PPeSse/WhiteRabbitNeo-v2-q4f32-transcoded-MLC
012
1{2 "metadata": {3 "ParamSize": 325,4 "ParamBytes": 5019811840.0,5 "BitsPerParam": 5.0008951738652076 },7 "records": [8 {9 "dataPath": "params_shard_0.bin",10 "format": "raw-shard",11 "nbytes": 262668288,12 "records": [13 {14 "name": "lm_head.q_weight",15 "shape": [16 128256,17 51218 ],19 "dtype": "uint32",20 "format": "f32-to-bf16",21 "nbytes": 262668288,22 "byteOffset": 023 }24 ],25 "md5sum": "66876da5b9089e2e762382a9b42f46dd"26 },27 {28 "dataPath": "params_shard_1.bin",29 "format": "raw-shard",30 "nbytes": 29360128,31 "records": [32 {33 "name": "model.layers.31.mlp.down_proj.q_weight",34 "shape": [35 4096,36 179237 ],38 "dtype": "uint32",39 "format": "f32-to-bf16",40 "nbytes": 29360128,41 "byteOffset": 042 }43 ],44 "md5sum": "7297979f3f6d7b6e290d816e5131b7e8"45 },46 {47 "dataPath": "params_shard_2.bin",48 "format": "raw-shard",49 "nbytes": 32841728,50 "records": [51 {52 "name": "lm_head.q_scale",53 "shape": [54 128256,55 12856 ],57 "dtype": "float32",58 "format": "f32-to-bf16",59 "nbytes": 32833536,60 "byteOffset": 061 },62 {63 "name": "model.layers.31.input_layernorm.weight",64 "shape": [65 409666 ],67 "dtype": "float32",68 "format": "f32-to-bf16",69 "nbytes": 8192,70 "byteOffset": 3283353671 }72 ],73 "md5sum": "842b8aa3675fee475d27176b9bc843b0"74 },75 {76 "dataPath": "params_shard_3.bin",77 "format": "raw-shard",78 "nbytes": 262668288,79 "records": [80 {81 "name": "model.embed_tokens.q_weight",82 "shape": [83 128256,84 51285 ],86 "dtype": "uint32",87 "format": "f32-to-bf16",88 "nbytes": 262668288,89 "byteOffset": 090 }91 ],92 "md5sum": "e87d008737acad3921acb09c89edeb68"93 },94 {95 "dataPath": "params_shard_4.bin",96 "format": "raw-shard",97 "nbytes": 32833536,98 "records": [99 {100 "name": "model.embed_tokens.q_scale",101 "shape": [102 128256,103 128104 ],105 "dtype": "float32",106 "format": "f32-to-bf16",107 "nbytes": 32833536,108 "byteOffset": 0109 }110 ],111 "md5sum": "ec84b9640563169ca5f8a9ded6123bf1"112 },113 {114 "dataPath": "params_shard_5.bin",115 "format": "raw-shard",116 "nbytes": 33054720,117 "records": [118 {119 "name": "model.layers.31.mlp.down_proj.q_scale",120 "shape": [121 4096,122 448123 ],124 "dtype": "float32",125 "format": "f32-to-bf16",126 "nbytes": 3670016,127 "byteOffset": 0128 },129 {130 "name": "model.layers.31.post_attention_layernorm.weight",131 "shape": [132 4096133 ],134 "dtype": "float32",135 "format": "f32-to-bf16",136 "nbytes": 8192,137 "byteOffset": 3670016138 },139 {140 "name": "model.norm.weight",141 "shape": [142 4096143 ],144 "dtype": "float32",145 "format": "f32-to-bf16",146 "nbytes": 8192,147 "byteOffset": 3678208148 },149 {150 "name": "model.layers.0.input_layernorm.weight",151 "shape": [152 4096153 ],154 "dtype": "float32",155 "format": "f32-to-bf16",156 "nbytes": 8192,157 "byteOffset": 3686400158 },159 {160 "name": "model.layers.0.mlp.down_proj.q_weight",161 "shape": [162 4096,163 1792164 ],165 "dtype": "uint32",166 "format": "f32-to-bf16",167 "nbytes": 29360128,168 "byteOffset": 3694592169 }170 ],171 "md5sum": "a3d2dc2dd9ae1a9091f6821a30cc9377"172 },173 {174 "dataPath": "params_shard_6.bin",175 "format": "raw-shard",176 "nbytes": 58720256,177 "records": [178 {179 "name": "model.layers.0.mlp.gate_up_proj.q_weight",180 "shape": [181 28672,182 512183 ],184 "dtype": "uint32",185 "format": "f32-to-bf16",186 "nbytes": 58720256,187 "byteOffset": 0188 }189 ],190 "md5sum": "3e01ed93f05e835d92d9df248e7533c7"191 },192 {193 "dataPath": "params_shard_7.bin",194 "format": "raw-shard",195 "nbytes": 25174016,196 "records": [197 {198 "name": "model.layers.0.mlp.down_proj.q_scale",199 "shape": [200 4096,201 448202 ],203 "dtype": "float32",204 "format": "f32-to-bf16",205 "nbytes": 3670016,206 "byteOffset": 0207 },208 {209 "name": "model.layers.0.mlp.gate_up_proj.q_scale",210 "shape": [211 28672,212 128213 ],214 "dtype": "float32",215 "format": "f32-to-bf16",216 "nbytes": 7340032,217 "byteOffset": 3670016218 },219 {220 "name": "model.layers.0.post_attention_layernorm.weight",221 "shape": [222 4096223 ],224 "dtype": "float32",225 "format": "f32-to-bf16",226 "nbytes": 8192,227 "byteOffset": 11010048228 },229 {230 "name": "model.layers.0.self_attn.qkv_proj.q_weight",231 "shape": [232 6144,233 512234 ],235 "dtype": "uint32",236 "format": "f32-to-bf16",237 "nbytes": 12582912,238 "byteOffset": 11018240239 },240 {241 "name": "model.layers.0.self_attn.qkv_proj.q_scale",242 "shape": [243 6144,244 128245 ],246 "dtype": "float32",247 "format": "f32-to-bf16",248 "nbytes": 1572864,249 "byteOffset": 23601152250 }251 ],252 "md5sum": "ab6895205b3387e335d7b44781a08260"253 },254 {255 "dataPath": "params_shard_8.bin",256 "format": "raw-shard",257 "nbytes": 29360128,258 "records": [259 {260 "name": "model.layers.1.mlp.down_proj.q_weight",261 "shape": [262 4096,263 1792264 ],265 "dtype": "uint32",266 "format": "f32-to-bf16",267 "nbytes": 29360128,268 "byteOffset": 0269 }270 ],271 "md5sum": "b8b772d5098aca60f7c3f6c8679ef83b"272 },273 {274 "dataPath": "params_shard_9.bin",275 "format": "raw-shard",276 "nbytes": 58720256,277 "records": [278 {279 "name": "model.layers.1.mlp.gate_up_proj.q_weight",280 "shape": [281 28672,282 512283 ],284 "dtype": "uint32",285 "format": "f32-to-bf16",286 "nbytes": 58720256,287 "byteOffset": 0288 }289 ],290 "md5sum": "ed0223b5671d1b649d500941c9190624"291 },292 {293 "dataPath": "params_shard_10.bin",294 "format": "raw-shard",295 "nbytes": 33046528,296 "records": [297 {298 "name": "model.layers.0.self_attn.o_proj.q_weight",299 "shape": [300 4096,301 512302 ],303 "dtype": "uint32",304 "format": "f32-to-bf16",305 "nbytes": 8388608,306 "byteOffset": 0307 },308 {309 "name": "model.layers.0.self_attn.o_proj.q_scale",310 "shape": [311 4096,312 128313 ],314 "dtype": "float32",315 "format": "f32-to-bf16",316 "nbytes": 1048576,317 "byteOffset": 8388608318 },319 {320 "name": "model.layers.1.input_layernorm.weight",321 "shape": [322 4096323 ],324 "dtype": "float32",325 "format": "f32-to-bf16",326 "nbytes": 8192,327 "byteOffset": 9437184328 },329 {330 "name": "model.layers.1.mlp.down_proj.q_scale",331 "shape": [332 4096,333 448334 ],335 "dtype": "float32",336 "format": "f32-to-bf16",337 "nbytes": 3670016,338 "byteOffset": 9445376339 },340 {341 "name": "model.layers.1.mlp.gate_up_proj.q_scale",342 "shape": [343 28672,344 128345 ],346 "dtype": "float32",347 "format": "f32-to-bf16",348 "nbytes": 7340032,349 "byteOffset": 13115392350 },351 {352 "name": "model.layers.1.post_attention_layernorm.weight",353 "shape": [354 4096355 ],356 "dtype": "float32",357 "format": "f32-to-bf16",358 "nbytes": 8192,359 "byteOffset": 20455424360 },361 {362 "name": "model.layers.1.self_attn.qkv_proj.q_weight",363 "shape": [364 6144,365 512366 ],367 "dtype": "uint32",368 "format": "f32-to-bf16",369 "nbytes": 12582912,370 "byteOffset": 20463616371 }372 ],373 "md5sum": "dc6269103cff054af8c8f5a9cad7a1e4"374 },375 {376 "dataPath": "params_shard_11.bin",377 "format": "raw-shard",378 "nbytes": 29360128,379 "records": [380 {381 "name": "model.layers.2.mlp.down_proj.q_weight",382 "shape": [383 4096,384 1792385 ],386 "dtype": "uint32",387 "format": "f32-to-bf16",388 "nbytes": 29360128,389 "byteOffset": 0390 }391 ],392 "md5sum": "224bbe5a8da92b73344dd0ac21fd4383"393 },394 {395 "dataPath": "params_shard_12.bin",396 "format": "raw-shard",397 "nbytes": 58720256,398 "records": [399 {400 "name": "model.layers.2.mlp.gate_up_proj.q_weight",401 "shape": [402 28672,403 512404 ],405 "dtype": "uint32",406 "format": "f32-to-bf16",407 "nbytes": 58720256,408 "byteOffset": 0409 }410 ],411 "md5sum": "bf8c8655648afeda0c50ce1e3dcea8ca"412 },413 {414 "dataPath": "params_shard_13.bin",415 "format": "raw-shard",416 "nbytes": 22036480,417 "records": [418 {419 "name": "model.layers.1.self_attn.qkv_proj.q_scale",420 "shape": [421 6144,422 128423 ],424 "dtype": "float32",425 "format": "f32-to-bf16",426 "nbytes": 1572864,427 "byteOffset": 0428 },429 {430 "name": "model.layers.1.self_attn.o_proj.q_weight",431 "shape": [432 4096,433 512434 ],435 "dtype": "uint32",436 "format": "f32-to-bf16",437 "nbytes": 8388608,438 "byteOffset": 1572864439 },440 {441 "name": "model.layers.1.self_attn.o_proj.q_scale",442 "shape": [443 4096,444 128445 ],446 "dtype": "float32",447 "format": "f32-to-bf16",448 "nbytes": 1048576,449 "byteOffset": 9961472450 },451 {452 "name": "model.layers.2.input_layernorm.weight",453 "shape": [454 4096455 ],456 "dtype": "float32",457 "format": "f32-to-bf16",458 "nbytes": 8192,459 "byteOffset": 11010048460 },461 {462 "name": "model.layers.2.mlp.down_proj.q_scale",463 "shape": [464 4096,465 448466 ],467 "dtype": "float32",468 "format": "f32-to-bf16",469 "nbytes": 3670016,470 "byteOffset": 11018240471 },472 {473 "name": "model.layers.2.mlp.gate_up_proj.q_scale",474 "shape": [475 28672,476 128477 ],478 "dtype": "float32",479 "format": "f32-to-bf16",480 "nbytes": 7340032,481 "byteOffset": 14688256482 },483 {484 "name": "model.layers.2.post_attention_layernorm.weight",485 "shape": [486 4096487 ],488 "dtype": "float32",489 "format": "f32-to-bf16",490 "nbytes": 8192,491 "byteOffset": 22028288492 }493 ],494 "md5sum": "28434c5e7fe0620dc8841a1db074f624"495 },496 {497 "dataPath": "params_shard_14.bin",498 "format": "raw-shard",499 "nbytes": 29360128,500 "records": [501 {502 "name": "model.layers.3.mlp.down_proj.q_weight",503 "shape": [504 4096,505 1792506 ],507 "dtype": "uint32",508 "format": "f32-to-bf16",509 "nbytes": 29360128,510 "byteOffset": 0511 }512 ],513 "md5sum": "a96d69a1fde5015159acf7dbbb615165"514 },515 {516 "dataPath": "params_shard_15.bin",517 "format": "raw-shard",518 "nbytes": 58720256,519 "records": [520 {521 "name": "model.layers.3.mlp.gate_up_proj.q_weight",522 "shape": [523 28672,524 512525 ],526 "dtype": "uint32",527 "format": "f32-to-bf16",528 "nbytes": 58720256,529 "byteOffset": 0530 }531 ],532 "md5sum": "9b3ea43d6ee8e0ab9f604e71639eecda"533 },534 {535 "dataPath": "params_shard_16.bin",536 "format": "raw-shard",537 "nbytes": 27271168,538 "records": [539 {540 "name": "model.layers.2.self_attn.qkv_proj.q_weight",541 "shape": [542 6144,543 512544 ],545 "dtype": "uint32",546 "format": "f32-to-bf16",547 "nbytes": 12582912,548 "byteOffset": 0549 },550 {551 "name": "model.layers.2.self_attn.qkv_proj.q_scale",552 "shape": [553 6144,554 128555 ],556 "dtype": "float32",557 "format": "f32-to-bf16",558 "nbytes": 1572864,559 "byteOffset": 12582912560 },561 {562 "name": "model.layers.2.self_attn.o_proj.q_weight",563 "shape": [564 4096,565 512566 ],567 "dtype": "uint32",568 "format": "f32-to-bf16",569 "nbytes": 8388608,570 "byteOffset": 14155776571 },572 {573 "name": "model.layers.2.self_attn.o_proj.q_scale",574 "shape": [575 4096,576 128577 ],578 "dtype": "float32",579 "format": "f32-to-bf16",580 "nbytes": 1048576,581 "byteOffset": 22544384582 },583 {584 "name": "model.layers.3.input_layernorm.weight",585 "shape": [586 4096587 ],588 "dtype": "float32",589 "format": "f32-to-bf16",590 "nbytes": 8192,591 "byteOffset": 23592960592 },593 {594 "name": "model.layers.3.mlp.down_proj.q_scale",595 "shape": [596 4096,597 448598 ],599 "dtype": "float32",600 "format": "f32-to-bf16",601 "nbytes": 3670016,602 "byteOffset": 23601152603 }604 ],605 "md5sum": "3914cb8d5875063c825b93bdd7e192f5"606 },607 {608 "dataPath": "params_shard_17.bin",609 "format": "raw-shard",610 "nbytes": 29360128,611 "records": [612 {613 "name": "model.layers.4.mlp.down_proj.q_weight",614 "shape": [615 4096,616 1792617 ],618 "dtype": "uint32",619 "format": "f32-to-bf16",620 "nbytes": 29360128,621 "byteOffset": 0622 }623 ],624 "md5sum": "3937a5985999868216deac5d7a589899"625 },626 {627 "dataPath": "params_shard_18.bin",628 "format": "raw-shard",629 "nbytes": 30949376,630 "records": [631 {632 "name": "model.layers.3.mlp.gate_up_proj.q_scale",633 "shape": [634 28672,635 128636 ],637 "dtype": "float32",638 "format": "f32-to-bf16",639 "nbytes": 7340032,640 "byteOffset": 0641 },642 {643 "name": "model.layers.3.post_attention_layernorm.weight",644 "shape": [645 4096646 ],647 "dtype": "float32",648 "format": "f32-to-bf16",649 "nbytes": 8192,650 "byteOffset": 7340032651 },652 {653 "name": "model.layers.3.self_attn.qkv_proj.q_weight",654 "shape": [655 6144,656 512657 ],658 "dtype": "uint32",659 "format": "f32-to-bf16",660 "nbytes": 12582912,661 "byteOffset": 7348224662 },663 {664 "name": "model.layers.3.self_attn.qkv_proj.q_scale",665 "shape": [666 6144,667 128668 ],669 "dtype": "float32",670 "format": "f32-to-bf16",671 "nbytes": 1572864,672 "byteOffset": 19931136673 },674 {675 "name": "model.layers.3.self_attn.o_proj.q_weight",676 "shape": [677 4096,678 512679 ],680 "dtype": "uint32",681 "format": "f32-to-bf16",682 "nbytes": 8388608,683 "byteOffset": 21504000684 },685 {686 "name": "model.layers.3.self_attn.o_proj.q_scale",687 "shape": [688 4096,689 128690 ],691 "dtype": "float32",692 "format": "f32-to-bf16",693 "nbytes": 1048576,694 "byteOffset": 29892608695 },696 {697 "name": "model.layers.4.input_layernorm.weight",698 "shape": [699 4096700 ],701 "dtype": "float32",702 "format": "f32-to-bf16",703 "nbytes": 8192,704 "byteOffset": 30941184705 }706 ],707 "md5sum": "e58c146e4081f0499094daf482abf01d"708 },709 {710 "dataPath": "params_shard_19.bin",711 "format": "raw-shard",712 "nbytes": 58720256,713 "records": [714 {715 "name": "model.layers.4.mlp.gate_up_proj.q_weight",716 "shape": [717 28672,718 512719 ],720 "dtype": "uint32",721 "format": "f32-to-bf16",722 "nbytes": 58720256,723 "byteOffset": 0724 }725 ],726 "md5sum": "230472af22d815f03b365d8041a0733a"727 },728 {729 "dataPath": "params_shard_20.bin",730 "format": "raw-shard",731 "nbytes": 25174016,732 "records": [733 {734 "name": "model.layers.4.mlp.down_proj.q_scale",735 "shape": [736 4096,737 448738 ],739 "dtype": "float32",740 "format": "f32-to-bf16",741 "nbytes": 3670016,742 "byteOffset": 0743 },744 {745 "name": "model.layers.4.mlp.gate_up_proj.q_scale",746 "shape": [747 28672,748 128749 ],750 "dtype": "float32",751 "format": "f32-to-bf16",752 "nbytes": 7340032,753 "byteOffset": 3670016754 },755 {756 "name": "model.layers.4.post_attention_layernorm.weight",757 "shape": [758 4096759 ],760 "dtype": "float32",761 "format": "f32-to-bf16",762 "nbytes": 8192,763 "byteOffset": 11010048764 },765 {766 "name": "model.layers.4.self_attn.qkv_proj.q_weight",767 "shape": [768 6144,769 512770 ],771 "dtype": "uint32",772 "format": "f32-to-bf16",773 "nbytes": 12582912,774 "byteOffset": 11018240775 },776 {777 "name": "model.layers.4.self_attn.qkv_proj.q_scale",778 "shape": [779 6144,780 128781 ],782 "dtype": "float32",783 "format": "f32-to-bf16",784 "nbytes": 1572864,785 "byteOffset": 23601152786 }787 ],788 "md5sum": "c0bae7d2b6ffdad6d8fda3c47faf3b19"789 },790 {791 "dataPath": "params_shard_21.bin",792 "format": "raw-shard",793 "nbytes": 29360128,794 "records": [795 {796 "name": "model.layers.5.mlp.down_proj.q_weight",797 "shape": [798 4096,799 1792800 ],801 "dtype": "uint32",802 "format": "f32-to-bf16",803 "nbytes": 29360128,804 "byteOffset": 0805 }806 ],807 "md5sum": "75d6eabb9ee443b83d48642c91ca37bc"808 },809 {810 "dataPath": "params_shard_22.bin",811 "format": "raw-shard",812 "nbytes": 58720256,813 "records": [814 {815 "name": "model.layers.5.mlp.gate_up_proj.q_weight",816 "shape": [817 28672,818 512819 ],820 "dtype": "uint32",821 "format": "f32-to-bf16",822 "nbytes": 58720256,823 "byteOffset": 0824 }825 ],826 "md5sum": "dc63346861165d246e48e10bb266688f"827 },828 {829 "dataPath": "params_shard_23.bin",830 "format": "raw-shard",831 "nbytes": 33046528,832 "records": [833 {834 "name": "model.layers.4.self_attn.o_proj.q_weight",835 "shape": [836 4096,837 512838 ],839 "dtype": "uint32",840 "format": "f32-to-bf16",841 "nbytes": 8388608,842 "byteOffset": 0843 },844 {845 "name": "model.layers.4.self_attn.o_proj.q_scale",846 "shape": [847 4096,848 128849 ],850 "dtype": "float32",851 "format": "f32-to-bf16",852 "nbytes": 1048576,853 "byteOffset": 8388608854 },855 {856 "name": "model.layers.5.input_layernorm.weight",857 "shape": [858 4096859 ],860 "dtype": "float32",861 "format": "f32-to-bf16",862 "nbytes": 8192,863 "byteOffset": 9437184864 },865 {866 "name": "model.layers.5.mlp.down_proj.q_scale",867 "shape": [868 4096,869 448870 ],871 "dtype": "float32",872 "format": "f32-to-bf16",873 "nbytes": 3670016,874 "byteOffset": 9445376875 },876 {877 "name": "model.layers.5.mlp.gate_up_proj.q_scale",878 "shape": [879 28672,880 128881 ],882 "dtype": "float32",883 "format": "f32-to-bf16",884 "nbytes": 7340032,885 "byteOffset": 13115392886 },887 {888 "name": "model.layers.5.post_attention_layernorm.weight",889 "shape": [890 4096891 ],892 "dtype": "float32",893 "format": "f32-to-bf16",894 "nbytes": 8192,895 "byteOffset": 20455424896 },897 {898 "name": "model.layers.5.self_attn.qkv_proj.q_weight",899 "shape": [900 6144,901 512902 ],903 "dtype": "uint32",904 "format": "f32-to-bf16",905 "nbytes": 12582912,906 "byteOffset": 20463616907 }908 ],909 "md5sum": "2b5e01e65f534627189776a7c28b4096"910 },911 {912 "dataPath": "params_shard_24.bin",913 "format": "raw-shard",914 "nbytes": 29360128,915 "records": [916 {917 "name": "model.layers.6.mlp.down_proj.q_weight",918 "shape": [919 4096,920 1792921 ],922 "dtype": "uint32",923 "format": "f32-to-bf16",924 "nbytes": 29360128,925 "byteOffset": 0926 }927 ],928 "md5sum": "0b390dee49201283b10aa6d36623dd6d"929 },930 {931 "dataPath": "params_shard_25.bin",932 "format": "raw-shard",933 "nbytes": 58720256,934 "records": [935 {936 "name": "model.layers.6.mlp.gate_up_proj.q_weight",937 "shape": [938 28672,939 512940 ],941 "dtype": "uint32",942 "format": "f32-to-bf16",943 "nbytes": 58720256,944 "byteOffset": 0945 }946 ],947 "md5sum": "37ece762e45c87ee1298b49dc1d03fd6"948 },949 {950 "dataPath": "params_shard_26.bin",951 "format": "raw-shard",952 "nbytes": 22036480,953 "records": [954 {955 "name": "model.layers.5.self_attn.qkv_proj.q_scale",956 "shape": [957 6144,958 128959 ],960 "dtype": "float32",961 "format": "f32-to-bf16",962 "nbytes": 1572864,963 "byteOffset": 0964 },965 {966 "name": "model.layers.5.self_attn.o_proj.q_weight",967 "shape": [968 4096,969 512970 ],971 "dtype": "uint32",972 "format": "f32-to-bf16",973 "nbytes": 8388608,974 "byteOffset": 1572864975 },976 {977 "name": "model.layers.5.self_attn.o_proj.q_scale",978 "shape": [979 4096,980 128981 ],982 "dtype": "float32",983 "format": "f32-to-bf16",984 "nbytes": 1048576,985 "byteOffset": 9961472986 },987 {988 "name": "model.layers.6.input_layernorm.weight",989 "shape": [990 4096991 ],992 "dtype": "float32",993 "format": "f32-to-bf16",994 "nbytes": 8192,995 "byteOffset": 11010048996 },997 {998 "name": "model.layers.6.mlp.down_proj.q_scale",999 "shape": [1000 4096,1001 4481002 ],1003 "dtype": "float32",1004 "format": "f32-to-bf16",1005 "nbytes": 3670016,1006 "byteOffset": 110182401007 },1008 {1009 "name": "model.layers.6.mlp.gate_up_proj.q_scale",1010 "shape": [1011 28672,1012 1281013 ],1014 "dtype": "float32",1015 "format": "f32-to-bf16",1016 "nbytes": 7340032,1017 "byteOffset": 146882561018 },1019 {1020 "name": "model.layers.6.post_attention_layernorm.weight",1021 "shape": [1022 40961023 ],1024 "dtype": "float32",1025 "format": "f32-to-bf16",1026 "nbytes": 8192,1027 "byteOffset": 220282881028 }1029 ],1030 "md5sum": "d1be81ab76d23231f5f1beffc8c889df"1031 },1032 {1033 "dataPath": "params_shard_27.bin",1034 "format": "raw-shard",1035 "nbytes": 29360128,1036 "records": [1037 {1038 "name": "model.layers.7.mlp.down_proj.q_weight",1039 "shape": [1040 4096,1041 17921042 ],1043 "dtype": "uint32",1044 "format": "f32-to-bf16",1045 "nbytes": 29360128,1046 "byteOffset": 01047 }1048 ],1049 "md5sum": "cb8e2a10503eabdb2c0990686f68cffa"1050 },1051 {1052 "dataPath": "params_shard_28.bin",1053 "format": "raw-shard",1054 "nbytes": 58720256,1055 "records": [1056 {1057 "name": "model.layers.7.mlp.gate_up_proj.q_weight",1058 "shape": [1059 28672,1060 5121061 ],1062 "dtype": "uint32",1063 "format": "f32-to-bf16",1064 "nbytes": 58720256,1065 "byteOffset": 01066 }1067 ],1068 "md5sum": "7485661b2b16dc44a95a0b9edd01fd30"1069 },1070 {1071 "dataPath": "params_shard_29.bin",1072 "format": "raw-shard",1073 "nbytes": 27271168,1074 "records": [1075 {1076 "name": "model.layers.6.self_attn.qkv_proj.q_weight",1077 "shape": [1078 6144,1079 5121080 ],1081 "dtype": "uint32",1082 "format": "f32-to-bf16",1083 "nbytes": 12582912,1084 "byteOffset": 01085 },1086 {1087 "name": "model.layers.6.self_attn.qkv_proj.q_scale",1088 "shape": [1089 6144,1090 1281091 ],1092 "dtype": "float32",1093 "format": "f32-to-bf16",1094 "nbytes": 1572864,1095 "byteOffset": 125829121096 },1097 {1098 "name": "model.layers.6.self_attn.o_proj.q_weight",1099 "shape": [1100 4096,1101 5121102 ],1103 "dtype": "uint32",1104 "format": "f32-to-bf16",1105 "nbytes": 8388608,1106 "byteOffset": 141557761107 },1108 {1109 "name": "model.layers.6.self_attn.o_proj.q_scale",1110 "shape": [1111 4096,1112 1281113 ],1114 "dtype": "float32",1115 "format": "f32-to-bf16",1116 "nbytes": 1048576,1117 "byteOffset": 225443841118 },1119 {1120 "name": "model.layers.7.input_layernorm.weight",1121 "shape": [1122 40961123 ],1124 "dtype": "float32",1125 "format": "f32-to-bf16",1126 "nbytes": 8192,1127 "byteOffset": 235929601128 },1129 {1130 "name": "model.layers.7.mlp.down_proj.q_scale",1131 "shape": [1132 4096,1133 4481134 ],1135 "dtype": "float32",1136 "format": "f32-to-bf16",1137 "nbytes": 3670016,1138 "byteOffset": 236011521139 }1140 ],1141 "md5sum": "6d825f646e376ffb653a030910ff2f06"1142 },1143 {1144 "dataPath": "params_shard_30.bin",1145 "format": "raw-shard",1146 "nbytes": 29360128,1147 "records": [1148 {1149 "name": "model.layers.8.mlp.down_proj.q_weight",1150 "shape": [1151 4096,1152 17921153 ],1154 "dtype": "uint32",1155 "format": "f32-to-bf16",1156 "nbytes": 29360128,1157 "byteOffset": 01158 }1159 ],1160 "md5sum": "aad7215080ddae0304b35a341dc511a4"1161 },1162 {1163 "dataPath": "params_shard_31.bin",1164 "format": "raw-shard",1165 "nbytes": 30949376,1166 "records": [1167 {1168 "name": "model.layers.7.mlp.gate_up_proj.q_scale",1169 "shape": [1170 28672,1171 1281172 ],1173 "dtype": "float32",1174 "format": "f32-to-bf16",1175 "nbytes": 7340032,1176 "byteOffset": 01177 },1178 {1179 "name": "model.layers.7.post_attention_layernorm.weight",1180 "shape": [1181 40961182 ],1183 "dtype": "float32",1184 "format": "f32-to-bf16",1185 "nbytes": 8192,1186 "byteOffset": 73400321187 },1188 {1189 "name": "model.layers.7.self_attn.qkv_proj.q_weight",1190 "shape": [1191 6144,1192 5121193 ],1194 "dtype": "uint32",1195 "format": "f32-to-bf16",1196 "nbytes": 12582912,1197 "byteOffset": 73482241198 },1199 {1200 "name": "model.layers.7.self_attn.qkv_proj.q_scale",