jondurbin commited on
Commit
9ff7c62
·
verified ·
1 Parent(s): 5b45690

Add export 165B-tokens_step15861 (step 15861, 165221498880 tokens)

Browse files
Files changed (34) hide show
  1. .gitattributes +1 -0
  2. exports/165B-tokens_step15861/SHA256SUMS +0 -0
  3. exports/165B-tokens_step15861/coverage.json +0 -0
  4. exports/165B-tokens_step15861/export_info.json +24 -0
  5. exports/165B-tokens_step15861/indexer.bundle +3 -0
  6. exports/165B-tokens_step15861/layouts.json +1 -0
  7. exports/165B-tokens_step15861/manifest.json +3 -0
  8. exports/165B-tokens_step15861/model_config.json +1 -0
  9. exports/165B-tokens_step15861/packed_experts.tar +3 -0
  10. exports/165B-tokens_step15861/relay_pack/manifest.json +0 -0
  11. exports/165B-tokens_step15861/relay_pack/trunk/f0.bin +3 -0
  12. exports/165B-tokens_step15861/relay_pack/trunk/f1.bin +3 -0
  13. exports/165B-tokens_step15861/relay_pack/trunk/f10.bin +3 -0
  14. exports/165B-tokens_step15861/relay_pack/trunk/f11.bin +3 -0
  15. exports/165B-tokens_step15861/relay_pack/trunk/f12.bin +3 -0
  16. exports/165B-tokens_step15861/relay_pack/trunk/f13.bin +3 -0
  17. exports/165B-tokens_step15861/relay_pack/trunk/f14.bin +3 -0
  18. exports/165B-tokens_step15861/relay_pack/trunk/f15.bin +3 -0
  19. exports/165B-tokens_step15861/relay_pack/trunk/f16.bin +3 -0
  20. exports/165B-tokens_step15861/relay_pack/trunk/f17.bin +3 -0
  21. exports/165B-tokens_step15861/relay_pack/trunk/f18.bin +3 -0
  22. exports/165B-tokens_step15861/relay_pack/trunk/f19.bin +3 -0
  23. exports/165B-tokens_step15861/relay_pack/trunk/f2.bin +3 -0
  24. exports/165B-tokens_step15861/relay_pack/trunk/f20.bin +3 -0
  25. exports/165B-tokens_step15861/relay_pack/trunk/f21.bin +3 -0
  26. exports/165B-tokens_step15861/relay_pack/trunk/f22.bin +3 -0
  27. exports/165B-tokens_step15861/relay_pack/trunk/f23.bin +3 -0
  28. exports/165B-tokens_step15861/relay_pack/trunk/f3.bin +3 -0
  29. exports/165B-tokens_step15861/relay_pack/trunk/f4.bin +3 -0
  30. exports/165B-tokens_step15861/relay_pack/trunk/f5.bin +3 -0
  31. exports/165B-tokens_step15861/relay_pack/trunk/f6.bin +3 -0
  32. exports/165B-tokens_step15861/relay_pack/trunk/f7.bin +3 -0
  33. exports/165B-tokens_step15861/relay_pack/trunk/f8.bin +3 -0
  34. exports/165B-tokens_step15861/relay_pack/trunk/f9.bin +3 -0
.gitattributes CHANGED
@@ -79,3 +79,4 @@ exports/150B-tokens_step14880/manifest.json filter=lfs diff=lfs merge=lfs -text
79
  exports/155B-tokens_step15179/manifest.json filter=lfs diff=lfs merge=lfs -text
80
  exports/157B-tokens_step15384/manifest.json filter=lfs diff=lfs merge=lfs -text
81
  exports/162B-tokens_step15667/manifest.json filter=lfs diff=lfs merge=lfs -text
 
 
79
  exports/155B-tokens_step15179/manifest.json filter=lfs diff=lfs merge=lfs -text
80
  exports/157B-tokens_step15384/manifest.json filter=lfs diff=lfs merge=lfs -text
81
  exports/162B-tokens_step15667/manifest.json filter=lfs diff=lfs merge=lfs -text
82
+ exports/165B-tokens_step15861/manifest.json filter=lfs diff=lfs merge=lfs -text
exports/165B-tokens_step15861/SHA256SUMS ADDED
The diff for this file is too large to render. See raw diff
 
exports/165B-tokens_step15861/coverage.json ADDED
The diff for this file is too large to render. See raw diff
 
exports/165B-tokens_step15861/export_info.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bytes": 2451934703,
3
+ "digests_sha256": "7d3630976c919d47d700ae27dd23dfaed1344ce58fbbc709cad29efd90f2a5c2",
4
+ "expert_files_in_tar": 4096,
5
+ "export": "model-a5d05fe5c4e9c7010ce3c2fd",
6
+ "files": 4126,
7
+ "format": "Parallax native compact export (mesh-relay-compact-export:1), inference only",
8
+ "generation": "51fb20261004b140",
9
+ "logit_scale": {
10
+ "above_bound": true,
11
+ "effective": 3.0000000000000004,
12
+ "log_raw": 1.1015625,
13
+ "max": 3.0,
14
+ "raw": 3.0088637024644465
15
+ },
16
+ "manifest_sha256": "2645379a93ed54062a38d17002fdd64d9311c05c8df00f458d47f0b75213327b",
17
+ "packed_experts_tar_sha256": "1b3c8a7217efd8d1656f5d7c3334457c729de26e26574c810e8fd83f371cc1d1",
18
+ "run_id": "p8b-51fb20261004b140",
19
+ "saved_at": 1791184251,
20
+ "saved_at_utc": "2026-10-05T07:10:51Z",
21
+ "step": 15861,
22
+ "tokens": 165221498880,
23
+ "unpack": "tar -xf packed_experts.tar && sha256sum -c SHA256SUMS"
24
+ }
exports/165B-tokens_step15861/indexer.bundle ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:de59bd44402befac48ba73f79fad0c194fc9df423344fd96769cd87c8943ad85
3
+ size 10175800
exports/165B-tokens_step15861/layouts.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"0":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":0,"key_order":["embed.weight::last[0:48]","layers.0.o_proj.weight","layers.1.shared_experts.0.down_proj.weight","layers.14._fast_impl.f_proj.0.weight","layers.14._fast_impl.k_proj.weight","layers.17.latent_down.weight","layers.2._fast_impl.g_proj.0.weight","layers.22._fast_impl.b_proj.weight","layers.30._fast_impl.k_proj.weight","layers.31.shared_experts.0.down_proj.weight","layers.32._fast_impl.f_proj.0.weight","layers.39.latent_down.weight","layers.44._fast_impl.v_proj.weight","layers.46._fast_impl.k_proj.weight","layers.52._fast_impl.f_proj.0.weight","layers.52._fast_impl.g_proj.0.weight","layers.53.shared_experts.0.down_proj.weight","layers.54._fast_impl.b_proj.weight","layers.61.latent_down.weight"],"numel":22753536,"offsets":[0,6162432,7342080,9996288,10162176,11489280,11931648,12079104,12089472,13416576,16070784,16236672,16679040,18006144,19333248,19499136,19646592,22300800,22311168],"shapes":[[128384,48],[1152,1024],[1152,2304],[144,1152],[1152,1152],[384,1152],[128,1152],[9,1152],[1152,1152],[1152,2304],[144,1152],[384,1152],[1152,1152],[1152,1152],[144,1152],[128,1152],[1152,2304],[9,1152],[384,1152]]},"1":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":1,"key_order":["embed.weight::last[48:96]","layers.0.q_proj.weight","layers.1.shared_experts.0.up_proj.weight","layers.14._fast_impl.f_proj.1.weight","layers.14._fast_impl.o_proj.weight","layers.17.latent_up.weight","layers.2._fast_impl.g_proj.1.weight","layers.22._fast_impl.c_proj.weight","layers.30._fast_impl.o_proj.weight","layers.31.shared_experts.0.up_proj.weight","layers.32._fast_impl.f_proj.1.weight","layers.39.latent_up.weight","layers.46._fast_impl.o_proj.weight","layers.48.q_proj.weight","layers.52._fast_impl.f_proj.1.weight","layers.52._fast_impl.g_proj.1.weight","layers.53.shared_experts.0.up_proj.weight","layers.54._fast_impl.c_proj.weight","layers.61.latent_up.weight"],"numel":23785728,"offsets":[0,6162432,7342080,9996288,10162176,11489280,11931648,12079104,12089472,13416576,16070784,16236672,16679040,18006144,20365440,20531328,20678784,23332992,23343360],"shapes":[[128384,48],[1024,1152],[2304,1152],[1152,144],[1152,1152],[1152,384],[1152,128],[9,1152],[1152,1152],[2304,1152],[1152,144],[1152,384],[1152,1152],[2048,1152],[1152,144],[1152,128],[2304,1152],[9,1152],[1152,384]]},"10":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":10,"key_order":["embed.weight::last[480:528]","layers.12.k_proj.weight","layers.14._fast_impl.b_proj.weight","layers.18.k_up_proj.weight","layers.19.shared_experts.0.down_proj.weight","layers.2._fast_impl.b_proj.weight","layers.2._fast_impl.q_proj.weight","layers.24._fast_impl.f_proj.0.weight","layers.27.latent_down.weight","layers.36._fast_impl.q_proj.weight","layers.40._fast_impl.f_proj.1.weight","layers.41.shared_experts.0.down_proj.weight","layers.42.o_proj.weight","layers.42.v_proj.weight","layers.44._fast_impl.f_proj.0.weight","layers.49.latent_down.weight","layers.52._fast_impl.b_proj.weight","layers.54._fast_impl.q_proj.weight","layers.60._fast_impl.f_proj.0.weight","layers.63.shared_experts.0.down_proj.weight"],"numel":21586304,"offsets":[0,6162432,6457344,6467712,6598784,9252992,9263360,10590464,10756352,11198720,12525824,12691712,15345920,16525568,16820480,16986368,17428736,17439104,18766208,18932096],"shapes":[[128384,48],[256,1152],[9,1152],[512,256],[1152,2304],[9,1152],[1152,1152],[144,1152],[384,1152],[1152,1152],[1152,144],[1152,2304],[1152,1024],[256,1152],[144,1152],[384,1152],[9,1152],[1152,1152],[144,1152],[1152,2304]]},"11":{"classes":["avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","rda","rda","avg","avg","avg","avg","avg","rda","rda","rda","avg","avg","rda","avg","avg","avg","rda","avg","avg","avg","rda","avg","avg","avg","avg","avg","avg","avg","rda","avg","avg","avg","avg","avg","rda","rda","avg","rda","rda","avg","avg","avg","avg","rda","rda","avg","avg","rda","avg","avg","rda","avg","avg","avg","avg","rda","avg","avg","rda"],"fragment_id":11,"key_order":["block_attn_res.14.norm.weight","block_attn_res.17.norm.weight","block_attn_res.2.norm.weight","block_attn_res.22.norm.weight","block_attn_res.25.norm.weight","block_attn_res.28.norm.weight","block_attn_res.30.norm.weight","block_attn_res.33.norm.weight","block_attn_res.36.norm.weight","block_attn_res.39.norm.weight","block_attn_res.41.norm.weight","block_attn_res.44.norm.weight","block_attn_res.47.norm.weight","block_attn_res.5.norm.weight","block_attn_res.52.norm.weight","block_attn_res.55.norm.weight","block_attn_res.58.norm.weight","block_attn_res.60.norm.weight","block_attn_res.63.norm.weight","block_attn_res.9.norm.weight","embed.weight::last[528:576]","layers.10._fast_impl.dt_bias","layers.10._fast_impl.q_conv1d.weight","layers.12.v_proj.weight","layers.14._fast_impl.c_proj.weight","layers.14._fast_impl.dt_bias","layers.14._fast_impl.k_conv1d.weight","layers.14._fast_impl.v_conv1d.weight","layers.16._fast_impl.q_conv1d.weight","layers.17.norm.weight","layers.18.v_up_proj.weight","layers.19.shared_experts.0.up_proj.weight","layers.2._fast_impl.c_proj.weight","layers.2._fast_impl.k_conv1d.weight","layers.2._fast_impl.v_conv1d.weight","layers.2._fast_impl.v_proj.weight","layers.21.norm.weight","layers.22._fast_impl.q_conv1d.weight","layers.24._fast_impl.c_proj.bias","layers.24._fast_impl.f_proj.1.weight","layers.24._fast_impl.k_conv1d.weight","layers.24._fast_impl.v_conv1d.weight","layers.24.norm.weight","layers.27.latent_up.weight","layers.30._fast_impl.dt_bias","layers.30._fast_impl.q_conv1d.weight","layers.32._fast_impl.k_conv1d.weight","layers.32._fast_impl.v_conv1d.weight","layers.33.norm.weight","layers.36._fast_impl.b_proj.bias","layers.36._fast_impl.q_conv1d.weight","layers.36._fast_impl.v_proj.weight","layers.38.norm.weight","layers.4._fast_impl.k_conv1d.weight","layers.4._fast_impl.v_conv1d.weight","layers.40._fast_impl.q_conv1d.weight","layers.40.norm.weight","layers.41.shared_experts.0.up_proj.weight","layers.42.q_proj.weight","layers.44._fast_impl.A_log","layers.44._fast_impl.c_proj.weight","layers.44._fast_impl.f_proj.1.weight","layers.44._fast_impl.k_conv1d.weight","layers.44._fast_impl.v_conv1d.weight","layers.45.norm.weight","layers.46._fast_impl.q_conv1d.weight","layers.48.kv_down_proj.weight","layers.49.latent_up.weight","layers.50.norm.weight","layers.52._fast_impl.c_proj.bias","layers.52._fast_impl.c_proj.weight","layers.52._fast_impl.k_conv1d.weight","layers.52._fast_impl.o_norm.weight","layers.54._fast_impl.v_proj.weight","layers.54.norm.weight","layers.56._fast_impl.k_conv1d.weight","layers.6._fast_impl.b_proj.bias","layers.6._fast_impl.dt_bias","layers.60._fast_impl.f_proj.1.weight","layers.60._fast_impl.k_conv1d.weight","layers.63.norm.weight","layers.63.shared_experts.0.up_proj.weight"],"numel":21571501,"offsets":[0,1152,2304,3456,4608,5760,6912,8064,9216,10368,11520,12672,13824,14976,16128,17280,18432,19584,20736,21888,23040,6185472,6186624,6191232,6486144,6496512,6497664,6502272,6506880,6511488,6512640,6643712,9297920,9308288,9312896,9317504,10644608,10645760,10650368,10650377,10816265,10820873,10825481,10826633,11269001,11270153,11274761,11279369,11283977,11285129,11285138,11289746,12616850,12618002,12622610,12627218,12631826,12632978,15287186,16466834,16466843,16477211,16643099,16647707,16652315,16653467,16658075,16952987,17395355,17396507,17396516,17406884,17411492,17411620,18738724,18739876,18744484,18744493,18745645,18911533,18916141,18917293],"shapes":[[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[128384,48],[1152],[1152,1,4],[256,1152],[9,1152],[1152],[1152,1,4],[1152,1,4],[1152,1,4],[1152],[512,256],[2304,1152],[9,1152],[1152,1,4],[1152,1,4],[1152,1152],[1152],[1152,1,4],[9],[1152,144],[1152,1,4],[1152,1,4],[1152],[1152,384],[1152],[1152,1,4],[1152,1,4],[1152,1,4],[1152],[9],[1152,1,4],[1152,1152],[1152],[1152,1,4],[1152,1,4],[1152,1,4],[1152],[2304,1152],[1024,1152],[9],[9,1152],[1152,144],[1152,1,4],[1152,1,4],[1152],[1152,1,4],[256,1152],[1152,384],[1152],[9],[9,1152],[1152,1,4],[128],[1152,1152],[1152],[1152,1,4],[9],[1152],[1152,144],[1152,1,4],[1152],[2304,1152]]},"12":{"classes":["avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","rda","avg","avg","rda","avg","rda","avg","rda","avg","rda","avg","rda","avg","avg","avg","avg","rda","rda","avg","rda","avg","avg","rda","rda","rda","rda","avg","avg","avg","avg","avg","rda","avg","rda","avg","avg","avg","avg","rda"],"fragment_id":12,"key_order":["block_attn_res.14.proj.weight","block_attn_res.17.proj.weight","block_attn_res.2.proj.weight","block_attn_res.22.proj.weight","block_attn_res.25.proj.weight","block_attn_res.28.proj.weight","block_attn_res.30.proj.weight","block_attn_res.33.proj.weight","block_attn_res.36.proj.weight","block_attn_res.39.proj.weight","block_attn_res.41.proj.weight","block_attn_res.44.proj.weight","block_attn_res.47.proj.weight","block_attn_res.5.proj.weight","block_attn_res.52.proj.weight","block_attn_res.55.proj.weight","block_attn_res.58.proj.weight","block_attn_res.60.proj.weight","block_attn_res.63.proj.weight","block_attn_res.9.proj.weight","embed.weight::last[576:624]","layers.10._fast_impl.g_proj.0.weight","layers.10._fast_impl.g_proj.1.bias","layers.14._fast_impl.g_proj.1.bias","layers.18.kv_down_proj.weight","layers.18.norm.weight","layers.21.shared_experts.0.down_proj.weight","layers.22._fast_impl.dt_bias","layers.22._fast_impl.k_proj.weight","layers.25.norm.weight","layers.29.latent_down.weight","layers.30._fast_impl.A_log","layers.30._fast_impl.e_proj.0.weight","layers.30._fast_impl.g_proj.1.bias","layers.34.norm.weight","layers.36._fast_impl.c_proj.bias","layers.39.norm.weight","layers.4._fast_impl.k_proj.weight","layers.40._fast_impl.g_proj.0.weight","layers.41.norm.weight","layers.43.shared_experts.0.down_proj.weight","layers.44._fast_impl.b_proj.bias","layers.46._fast_impl.dt_bias","layers.46._fast_impl.e_proj.0.weight","layers.5.latent_down.weight","layers.50.k_proj.weight","layers.50.o_proj.weight","layers.51.norm.weight","layers.52._fast_impl.q_conv1d.weight","layers.54._fast_impl.A_log","layers.54._fast_impl.o_norm.weight","layers.55.norm.weight","layers.56._fast_impl.k_proj.weight","layers.56._fast_impl.q_conv1d.weight","layers.58.k_up_proj.weight","layers.6._fast_impl.c_proj.bias","layers.6._fast_impl.g_proj.1.bias","layers.60._fast_impl.q_conv1d.weight","layers.7.norm.weight","layers.7.shared_experts.0.down_proj.weight"],"numel":21571501,"offsets":[0,1152,2304,3456,4608,5760,6912,8064,9216,10368,11520,12672,13824,14976,16128,17280,18432,19584,20736,21888,23040,6185472,6332928,6334080,6335232,6630144,6631296,9285504,9286656,10613760,10614912,11057280,11057289,11223177,11224329,11225481,11225490,11226642,12553746,12701202,12702354,15356562,15356571,15357723,15523611,15965979,16260891,17440539,17441691,17446299,17446308,17446436,17447588,18774692,18779300,18910372,18910381,18911533,18916141,18917293],"shapes":[[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[128384,48],[128,1152],[1152],[1152],[256,1152],[1152],[1152,2304],[1152],[1152,1152],[1152],[384,1152],[9],[144,1152],[1152],[1152],[9],[1152],[1152,1152],[128,1152],[1152],[1152,2304],[9],[1152],[144,1152],[384,1152],[256,1152],[1152,1024],[1152],[1152,1,4],[9],[128],[1152],[1152,1152],[1152,1,4],[512,256],[9],[1152],[1152,1,4],[1152],[1152,2304]]},"13":{"classes":["avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","rda","avg","avg","avg","rda","rda","avg","rda","avg","rda","avg","rda","avg","avg","avg","avg","rda","rda","avg","rda","avg","rda","avg","rda","rda","rda","avg","avg","avg","avg","avg","rda","avg","rda","avg","avg","avg","rda","avg"],"fragment_id":13,"key_order":["block_attn_res.15.norm.weight","block_attn_res.18.norm.weight","block_attn_res.20.norm.weight","block_attn_res.23.norm.weight","block_attn_res.26.norm.weight","block_attn_res.29.norm.weight","block_attn_res.31.norm.weight","block_attn_res.34.norm.weight","block_attn_res.37.norm.weight","block_attn_res.4.norm.weight","block_attn_res.42.norm.weight","block_attn_res.45.norm.weight","block_attn_res.48.norm.weight","block_attn_res.50.norm.weight","block_attn_res.53.norm.weight","block_attn_res.56.norm.weight","block_attn_res.59.norm.weight","block_attn_res.61.norm.weight","block_attn_res.7.norm.weight","block_attn_res_head.norm.weight","embed.weight::last[624:672]","layers.10._fast_impl.g_proj.1.weight","layers.10.norm.weight","layers.14.norm.weight","layers.19.norm.weight","layers.20.k_proj.weight","layers.21.shared_experts.0.up_proj.weight","layers.22._fast_impl.g_proj.1.bias","layers.22._fast_impl.o_proj.weight","layers.26.norm.weight","layers.29.latent_up.weight","layers.30._fast_impl.b_proj.bias","layers.30._fast_impl.e_proj.1.weight","layers.30.norm.weight","layers.35.norm.weight","layers.4._fast_impl.A_log","layers.4._fast_impl.dt_bias","layers.4._fast_impl.o_proj.weight","layers.40._fast_impl.g_proj.1.weight","layers.42.norm.weight","layers.43.shared_experts.0.up_proj.weight","layers.44._fast_impl.c_proj.bias","layers.46._fast_impl.e_proj.1.weight","layers.46._fast_impl.g_proj.1.bias","layers.5.latent_up.weight","layers.50.q_proj.weight","layers.50.v_proj.weight","layers.52._fast_impl.dt_bias","layers.52._fast_impl.v_conv1d.weight","layers.54._fast_impl.b_proj.bias","layers.56._fast_impl.dt_bias","layers.56._fast_impl.o_norm.weight","layers.56._fast_impl.o_proj.weight","layers.56._fast_impl.v_conv1d.weight","layers.58.v_up_proj.weight","layers.6.norm.weight","layers.60._fast_impl.A_log","layers.60._fast_impl.v_conv1d.weight","layers.7.shared_experts.0.up_proj.weight","layers.8.norm.weight"],"numel":21571501,"offsets":[0,1152,2304,3456,4608,5760,6912,8064,9216,10368,11520,12672,13824,14976,16128,17280,18432,19584,20736,21888,23040,6185472,6332928,6334080,6335232,6336384,6631296,9285504,9286656,10613760,10614912,11057280,11057289,11223177,11224329,11225481,11225490,11226642,12553746,12701202,12702354,15356562,15356571,15522459,15523611,15965979,17145627,17440539,17441691,17446299,17446308,17447460,17447588,18774692,18779300,18910372,18911524,18911533,18916141,21570349],"shapes":[[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[128384,48],[1152,128],[1152],[1152],[1152],[256,1152],[2304,1152],[1152],[1152,1152],[1152],[1152,384],[9],[1152,144],[1152],[1152],[9],[1152],[1152,1152],[1152,128],[1152],[2304,1152],[9],[1152,144],[1152],[1152,384],[1024,1152],[256,1152],[1152],[1152,1,4],[9],[1152],[128],[1152,1152],[1152,1,4],[512,256],[1152],[9],[1152,1,4],[2304,1152],[1152]]},"14":{"classes":["avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","rda","avg","avg","rda","rda","avg","rda","avg","rda","avg","rda","avg","avg","avg","avg","rda","avg","rda","rda","avg","rda","avg","rda","avg","avg","avg","avg","rda","rda","avg","avg","avg","avg","rda","rda","avg","rda"],"fragment_id":14,"key_order":["block_attn_res.0.norm.weight","block_attn_res.1.proj.weight","block_attn_res.11.norm.weight","block_attn_res.12.proj.weight","block_attn_res.15.proj.weight","block_attn_res.18.proj.weight","block_attn_res.20.proj.weight","block_attn_res.23.proj.weight","block_attn_res.26.proj.weight","block_attn_res.29.proj.weight","block_attn_res.31.proj.weight","block_attn_res.34.proj.weight","block_attn_res.37.proj.weight","block_attn_res.4.proj.weight","block_attn_res.42.proj.weight","block_attn_res.45.proj.weight","block_attn_res.48.proj.weight","block_attn_res.50.proj.weight","block_attn_res.53.proj.weight","block_attn_res.56.proj.weight","block_attn_res.59.proj.weight","block_attn_res.61.proj.weight","block_attn_res.7.proj.weight","embed.weight::last[672:720]","final_norm.weight","layers.11.norm.weight","layers.14._fast_impl.g_proj.0.weight","layers.15.norm.weight","layers.2._fast_impl.dt_bias","layers.20.v_proj.weight","layers.22._fast_impl.q_proj.weight","layers.22.norm.weight","layers.23.shared_experts.0.down_proj.weight","layers.27.norm.weight","layers.3.latent_down.weight","layers.30._fast_impl.c_proj.bias","layers.30._fast_impl.f_proj.0.weight","layers.31.norm.weight","layers.36._fast_impl.dt_bias","layers.4._fast_impl.b_proj.bias","layers.4._fast_impl.g_proj.1.bias","layers.4._fast_impl.q_proj.weight","layers.43.norm.weight","layers.44._fast_impl.g_proj.0.weight","layers.45.shared_experts.0.down_proj.weight","layers.46._fast_impl.A_log","layers.46._fast_impl.f_proj.0.weight","layers.46.norm.weight","layers.51.latent_down.weight","layers.52._fast_impl.g_proj.1.bias","layers.54._fast_impl.c_proj.bias","layers.54._fast_impl.k_conv1d.weight","layers.56._fast_impl.g_proj.1.bias","layers.56._fast_impl.q_proj.weight","layers.58.kv_down_proj.weight","layers.6._fast_impl.k_conv1d.weight","layers.6._fast_impl.o_norm.weight","layers.60._fast_impl.b_proj.bias","layers.60._fast_impl.dt_bias","layers.62.o_proj.weight","layers.8.k_up_proj.weight","layers.9.norm.weight","layers.9.shared_experts.0.down_proj.weight"],"numel":21571501,"offsets":[0,1152,2304,3456,4608,5760,6912,8064,9216,10368,11520,12672,13824,14976,16128,17280,18432,19584,20736,21888,23040,24192,25344,26496,6188928,6190080,6191232,6338688,6339840,6340992,6635904,7963008,7964160,10618368,10619520,11061888,11061897,11227785,11228937,11230089,11230098,11231250,12558354,12559506,12706962,15361170,15361179,15527067,15528219,15970587,15971739,15971748,15976356,15977508,17304612,17599524,17604132,17604260,17604269,17605421,18785069,18916141,18917293],"shapes":[[1152],[1,1152],[1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[128384,48],[1152],[1152],[128,1152],[1152],[1152],[256,1152],[1152,1152],[1152],[1152,2304],[1152],[384,1152],[9],[144,1152],[1152],[1152],[9],[1152],[1152,1152],[1152],[128,1152],[1152,2304],[9],[144,1152],[1152],[384,1152],[1152],[9],[1152,1,4],[1152],[1152,1152],[256,1152],[1152,1,4],[128],[9],[1152],[1152,1024],[512,256],[1152],[1152,2304]]},"15":{"classes":["avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","rda","avg","avg","avg","avg","rda","avg","rda","rda","avg","rda","rda","avg","avg","avg","avg","avg","avg","avg","rda","avg","avg","rda","rda","avg","rda","avg","rda","avg","avg","avg","rda","avg","avg","avg","avg","avg","rda","rda","avg","rda","rda"],"fragment_id":15,"key_order":["block_attn_res.0.proj.weight","block_attn_res.10.norm.weight","block_attn_res.11.proj.weight","block_attn_res.13.norm.weight","block_attn_res.16.norm.weight","block_attn_res.19.norm.weight","block_attn_res.21.norm.weight","block_attn_res.24.norm.weight","block_attn_res.27.norm.weight","block_attn_res.3.norm.weight","block_attn_res.32.norm.weight","block_attn_res.35.norm.weight","block_attn_res.38.norm.weight","block_attn_res.40.norm.weight","block_attn_res.43.norm.weight","block_attn_res.46.norm.weight","block_attn_res.49.norm.weight","block_attn_res.51.norm.weight","block_attn_res.54.norm.weight","block_attn_res.57.norm.weight","block_attn_res.6.norm.weight","block_attn_res.62.norm.weight","block_attn_res.8.norm.weight","embed.weight::last[720:768]","layers.0.norm.weight","layers.12.norm.weight","layers.14._fast_impl.g_proj.1.weight","layers.16._fast_impl.dt_bias","layers.18.kv_latent_norm.weight","layers.2._fast_impl.g_proj.1.bias","layers.2._fast_impl.o_norm.weight","layers.22._fast_impl.v_proj.weight","layers.23.norm.weight","layers.23.shared_experts.0.up_proj.weight","layers.26.k_proj.weight","layers.28.norm.weight","layers.3.latent_up.weight","layers.30._fast_impl.f_proj.1.weight","layers.30._fast_impl.o_norm.weight","layers.32._fast_impl.A_log","layers.32._fast_impl.dt_bias","layers.36._fast_impl.g_proj.1.bias","layers.38.kv_latent_norm.weight","layers.4._fast_impl.c_proj.bias","layers.4._fast_impl.o_norm.weight","layers.4._fast_impl.v_proj.weight","layers.4.norm.weight","layers.44._fast_impl.dt_bias","layers.44._fast_impl.g_proj.1.weight","layers.45.shared_experts.0.up_proj.weight","layers.46._fast_impl.b_proj.bias","layers.46._fast_impl.f_proj.1.weight","layers.47.norm.weight","layers.51.latent_up.weight","layers.52.norm.weight","layers.54._fast_impl.q_conv1d.weight","layers.56._fast_impl.A_log","layers.56._fast_impl.v_proj.weight","layers.56.norm.weight","layers.6._fast_impl.q_conv1d.weight","layers.60._fast_impl.c_proj.bias","layers.60._fast_impl.g_proj.1.bias","layers.60._fast_impl.o_norm.weight","layers.62.k_proj.weight","layers.62.q_proj.weight","layers.8.kv_latent_norm.weight","layers.8.v_up_proj.weight","layers.9.shared_experts.0.up_proj.weight"],"numel":21571501,"offsets":[0,1152,2304,3456,4608,5760,6912,8064,9216,10368,11520,12672,13824,14976,16128,17280,18432,19584,20736,21888,23040,24192,25344,26496,6188928,6190080,6191232,6338688,6339840,6340096,6341248,6341376,7668480,7669632,10323840,10618752,10619904,11062272,11228160,11228288,11228297,11229449,11230601,11230857,11230866,11230994,12558098,12559250,12560402,12707858,15362066,15362075,15527963,15529115,15971483,15972635,15977243,15977252,17304356,17305508,17310116,17310125,17311277,17311405,17606317,18785965,18786221,18917293],"shapes":[[1,1152],[1152],[1,1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[1152],[128384,48],[1152],[1152],[1152,128],[1152],[256],[1152],[128],[1152,1152],[1152],[2304,1152],[256,1152],[1152],[1152,384],[1152,144],[128],[9],[1152],[1152],[256],[9],[128],[1152,1152],[1152],[1152],[1152,128],[2304,1152],[9],[1152,144],[1152],[1152,384],[1152],[1152,1,4],[9],[1152,1152],[1152],[1152,1,4],[9],[1152],[128],[256,1152],[1024,1152],[256],[512,256],[2304,1152]]},"16":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":16,"key_order":["embed.weight::last[768:816]","layers.1.latent_down.weight","layers.10._fast_impl.b_proj.weight","layers.16._fast_impl.b_proj.weight","layers.16._fast_impl.g_proj.0.weight","layers.18.o_proj.weight","layers.25.shared_experts.0.down_proj.weight","layers.26.v_proj.weight","layers.31.latent_down.weight","layers.32._fast_impl.e_proj.0.weight","layers.37.latent_up.weight","layers.40._fast_impl.k_proj.weight","layers.46._fast_impl.b_proj.weight","layers.46._fast_impl.g_proj.0.weight","layers.47.shared_experts.0.down_proj.weight","layers.52._fast_impl.e_proj.0.weight","layers.53.latent_down.weight","layers.58.o_proj.weight","layers.6._fast_impl.k_proj.weight","layers.62.v_proj.weight"],"numel":21860736,"offsets":[0,6162432,6604800,6615168,6625536,6772992,9132288,11786496,12081408,12523776,12689664,13132032,14459136,14469504,14616960,17271168,17437056,17879424,20238720,21565824],"shapes":[[128384,48],[384,1152],[9,1152],[9,1152],[128,1152],[1152,2048],[1152,2304],[256,1152],[384,1152],[144,1152],[1152,384],[1152,1152],[9,1152],[128,1152],[1152,2304],[144,1152],[384,1152],[1152,2048],[1152,1152],[256,1152]]},"17":{"classes":["avg","avg","rda","avg","avg","avg","rda","avg","avg","avg","avg","avg","avg","avg","avg","rda","rda","rda","avg","avg","avg","avg","avg","avg","avg","rda","rda","rda","avg","rda","avg","rda","avg","avg","rda","avg","rda","rda","rda","avg","avg","rda","rda","avg","avg","avg","rda","rda","avg","rda"],"fragment_id":17,"key_order":["embed.weight::last[816:864]","final_norm.logit_scale_log","layers.1.latent_up.weight","layers.10._fast_impl.A_log","layers.10._fast_impl.b_proj.bias","layers.10._fast_impl.c_proj.bias","layers.10._fast_impl.c_proj.weight","layers.10._fast_impl.o_norm.weight","layers.14._fast_impl.A_log","layers.14._fast_impl.b_proj.bias","layers.14._fast_impl.c_proj.bias","layers.14._fast_impl.o_norm.weight","layers.16._fast_impl.A_log","layers.16._fast_impl.b_proj.bias","layers.16._fast_impl.c_proj.bias","layers.16._fast_impl.c_proj.weight","layers.16._fast_impl.g_proj.1.weight","layers.18.q_proj.weight","layers.2._fast_impl.A_log","layers.2._fast_impl.b_proj.bias","layers.2._fast_impl.c_proj.bias","layers.22._fast_impl.A_log","layers.22._fast_impl.b_proj.bias","layers.22._fast_impl.c_proj.bias","layers.22._fast_impl.o_norm.weight","layers.25.shared_experts.0.up_proj.weight","layers.28.kv_down_proj.weight","layers.31.latent_up.weight","layers.32._fast_impl.b_proj.bias","layers.32._fast_impl.e_proj.1.weight","layers.32._fast_impl.o_norm.weight","layers.4._fast_impl.g_proj.1.weight","layers.40._fast_impl.A_log","layers.40._fast_impl.o_norm.weight","layers.40._fast_impl.o_proj.weight","layers.46._fast_impl.c_proj.bias","layers.46._fast_impl.c_proj.weight","layers.46._fast_impl.g_proj.1.weight","layers.47.shared_experts.0.up_proj.weight","layers.48.kv_latent_norm.weight","layers.49.norm.weight","layers.52._fast_impl.e_proj.1.weight","layers.53.latent_up.weight","layers.54._fast_impl.dt_bias","layers.56._fast_impl.b_proj.bias","layers.58.norm.weight","layers.58.q_proj.weight","layers.6._fast_impl.o_proj.weight","layers.61.norm.weight","layers.8.kv_down_proj.weight"],"numel":21571500,"offsets":[0,6162432,6162433,6604801,6604810,6604819,6604828,6615196,6615324,6615333,6615342,6615351,6615479,6615488,6615497,6615506,6625874,6773330,9132626,9132635,9132644,9132653,9132662,9132671,9132680,9132808,11787016,12081928,12524296,12524305,12690193,12690321,12837777,12837786,12837914,14165018,14165027,14175395,14322851,16977059,16977315,16978467,17144355,17586723,17587875,17587884,17589036,19948332,21275436,21276588],"shapes":[[128384,48],[1],[1152,384],[9],[9],[9],[9,1152],[128],[9],[9],[9],[128],[9],[9],[9],[9,1152],[1152,128],[2048,1152],[9],[9],[9],[9],[9],[9],[128],[2304,1152],[256,1152],[1152,384],[9],[1152,144],[128],[1152,128],[9],[128],[1152,1152],[9],[9,1152],[1152,128],[2304,1152],[256],[1152],[1152,144],[1152,384],[1152],[9],[1152],[2048,1152],[1152,1152],[1152],[256,1152]]},"18":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":18,"key_order":["embed.weight::last[864:912]","layers.10._fast_impl.e_proj.0.weight","layers.11.latent_down.weight","layers.14._fast_impl.e_proj.1.weight","layers.2._fast_impl.f_proj.0.weight","layers.27.shared_experts.0.down_proj.weight","layers.28.k_up_proj.weight","layers.28.o_proj.weight","layers.32._fast_impl.g_proj.0.weight","layers.33.latent_down.weight","layers.34.k_proj.weight","layers.4._fast_impl.b_proj.weight","layers.4._fast_impl.f_proj.0.weight","layers.40._fast_impl.q_proj.weight","layers.49.shared_experts.0.down_proj.weight","layers.55.latent_down.weight","layers.56._fast_impl.f_proj.0.weight","layers.6._fast_impl.q_proj.weight","layers.8.o_proj.weight"],"numel":21584000,"offsets":[0,6162432,6328320,6770688,6936576,7102464,9756672,9887744,12247040,12394496,12836864,13131776,13142144,13308032,14635136,17289344,17731712,17897600,19224704],"shapes":[[128384,48],[144,1152],[384,1152],[1152,144],[144,1152],[1152,2304],[512,256],[1152,2048],[128,1152],[384,1152],[256,1152],[9,1152],[144,1152],[1152,1152],[1152,2304],[384,1152],[144,1152],[1152,1152],[1152,2048]]},"19":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":19,"key_order":["embed.weight::last[912:960]","layers.10._fast_impl.e_proj.1.weight","layers.11.latent_up.weight","layers.15.latent_up.weight","layers.2._fast_impl.f_proj.1.weight","layers.27.shared_experts.0.up_proj.weight","layers.28.q_proj.weight","layers.28.v_up_proj.weight","layers.32._fast_impl.g_proj.1.weight","layers.33.latent_up.weight","layers.34.v_proj.weight","layers.4._fast_impl.c_proj.weight","layers.4._fast_impl.f_proj.1.weight","layers.40._fast_impl.v_proj.weight","layers.49.shared_experts.0.up_proj.weight","layers.55.latent_up.weight","layers.56._fast_impl.f_proj.1.weight","layers.6._fast_impl.v_proj.weight","layers.8.q_proj.weight"],"numel":21860480,"offsets":[0,6162432,6328320,6770688,7213056,7378944,10033152,12392448,12523520,12670976,13113344,13408256,13418624,13584512,14911616,17565824,18008192,18174080,19501184],"shapes":[[128384,48],[1152,144],[1152,384],[1152,384],[1152,144],[2304,1152],[2048,1152],[512,256],[1152,128],[1152,384],[256,1152],[9,1152],[1152,144],[1152,1152],[2304,1152],[1152,384],[1152,144],[1152,1152],[2048,1152]]},"2":{"classes":["avg","rda","rda","rda","rda","avg","rda","avg","rda","avg","rda","avg","avg","rda","avg","avg","rda","rda","avg","avg","avg","rda","avg","avg","rda","rda","avg","avg","rda","rda","avg","rda","rda","avg","avg","avg","rda"],"fragment_id":2,"key_order":["embed.weight::last[96:144]","layers.11.shared_experts.0.down_proj.weight","layers.12.o_proj.weight","layers.14._fast_impl.q_proj.weight","layers.16._fast_impl.e_proj.0.weight","layers.16.norm.weight","layers.19.latent_down.weight","layers.20.norm.weight","layers.22._fast_impl.g_proj.0.weight","layers.24._fast_impl.A_log","layers.24._fast_impl.b_proj.weight","layers.24._fast_impl.g_proj.1.bias","layers.3.norm.weight","layers.30._fast_impl.q_proj.weight","layers.32._fast_impl.c_proj.bias","layers.32.norm.weight","layers.33.shared_experts.0.down_proj.weight","layers.36._fast_impl.e_proj.0.weight","layers.37.norm.weight","layers.40._fast_impl.b_proj.bias","layers.40._fast_impl.g_proj.1.bias","layers.41.latent_down.weight","layers.44._fast_impl.o_norm.weight","layers.44.norm.weight","layers.46._fast_impl.q_proj.weight","layers.48.v_up_proj.weight","layers.5.norm.weight","layers.52._fast_impl.A_log","layers.54._fast_impl.e_proj.0.weight","layers.54._fast_impl.g_proj.0.weight","layers.54._fast_impl.g_proj.1.bias","layers.55.shared_experts.0.down_proj.weight","layers.56._fast_impl.b_proj.weight","layers.56._fast_impl.c_proj.bias","layers.59.norm.weight","layers.62.norm.weight","layers.63.latent_down.weight"],"numel":21571501,"offsets":[0,6162432,8816640,9996288,11323392,11489280,11490432,11932800,11933952,12081408,12081417,12091785,12092937,12094089,13421193,13421202,13422354,16076562,16242450,16243602,16243611,16244763,16687131,16687259,16688411,18015515,18146587,18147739,18147748,18313636,18461092,18462244,21116452,21126820,21126829,21127981,21129133],"shapes":[[128384,48],[1152,2304],[1152,1024],[1152,1152],[144,1152],[1152],[384,1152],[1152],[128,1152],[9],[9,1152],[1152],[1152],[1152,1152],[9],[1152],[1152,2304],[144,1152],[1152],[9],[1152],[384,1152],[128],[1152],[1152,1152],[512,256],[1152],[9],[144,1152],[128,1152],[1152],[1152,2304],[9,1152],[9],[1152],[1152],[384,1152]]},"20":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":20,"key_order":["embed.weight::last[960:1008]","layers.10._fast_impl.f_proj.0.weight","layers.10._fast_impl.k_proj.weight","layers.13.latent_down.weight","layers.22._fast_impl.e_proj.0.weight","layers.22._fast_impl.f_proj.1.weight","layers.24._fast_impl.k_proj.weight","layers.29.shared_experts.0.down_proj.weight","layers.35.latent_down.weight","layers.36._fast_impl.g_proj.0.weight","layers.38.k_up_proj.weight","layers.38.o_proj.weight","layers.40._fast_impl.b_proj.weight","layers.40._fast_impl.e_proj.0.weight","layers.44._fast_impl.k_proj.weight","layers.5.shared_experts.0.down_proj.weight","layers.57.latent_down.weight","layers.6._fast_impl.e_proj.0.weight","layers.60._fast_impl.k_proj.weight"],"numel":21584000,"offsets":[0,6162432,6328320,7655424,8097792,8263680,8429568,9756672,12410880,12853248,13000704,13131776,15491072,15501440,15667328,16994432,19648640,20091008,20256896],"shapes":[[128384,48],[144,1152],[1152,1152],[384,1152],[144,1152],[1152,144],[1152,1152],[1152,2304],[384,1152],[128,1152],[512,256],[1152,2048],[9,1152],[144,1152],[1152,1152],[1152,2304],[384,1152],[144,1152],[1152,1152]]},"21":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":21,"key_order":["embed.weight::last[1008:1056]","layers.10._fast_impl.f_proj.1.weight","layers.10._fast_impl.o_proj.weight","layers.13.latent_up.weight","layers.22._fast_impl.e_proj.1.weight","layers.24._fast_impl.o_proj.weight","layers.24._fast_impl.v_proj.weight","layers.29.shared_experts.0.up_proj.weight","layers.35.latent_up.weight","layers.36._fast_impl.g_proj.1.weight","layers.38.q_proj.weight","layers.38.v_up_proj.weight","layers.40._fast_impl.c_proj.weight","layers.40._fast_impl.e_proj.1.weight","layers.44._fast_impl.o_proj.weight","layers.5.shared_experts.0.up_proj.weight","layers.57.latent_up.weight","layers.6._fast_impl.e_proj.1.weight","layers.60._fast_impl.o_proj.weight"],"numel":22745216,"offsets":[0,6162432,6328320,7655424,8097792,8263680,9590784,10917888,13572096,14014464,14161920,16521216,16652288,16662656,16828544,18155648,20809856,21252224,21418112],"shapes":[[128384,48],[1152,144],[1152,1152],[1152,384],[1152,144],[1152,1152],[1152,1152],[2304,1152],[1152,384],[1152,128],[2048,1152],[512,256],[9,1152],[1152,144],[1152,1152],[2304,1152],[1152,384],[1152,144],[1152,1152]]},"22":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":22,"key_order":["embed.weight::last[1056:1104]","layers.10._fast_impl.q_proj.weight","layers.14._fast_impl.e_proj.0.weight","layers.15.latent_down.weight","layers.22._fast_impl.f_proj.0.weight","layers.24._fast_impl.q_proj.weight","layers.3.shared_experts.0.down_proj.weight","layers.3.shared_experts.0.up_proj.weight","layers.37.latent_down.weight","layers.4._fast_impl.g_proj.0.weight","layers.40._fast_impl.f_proj.0.weight","layers.44._fast_impl.b_proj.weight","layers.44._fast_impl.q_proj.weight","layers.48.k_up_proj.weight","layers.48.o_proj.weight","layers.51.shared_experts.0.down_proj.weight","layers.59.latent_down.weight","layers.6._fast_impl.f_proj.0.weight","layers.60._fast_impl.q_proj.weight"],"numel":24072320,"offsets":[0,6162432,7489536,7655424,8097792,8263680,9590784,12244992,14899200,15341568,15489024,15654912,15665280,16992384,17123456,19482752,22136960,22579328,22745216],"shapes":[[128384,48],[1152,1152],[144,1152],[384,1152],[144,1152],[1152,1152],[1152,2304],[2304,1152],[384,1152],[128,1152],[144,1152],[9,1152],[1152,1152],[512,256],[1152,2048],[1152,2304],[384,1152],[144,1152],[1152,1152]]},"23":{"classes":["router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg","router_avg"],"fragment_id":23,"key_order":["layers.1.router.expert_bias","layers.1.router.gate.weight","layers.11.router.expert_bias","layers.11.router.gate.weight","layers.13.router.expert_bias","layers.13.router.gate.weight","layers.15.router.expert_bias","layers.15.router.gate.weight","layers.17.router.expert_bias","layers.17.router.gate.weight","layers.19.router.expert_bias","layers.19.router.gate.weight","layers.21.router.expert_bias","layers.21.router.gate.weight","layers.23.router.expert_bias","layers.23.router.gate.weight","layers.25.router.expert_bias","layers.25.router.gate.weight","layers.27.router.expert_bias","layers.27.router.gate.weight","layers.29.router.expert_bias","layers.29.router.gate.weight","layers.3.router.expert_bias","layers.3.router.gate.weight","layers.31.router.expert_bias","layers.31.router.gate.weight","layers.33.router.expert_bias","layers.33.router.gate.weight","layers.35.router.expert_bias","layers.35.router.gate.weight","layers.37.router.expert_bias","layers.37.router.gate.weight","layers.39.router.expert_bias","layers.39.router.gate.weight","layers.41.router.expert_bias","layers.41.router.gate.weight","layers.43.router.expert_bias","layers.43.router.gate.weight","layers.45.router.expert_bias","layers.45.router.gate.weight","layers.47.router.expert_bias","layers.47.router.gate.weight","layers.49.router.expert_bias","layers.49.router.gate.weight","layers.5.router.expert_bias","layers.5.router.gate.weight","layers.51.router.expert_bias","layers.51.router.gate.weight","layers.53.router.expert_bias","layers.53.router.gate.weight","layers.55.router.expert_bias","layers.55.router.gate.weight","layers.57.router.expert_bias","layers.57.router.gate.weight","layers.59.router.expert_bias","layers.59.router.gate.weight","layers.61.router.expert_bias","layers.61.router.gate.weight","layers.63.router.expert_bias","layers.63.router.gate.weight","layers.7.router.expert_bias","layers.7.router.gate.weight","layers.9.router.expert_bias","layers.9.router.gate.weight","qb_beta.moe0","qb_beta.moe1","qb_beta.moe10","qb_beta.moe11","qb_beta.moe12","qb_beta.moe13","qb_beta.moe14","qb_beta.moe15","qb_beta.moe16","qb_beta.moe17","qb_beta.moe18","qb_beta.moe19","qb_beta.moe2","qb_beta.moe20","qb_beta.moe21","qb_beta.moe22","qb_beta.moe23","qb_beta.moe24","qb_beta.moe25","qb_beta.moe26","qb_beta.moe27","qb_beta.moe28","qb_beta.moe29","qb_beta.moe3","qb_beta.moe30","qb_beta.moe31","qb_beta.moe4","qb_beta.moe5","qb_beta.moe6","qb_beta.moe7","qb_beta.moe8","qb_beta.moe9"],"numel":4726784,"offsets":[0,128,147584,147712,295168,295296,442752,442880,590336,590464,737920,738048,885504,885632,1033088,1033216,1180672,1180800,1328256,1328384,1475840,1475968,1623424,1623552,1771008,1771136,1918592,1918720,2066176,2066304,2213760,2213888,2361344,2361472,2508928,2509056,2656512,2656640,2804096,2804224,2951680,2951808,3099264,3099392,3246848,3246976,3394432,3394560,3542016,3542144,3689600,3689728,3837184,3837312,3984768,3984896,4132352,4132480,4279936,4280064,4427520,4427648,4575104,4575232,4722688,4722816,4722944,4723072,4723200,4723328,4723456,4723584,4723712,4723840,4723968,4724096,4724224,4724352,4724480,4724608,4724736,4724864,4724992,4725120,4725248,4725376,4725504,4725632,4725760,4725888,4726016,4726144,4726272,4726400,4726528,4726656],"shapes":[[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128,1152],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128],[128]]},"3":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":3,"key_order":["embed.weight::last[144:192]","layers.11.shared_experts.0.up_proj.weight","layers.12.q_proj.weight","layers.14._fast_impl.v_proj.weight","layers.16._fast_impl.e_proj.1.weight","layers.19.latent_up.weight","layers.22._fast_impl.g_proj.1.weight","layers.24._fast_impl.c_proj.weight","layers.30._fast_impl.v_proj.weight","layers.33.shared_experts.0.up_proj.weight","layers.36._fast_impl.e_proj.1.weight","layers.41.latent_up.weight","layers.46._fast_impl.v_proj.weight","layers.51.shared_experts.0.up_proj.weight","layers.54._fast_impl.e_proj.1.weight","layers.54._fast_impl.g_proj.1.weight","layers.55.shared_experts.0.up_proj.weight","layers.56._fast_impl.c_proj.weight","layers.63.latent_up.weight"],"numel":24080640,"offsets":[0,6162432,8816640,9996288,11323392,11489280,11931648,12079104,12089472,13416576,16070784,16236672,16679040,18006144,20660352,20826240,20973696,23627904,23638272],"shapes":[[128384,48],[2304,1152],[1024,1152],[1152,1152],[1152,144],[1152,384],[1152,128],[9,1152],[1152,1152],[2304,1152],[1152,144],[1152,384],[1152,1152],[2304,1152],[1152,144],[1152,128],[2304,1152],[9,1152],[1152,384]]},"4":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":4,"key_order":["embed.weight::last[192:240]","layers.13.shared_experts.0.down_proj.weight","layers.16._fast_impl.f_proj.0.weight","layers.16._fast_impl.k_proj.weight","layers.20.o_proj.weight","layers.21.latent_down.weight","layers.24._fast_impl.g_proj.0.weight","layers.30._fast_impl.b_proj.weight","layers.32._fast_impl.k_proj.weight","layers.35.shared_experts.0.down_proj.weight","layers.36._fast_impl.f_proj.0.weight","layers.43.latent_down.weight","layers.52._fast_impl.k_proj.weight","layers.54._fast_impl.f_proj.0.weight","layers.56._fast_impl.g_proj.0.weight","layers.57.shared_experts.0.down_proj.weight","layers.59.latent_up.weight","layers.6._fast_impl.b_proj.weight","layers.7.latent_down.weight"],"numel":21868800,"offsets":[0,6162432,8816640,8982528,10309632,11489280,11931648,12079104,12089472,13416576,16070784,16236672,16679040,18006144,18172032,18319488,20973696,21416064,21426432],"shapes":[[128384,48],[1152,2304],[144,1152],[1152,1152],[1152,1024],[384,1152],[128,1152],[9,1152],[1152,1152],[1152,2304],[144,1152],[384,1152],[1152,1152],[144,1152],[128,1152],[1152,2304],[1152,384],[9,1152],[384,1152]]},"5":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":5,"key_order":["embed.weight::last[240:288]","layers.13.shared_experts.0.up_proj.weight","layers.16._fast_impl.f_proj.1.weight","layers.16._fast_impl.o_proj.weight","layers.20.q_proj.weight","layers.21.latent_up.weight","layers.24._fast_impl.g_proj.1.weight","layers.30._fast_impl.c_proj.weight","layers.32._fast_impl.o_proj.weight","layers.35.shared_experts.0.up_proj.weight","layers.36._fast_impl.f_proj.1.weight","layers.43.latent_up.weight","layers.52._fast_impl.o_proj.weight","layers.54._fast_impl.f_proj.1.weight","layers.56._fast_impl.g_proj.1.weight","layers.57.shared_experts.0.up_proj.weight","layers.6._fast_impl.c_proj.weight","layers.6._fast_impl.f_proj.1.weight","layers.7.latent_up.weight"],"numel":21592320,"offsets":[0,6162432,8816640,8982528,10309632,11489280,11931648,12079104,12089472,13416576,16070784,16236672,16679040,18006144,18172032,18319488,20973696,20984064,21149952],"shapes":[[128384,48],[2304,1152],[1152,144],[1152,1152],[1024,1152],[1152,384],[1152,128],[9,1152],[1152,1152],[2304,1152],[1152,144],[1152,384],[1152,1152],[1152,144],[1152,128],[2304,1152],[9,1152],[1152,144],[1152,384]]},"6":{"classes":["avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":6,"key_order":["embed.weight::last[288:336]","layers.15.shared_experts.0.down_proj.weight","layers.16._fast_impl.q_proj.weight","layers.2._fast_impl.e_proj.0.weight","layers.23.latent_down.weight","layers.26.o_proj.weight","layers.30._fast_impl.g_proj.0.weight","layers.32._fast_impl.b_proj.weight","layers.32._fast_impl.q_proj.weight","layers.37.shared_experts.0.down_proj.weight","layers.4._fast_impl.e_proj.0.weight","layers.45.latent_down.weight","layers.52._fast_impl.q_proj.weight","layers.56._fast_impl.e_proj.0.weight","layers.59.shared_experts.0.down_proj.weight","layers.6._fast_impl.g_proj.0.weight","layers.60._fast_impl.b_proj.weight","layers.60._fast_impl.v_proj.weight","layers.9.latent_down.weight"],"numel":22753536,"offsets":[0,6162432,8816640,10143744,10309632,10752000,11931648,12079104,12089472,13416576,16070784,16236672,16679040,18006144,18172032,20826240,20973696,20984064,22311168],"shapes":[[128384,48],[1152,2304],[1152,1152],[144,1152],[384,1152],[1152,1024],[128,1152],[9,1152],[1152,1152],[1152,2304],[144,1152],[384,1152],[1152,1152],[144,1152],[1152,2304],[128,1152],[9,1152],[1152,1152],[384,1152]]},"7":{"classes":["avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","avg","rda","avg","avg","avg","avg","rda","rda","avg","avg","avg","avg","rda","avg","avg","avg","avg","rda","avg","avg","rda","avg","avg","rda","avg","avg","rda","avg","avg","avg","avg","avg","rda","rda","avg","avg","avg","avg","avg","avg","avg","rda","avg","avg","avg","avg","avg","rda","avg","avg","rda","avg","avg","rda","avg","rda","avg","rda","avg","rda"],"fragment_id":7,"key_order":["block_attn_res.1.norm.weight","block_attn_res.10.proj.weight","block_attn_res.12.norm.weight","block_attn_res.13.proj.weight","block_attn_res.16.proj.weight","block_attn_res.19.proj.weight","block_attn_res.21.proj.weight","block_attn_res.24.proj.weight","block_attn_res.27.proj.weight","block_attn_res.3.proj.weight","block_attn_res.32.proj.weight","block_attn_res.35.proj.weight","block_attn_res.38.proj.weight","block_attn_res.40.proj.weight","block_attn_res.43.proj.weight","block_attn_res.46.proj.weight","block_attn_res.49.proj.weight","block_attn_res.51.proj.weight","block_attn_res.54.proj.weight","block_attn_res.57.proj.weight","block_attn_res.6.proj.weight","block_attn_res.62.proj.weight","block_attn_res.8.proj.weight","embed.weight::last[336:384]","layers.1.norm.weight","layers.10._fast_impl.k_conv1d.weight","layers.10._fast_impl.v_conv1d.weight","layers.13.norm.weight","layers.14._fast_impl.q_conv1d.weight","layers.15.shared_experts.0.up_proj.weight","layers.16._fast_impl.g_proj.1.bias","layers.16._fast_impl.k_conv1d.weight","layers.16._fast_impl.o_norm.weight","layers.16._fast_impl.v_conv1d.weight","layers.16._fast_impl.v_proj.weight","layers.2._fast_impl.e_proj.1.weight","layers.2._fast_impl.q_conv1d.weight","layers.2.norm.weight","layers.22._fast_impl.k_conv1d.weight","layers.22._fast_impl.v_conv1d.weight","layers.23.latent_up.weight","layers.24._fast_impl.b_proj.bias","layers.24._fast_impl.dt_bias","layers.24._fast_impl.o_norm.weight","layers.24._fast_impl.q_conv1d.weight","layers.26.q_proj.weight","layers.28.kv_latent_norm.weight","layers.29.norm.weight","layers.30._fast_impl.g_proj.1.weight","layers.30._fast_impl.k_conv1d.weight","layers.30._fast_impl.v_conv1d.weight","layers.32._fast_impl.c_proj.weight","layers.32._fast_impl.g_proj.1.bias","layers.32._fast_impl.q_conv1d.weight","layers.32._fast_impl.v_proj.weight","layers.36._fast_impl.A_log","layers.36._fast_impl.k_conv1d.weight","layers.36._fast_impl.o_norm.weight","layers.36._fast_impl.v_conv1d.weight","layers.36.norm.weight","layers.37.shared_experts.0.up_proj.weight","layers.4._fast_impl.e_proj.1.weight","layers.4._fast_impl.q_conv1d.weight","layers.40._fast_impl.c_proj.bias","layers.40._fast_impl.dt_bias","layers.40._fast_impl.k_conv1d.weight","layers.40._fast_impl.v_conv1d.weight","layers.44._fast_impl.g_proj.1.bias","layers.44._fast_impl.q_conv1d.weight","layers.45.latent_up.weight","layers.46._fast_impl.k_conv1d.weight","layers.46._fast_impl.o_norm.weight","layers.46._fast_impl.v_conv1d.weight","layers.48.norm.weight","layers.52._fast_impl.b_proj.bias","layers.52._fast_impl.v_proj.weight","layers.53.norm.weight","layers.54._fast_impl.v_conv1d.weight","layers.56._fast_impl.e_proj.1.weight","layers.57.norm.weight","layers.58.kv_latent_norm.weight","layers.59.shared_experts.0.up_proj.weight","layers.6._fast_impl.A_log","layers.6._fast_impl.g_proj.1.weight","layers.6._fast_impl.v_conv1d.weight","layers.60._fast_impl.c_proj.weight","layers.60.norm.weight","layers.9.latent_up.weight"],"numel":21571501,"offsets":[0,1152,2304,3456,4608,5760,6912,8064,9216,10368,11520,12672,13824,14976,16128,17280,18432,19584,20736,21888,23040,24192,25344,26496,6188928,6190080,6194688,6199296,6200448,6205056,8859264,8860416,8865024,8865152,8869760,10196864,10362752,10367360,10368512,10373120,10377728,10820096,10820105,10821257,10821385,10825993,12005641,12005897,12007049,12154505,12159113,12163721,12174089,12175241,12179849,13506953,13506962,13511570,13511698,13516306,13517458,16171666,16337554,16342162,16342171,16343323,16347931,16352539,16353691,16358299,16800667,16805275,16805403,16810011,16811163,16811172,18138276,18139428,18144036,18309924,18311076,18311332,20965540,20965549,21113005,21117613,21127981,21129133],"shapes":[[1152],[1,1152],[1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[1,1152],[128384,48],[1152],[1152,1,4],[1152,1,4],[1152],[1152,1,4],[2304,1152],[1152],[1152,1,4],[128],[1152,1,4],[1152,1152],[1152,144],[1152,1,4],[1152],[1152,1,4],[1152,1,4],[1152,384],[9],[1152],[128],[1152,1,4],[1024,1152],[256],[1152],[1152,128],[1152,1,4],[1152,1,4],[9,1152],[1152],[1152,1,4],[1152,1152],[9],[1152,1,4],[128],[1152,1,4],[1152],[2304,1152],[1152,144],[1152,1,4],[9],[1152],[1152,1,4],[1152,1,4],[1152],[1152,1,4],[1152,384],[1152,1,4],[128],[1152,1,4],[1152],[9],[1152,1152],[1152],[1152,1,4],[1152,144],[1152],[256],[2304,1152],[9],[1152,128],[1152,1,4],[9,1152],[1152],[1152,384]]},"8":{"classes":["rda","avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":8,"key_order":["block_attn_res_head.proj.weight","embed.weight::last[384:432]","layers.0.k_proj.weight","layers.10._fast_impl.v_proj.weight","layers.17.shared_experts.0.down_proj.weight","layers.2._fast_impl.k_proj.weight","layers.24._fast_impl.e_proj.0.weight","layers.25.latent_down.weight","layers.34.o_proj.weight","layers.36._fast_impl.b_proj.weight","layers.36._fast_impl.k_proj.weight","layers.38.kv_down_proj.weight","layers.39.shared_experts.0.down_proj.weight","layers.44._fast_impl.e_proj.0.weight","layers.47.latent_down.weight","layers.54._fast_impl.k_proj.weight","layers.60._fast_impl.e_proj.0.weight","layers.60._fast_impl.g_proj.0.weight","layers.61.shared_experts.0.down_proj.weight"],"numel":22744320,"offsets":[0,1152,6163584,6458496,7785600,10439808,11766912,11932800,12375168,13554816,13565184,14892288,15187200,17841408,18007296,18449664,19776768,19942656,20090112],"shapes":[[1,1152],[128384,48],[256,1152],[1152,1152],[1152,2304],[1152,1152],[144,1152],[384,1152],[1152,1024],[9,1152],[1152,1152],[256,1152],[1152,2304],[144,1152],[384,1152],[1152,1152],[144,1152],[128,1152],[1152,2304]]},"9":{"classes":["avg","avg","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda","rda"],"fragment_id":9,"key_order":["embed.weight::last[1104:1152]","embed.weight::last[432:480]","layers.0.v_proj.weight","layers.17.shared_experts.0.up_proj.weight","layers.2._fast_impl.o_proj.weight","layers.24._fast_impl.e_proj.1.weight","layers.25.latent_up.weight","layers.34.q_proj.weight","layers.36._fast_impl.c_proj.weight","layers.36._fast_impl.o_proj.weight","layers.39.shared_experts.0.up_proj.weight","layers.42.k_proj.weight","layers.44._fast_impl.e_proj.1.weight","layers.47.latent_up.weight","layers.54._fast_impl.o_proj.weight","layers.60._fast_impl.e_proj.1.weight","layers.60._fast_impl.g_proj.1.weight","layers.61.shared_experts.0.up_proj.weight"],"numel":27578496,"offsets":[0,6162432,12324864,12619776,15273984,16601088,16766976,17209344,18388992,18399360,19726464,22380672,22675584,22841472,23283840,24610944,24776832,24924288],"shapes":[[128384,48],[128384,48],[256,1152],[2304,1152],[1152,1152],[1152,144],[1152,384],[1024,1152],[9,1152],[1152,1152],[2304,1152],[256,1152],[1152,144],[1152,384],[1152,1152],[1152,144],[1152,128],[2304,1152]]}}
exports/165B-tokens_step15861/manifest.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2645379a93ed54062a38d17002fdd64d9311c05c8df00f458d47f0b75213327b
3
+ size 181689805
exports/165B-tokens_step15861/model_config.json ADDED
@@ -0,0 +1 @@
 
 
1
+ {"adaptive_loop_after":"R","adaptive_loop_block_depth":1,"adaptive_loop_block_positions":null,"adaptive_loop_count":0,"adaptive_loop_depth":1,"adaptive_loop_positions":null,"atlas_memory_after":"R*W","atlas_memory_count":0,"atlas_memory_dim":null,"atlas_memory_gate_init":-5.0,"atlas_memory_positions":null,"atlas_memory_segment_size":64,"block_attn_res_block_size":null,"block_attn_res_n_blocks":8,"csa_compress_ratio":4,"csa_index_chunk_size":8192,"csa_index_head_dim":128,"csa_index_n_heads":64,"csa_index_topk":512,"d_model":1152,"d_state":64,"dense_activation":"relu2","dense_intermediate_size":null,"dsa_attention_chunk_size":32,"dsa_index_chunk_size":64,"dsa_index_head_dim":128,"dsa_index_n_heads":32,"dsa_index_topk":2048,"eda_decay_rank":null,"eda_erase_rank":16,"eda_gate_lower":-5.0,"eda_kernel_backend":"flashqla_eda","eda_onorm_eps":1e-06,"eda_scale_mode":"dk^-0.5","expand":2,"expert_activation":"relu2","expert_clip_val":8.0,"expert_family_core_gamma":0.0,"expert_family_core_intermediate_size":0,"expert_family_core_lanes":0,"expert_int8":false,"expert_ternary":true,"final_norm_unit_gain":true,"gdn2_beta_max":1.0,"gdn2_chunk_size":64,"gdn2_decay_bias":2.0,"gdn2_erase_bias":0.0,"gdn2_factored_projections":"","gdn2_factored_rank":0,"gdn2_factored_residual":false,"gdn2_head_dim":128,"gdn2_min_decay":0.0,"gdn2_n_heads":9,"gdn2_onorm_eps":1e-06,"gdn2_onorm_rms_floor":0.0,"gdn2_write_bias":0.0,"gdn2_write_gate_groups":0,"global_attention_rope":true,"global_attention_schedule":"","global_attention_type":"msa","hca_compress_ratio":128,"hca_local_window_size":null,"hca_o_groups":8,"hca_o_lora_rank":null,"hca_q_lora_rank":null,"hca_rope_head_dim":null,"hca_use_tilelang":true,"head_dim":64,"headdim":64,"hybrid_override_pattern":"WERERERE*EREWERERE*EWEREREWE*EREREWERE*EREWERERE*EWERERERE*EREWE","latent_shared_gdn2_chunk_size":64,"latent_shared_gdn2_head_dim":64,"latent_shared_mixer":"none","latent_shared_mixer_gate_init":-4.0,"logit_scale_learnable":true,"m2rnn_gradient_clipping":1.0,"m2rnn_kernel_size":4,"m2rnn_key_head_dim":64,"m2rnn_num_heads":null,"m2rnn_value_head_dim":16,"mamba3_chunk_size":16,"mamba3_mimo":false,"mamba3_mimo_pattern":null,"mamba3_mimo_rank":4,"mamba3_rope_fraction":0.5,"mamba3_siso_chunk_size":64,"max_seq_len":262144,"mhc_init_mode":"identity","mhc_n_streams":4,"mhc_sinkhorn_iters":20,"mla_kv_latent_dim":null,"moe_hash_counts_path":null,"moe_hash_counts_sha256":null,"moe_hash_seed":42,"moe_intermediate_size":2304,"moe_latent_dim":384,"moe_router_bias_update_rate":0.001,"moe_router_dim":null,"moe_router_fp32":true,"moe_router_input_source":"prelatent","moe_router_load_balancing":"quantile","moe_router_min_owned_selections":0,"moe_router_normalize_input":false,"moe_router_prototype_temperature":1.0,"moe_router_prototypes_per_expert":1,"moe_router_quantile_ema":0.9,"moe_router_selection_mode":"global_topk","moe_router_semantic_aligned_rows":0,"moe_router_semantic_free_rows":0,"moe_router_semantic_lanes":0,"moe_router_type":"linear","moe_router_z_loss_coef":0.0001,"msa_block_size":128,"msa_force_local_blocks":1,"msa_head_dim":128,"msa_index_loss_every":1,"msa_index_loss_tokens":256,"msa_index_loss_weight":0.01,"msa_index_query_chunk":256,"msa_index_rank":null,"msa_index_warmup_steps":1000000000,"msa_kv_latent_dim":256,"msa_sparse_topk":16,"mtp_loss_weight":0.3,"mtp_n_heads":0,"mtp_shared_weights":true,"n_kv_heads":4,"n_q_heads":16,"n_routed_experts":128,"n_shared_experts":1,"num_experts_per_tok":12,"pinned_swa_windows":[[0,2048],[62,2048]],"recurrent_backend":"eda","rope_theta":1000000.0,"routed_scaling_factor":1.0,"shared_expert_intermediate_size":null,"shared_expert_output_scale":1.0,"shared_expert_use_routed_latent":false,"shared_expert_zero_init_output":true,"shared_latent_dim":null,"sliding_attention_rope":true,"sliding_attention_type":"gqa","sliding_window_size":2048,"tie_word_embeddings":true,"tied_input_scale":null,"use_block_attn_res":true,"use_mamba3":false,"use_mhc":false,"vocab_size":128384}
exports/165B-tokens_step15861/packed_experts.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1b3c8a7217efd8d1656f5d7c3334457c729de26e26574c810e8fd83f371cc1d1
3
+ size 1210705920
exports/165B-tokens_step15861/relay_pack/manifest.json ADDED
The diff for this file is too large to render. See raw diff
 
exports/165B-tokens_step15861/relay_pack/trunk/f0.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4898282dc966cf4a02cc312f58f9d1e0107f49473e08bb7392e039af93a53b68
3
+ size 45507072
exports/165B-tokens_step15861/relay_pack/trunk/f1.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ed38ef0bd9a12c51a63343df7e9ce50c09e24ba1548b7fc973df8e385eb1c1ec
3
+ size 47571456
exports/165B-tokens_step15861/relay_pack/trunk/f10.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:449294c8d34d2fbe6dc801ff0077c4c77af1bc9dedca63a490ee7616cefcc45d
3
+ size 43172608
exports/165B-tokens_step15861/relay_pack/trunk/f11.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dfe0020e2cc043efa15bec01615d15f26bddf70cf28452cfa29fb63541636238
3
+ size 43143002
exports/165B-tokens_step15861/relay_pack/trunk/f12.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b0b856e239d0539985ce9c1c03f272705e58ca52303016fbf3c87f1fe88d1773
3
+ size 43143002
exports/165B-tokens_step15861/relay_pack/trunk/f13.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:859acbff6e94292478da588de828bf9256ba9ff753145da1b57d4061f8e25fae
3
+ size 43143002
exports/165B-tokens_step15861/relay_pack/trunk/f14.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6325b130fc708b510a9a3911bef4799ed5276653c912089474b371395d3aa95c
3
+ size 43143002
exports/165B-tokens_step15861/relay_pack/trunk/f15.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:538ad8ba73a7b15d8ca7fb3883ddfe721914faa83ddfa3f4f74ed3202262b9d3
3
+ size 43143002
exports/165B-tokens_step15861/relay_pack/trunk/f16.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4ed154e7e611d371233cff34e9480e07f8fee7a485b38f7efb8397bf16d3743
3
+ size 43721472
exports/165B-tokens_step15861/relay_pack/trunk/f17.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b7a77b6bce8f4bc89c46c9499f8e444b740964f97d8898e3b338e843a7dea7e1
3
+ size 43143000
exports/165B-tokens_step15861/relay_pack/trunk/f18.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3056074928dad28b4685a7eff8e6cf1296bbe94e17e1d0cf0e32b46e6d27c2bc
3
+ size 43168000
exports/165B-tokens_step15861/relay_pack/trunk/f19.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e431dc029281c3d3b9117a35e5052d391b4566dedbe6af2c1246bdf4119b40ec
3
+ size 43720960
exports/165B-tokens_step15861/relay_pack/trunk/f2.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:68e8e9cb94b1c06e5ab8200da1d5e62543f9398b373afce31b52883d15cd4f2a
3
+ size 43143002
exports/165B-tokens_step15861/relay_pack/trunk/f20.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4b0df5bb6aab1c5745a1c0c1ee0d8deb39a8ddaceb97ffa20d9da87ee2810434
3
+ size 43168000
exports/165B-tokens_step15861/relay_pack/trunk/f21.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:263f0826e4bf06530d0eab1839c41bd9053a565dc2481699ea4c7b37dd402239
3
+ size 45490432
exports/165B-tokens_step15861/relay_pack/trunk/f22.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cde6d22a7e7cd6f4f6d0240894459fce036093f6a0be00ebe624375d95d70b2d
3
+ size 48144640
exports/165B-tokens_step15861/relay_pack/trunk/f23.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6c9dba952e535f46975be3e1d2b93275b2770abb8573d583b13fc4116d29c5fb
3
+ size 18907136
exports/165B-tokens_step15861/relay_pack/trunk/f3.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc0e013cb02177fba442b89932aa52bc51380fc8e4d39c0551afc9ee3d9ee5f6
3
+ size 48161280
exports/165B-tokens_step15861/relay_pack/trunk/f4.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ff1ef1bab22278d48d52a308e2c9f90335e603c19fed94d06421e0bdc66e59d
3
+ size 43737600
exports/165B-tokens_step15861/relay_pack/trunk/f5.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2638c8289e3e0b200e5a6012339116efb50bbc64772c8d222cd52ee8588be954
3
+ size 43184640
exports/165B-tokens_step15861/relay_pack/trunk/f6.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2d952ce1ebb65c5f9ab1075bb984701cf7c8ed11d5f38d1e2f05136c03ff7ba1
3
+ size 45507072
exports/165B-tokens_step15861/relay_pack/trunk/f7.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e43feb78456c6cefe42b4936731527abb670d47881648e740e8673a244ef25cc
3
+ size 43143002
exports/165B-tokens_step15861/relay_pack/trunk/f8.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bb7143b83e906e10218c82d8921cefae4a99a483cf39cfb0d691a85c5e3f7e05
3
+ size 45488640
exports/165B-tokens_step15861/relay_pack/trunk/f9.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5336d45b00227b2ac5ac4e8439d249228c8b5fd0448ab3fc5535cee3a3b04cd1
3
+ size 55156992