, "size_mb": 177.7 }