BioPhy/match_repo_o3_prod

modernbert-llm-router

This model is a fine-tuned version of answerdotai/ModernBERT-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.7888
• F1: 0.7785
• Macro F1: 0.7785
• Precision: 0.7832
• Cross Entropy: 0.5213
• Min Class Accuracy: 0.658
• Confusion Matrix: [[938, 56, 6], [232, 658, 110], [20, 232, 748]]
• Accuracy Class 0: 0.938
• Accuracy Class 1: 0.658
• Accuracy Class 2: 0.748

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• train_batch_size: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 16
• total_train_batch_size: 64
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Macro F1	Precision	Cross Entropy	Min Class Accuracy	Confusion Matrix	Accuracy Class 0	Accuracy Class 1	Accuracy Class 2
67.1584	0.0939	4000	4.3133	0.6905	0.6905	0.7174	0.6445	0.49	[[964, 32, 4], [427, 490, 83], [80, 274, 646]]	0.964	0.49	0.646
66.8186	0.0986	4200	4.2763	0.7056	0.7056	0.7332	0.6243	0.57	[[951, 45, 4], [370, 570, 60], [62, 325, 613]]	0.951	0.57	0.613
66.666	0.1033	4400	4.3091	0.6693	0.6693	0.7083	0.6652	0.534	[[959, 39, 2], [400, 542, 58], [71, 395, 534]]	0.959	0.542	0.534
66.5493	0.1080	4600	4.2950	0.6930	0.6930	0.7279	0.6567	0.531	[[937, 58, 5], [314, 628, 58], [29, 440, 531]]	0.937	0.628	0.531
66.4201	0.1127	4800	4.2264	0.7317	0.7317	0.7578	0.5987	0.583	[[935, 62, 3], [250, 687, 63], [23, 394, 583]]	0.935	0.687	0.583
66.2139	0.1174	5000	4.3305	0.6664	0.6664	0.6993	0.6979	0.501	[[969, 28, 3], [423, 501, 76], [73, 366, 561]]	0.969	0.501	0.561
65.8371	0.1221	5200	4.2472	0.6927	0.6927	0.7244	0.6322	0.562	[[968, 30, 2], [375, 562, 63], [60, 368, 572]]	0.968	0.562	0.572
65.7941	0.1268	5400	4.2528	0.7054	0.7054	0.7398	0.6475	0.553	[[958, 40, 2], [323, 625, 52], [53, 394, 553]]	0.958	0.625	0.553
65.5811	0.1314	5600	4.2208	0.7232	0.7232	0.7341	0.6087	0.57	[[941, 52, 7], [320, 570, 110], [27, 298, 675]]	0.941	0.57	0.675
65.4718	0.1361	5800	4.1907	0.7218	0.7218	0.7446	0.5945	0.589	[[948, 48, 4], [282, 643, 75], [36, 375, 589]]	0.948	0.643	0.589
65.2497	0.1408	6000	4.2049	0.7034	0.7034	0.7274	0.6252	0.591	[[947, 47, 6], [330, 591, 79], [69, 340, 591]]	0.947	0.591	0.591
65.1997	0.1455	6200	4.2574	0.6679	0.6679	0.7118	0.6690	0.526	[[968, 29, 3], [431, 526, 43], [69, 393, 538]]	0.968	0.526	0.538
65.0097	0.1502	6400	4.1857	0.7235	0.7235	0.7612	0.6106	0.552	[[950, 50, 0], [277, 685, 38], [59, 389, 552]]	0.95	0.685	0.552
64.8219	0.1549	6600	4.2046	0.7092	0.7092	0.7356	0.6392	0.584	[[958, 38, 4], [329, 604, 67], [40, 376, 584]]	0.958	0.604	0.584
64.7263	0.1596	6800	4.2920	0.6220	0.6220	0.7073	0.7250	0.355	[[955, 45, 0], [360, 615, 25], [67, 578, 355]]	0.955	0.615	0.355
64.6495	0.1643	7000	4.1895	0.6843	0.6843	0.7222	0.6356	0.537	[[957, 42, 1], [364, 582, 54], [57, 406, 537]]	0.957	0.582	0.537
64.3158	0.1690	7200	4.2260	0.6940	0.6940	0.7358	0.6636	0.521	[[959, 38, 3], [331, 626, 43], [65, 414, 521]]	0.959	0.626	0.521
64.3062	0.1737	7400	4.1442	0.7167	0.7167	0.7442	0.6048	0.599	[[954, 41, 5], [331, 613, 56], [56, 345, 599]]	0.954	0.613	0.599
64.0348	0.1784	7600	4.1174	0.7345	0.7345	0.7454	0.5703	0.583	[[955, 36, 9], [304, 583, 113], [46, 270, 684]]	0.955	0.583	0.684
64.0196	0.1831	7800	4.1144	0.7247	0.7247	0.7462	0.5859	0.613	[[952, 46, 2], [301, 624, 75], [50, 337, 613]]	0.952	0.624	0.613
63.8908	0.1878	8000	4.1876	0.6717	0.6717	0.7164	0.6808	0.5	[[949, 49, 2], [361, 591, 48], [46, 454, 500]]	0.949	0.591	0.5
63.7917	0.1925	8200	4.1457	0.6920	0.6920	0.7347	0.6319	0.506	[[938, 59, 3], [302, 652, 46], [40, 454, 506]]	0.938	0.652	0.506
63.6911	0.1972	8400	4.1293	0.7201	0.7201	0.7390	0.6186	0.619	[[935, 61, 4], [297, 619, 84], [49, 332, 619]]	0.935	0.619	0.619
63.4114	0.2019	8600	4.1279	0.6899	0.6899	0.7372	0.6263	0.501	[[959, 41, 0], [327, 635, 38], [52, 447, 501]]	0.959	0.635	0.501
63.312	0.2066	8800	4.1076	0.7273	0.7273	0.7630	0.6074	0.559	[[944, 56, 0], [267, 692, 41], [34, 407, 559]]	0.944	0.692	0.559
63.2253	0.2113	9000	4.0827	0.7207	0.7207	0.7399	0.5908	0.587	[[955, 42, 3], [331, 587, 82], [45, 318, 637]]	0.955	0.587	0.637
63.1791	0.2159	9200	4.1021	0.7144	0.7144	0.7459	0.6179	0.582	[[957, 42, 1], [328, 621, 51], [58, 360, 582]]	0.957	0.621	0.582
62.9892	0.2206	9400	4.0988	0.6994	0.6994	0.7354	0.6213	0.568	[[965, 33, 2], [368, 586, 46], [59, 373, 568]]	0.965	0.586	0.568
62.9596	0.2253	9600	4.0299	0.7349	0.7349	0.7547	0.5633	0.622	[[953, 46, 1], [309, 622, 69], [33, 325, 642]]	0.953	0.622	0.642
62.7207	0.2300	9800	4.1452	0.6911	0.6911	0.7252	0.6585	0.543	[[966, 32, 2], [404, 543, 53], [64, 349, 587]]	0.966	0.543	0.587
62.6702	0.2347	10000	4.0368	0.7287	0.7287	0.7537	0.5715	0.602	[[947, 50, 3], [292, 649, 59], [32, 366, 602]]	0.947	0.649	0.602
62.5446	0.2394	10200	4.0780	0.7032	0.7032	0.7319	0.6159	0.576	[[962, 37, 1], [362, 576, 62], [44, 365, 591]]	0.962	0.576	0.591
62.3668	0.2441	10400	4.1327	0.6841	0.6841	0.7314	0.6766	0.491	[[959, 38, 3], [331, 629, 40], [39, 470, 491]]	0.959	0.629	0.491
62.2441	0.2488	10600	4.0393	0.7446	0.7446	0.7581	0.5807	0.645	[[951, 43, 6], [265, 645, 90], [30, 320, 650]]	0.951	0.645	0.65
62.2008	0.2535	10800	3.9862	0.7569	0.7569	0.7644	0.5375	0.612	[[952, 44, 4], [273, 612, 115], [30, 249, 721]]	0.952	0.612	0.721
62.1253	0.2582	11000	4.0344	0.7083	0.7083	0.7336	0.6014	0.558	[[964, 32, 4], [376, 558, 66], [49, 328, 623]]	0.964	0.558	0.623
61.9099	0.2629	11200	4.0261	0.7336	0.7336	0.7616	0.5951	0.583	[[937, 58, 5], [256, 692, 52], [40, 377, 583]]	0.937	0.692	0.583
61.7465	0.2676	11400	4.1156	0.6506	0.6506	0.7096	0.6897	0.457	[[967, 33, 0], [403, 563, 34], [41, 502, 457]]	0.967	0.563	0.457
61.7948	0.2723	11600	4.0768	0.6828	0.6828	0.7276	0.6630	0.526	[[963, 36, 1], [380, 583, 37], [49, 425, 526]]	0.963	0.583	0.526
61.583	0.2770	11800	4.0815	0.6942	0.6942	0.7298	0.6597	0.57	[[962, 37, 1], [379, 570, 51], [71, 357, 572]]	0.962	0.57	0.572
61.4462	0.2817	12000	4.0190	0.7266	0.7266	0.7531	0.6087	0.596	[[952, 47, 1], [297, 645, 58], [36, 368, 596]]	0.952	0.645	0.596
61.1997	0.2864	12200	3.9755	0.7559	0.7559	0.7765	0.5705	0.623	[[949, 48, 3], [234, 704, 62], [15, 362, 623]]	0.949	0.704	0.623
61.2348	0.2911	12400	3.9294	0.7562	0.7562	0.7719	0.5367	0.657	[[955, 43, 2], [271, 657, 72], [34, 299, 667]]	0.955	0.657	0.667
61.0238	0.2958	12600	4.0619	0.6856	0.6856	0.7384	0.6624	0.473	[[951, 49, 0], [305, 658, 37], [17, 510, 473]]	0.951	0.658	0.473
61.0242	0.3005	12800	3.9927	0.6979	0.6979	0.7304	0.6148	0.571	[[967, 31, 2], [368, 577, 55], [40, 389, 571]]	0.967	0.577	0.571
60.8857	0.3051	13000	3.9335	0.7601	0.7601	0.7726	0.5507	0.668	[[937, 59, 4], [234, 682, 84], [22, 310, 668]]	0.937	0.682	0.668
60.7716	0.3098	13200	4.0305	0.7096	0.7096	0.7453	0.6505	0.565	[[960, 39, 1], [337, 621, 42], [39, 396, 565]]	0.96	0.621	0.565
60.6762	0.3145	13400	3.9114	0.7727	0.7727	0.7840	0.5417	0.686	[[921, 77, 2], [203, 713, 84], [15, 299, 686]]	0.921	0.713	0.686
60.5466	0.3192	13600	3.9843	0.7247	0.7247	0.7476	0.6131	0.601	[[958, 41, 1], [296, 631, 73], [42, 357, 601]]	0.958	0.631	0.601
60.5036	0.3239	13800	3.9504	0.7298	0.7298	0.7534	0.5888	0.593	[[948, 49, 3], [271, 661, 68], [28, 379, 593]]	0.948	0.661	0.593
60.3629	0.3286	14000	3.9785	0.7168	0.7168	0.7382	0.6287	0.544	[[965, 33, 2], [381, 544, 75], [44, 294, 662]]	0.965	0.544	0.662
60.259	0.3333	14200	3.8927	0.7358	0.7358	0.7532	0.5535	0.631	[[943, 55, 2], [292, 631, 77], [30, 326, 644]]	0.943	0.631	0.644
60.0233	0.3380	14400	3.9874	0.7001	0.7001	0.7249	0.6405	0.536	[[968, 28, 4], [394, 536, 70], [32, 349, 619]]	0.968	0.536	0.619
59.9092	0.3427	14600	3.9597	0.7173	0.7173	0.7462	0.6215	0.571	[[943, 53, 4], [291, 651, 58], [23, 406, 571]]	0.943	0.651	0.571
59.827	0.3474	14800	3.9515	0.7201	0.7201	0.7480	0.6193	0.589	[[952, 46, 2], [309, 634, 57], [44, 367, 589]]	0.952	0.634	0.589
59.7204	0.3521	15000	3.8741	0.7626	0.7626	0.7738	0.5534	0.666	[[929, 65, 6], [210, 698, 92], [20, 314, 666]]	0.929	0.698	0.666
59.667	0.3568	15200	3.9475	0.7208	0.7208	0.7517	0.6281	0.553	[[940, 57, 3], [257, 683, 60], [23, 424, 553]]	0.94	0.683	0.553
59.542	0.3615	15400	3.9530	0.6913	0.6913	0.7331	0.6452	0.532	[[961, 39, 0], [357, 602, 41], [32, 436, 532]]	0.961	0.602	0.532
59.4635	0.3662	15600	3.9513	0.7037	0.7037	0.7344	0.6504	0.582	[[965, 32, 3], [360, 582, 58], [61, 354, 585]]	0.965	0.582	0.585
59.2113	0.3709	15800	3.8991	0.7461	0.7461	0.7663	0.5940	0.632	[[947, 51, 2], [269, 668, 63], [21, 347, 632]]	0.947	0.668	0.632
59.2237	0.3756	16000	3.9010	0.7422	0.7422	0.7663	0.5978	0.611	[[943, 55, 2], [263, 681, 56], [17, 372, 611]]	0.943	0.681	0.611
59.0031	0.3803	16200	3.9380	0.7139	0.7139	0.7373	0.6298	0.573	[[961, 37, 2], [357, 573, 70], [43, 331, 626]]	0.961	0.573	0.626
59.0405	0.3850	16400	3.8858	0.7261	0.7261	0.7552	0.6002	0.583	[[950, 48, 2], [290, 658, 52], [25, 392, 583]]	0.95	0.658	0.583
58.8338	0.3896	16600	3.8715	0.7270	0.7270	0.7510	0.5955	0.595	[[946, 51, 3], [280, 653, 67], [38, 367, 595]]	0.946	0.653	0.595
58.8746	0.3943	16800	3.9528	0.7002	0.7002	0.7382	0.6745	0.553	[[956, 42, 2], [349, 610, 41], [45, 402, 553]]	0.956	0.61	0.553
58.6048	0.3990	17000	3.8640	0.7368	0.7368	0.7687	0.5965	0.575	[[938, 61, 1], [248, 707, 45], [23, 402, 575]]	0.938	0.707	0.575
58.4862	0.4037	17200	3.9524	0.6945	0.6945	0.7404	0.6891	0.511	[[952, 46, 2], [323, 642, 35], [40, 449, 511]]	0.952	0.642	0.511
58.4176	0.4084	17400	3.7888	0.7785	0.7785	0.7832	0.5213	0.658	[[938, 56, 6], [232, 658, 110], [20, 232, 748]]	0.938	0.658	0.748
58.3256	0.4131	17600	3.8355	0.7427	0.7427	0.7628	0.5795	0.626	[[946, 52, 2], [266, 666, 68], [28, 346, 626]]	0.946	0.666	0.626
58.193	0.4178	17800	3.9014	0.7114	0.7114	0.7458	0.6464	0.551	[[951, 48, 1], [300, 649, 51], [37, 412, 551]]	0.951	0.649	0.551
58.0348	0.4225	18000	3.9505	0.6624	0.6624	0.7268	0.7058	0.441	[[961, 39, 0], [354, 620, 26], [35, 524, 441]]	0.961	0.62	0.441
57.9114	0.4272	18200	3.9220	0.7076	0.7076	0.7400	0.6741	0.577	[[964, 35, 1], [350, 600, 50], [37, 386, 577]]	0.964	0.6	0.577
57.9024	0.4319	18400	3.8897	0.6970	0.6970	0.7320	0.6580	0.551	[[964, 34, 2], [351, 597, 52], [35, 414, 551]]	0.964	0.597	0.551
57.819	0.4366	18600	3.8781	0.7186	0.7186	0.7462	0.6470	0.587	[[959, 39, 2], [315, 626, 59], [35, 378, 587]]	0.959	0.626	0.587
57.7407	0.4413	18800	3.9488	0.6797	0.6797	0.7243	0.7260	0.498	[[960, 38, 2], [345, 608, 47], [44, 458, 498]]	0.96	0.608	0.498
57.5802	0.4460	19000	3.8100	0.7301	0.7301	0.7501	0.6038	0.619	[[952, 46, 2], [293, 632, 75], [22, 359, 619]]	0.952	0.632	0.619
57.4922	0.4507	19200	3.8570	0.7275	0.7275	0.7461	0.6478	0.616	[[953, 46, 1], [301, 616, 83], [39, 333, 628]]	0.953	0.616	0.628
57.3413	0.4554	19400	3.8799	0.6979	0.6979	0.7352	0.6764	0.539	[[951, 47, 2], [328, 623, 49], [44, 417, 539]]	0.951	0.623	0.539
57.2835	0.4601	19600	3.7841	0.7574	0.7574	0.7780	0.5797	0.618	[[928, 69, 3], [200, 732, 68], [24, 358, 618]]	0.928	0.732	0.618
57.0308	0.4648	19800	3.8292	0.7367	0.7367	0.7564	0.6354	0.617	[[942, 54, 4], [266, 661, 73], [21, 362, 617]]	0.942	0.661	0.617
56.9914	0.4695	20000	3.8292	0.7219	0.7219	0.7428	0.6366	0.601	[[951, 46, 3], [324, 601, 75], [52, 319, 629]]	0.951	0.601	0.629
56.868	0.4741	20200	3.8733	0.7062	0.7062	0.7367	0.6846	0.553	[[946, 53, 1], [299, 636, 65], [34, 413, 553]]	0.946	0.636	0.553
56.6672	0.4788	20400	3.8836	0.7048	0.7048	0.7404	0.6927	0.548	[[955, 44, 1], [318, 631, 51], [58, 394, 548]]	0.955	0.631	0.548
56.7036	0.4835	20600	3.8409	0.7219	0.7219	0.7546	0.6626	0.562	[[952, 47, 1], [282, 668, 50], [49, 389, 562]]	0.952	0.668	0.562
56.5975	0.4882	20800	3.8347	0.7205	0.7205	0.7446	0.6557	0.565	[[968, 29, 3], [368, 565, 67], [61, 291, 648]]	0.968	0.565	0.648
56.4269	0.4929	21000	3.7917	0.7427	0.7427	0.7637	0.6239	0.63	[[954, 44, 2], [283, 655, 62], [26, 344, 630]]	0.954	0.655	0.63
56.4064	0.4976	21200	3.8596	0.6763	0.6763	0.7197	0.7084	0.531	[[979, 20, 1], [426, 531, 43], [73, 379, 548]]	0.979	0.531	0.548
56.3241	0.5023	21400	3.7867	0.7141	0.7141	0.7496	0.6337	0.545	[[952, 48, 0], [288, 662, 50], [31, 424, 545]]	0.952	0.662	0.545

Framework versions

• Transformers 4.51.3
• Pytorch 2.6.0+cu126
• Datasets 3.5.1
• Tokenizers 0.21.1