{ "capsule_id": "quantal-1786648501", "address": { "role": "matrix.ternary.quantal", "tags": [ "quantal", "bitnet", "ternary", "ayeos" ] }, "payload_type": "ternary_matrix", "metadata": { "base_model": "Qwen/Qwen2.5-0.5B", "reconstruction_model": "Qwen/Qwen2.5-0.5B", "binary": false, "group_size": 64, "epochs": 7, "final_loss": null, "val_loss": 2.1469, "checkpoint_sha256": "21294c68f05f36fcf72a25246caea92cac6664173f684328a2aad72f9a988285", "checkpoint_size_bytes": 988097722, "checkpoint_source": "vast.ai remote GPU run", "export_tool": "MLX-QUANT fork mlx 0.32.0 (ternary quantize)", "export_complete": true, "best_val": 2.1469, "quantizer": "thresholded-ternary (per-group 64, zero state)", "deployed_forward": true }, "timestamp": 1786648501, "matrices": [ { "file": "m000.json", "name": "model.layers.23.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4283849 }, { "file": "m001.json", "name": "model.layers.23.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4291889 }, { "file": "m002.json", "name": "model.layers.23.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4285700 }, { "file": "m003.json", "name": "model.layers.23.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 791246 }, { "file": "m004.json", "name": "model.layers.23.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112940 }, { "file": "m005.json", "name": "model.layers.23.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112976 }, { "file": "m006.json", "name": "model.layers.23.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789248 }, { "file": "m007.json", "name": "model.layers.22.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4283364 }, { "file": "m008.json", "name": "model.layers.22.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4284475 }, { "file": "m009.json", "name": "model.layers.22.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4286000 }, { "file": "m010.json", "name": "model.layers.22.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789446 }, { "file": "m011.json", "name": "model.layers.22.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113063 }, { "file": "m012.json", "name": "model.layers.22.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112856 }, { "file": "m013.json", "name": "model.layers.22.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789687 }, { "file": "m014.json", "name": "model.layers.21.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4284121 }, { "file": "m015.json", "name": "model.layers.21.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4284953 }, { "file": "m016.json", "name": "model.layers.21.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4283901 }, { "file": "m017.json", "name": "model.layers.21.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789628 }, { "file": "m018.json", "name": "model.layers.21.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112950 }, { "file": "m019.json", "name": "model.layers.21.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113017 }, { "file": "m020.json", "name": "model.layers.21.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789474 }, { "file": "m021.json", "name": "model.layers.20.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4284952 }, { "file": "m022.json", "name": "model.layers.20.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4287399 }, { "file": "m023.json", "name": "model.layers.20.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4285130 }, { "file": "m024.json", "name": "model.layers.20.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790081 }, { "file": "m025.json", "name": "model.layers.20.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113142 }, { "file": "m026.json", "name": "model.layers.20.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112983 }, { "file": "m027.json", "name": "model.layers.20.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789223 }, { "file": "m028.json", "name": "model.layers.19.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4284935 }, { "file": "m029.json", "name": "model.layers.19.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4286460 }, { "file": "m030.json", "name": "model.layers.19.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4285327 }, { "file": "m031.json", "name": "model.layers.19.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790102 }, { "file": "m032.json", "name": "model.layers.19.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112852 }, { "file": "m033.json", "name": "model.layers.19.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112820 }, { "file": "m034.json", "name": "model.layers.19.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 788718 }, { "file": "m035.json", "name": "model.layers.18.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4284249 }, { "file": "m036.json", "name": "model.layers.18.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4285141 }, { "file": "m037.json", "name": "model.layers.18.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4284392 }, { "file": "m038.json", "name": "model.layers.18.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789565 }, { "file": "m039.json", "name": "model.layers.18.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112943 }, { "file": "m040.json", "name": "model.layers.18.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112860 }, { "file": "m041.json", "name": "model.layers.18.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789324 }, { "file": "m042.json", "name": "model.layers.17.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4286501 }, { "file": "m043.json", "name": "model.layers.17.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4287563 }, { "file": "m044.json", "name": "model.layers.17.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4283296 }, { "file": "m045.json", "name": "model.layers.17.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789650 }, { "file": "m046.json", "name": "model.layers.17.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113007 }, { "file": "m047.json", "name": "model.layers.17.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112789 }, { "file": "m048.json", "name": "model.layers.17.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 788791 }, { "file": "m049.json", "name": "model.layers.16.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4286106 }, { "file": "m050.json", "name": "model.layers.16.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4287227 }, { "file": "m051.json", "name": "model.layers.16.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4287373 }, { "file": "m052.json", "name": "model.layers.16.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790179 }, { "file": "m053.json", "name": "model.layers.16.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113393 }, { "file": "m054.json", "name": "model.layers.16.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112940 }, { "file": "m055.json", "name": "model.layers.16.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 788518 }, { "file": "m056.json", "name": "model.layers.15.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4284568 }, { "file": "m057.json", "name": "model.layers.15.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4286564 }, { "file": "m058.json", "name": "model.layers.15.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4287605 }, { "file": "m059.json", "name": "model.layers.15.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 791086 }, { "file": "m060.json", "name": "model.layers.15.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113029 }, { "file": "m061.json", "name": "model.layers.15.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112838 }, { "file": "m062.json", "name": "model.layers.15.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789211 }, { "file": "m063.json", "name": "model.layers.14.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4286014 }, { "file": "m064.json", "name": "model.layers.14.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4287565 }, { "file": "m065.json", "name": "model.layers.14.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4289299 }, { "file": "m066.json", "name": "model.layers.14.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790126 }, { "file": "m067.json", "name": "model.layers.14.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113008 }, { "file": "m068.json", "name": "model.layers.14.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112760 }, { "file": "m069.json", "name": "model.layers.14.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 788672 }, { "file": "m070.json", "name": "model.layers.13.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4286759 }, { "file": "m071.json", "name": "model.layers.13.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4287817 }, { "file": "m072.json", "name": "model.layers.13.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4290133 }, { "file": "m073.json", "name": "model.layers.13.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 791172 }, { "file": "m074.json", "name": "model.layers.13.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113060 }, { "file": "m075.json", "name": "model.layers.13.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113010 }, { "file": "m076.json", "name": "model.layers.13.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789360 }, { "file": "m077.json", "name": "model.layers.12.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4286455 }, { "file": "m078.json", "name": "model.layers.12.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4288611 }, { "file": "m079.json", "name": "model.layers.12.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4288963 }, { "file": "m080.json", "name": "model.layers.12.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790874 }, { "file": "m081.json", "name": "model.layers.12.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112930 }, { "file": "m082.json", "name": "model.layers.12.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112910 }, { "file": "m083.json", "name": "model.layers.12.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789234 }, { "file": "m084.json", "name": "model.layers.11.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4288316 }, { "file": "m085.json", "name": "model.layers.11.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4289811 }, { "file": "m086.json", "name": "model.layers.11.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4289740 }, { "file": "m087.json", "name": "model.layers.11.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790900 }, { "file": "m088.json", "name": "model.layers.11.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113203 }, { "file": "m089.json", "name": "model.layers.11.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112960 }, { "file": "m090.json", "name": "model.layers.11.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789600 }, { "file": "m091.json", "name": "model.layers.10.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4288354 }, { "file": "m092.json", "name": "model.layers.10.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4290137 }, { "file": "m093.json", "name": "model.layers.10.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4286596 }, { "file": "m094.json", "name": "model.layers.10.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790474 }, { "file": "m095.json", "name": "model.layers.10.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113058 }, { "file": "m096.json", "name": "model.layers.10.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112780 }, { "file": "m097.json", "name": "model.layers.10.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 788981 }, { "file": "m098.json", "name": "model.layers.9.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4287611 }, { "file": "m099.json", "name": "model.layers.9.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4288391 }, { "file": "m100.json", "name": "model.layers.9.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4286311 }, { "file": "m101.json", "name": "model.layers.9.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 791026 }, { "file": "m102.json", "name": "model.layers.9.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113140 }, { "file": "m103.json", "name": "model.layers.9.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112931 }, { "file": "m104.json", "name": "model.layers.9.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789139 }, { "file": "m105.json", "name": "model.layers.8.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4286962 }, { "file": "m106.json", "name": "model.layers.8.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4288661 }, { "file": "m107.json", "name": "model.layers.8.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4284535 }, { "file": "m108.json", "name": "model.layers.8.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790995 }, { "file": "m109.json", "name": "model.layers.8.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113006 }, { "file": "m110.json", "name": "model.layers.8.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112708 }, { "file": "m111.json", "name": "model.layers.8.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 788538 }, { "file": "m112.json", "name": "model.layers.7.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4286897 }, { "file": "m113.json", "name": "model.layers.7.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4289267 }, { "file": "m114.json", "name": "model.layers.7.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4286201 }, { "file": "m115.json", "name": "model.layers.7.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790112 }, { "file": "m116.json", "name": "model.layers.7.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113157 }, { "file": "m117.json", "name": "model.layers.7.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112737 }, { "file": "m118.json", "name": "model.layers.7.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 788546 }, { "file": "m119.json", "name": "model.layers.6.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4287456 }, { "file": "m120.json", "name": "model.layers.6.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4288495 }, { "file": "m121.json", "name": "model.layers.6.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4283248 }, { "file": "m122.json", "name": "model.layers.6.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790785 }, { "file": "m123.json", "name": "model.layers.6.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113059 }, { "file": "m124.json", "name": "model.layers.6.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112700 }, { "file": "m125.json", "name": "model.layers.6.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 788944 }, { "file": "m126.json", "name": "model.layers.5.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4289282 }, { "file": "m127.json", "name": "model.layers.5.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4290870 }, { "file": "m128.json", "name": "model.layers.5.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4282189 }, { "file": "m129.json", "name": "model.layers.5.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790733 }, { "file": "m130.json", "name": "model.layers.5.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113088 }, { "file": "m131.json", "name": "model.layers.5.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112822 }, { "file": "m132.json", "name": "model.layers.5.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 788908 }, { "file": "m133.json", "name": "model.layers.4.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4287369 }, { "file": "m134.json", "name": "model.layers.4.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4288544 }, { "file": "m135.json", "name": "model.layers.4.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4283087 }, { "file": "m136.json", "name": "model.layers.4.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 791090 }, { "file": "m137.json", "name": "model.layers.4.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113054 }, { "file": "m138.json", "name": "model.layers.4.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112876 }, { "file": "m139.json", "name": "model.layers.4.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 788881 }, { "file": "m140.json", "name": "model.layers.3.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4288778 }, { "file": "m141.json", "name": "model.layers.3.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4288291 }, { "file": "m142.json", "name": "model.layers.3.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4284474 }, { "file": "m143.json", "name": "model.layers.3.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790879 }, { "file": "m144.json", "name": "model.layers.3.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113253 }, { "file": "m145.json", "name": "model.layers.3.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112695 }, { "file": "m146.json", "name": "model.layers.3.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 789399 }, { "file": "m147.json", "name": "model.layers.2.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4288930 }, { "file": "m148.json", "name": "model.layers.2.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4289524 }, { "file": "m149.json", "name": "model.layers.2.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4285146 }, { "file": "m150.json", "name": "model.layers.2.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790559 }, { "file": "m151.json", "name": "model.layers.2.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113260 }, { "file": "m152.json", "name": "model.layers.2.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112665 }, { "file": "m153.json", "name": "model.layers.2.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 788637 }, { "file": "m154.json", "name": "model.layers.1.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4287371 }, { "file": "m155.json", "name": "model.layers.1.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4288802 }, { "file": "m156.json", "name": "model.layers.1.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4285779 }, { "file": "m157.json", "name": "model.layers.1.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 790705 }, { "file": "m158.json", "name": "model.layers.1.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113107 }, { "file": "m159.json", "name": "model.layers.1.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112283 }, { "file": "m160.json", "name": "model.layers.1.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 787914 }, { "file": "m161.json", "name": "model.layers.0.mlp.up_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4282928 }, { "file": "m162.json", "name": "model.layers.0.mlp.down_proj", "dim": 896, "in_features": 4864, "group_size": 64, "bytes": 4286473 }, { "file": "m163.json", "name": "model.layers.0.mlp.gate_proj", "dim": 4864, "in_features": 896, "group_size": 64, "bytes": 4274228 }, { "file": "m164.json", "name": "model.layers.0.self_attn.o_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 791378 }, { "file": "m165.json", "name": "model.layers.0.self_attn.v_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 113281 }, { "file": "m166.json", "name": "model.layers.0.self_attn.k_proj", "dim": 128, "in_features": 896, "group_size": 64, "bytes": 112192 }, { "file": "m167.json", "name": "model.layers.0.self_attn.q_proj", "dim": 896, "in_features": 896, "group_size": 64, "bytes": 784160 } ], "assets": { "embeddings.f16": { "dtype": "f16", "shape": [ 151936, 896 ], "bytes": 272269312, "sha256": "9e62c98c3b6394db95596181c17f4c78afb490da7e1933d693c839b3c472d696", "source": "/tmp/nightly-best.safetensors model.embed_tokens.weight (BF16->FP16)" }, "norms.f32": { "dtype": "f32", "shape": [ 49, 896 ], "bytes": 175616, "sha256": "cd984997de4f99d1d57a23e4add309272615ee079f7c0daf1881fb9f7bda6c44", "source": "/tmp/nightly-best.safetensors RMSNorm weights (BF16->FP32)" } } }