eagle_offline

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 11.0588
• Loss Class 1: 11.0144
• Loss Class 2: 11.0493
• Loss Class 3: 11.0751
• Loss Class 4: 11.0964
• Top 1 Acc: 0.1509
• Top 2 Acc: 0.2279
• Top 3 Acc: 0.2835

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: 0.0001
• train_batch_size: 1
• eval_batch_size: 1
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• total_eval_batch_size: 8
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.95) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: constant
• lr_scheduler_warmup_ratio: 0.01
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Loss Class 1	Loss Class 2	Loss Class 3	Loss Class 4	Top 1 Acc	Top 2 Acc	Top 3 Acc
2.8296	0.1268	256	11.4931	11.4932	11.4766	11.4933	11.5095	0.0201	0.0336	0.0411
2.8011	0.2536	512	11.3786	11.3594	11.3658	11.3854	11.4040	0.0426	0.0650	0.0810
2.7246	0.3804	768	11.3304	11.2980	11.3166	11.3418	11.3651	0.0527	0.0792	0.1006
2.8211	0.5072	1024	11.2962	11.2578	11.2831	11.3108	11.3331	0.0634	0.0953	0.1188
2.7285	0.6341	1280	11.2761	11.2310	11.2628	11.2930	11.3177	0.0688	0.1062	0.1319
2.8096	0.7609	1536	11.2537	11.2060	11.2392	11.2715	11.2982	0.0769	0.1182	0.1458
2.7645	0.8877	1792	11.2351	11.1856	11.2213	11.2531	11.2803	0.0792	0.1240	0.1564
2.7203	1.0144	2048	11.2247	11.1752	11.2115	11.2434	11.2689	0.0853	0.1269	0.1610
2.7811	1.1412	2304	11.2101	11.1601	11.1972	11.2287	11.2546	0.0892	0.1367	0.1696
2.7113	1.2680	2560	11.2020	11.1495	11.1886	11.2218	11.2482	0.0953	0.1450	0.1819
2.7748	1.3948	2816	11.1913	11.1390	11.1780	11.2108	11.2375	0.0965	0.1479	0.1867
2.7932	1.5216	3072	11.1860	11.1336	11.1727	11.2058	11.2319	0.0978	0.1503	0.1886
2.7951	1.6484	3328	11.1811	11.1283	11.1688	11.2006	11.2266	0.1011	0.1532	0.1921
2.7271	1.7752	3584	11.1716	11.1191	11.1593	11.1912	11.2167	0.1041	0.1613	0.2000
2.7682	1.9020	3840	11.1704	11.1156	11.1571	11.1908	11.2178	0.1042	0.1638	0.2044
2.7439	2.0287	4096	11.1628	11.1088	11.1508	11.1828	11.2088	0.1093	0.1698	0.2080
2.764	2.1555	4352	11.1579	11.1041	11.1459	11.1779	11.2037	0.1093	0.1685	0.2119
2.7638	2.2824	4608	11.1517	11.0984	11.1395	11.1717	11.1972	0.1117	0.1741	0.2151
2.7852	2.4092	4864	11.1495	11.0953	11.1372	11.1695	11.1961	0.1106	0.1754	0.2204
2.7414	2.5360	5120	11.1444	11.0918	11.1324	11.1642	11.1892	0.1171	0.1810	0.2247
2.7618	2.6628	5376	11.1400	11.0874	11.1279	11.1595	11.1850	0.1180	0.1816	0.2240
2.7808	2.7896	5632	11.1377	11.0842	11.1254	11.1578	11.1835	0.1208	0.1858	0.2308
2.6931	2.9164	5888	11.1345	11.0807	11.1222	11.1551	11.1801	0.1227	0.1825	0.2265
2.7398	3.0431	6144	11.1328	11.0783	11.1210	11.1532	11.1785	0.1254	0.1865	0.2295
2.7563	3.1699	6400	11.1335	11.0756	11.1197	11.1551	11.1835	0.1249	0.1896	0.2331
2.7295	3.2967	6656	11.1267	11.0740	11.1148	11.1465	11.1715	0.1277	0.1916	0.2379
2.7348	3.4235	6912	11.1250	11.0707	11.1126	11.1454	11.1712	0.1268	0.1924	0.2376
2.6711	3.5503	7168	11.1212	11.0678	11.1088	11.1410	11.1672	0.1272	0.1920	0.2400
2.6863	3.6772	7424	11.1202	11.0667	11.1087	11.1400	11.1652	0.1281	0.1977	0.2402
2.7103	3.8040	7680	11.1163	11.0631	11.1043	11.1362	11.1617	0.1306	0.1978	0.2448
2.7503	3.9308	7936	11.1173	11.0637	11.1057	11.1370	11.1627	0.1299	0.1992	0.2443
2.7507	4.0575	8192	11.1110	11.0586	11.0995	11.1305	11.1553	0.1305	0.1961	0.2407
2.7319	4.1843	8448	11.1104	11.0576	11.0987	11.1302	11.1550	0.1352	0.2012	0.2463
2.7322	4.3111	8704	11.1089	11.0555	11.0968	11.1290	11.1544	0.1323	0.2033	0.2494
2.7094	4.4379	8960	11.1082	11.0546	11.0956	11.1283	11.1542	0.1322	0.2026	0.2507
2.7303	4.5647	9216	11.1037	11.0515	11.0925	11.1232	11.1477	0.1355	0.2046	0.2502
2.7736	4.6915	9472	11.1032	11.0508	11.0918	11.1226	11.1474	0.1375	0.2067	0.2567
2.685	4.8183	9728	11.1020	11.0491	11.0901	11.1215	11.1473	0.1339	0.2044	0.2527
2.7081	4.9451	9984	11.0981	11.0475	11.0863	11.1168	11.1418	0.1354	0.2070	0.2532
2.7004	5.0718	10240	11.0967	11.0462	11.0853	11.1154	11.1397	0.1384	0.2090	0.2568
2.7702	5.1986	10496	11.0953	11.0451	11.0837	11.1136	11.1386	0.1379	0.2100	0.2575
2.7267	5.3254	10752	11.0952	11.0433	11.0838	11.1142	11.1396	0.1391	0.2107	0.2601
2.7498	5.4523	11008	11.0931	11.0421	11.0820	11.1120	11.1362	0.1407	0.2102	0.2593
2.7271	5.5791	11264	11.0919	11.0407	11.0807	11.1108	11.1354	0.1383	0.2114	0.2620
2.7056	5.7059	11520	11.0921	11.0403	11.0801	11.1113	11.1366	0.1388	0.2084	0.2592
2.7489	5.8327	11776	11.0899	11.0400	11.0783	11.1085	11.1328	0.1395	0.2144	0.2636
2.7454	5.9595	12032	11.0901	11.0388	11.0793	11.1089	11.1333	0.1412	0.2126	0.2648
2.7009	6.0862	12288	11.0841	11.0349	11.0731	11.1024	11.1260	0.1423	0.2180	0.2681
2.7442	6.2130	12544	11.0846	11.0353	11.0733	11.1027	11.1272	0.1439	0.2174	0.2668
2.7453	6.3398	12800	11.0837	11.0345	11.0730	11.1019	11.1255	0.1416	0.2169	0.2654
2.7436	6.4666	13056	11.0828	11.0332	11.0716	11.1011	11.1252	0.1427	0.2193	0.2681
2.6979	6.5934	13312	11.0820	11.0324	11.0708	11.1002	11.1246	0.1476	0.2199	0.2692
2.6779	6.7202	13568	11.0817	11.0326	11.0704	11.0997	11.1242	0.1428	0.2179	0.2702
2.7241	6.8471	13824	11.0789	11.0299	11.0680	11.0970	11.1209	0.1466	0.2184	0.2682
2.721	6.9739	14080	11.0756	11.0283	11.0648	11.0931	11.1161	0.1468	0.2179	0.2694
2.7441	7.1006	14336	11.0788	11.0301	11.0678	11.0968	11.1205	0.1423	0.2156	0.2642
2.7423	7.2274	14592	11.0772	11.0281	11.0661	11.0951	11.1194	0.1440	0.2195	0.2722
2.6967	7.3542	14848	11.0735	11.0266	11.0635	11.0909	11.1131	0.1451	0.2210	0.2721
2.7645	7.4810	15104	11.0721	11.0250	11.0616	11.0897	11.1120	0.1483	0.2223	0.2727
2.7416	7.6078	15360	11.0740	11.0253	11.0627	11.0922	11.1160	0.1466	0.2191	0.2742
2.7447	7.7346	15616	11.0721	11.0241	11.0619	11.0898	11.1127	0.1473	0.2182	0.2737
2.7438	7.8614	15872	11.0709	11.0240	11.0602	11.0883	11.1110	0.1498	0.2248	0.2749
2.7192	7.9882	16128	11.0699	11.0230	11.0596	11.0871	11.1101	0.1499	0.2236	0.2761
2.7192	8.1149	16384	11.0685	11.0220	11.0584	11.0855	11.1080	0.1470	0.2224	0.2751
2.7207	8.2417	16640	11.0683	11.0220	11.0579	11.0855	11.1078	0.1459	0.2194	0.2745
2.7204	8.3685	16896	11.0673	11.0204	11.0570	11.0844	11.1074	0.1483	0.2226	0.2739
2.7655	8.4954	17152	11.0665	11.0199	11.0564	11.0838	11.1058	0.1487	0.2276	0.2769
2.7396	8.6222	17408	11.0668	11.0198	11.0559	11.0842	11.1072	0.1480	0.2256	0.2781
2.6544	8.7490	17664	11.0666	11.0195	11.0564	11.0840	11.1066	0.1490	0.2264	0.2793
2.7408	8.8758	17920	11.0658	11.0191	11.0558	11.0829	11.1054	0.1497	0.2268	0.2782
2.7409	9.0025	18176	11.0651	11.0183	11.0549	11.0821	11.1052	0.1460	0.2253	0.2779
2.7403	9.1293	18432	11.0641	11.0177	11.0542	11.0811	11.1033	0.1501	0.2237	0.2775
2.737	9.2561	18688	11.0629	11.0172	11.0529	11.0797	11.1018	0.1509	0.2250	0.2807
2.6977	9.3829	18944	11.0628	11.0169	11.0529	11.0794	11.1018	0.1491	0.2252	0.2794
2.7396	9.5097	19200	11.0633	11.0167	11.0529	11.0804	11.1031	0.1502	0.2278	0.2785
2.7177	9.6365	19456	11.0603	11.0158	11.0504	11.0765	11.0984	0.1533	0.2296	0.2818
2.7369	9.7633	19712	11.0593	11.0149	11.0498	11.0755	11.0970	0.1527	0.2299	0.2852
2.7607	9.8902	19968	11.0588	11.0144	11.0493	11.0751	11.0964	0.1509	0.2279	0.2835

Framework versions

• Transformers 4.53.0
• Pytorch 2.7.1+cu126
• Datasets 4.0.0
• Tokenizers 0.21.2