logitsml/drowse-web-qwen35-2b-base-candidate
0
1{2 "metadata": {3 "ParamSize": 435,4 "ParamBytes": 1059316480.0,5 "BitsPerParam": 3.7265926272631466 },7 "records": [8 {9 "dataPath": "params_shard_0.bin",10 "format": "raw-shard",11 "nbytes": 254279680,12 "records": [13 {14 "name": "model.embed_tokens.q_weight",15 "shape": [16 248320,17 25618 ],19 "dtype": "uint32",20 "format": "f32-to-bf16",21 "nbytes": 254279680,22 "byteOffset": 023 }24 ],25 "md5sum": "422f68032ef81a1dbd977a771b95e7fd"26 },27 {28 "dataPath": "params_shard_1.bin",29 "format": "raw-shard",30 "nbytes": 31784960,31 "records": [32 {33 "name": "model.embed_tokens.q_scale",34 "shape": [35 248320,36 6437 ],38 "dtype": "float16",39 "format": "f32-to-bf16",40 "nbytes": 31784960,41 "byteOffset": 042 }43 ],44 "md5sum": "214b6ded9f4b085cef1ad15823704172"45 },46 {47 "dataPath": "params_shard_2.bin",48 "format": "raw-shard",49 "nbytes": 28311552,50 "records": [51 {52 "name": "model.layers.11.mlp.down_proj.q_weight",53 "shape": [54 2048,55 76856 ],57 "dtype": "uint32",58 "format": "f32-to-bf16",59 "nbytes": 6291456,60 "byteOffset": 061 },62 {63 "name": "model.layers.11.mlp.down_proj.q_scale",64 "shape": [65 2048,66 19267 ],68 "dtype": "float16",69 "format": "f32-to-bf16",70 "nbytes": 786432,71 "byteOffset": 629145672 },73 {74 "name": "model.layers.11.mlp.gate_up_proj.q_weight",75 "shape": [76 12288,77 25678 ],79 "dtype": "uint32",80 "format": "f32-to-bf16",81 "nbytes": 12582912,82 "byteOffset": 707788883 },84 {85 "name": "model.layers.11.mlp.gate_up_proj.q_scale",86 "shape": [87 12288,88 6489 ],90 "dtype": "float16",91 "format": "f32-to-bf16",92 "nbytes": 1572864,93 "byteOffset": 1966080094 },95 {96 "name": "model.layers.1.mlp.down_proj.q_weight",97 "shape": [98 2048,99 768100 ],101 "dtype": "uint32",102 "format": "f32-to-bf16",103 "nbytes": 6291456,104 "byteOffset": 21233664105 },106 {107 "name": "model.layers.1.mlp.down_proj.q_scale",108 "shape": [109 2048,110 192111 ],112 "dtype": "float16",113 "format": "f32-to-bf16",114 "nbytes": 786432,115 "byteOffset": 27525120116 }117 ],118 "md5sum": "0da602feee28af6401a6aaa2c77eace3"119 },120 {121 "dataPath": "params_shard_3.bin",122 "format": "raw-shard",123 "nbytes": 21233664,124 "records": [125 {126 "name": "model.layers.1.mlp.gate_up_proj.q_weight",127 "shape": [128 12288,129 256130 ],131 "dtype": "uint32",132 "format": "f32-to-bf16",133 "nbytes": 12582912,134 "byteOffset": 0135 },136 {137 "name": "model.layers.1.mlp.gate_up_proj.q_scale",138 "shape": [139 12288,140 64141 ],142 "dtype": "float16",143 "format": "f32-to-bf16",144 "nbytes": 1572864,145 "byteOffset": 12582912146 },147 {148 "name": "model.layers.12.mlp.down_proj.q_weight",149 "shape": [150 2048,151 768152 ],153 "dtype": "uint32",154 "format": "f32-to-bf16",155 "nbytes": 6291456,156 "byteOffset": 14155776157 },158 {159 "name": "model.layers.12.mlp.down_proj.q_scale",160 "shape": [161 2048,162 192163 ],164 "dtype": "float16",165 "format": "f32-to-bf16",166 "nbytes": 786432,167 "byteOffset": 20447232168 }169 ],170 "md5sum": "8207c7df5e01917f8c83d8c4fc2947c0"171 },172 {173 "dataPath": "params_shard_4.bin",174 "format": "raw-shard",175 "nbytes": 21233664,176 "records": [177 {178 "name": "model.layers.12.mlp.gate_up_proj.q_weight",179 "shape": [180 12288,181 256182 ],183 "dtype": "uint32",184 "format": "f32-to-bf16",185 "nbytes": 12582912,186 "byteOffset": 0187 },188 {189 "name": "model.layers.12.mlp.gate_up_proj.q_scale",190 "shape": [191 12288,192 64193 ],194 "dtype": "float16",195 "format": "f32-to-bf16",196 "nbytes": 1572864,197 "byteOffset": 12582912198 },199 {200 "name": "model.layers.16.mlp.down_proj.q_weight",201 "shape": [202 2048,203 768204 ],205 "dtype": "uint32",206 "format": "f32-to-bf16",207 "nbytes": 6291456,208 "byteOffset": 14155776209 },210 {211 "name": "model.layers.16.mlp.down_proj.q_scale",212 "shape": [213 2048,214 192215 ],216 "dtype": "float16",217 "format": "f32-to-bf16",218 "nbytes": 786432,219 "byteOffset": 20447232220 }221 ],222 "md5sum": "3e184bac76a2c9fce27ede8a832fdefc"223 },224 {225 "dataPath": "params_shard_5.bin",226 "format": "raw-shard",227 "nbytes": 21233664,228 "records": [229 {230 "name": "model.layers.16.mlp.gate_up_proj.q_weight",231 "shape": [232 12288,233 256234 ],235 "dtype": "uint32",236 "format": "f32-to-bf16",237 "nbytes": 12582912,238 "byteOffset": 0239 },240 {241 "name": "model.layers.16.mlp.gate_up_proj.q_scale",242 "shape": [243 12288,244 64245 ],246 "dtype": "float16",247 "format": "f32-to-bf16",248 "nbytes": 1572864,249 "byteOffset": 12582912250 },251 {252 "name": "model.layers.15.mlp.down_proj.q_weight",253 "shape": [254 2048,255 768256 ],257 "dtype": "uint32",258 "format": "f32-to-bf16",259 "nbytes": 6291456,260 "byteOffset": 14155776261 },262 {263 "name": "model.layers.15.mlp.down_proj.q_scale",264 "shape": [265 2048,266 192267 ],268 "dtype": "float16",269 "format": "f32-to-bf16",270 "nbytes": 786432,271 "byteOffset": 20447232272 }273 ],274 "md5sum": "609567a03ecb7ee20e40a047a5662d21"275 },276 {277 "dataPath": "params_shard_6.bin",278 "format": "raw-shard",279 "nbytes": 21233664,280 "records": [281 {282 "name": "model.layers.15.mlp.gate_up_proj.q_weight",283 "shape": [284 12288,285 256286 ],287 "dtype": "uint32",288 "format": "f32-to-bf16",289 "nbytes": 12582912,290 "byteOffset": 0291 },292 {293 "name": "model.layers.15.mlp.gate_up_proj.q_scale",294 "shape": [295 12288,296 64297 ],298 "dtype": "float16",299 "format": "f32-to-bf16",300 "nbytes": 1572864,301 "byteOffset": 12582912302 },303 {304 "name": "model.layers.18.mlp.down_proj.q_weight",305 "shape": [306 2048,307 768308 ],309 "dtype": "uint32",310 "format": "f32-to-bf16",311 "nbytes": 6291456,312 "byteOffset": 14155776313 },314 {315 "name": "model.layers.18.mlp.down_proj.q_scale",316 "shape": [317 2048,318 192319 ],320 "dtype": "float16",321 "format": "f32-to-bf16",322 "nbytes": 786432,323 "byteOffset": 20447232324 }325 ],326 "md5sum": "9e1f9a2daec0814fe95064f3b6dba6d2"327 },328 {329 "dataPath": "params_shard_7.bin",330 "format": "raw-shard",331 "nbytes": 28311552,332 "records": [333 {334 "name": "model.layers.18.mlp.gate_up_proj.q_weight",335 "shape": [336 12288,337 256338 ],339 "dtype": "uint32",340 "format": "f32-to-bf16",341 "nbytes": 12582912,342 "byteOffset": 0343 },344 {345 "name": "model.layers.18.mlp.gate_up_proj.q_scale",346 "shape": [347 12288,348 64349 ],350 "dtype": "float16",351 "format": "f32-to-bf16",352 "nbytes": 1572864,353 "byteOffset": 12582912354 },355 {356 "name": "model.layers.19.mlp.gate_up_proj.q_weight",357 "shape": [358 12288,359 256360 ],361 "dtype": "uint32",362 "format": "f32-to-bf16",363 "nbytes": 12582912,364 "byteOffset": 14155776365 },366 {367 "name": "model.layers.19.mlp.gate_up_proj.q_scale",368 "shape": [369 12288,370 64371 ],372 "dtype": "float16",373 "format": "f32-to-bf16",374 "nbytes": 1572864,375 "byteOffset": 26738688376 }377 ],378 "md5sum": "d40abda00657776b6d27fd1141d2c7f2"379 },380 {381 "dataPath": "params_shard_8.bin",382 "format": "raw-shard",383 "nbytes": 28311552,384 "records": [385 {386 "name": "model.layers.0.linear_attn.in_proj_qkv.q_weight",387 "shape": [388 6144,389 256390 ],391 "dtype": "uint32",392 "format": "f32-to-bf16",393 "nbytes": 6291456,394 "byteOffset": 0395 },396 {397 "name": "model.layers.0.linear_attn.in_proj_qkv.q_scale",398 "shape": [399 6144,400 64401 ],402 "dtype": "float16",403 "format": "f32-to-bf16",404 "nbytes": 786432,405 "byteOffset": 6291456406 },407 {408 "name": "model.layers.0.mlp.down_proj.q_weight",409 "shape": [410 2048,411 768412 ],413 "dtype": "uint32",414 "format": "f32-to-bf16",415 "nbytes": 6291456,416 "byteOffset": 7077888417 },418 {419 "name": "model.layers.0.mlp.down_proj.q_scale",420 "shape": [421 2048,422 192423 ],424 "dtype": "float16",425 "format": "f32-to-bf16",426 "nbytes": 786432,427 "byteOffset": 13369344428 },429 {430 "name": "model.layers.0.mlp.gate_up_proj.q_weight",431 "shape": [432 12288,433 256434 ],435 "dtype": "uint32",436 "format": "f32-to-bf16",437 "nbytes": 12582912,438 "byteOffset": 14155776439 },440 {441 "name": "model.layers.0.mlp.gate_up_proj.q_scale",442 "shape": [443 12288,444 64445 ],446 "dtype": "float16",447 "format": "f32-to-bf16",448 "nbytes": 1572864,449 "byteOffset": 26738688450 }451 ],452 "md5sum": "607b9fd78c8486f639e78a463b753773"453 },454 {455 "dataPath": "params_shard_9.bin",456 "format": "raw-shard",457 "nbytes": 28311552,458 "records": [459 {460 "name": "model.layers.13.mlp.down_proj.q_weight",461 "shape": [462 2048,463 768464 ],465 "dtype": "uint32",466 "format": "f32-to-bf16",467 "nbytes": 6291456,468 "byteOffset": 0469 },470 {471 "name": "model.layers.13.mlp.down_proj.q_scale",472 "shape": [473 2048,474 192475 ],476 "dtype": "float16",477 "format": "f32-to-bf16",478 "nbytes": 786432,479 "byteOffset": 6291456480 },481 {482 "name": "model.layers.13.mlp.gate_up_proj.q_weight",483 "shape": [484 12288,485 256486 ],487 "dtype": "uint32",488 "format": "f32-to-bf16",489 "nbytes": 12582912,490 "byteOffset": 7077888491 },492 {493 "name": "model.layers.13.mlp.gate_up_proj.q_scale",494 "shape": [495 12288,496 64497 ],498 "dtype": "float16",499 "format": "f32-to-bf16",500 "nbytes": 1572864,501 "byteOffset": 19660800502 },503 {504 "name": "model.layers.1.linear_attn.in_proj_qkv.q_weight",505 "shape": [506 6144,507 256508 ],509 "dtype": "uint32",510 "format": "f32-to-bf16",511 "nbytes": 6291456,512 "byteOffset": 21233664513 },514 {515 "name": "model.layers.1.linear_attn.in_proj_qkv.q_scale",516 "shape": [517 6144,518 64519 ],520 "dtype": "float16",521 "format": "f32-to-bf16",522 "nbytes": 786432,523 "byteOffset": 27525120524 }525 ],526 "md5sum": "365ce463b6e863128da5766fa2fa464a"527 },528 {529 "dataPath": "params_shard_10.bin",530 "format": "raw-shard",531 "nbytes": 28311552,532 "records": [533 {534 "name": "model.layers.10.linear_attn.in_proj_qkv.q_weight",535 "shape": [536 6144,537 256538 ],539 "dtype": "uint32",540 "format": "f32-to-bf16",541 "nbytes": 6291456,542 "byteOffset": 0543 },544 {545 "name": "model.layers.10.linear_attn.in_proj_qkv.q_scale",546 "shape": [547 6144,548 64549 ],550 "dtype": "float16",551 "format": "f32-to-bf16",552 "nbytes": 786432,553 "byteOffset": 6291456554 },555 {556 "name": "model.layers.10.mlp.down_proj.q_weight",557 "shape": [558 2048,559 768560 ],561 "dtype": "uint32",562 "format": "f32-to-bf16",563 "nbytes": 6291456,564 "byteOffset": 7077888565 },566 {567 "name": "model.layers.10.mlp.down_proj.q_scale",568 "shape": [569 2048,570 192571 ],572 "dtype": "float16",573 "format": "f32-to-bf16",574 "nbytes": 786432,575 "byteOffset": 13369344576 },577 {578 "name": "model.layers.10.mlp.gate_up_proj.q_weight",579 "shape": [580 12288,581 256582 ],583 "dtype": "uint32",584 "format": "f32-to-bf16",585 "nbytes": 12582912,586 "byteOffset": 14155776587 },588 {589 "name": "model.layers.10.mlp.gate_up_proj.q_scale",590 "shape": [591 12288,592 64593 ],594 "dtype": "float16",595 "format": "f32-to-bf16",596 "nbytes": 1572864,597 "byteOffset": 26738688598 }599 ],600 "md5sum": "ec488164724c8e24d3daf8704d27e56b"601 },602 {603 "dataPath": "params_shard_11.bin",604 "format": "raw-shard",605 "nbytes": 28311552,606 "records": [607 {608 "name": "model.layers.14.linear_attn.in_proj_qkv.q_weight",609 "shape": [610 6144,611 256612 ],613 "dtype": "uint32",614 "format": "f32-to-bf16",615 "nbytes": 6291456,616 "byteOffset": 0617 },618 {619 "name": "model.layers.14.linear_attn.in_proj_qkv.q_scale",620 "shape": [621 6144,622 64623 ],624 "dtype": "float16",625 "format": "f32-to-bf16",626 "nbytes": 786432,627 "byteOffset": 6291456628 },629 {630 "name": "model.layers.14.mlp.down_proj.q_weight",631 "shape": [632 2048,633 768634 ],635 "dtype": "uint32",636 "format": "f32-to-bf16",637 "nbytes": 6291456,638 "byteOffset": 7077888639 },640 {641 "name": "model.layers.14.mlp.down_proj.q_scale",642 "shape": [643 2048,644 192645 ],646 "dtype": "float16",647 "format": "f32-to-bf16",648 "nbytes": 786432,649 "byteOffset": 13369344650 },651 {652 "name": "model.layers.14.mlp.gate_up_proj.q_weight",653 "shape": [654 12288,655 256656 ],657 "dtype": "uint32",658 "format": "f32-to-bf16",659 "nbytes": 12582912,660 "byteOffset": 14155776661 },662 {663 "name": "model.layers.14.mlp.gate_up_proj.q_scale",664 "shape": [665 12288,666 64667 ],668 "dtype": "float16",669 "format": "f32-to-bf16",670 "nbytes": 1572864,671 "byteOffset": 26738688672 }673 ],674 "md5sum": "d8120f1fb1d0b91ac4370be5f0a226f1"675 },676 {677 "dataPath": "params_shard_12.bin",678 "format": "raw-shard",679 "nbytes": 28311552,680 "records": [681 {682 "name": "model.layers.19.mlp.down_proj.q_weight",683 "shape": [684 2048,685 768686 ],687 "dtype": "uint32",688 "format": "f32-to-bf16",689 "nbytes": 6291456,690 "byteOffset": 0691 },692 {693 "name": "model.layers.19.mlp.down_proj.q_scale",694 "shape": [695 2048,696 192697 ],698 "dtype": "float16",699 "format": "f32-to-bf16",700 "nbytes": 786432,701 "byteOffset": 6291456702 },703 {704 "name": "model.layers.2.mlp.down_proj.q_weight",705 "shape": [706 2048,707 768708 ],709 "dtype": "uint32",710 "format": "f32-to-bf16",711 "nbytes": 6291456,712 "byteOffset": 7077888713 },714 {715 "name": "model.layers.2.mlp.down_proj.q_scale",716 "shape": [717 2048,718 192719 ],720 "dtype": "float16",721 "format": "f32-to-bf16",722 "nbytes": 786432,723 "byteOffset": 13369344724 },725 {726 "name": "model.layers.2.mlp.gate_up_proj.q_weight",727 "shape": [728 12288,729 256730 ],731 "dtype": "uint32",732 "format": "f32-to-bf16",733 "nbytes": 12582912,734 "byteOffset": 14155776735 },736 {737 "name": "model.layers.2.mlp.gate_up_proj.q_scale",738 "shape": [739 12288,740 64741 ],742 "dtype": "float16",743 "format": "f32-to-bf16",744 "nbytes": 1572864,745 "byteOffset": 26738688746 }747 ],748 "md5sum": "b3ae8a11f6503536aa410cb7d7c8ddc3"749 },750 {751 "dataPath": "params_shard_13.bin",752 "format": "raw-shard",753 "nbytes": 28311552,754 "records": [755 {756 "name": "model.layers.13.linear_attn.in_proj_qkv.q_weight",757 "shape": [758 6144,759 256760 ],761 "dtype": "uint32",762 "format": "f32-to-bf16",763 "nbytes": 6291456,764 "byteOffset": 0765 },766 {767 "name": "model.layers.13.linear_attn.in_proj_qkv.q_scale",768 "shape": [769 6144,770 64771 ],772 "dtype": "float16",773 "format": "f32-to-bf16",774 "nbytes": 786432,775 "byteOffset": 6291456776 },777 {778 "name": "model.layers.21.mlp.down_proj.q_weight",779 "shape": [780 2048,781 768782 ],783 "dtype": "uint32",784 "format": "f32-to-bf16",785 "nbytes": 6291456,786 "byteOffset": 7077888787 },788 {789 "name": "model.layers.21.mlp.down_proj.q_scale",790 "shape": [791 2048,792 192793 ],794 "dtype": "float16",795 "format": "f32-to-bf16",796 "nbytes": 786432,797 "byteOffset": 13369344798 },799 {800 "name": "model.layers.21.mlp.gate_up_proj.q_weight",801 "shape": [802 12288,803 256804 ],805 "dtype": "uint32",806 "format": "f32-to-bf16",807 "nbytes": 12582912,808 "byteOffset": 14155776809 },810 {811 "name": "model.layers.21.mlp.gate_up_proj.q_scale",812 "shape": [813 12288,814 64815 ],816 "dtype": "float16",817 "format": "f32-to-bf16",818 "nbytes": 1572864,819 "byteOffset": 26738688820 }821 ],822 "md5sum": "b6ba73a9385b3680ac682ca116e438a4"823 },824 {825 "dataPath": "params_shard_14.bin",826 "format": "raw-shard",827 "nbytes": 28311552,828 "records": [829 {830 "name": "model.layers.22.mlp.down_proj.q_weight",831 "shape": [832 2048,833 768834 ],835 "dtype": "uint32",836 "format": "f32-to-bf16",837 "nbytes": 6291456,838 "byteOffset": 0839 },840 {841 "name": "model.layers.22.mlp.down_proj.q_scale",842 "shape": [843 2048,844 192845 ],846 "dtype": "float16",847 "format": "f32-to-bf16",848 "nbytes": 786432,849 "byteOffset": 6291456850 },851 {852 "name": "model.layers.22.mlp.gate_up_proj.q_weight",853 "shape": [854 12288,855 256856 ],857 "dtype": "uint32",858 "format": "f32-to-bf16",859 "nbytes": 12582912,860 "byteOffset": 7077888861 },862 {863 "name": "model.layers.22.mlp.gate_up_proj.q_scale",864 "shape": [865 12288,866 64867 ],868 "dtype": "float16",869 "format": "f32-to-bf16",870 "nbytes": 1572864,871 "byteOffset": 19660800872 },873 {874 "name": "model.layers.23.mlp.down_proj.q_weight",875 "shape": [876 2048,877 768878 ],879 "dtype": "uint32",880 "format": "f32-to-bf16",881 "nbytes": 6291456,882 "byteOffset": 21233664883 },884 {885 "name": "model.layers.23.mlp.down_proj.q_scale",886 "shape": [887 2048,888 192889 ],890 "dtype": "float16",891 "format": "f32-to-bf16",892 "nbytes": 786432,893 "byteOffset": 27525120894 }895 ],896 "md5sum": "37172bfaa484e56c22e3f86e9ec24a0e"897 },898 {899 "dataPath": "params_shard_15.bin",900 "format": "raw-shard",901 "nbytes": 28311552,902 "records": [903 {904 "name": "model.layers.23.mlp.gate_up_proj.q_weight",905 "shape": [906 12288,907 256908 ],909 "dtype": "uint32",910 "format": "f32-to-bf16",911 "nbytes": 12582912,912 "byteOffset": 0913 },914 {915 "name": "model.layers.23.mlp.gate_up_proj.q_scale",916 "shape": [917 12288,918 64919 ],920 "dtype": "float16",921 "format": "f32-to-bf16",922 "nbytes": 1572864,923 "byteOffset": 12582912924 },925 {926 "name": "model.layers.12.linear_attn.in_proj_qkv.q_weight",927 "shape": [928 6144,929 256930 ],931 "dtype": "uint32",932 "format": "f32-to-bf16",933 "nbytes": 6291456,934 "byteOffset": 14155776935 },936 {937 "name": "model.layers.12.linear_attn.in_proj_qkv.q_scale",938 "shape": [939 6144,940 64941 ],942 "dtype": "float16",943 "format": "f32-to-bf16",944 "nbytes": 786432,945 "byteOffset": 20447232946 },947 {948 "name": "model.layers.17.mlp.down_proj.q_weight",949 "shape": [950 2048,951 768952 ],953 "dtype": "uint32",954 "format": "f32-to-bf16",955 "nbytes": 6291456,956 "byteOffset": 21233664957 },958 {959 "name": "model.layers.17.mlp.down_proj.q_scale",960 "shape": [961 2048,962 192963 ],964 "dtype": "float16",965 "format": "f32-to-bf16",966 "nbytes": 786432,967 "byteOffset": 27525120968 }969 ],970 "md5sum": "4b94993a286c3778ae8a6e34c8e32ae0"971 },972 {973 "dataPath": "params_shard_16.bin",974 "format": "raw-shard",975 "nbytes": 21233664,976 "records": [977 {978 "name": "model.layers.18.linear_attn.in_proj_qkv.q_weight",979 "shape": [980 6144,981 256982 ],983 "dtype": "uint32",984 "format": "f32-to-bf16",985 "nbytes": 6291456,986 "byteOffset": 0987 },988 {989 "name": "model.layers.18.linear_attn.in_proj_qkv.q_scale",990 "shape": [991 6144,992 64993 ],994 "dtype": "float16",995 "format": "f32-to-bf16",996 "nbytes": 786432,997 "byteOffset": 6291456998 },999 {1000 "name": "model.layers.16.linear_attn.in_proj_qkv.q_weight",1001 "shape": [1002 6144,1003 2561004 ],1005 "dtype": "uint32",1006 "format": "f32-to-bf16",1007 "nbytes": 6291456,1008 "byteOffset": 70778881009 },1010 {1011 "name": "model.layers.16.linear_attn.in_proj_qkv.q_scale",1012 "shape": [1013 6144,1014 641015 ],1016 "dtype": "float16",1017 "format": "f32-to-bf16",1018 "nbytes": 786432,1019 "byteOffset": 133693441020 },1021 {1022 "name": "model.layers.5.mlp.down_proj.q_weight",1023 "shape": [1024 2048,1025 7681026 ],1027 "dtype": "uint32",1028 "format": "f32-to-bf16",1029 "nbytes": 6291456,1030 "byteOffset": 141557761031 },1032 {1033 "name": "model.layers.5.mlp.down_proj.q_scale",1034 "shape": [1035 2048,1036 1921037 ],1038 "dtype": "float16",1039 "format": "f32-to-bf16",1040 "nbytes": 786432,1041 "byteOffset": 204472321042 }1043 ],1044 "md5sum": "b1e949707e48ce64cbfd714ba796d316"1045 },1046 {1047 "dataPath": "params_shard_17.bin",1048 "format": "raw-shard",1049 "nbytes": 28311552,1050 "records": [1051 {1052 "name": "model.layers.5.mlp.gate_up_proj.q_weight",1053 "shape": [1054 12288,1055 2561056 ],1057 "dtype": "uint32",1058 "format": "f32-to-bf16",1059 "nbytes": 12582912,1060 "byteOffset": 01061 },1062 {1063 "name": "model.layers.5.mlp.gate_up_proj.q_scale",1064 "shape": [1065 12288,1066 641067 ],1068 "dtype": "float16",1069 "format": "f32-to-bf16",1070 "nbytes": 1572864,1071 "byteOffset": 125829121072 },1073 {1074 "name": "model.layers.20.linear_attn.in_proj_qkv.q_weight",1075 "shape": [1076 6144,1077 2561078 ],1079 "dtype": "uint32",1080 "format": "f32-to-bf16",1081 "nbytes": 6291456,1082 "byteOffset": 141557761083 },1084 {1085 "name": "model.layers.20.linear_attn.in_proj_qkv.q_scale",1086 "shape": [1087 6144,1088 641089 ],1090 "dtype": "float16",1091 "format": "f32-to-bf16",1092 "nbytes": 786432,1093 "byteOffset": 204472321094 },1095 {1096 "name": "model.layers.20.mlp.down_proj.q_weight",1097 "shape": [1098 2048,1099 7681100 ],1101 "dtype": "uint32",1102 "format": "f32-to-bf16",1103 "nbytes": 6291456,1104 "byteOffset": 212336641105 },1106 {1107 "name": "model.layers.20.mlp.down_proj.q_scale",1108 "shape": [1109 2048,1110 1921111 ],1112 "dtype": "float16",1113 "format": "f32-to-bf16",1114 "nbytes": 786432,1115 "byteOffset": 275251201116 }1117 ],1118 "md5sum": "4d98409431fe2f6e9b3a053ccb99d350"1119 },1120 {1121 "dataPath": "params_shard_18.bin",1122 "format": "raw-shard",1123 "nbytes": 21233664,1124 "records": [1125 {1126 "name": "model.layers.20.mlp.gate_up_proj.q_weight",1127 "shape": [1128 12288,1129 2561130 ],1131 "dtype": "uint32",1132 "format": "f32-to-bf16",1133 "nbytes": 12582912,1134 "byteOffset": 01135 },1136 {1137 "name": "model.layers.20.mlp.gate_up_proj.q_scale",1138 "shape": [1139 12288,1140 641141 ],1142 "dtype": "float16",1143 "format": "f32-to-bf16",1144 "nbytes": 1572864,1145 "byteOffset": 125829121146 },1147 {1148 "name": "model.layers.3.mlp.down_proj.q_weight",1149 "shape": [1150 2048,1151 7681152 ],1153 "dtype": "uint32",1154 "format": "f32-to-bf16",1155 "nbytes": 6291456,1156 "byteOffset": 141557761157 },1158 {1159 "name": "model.layers.3.mlp.down_proj.q_scale",1160 "shape": [1161 2048,1162 1921163 ],1164 "dtype": "float16",1165 "format": "f32-to-bf16",1166 "nbytes": 786432,1167 "byteOffset": 204472321168 }1169 ],1170 "md5sum": "77e193769cabac297b88f2fc55227cdc"1171 },1172 {1173 "dataPath": "params_shard_19.bin",1174 "format": "raw-shard",1175 "nbytes": 28311552,1176 "records": [1177 {1178 "name": "model.layers.3.mlp.gate_up_proj.q_weight",1179 "shape": [1180 12288,1181 2561182 ],1183 "dtype": "uint32",1184 "format": "f32-to-bf16",1185 "nbytes": 12582912,1186 "byteOffset": 01187 },1188 {1189 "name": "model.layers.3.mlp.gate_up_proj.q_scale",1190 "shape": [1191 12288,1192 641193 ],1194 "dtype": "float16",1195 "format": "f32-to-bf16",1196 "nbytes": 1572864,1197 "byteOffset": 125829121198 },1199 {1200 "name": "model.layers.21.linear_attn.in_proj_qkv.q_weight",