File size: 213,962 Bytes
08276da |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 3075 3076 3077 3078 3079 3080 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125 3126 3127 3128 3129 3130 3131 3132 3133 3134 3135 3136 3137 3138 3139 3140 3141 3142 3143 3144 3145 3146 3147 3148 3149 3150 3151 3152 3153 3154 3155 3156 3157 3158 3159 3160 3161 3162 3163 3164 3165 3166 3167 3168 3169 3170 3171 3172 3173 3174 3175 3176 3177 3178 3179 3180 3181 3182 3183 3184 3185 3186 3187 3188 3189 3190 3191 3192 3193 3194 3195 3196 3197 3198 3199 3200 3201 3202 3203 3204 3205 3206 3207 3208 3209 3210 3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222 3223 3224 3225 3226 3227 3228 3229 3230 3231 3232 3233 3234 3235 3236 3237 3238 3239 3240 3241 3242 3243 3244 3245 3246 3247 3248 3249 3250 3251 3252 3253 3254 3255 3256 3257 3258 3259 3260 3261 3262 3263 3264 3265 3266 3267 3268 3269 3270 3271 3272 3273 3274 3275 3276 3277 3278 3279 3280 3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318 3319 3320 3321 3322 3323 3324 3325 3326 3327 3328 3329 3330 3331 3332 3333 3334 3335 3336 3337 3338 3339 3340 3341 3342 3343 3344 3345 3346 3347 3348 3349 3350 3351 3352 3353 3354 3355 3356 3357 3358 3359 3360 3361 3362 3363 3364 3365 3366 3367 3368 3369 3370 3371 3372 3373 3374 3375 3376 3377 3378 3379 3380 3381 3382 3383 3384 3385 3386 3387 3388 3389 3390 3391 3392 3393 3394 3395 3396 3397 3398 3399 3400 3401 3402 3403 3404 3405 3406 3407 3408 3409 3410 3411 3412 3413 3414 3415 3416 3417 3418 3419 3420 3421 3422 3423 3424 3425 3426 3427 3428 3429 3430 3431 3432 3433 3434 3435 3436 3437 3438 3439 3440 3441 3442 3443 3444 3445 3446 3447 3448 3449 3450 3451 3452 3453 3454 3455 |
2025-10-14 02:33:37,074 - train - INFO - ConformerModel(
(linear_layer_1): Linear(in_features=128, out_features=256, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
(conformer_blocks): Sequential(
(0): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(1): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(2): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(3): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(4): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(5): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(6): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(7): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(8): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(9): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
)
(linear_layer_2): Linear(in_features=256, out_features=28, bias=True)
)
2025-10-14 02:41:40,845 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 02:43:53,269 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 02:53:22,675 - train - INFO - epoch : 1
2025-10-14 02:53:22,676 - train - INFO - loss : 1.8004077727144414
2025-10-14 02:53:22,676 - train - INFO - grad_norm : 1.1968199670916857
2025-10-14 02:53:22,676 - train - INFO - val_loss : 1.467006968049442
2025-10-14 02:53:22,676 - train - INFO - val_CER_(Argmax): 0.46757403776657935
2025-10-14 02:53:22,676 - train - INFO - val_WER_(Argmax): 0.91038943818923
2025-10-14 02:53:22,676 - train - INFO - val_WER : 0.91038943818923
2025-10-14 02:53:22,677 - train - INFO - val_CER : 0.4676158105801009
2025-10-14 02:53:22,677 - train - INFO - test_loss : 1.430062472820282
2025-10-14 02:53:22,677 - train - INFO - test_CER_(Argmax): 0.4589772069767894
2025-10-14 02:53:22,677 - train - INFO - test_WER_(Argmax): 0.906752747897352
2025-10-14 02:53:22,677 - train - INFO - test_WER : 0.906752747897352
2025-10-14 02:53:22,677 - train - INFO - test_CER : 0.45909848600936254
2025-10-14 02:53:23,214 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 03:03:20,978 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 03:05:39,207 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 03:13:20,217 - train - INFO - epoch : 2
2025-10-14 03:13:20,217 - train - INFO - loss : 0.9497057100636276
2025-10-14 03:13:20,217 - train - INFO - grad_norm : 1.2964627569045253
2025-10-14 03:13:20,218 - train - INFO - val_loss : 0.9703210683429943
2025-10-14 03:13:20,218 - train - INFO - val_CER_(Argmax): 0.3104969995818503
2025-10-14 03:13:20,218 - train - INFO - val_WER_(Argmax): 0.7484133288798358
2025-10-14 03:13:20,218 - train - INFO - val_WER : 0.7484133288798358
2025-10-14 03:13:20,218 - train - INFO - val_CER : 0.30930462548815907
2025-10-14 03:13:20,218 - train - INFO - test_loss : 0.9425025051686822
2025-10-14 03:13:20,218 - train - INFO - test_CER_(Argmax): 0.30274019803097907
2025-10-14 03:13:20,218 - train - INFO - test_WER_(Argmax): 0.7372494591821148
2025-10-14 03:13:20,218 - train - INFO - test_WER : 0.7372494591821148
2025-10-14 03:13:20,219 - train - INFO - test_CER : 0.30157728834330844
2025-10-14 03:13:20,885 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 03:25:20,451 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 03:33:10,510 - train - INFO - epoch : 3
2025-10-14 03:33:10,510 - train - INFO - loss : 0.7622273603081703
2025-10-14 03:33:10,510 - train - INFO - grad_norm : 1.1402055078744888
2025-10-14 03:33:10,511 - train - INFO - val_loss : 0.8829054685199962
2025-10-14 03:33:10,511 - train - INFO - val_CER_(Argmax): 0.28195997427460867
2025-10-14 03:33:10,511 - train - INFO - val_WER_(Argmax): 0.686879415725692
2025-10-14 03:33:10,511 - train - INFO - val_WER : 0.686879415725692
2025-10-14 03:33:10,511 - train - INFO - val_CER : 0.28205835407842134
2025-10-14 03:33:10,511 - train - INFO - test_loss : 0.8557483960942525
2025-10-14 03:33:10,511 - train - INFO - test_CER_(Argmax): 0.2709563942448364
2025-10-14 03:33:10,511 - train - INFO - test_WER_(Argmax): 0.6692670294343694
2025-10-14 03:33:10,511 - train - INFO - test_WER : 0.6692670294343694
2025-10-14 03:33:10,512 - train - INFO - test_CER : 0.27099345931346924
2025-10-14 03:33:11,148 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 03:53:06,907 - train - INFO - epoch : 4
2025-10-14 03:53:06,908 - train - INFO - loss : 0.602354964017868
2025-10-14 03:53:06,908 - train - INFO - grad_norm : 0.8039740589261055
2025-10-14 03:53:06,908 - train - INFO - val_loss : 0.7007653839447919
2025-10-14 03:53:06,908 - train - INFO - val_CER_(Argmax): 0.22225459871101402
2025-10-14 03:53:06,908 - train - INFO - val_WER_(Argmax): 0.5930371723301392
2025-10-14 03:53:06,908 - train - INFO - val_WER : 0.5930371723301392
2025-10-14 03:53:06,908 - train - INFO - val_CER : 0.22210362264174635
2025-10-14 03:53:06,908 - train - INFO - test_loss : 0.6907694921260927
2025-10-14 03:53:06,908 - train - INFO - test_CER_(Argmax): 0.2163203087967603
2025-10-14 03:53:06,909 - train - INFO - test_WER_(Argmax): 0.5772301195235712
2025-10-14 03:53:06,909 - train - INFO - test_WER : 0.5772301195235712
2025-10-14 03:53:06,909 - train - INFO - test_CER : 0.21606124644889366
2025-10-14 03:53:07,539 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 03:58:05,821 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 04:13:08,317 - train - INFO - epoch : 5
2025-10-14 04:13:08,318 - train - INFO - loss : 0.5589919736981392
2025-10-14 04:13:08,318 - train - INFO - grad_norm : 0.6693760113418102
2025-10-14 04:13:08,318 - train - INFO - val_loss : 0.6744126432082233
2025-10-14 04:13:08,318 - train - INFO - val_CER_(Argmax): 0.21458751555183747
2025-10-14 04:13:08,319 - train - INFO - val_WER_(Argmax): 0.5772959000644565
2025-10-14 04:13:08,319 - train - INFO - val_WER : 0.5772959000644565
2025-10-14 04:13:08,319 - train - INFO - val_CER : 0.2145634022145982
2025-10-14 04:13:08,319 - train - INFO - test_loss : 0.6590640392245316
2025-10-14 04:13:08,319 - train - INFO - test_CER_(Argmax): 0.20839209387247756
2025-10-14 04:13:08,319 - train - INFO - test_WER_(Argmax): 0.5701414952271677
2025-10-14 04:13:08,319 - train - INFO - test_WER : 0.5701414952271677
2025-10-14 04:13:08,320 - train - INFO - test_CER : 0.20839523910387514
2025-10-14 04:13:08,988 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 04:14:24,248 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 04:24:28,763 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 04:33:06,709 - train - INFO - epoch : 6
2025-10-14 04:33:06,709 - train - INFO - loss : 0.510627674460411
2025-10-14 04:33:06,709 - train - INFO - grad_norm : 0.5528287927806378
2025-10-14 04:33:06,710 - train - INFO - val_loss : 0.6022388451239642
2025-10-14 04:33:06,710 - train - INFO - val_CER_(Argmax): 0.18296962699849365
2025-10-14 04:33:06,710 - train - INFO - val_WER_(Argmax): 0.5093214369543156
2025-10-14 04:33:06,710 - train - INFO - val_WER : 0.5093214369543156
2025-10-14 04:33:06,710 - train - INFO - val_CER : 0.18261087096328094
2025-10-14 04:33:06,710 - train - INFO - test_loss : 0.6027363142589244
2025-10-14 04:33:06,710 - train - INFO - test_CER_(Argmax): 0.1820291797987071
2025-10-14 04:33:06,710 - train - INFO - test_WER_(Argmax): 0.5047745633151853
2025-10-14 04:33:06,710 - train - INFO - test_WER : 0.5047745633151853
2025-10-14 04:33:06,710 - train - INFO - test_CER : 0.18158809112179367
2025-10-14 04:33:07,385 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 04:42:03,191 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 04:53:08,991 - train - INFO - epoch : 7
2025-10-14 04:53:08,992 - train - INFO - loss : 0.49359407886188833
2025-10-14 04:53:08,992 - train - INFO - grad_norm : 0.5190753033712282
2025-10-14 04:53:08,992 - train - INFO - val_loss : 0.6014730471021988
2025-10-14 04:53:08,992 - train - INFO - val_CER_(Argmax): 0.181704769539811
2025-10-14 04:53:08,993 - train - INFO - val_WER_(Argmax): 0.5059981667294224
2025-10-14 04:53:08,993 - train - INFO - val_WER : 0.5059981667294224
2025-10-14 04:53:08,993 - train - INFO - val_CER : 0.1815978525961797
2025-10-14 04:53:08,993 - train - INFO - test_loss : 0.5996043998293761
2025-10-14 04:53:08,993 - train - INFO - test_CER_(Argmax): 0.1791391732758246
2025-10-14 04:53:08,993 - train - INFO - test_WER_(Argmax): 0.4997126493658538
2025-10-14 04:53:08,993 - train - INFO - test_WER : 0.4997126493658538
2025-10-14 04:53:08,994 - train - INFO - test_CER : 0.179074904175568
2025-10-14 04:53:09,676 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 04:58:15,378 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 05:13:07,029 - train - INFO - epoch : 8
2025-10-14 05:13:07,030 - train - INFO - loss : 0.46707881450653077
2025-10-14 05:13:07,030 - train - INFO - grad_norm : 0.48182750105857847
2025-10-14 05:13:07,030 - train - INFO - val_loss : 0.6367673782741322
2025-10-14 05:13:07,030 - train - INFO - val_CER_(Argmax): 0.19087030186457948
2025-10-14 05:13:07,030 - train - INFO - val_WER_(Argmax): 0.5220166510054483
2025-10-14 05:13:07,030 - train - INFO - val_WER : 0.5220166510054483
2025-10-14 05:13:07,030 - train - INFO - val_CER : 0.190838427285644
2025-10-14 05:13:07,031 - train - INFO - test_loss : 0.6175299279573487
2025-10-14 05:13:07,031 - train - INFO - test_CER_(Argmax): 0.18446621558922244
2025-10-14 05:13:07,031 - train - INFO - test_WER_(Argmax): 0.5084007759551228
2025-10-14 05:13:07,031 - train - INFO - test_WER : 0.5084007759551228
2025-10-14 05:13:07,031 - train - INFO - test_CER : 0.184502159264417
2025-10-14 05:20:54,798 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 05:33:02,433 - train - INFO - epoch : 9
2025-10-14 05:33:02,434 - train - INFO - loss : 0.45376209877244195
2025-10-14 05:33:02,434 - train - INFO - grad_norm : 0.45064780400626026
2025-10-14 05:33:02,435 - train - INFO - val_loss : 0.5756007843157824
2025-10-14 05:33:02,435 - train - INFO - val_CER_(Argmax): 0.16914851214323548
2025-10-14 05:33:02,435 - train - INFO - val_WER_(Argmax): 0.47379946747058904
2025-10-14 05:33:02,435 - train - INFO - val_WER : 0.47379946747058904
2025-10-14 05:33:02,435 - train - INFO - val_CER : 0.16900721738636365
2025-10-14 05:33:02,435 - train - INFO - test_loss : 0.5721487871757368
2025-10-14 05:33:02,436 - train - INFO - test_CER_(Argmax): 0.16664663014275632
2025-10-14 05:33:02,436 - train - INFO - test_WER_(Argmax): 0.4681725557437241
2025-10-14 05:33:02,436 - train - INFO - test_WER : 0.4681725557437241
2025-10-14 05:33:02,436 - train - INFO - test_CER : 0.16649619330717766
2025-10-14 05:33:03,081 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 05:33:14,650 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 05:43:25,713 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 05:52:54,270 - train - INFO - epoch : 10
2025-10-14 05:52:54,271 - train - INFO - loss : 0.45315693199634555
2025-10-14 05:52:54,271 - train - INFO - grad_norm : 0.4713628640770912
2025-10-14 05:52:54,271 - train - INFO - val_loss : 0.5362565857522628
2025-10-14 05:52:54,271 - train - INFO - val_CER_(Argmax): 0.16329148533434748
2025-10-14 05:52:54,271 - train - INFO - val_WER_(Argmax): 0.46387223589891163
2025-10-14 05:52:54,271 - train - INFO - val_WER : 0.46387223589891163
2025-10-14 05:52:54,271 - train - INFO - val_CER : 0.16286558348739288
2025-10-14 05:52:54,271 - train - INFO - test_loss : 0.5317229174259233
2025-10-14 05:52:54,271 - train - INFO - test_CER_(Argmax): 0.16059709717785184
2025-10-14 05:52:54,271 - train - INFO - test_WER_(Argmax): 0.45693877851173287
2025-10-14 05:52:54,272 - train - INFO - test_WER : 0.45693877851173287
2025-10-14 05:52:54,272 - train - INFO - test_CER : 0.1603862774668369
2025-10-14 05:52:54,905 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 06:12:35,761 - train - INFO - epoch : 11
2025-10-14 06:12:35,762 - train - INFO - loss : 0.39052830457687376
2025-10-14 06:12:35,762 - train - INFO - grad_norm : 0.4450642728060484
2025-10-14 06:12:35,762 - train - INFO - val_loss : 0.51932592532214
2025-10-14 06:12:35,762 - train - INFO - val_CER_(Argmax): 0.15941955542976455
2025-10-14 06:12:35,762 - train - INFO - val_WER_(Argmax): 0.45788092649283985
2025-10-14 06:12:35,763 - train - INFO - val_WER : 0.45788092649283985
2025-10-14 06:12:35,763 - train - INFO - val_CER : 0.15936533181396875
2025-10-14 06:12:35,763 - train - INFO - test_loss : 0.5151232182979584
2025-10-14 06:12:35,763 - train - INFO - test_CER_(Argmax): 0.15455427775801173
2025-10-14 06:12:35,763 - train - INFO - test_WER_(Argmax): 0.44637718160637985
2025-10-14 06:12:35,763 - train - INFO - test_WER : 0.44637718160637985
2025-10-14 06:12:35,763 - train - INFO - test_CER : 0.15453517470721737
2025-10-14 06:12:36,429 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 06:13:48,904 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 06:23:40,844 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 06:32:20,518 - train - INFO - epoch : 12
2025-10-14 06:32:20,518 - train - INFO - loss : 0.34388998910784724
2025-10-14 06:32:20,518 - train - INFO - grad_norm : 0.40590107560157773
2025-10-14 06:32:20,518 - train - INFO - val_loss : 0.5381783243487863
2025-10-14 06:32:20,518 - train - INFO - val_CER_(Argmax): 0.16583726089869136
2025-10-14 06:32:20,519 - train - INFO - val_WER_(Argmax): 0.465543927183187
2025-10-14 06:32:20,519 - train - INFO - val_WER : 0.465543927183187
2025-10-14 06:32:20,519 - train - INFO - val_CER : 0.16531468827653925
2025-10-14 06:32:20,519 - train - INFO - test_loss : 0.5295760235408458
2025-10-14 06:32:20,519 - train - INFO - test_CER_(Argmax): 0.16174688828818035
2025-10-14 06:32:20,519 - train - INFO - test_WER_(Argmax): 0.4550225790980045
2025-10-14 06:32:20,519 - train - INFO - test_WER : 0.4550225790980045
2025-10-14 06:32:20,519 - train - INFO - test_CER : 0.16122770436209705
2025-10-14 06:43:29,714 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 06:52:09,495 - train - INFO - epoch : 13
2025-10-14 06:52:09,496 - train - INFO - loss : 0.34937265180051325
2025-10-14 06:52:09,496 - train - INFO - grad_norm : 0.4043967518210411
2025-10-14 06:52:09,496 - train - INFO - val_loss : 0.48908867941183204
2025-10-14 06:52:09,496 - train - INFO - val_CER_(Argmax): 0.14971865099250545
2025-10-14 06:52:09,497 - train - INFO - val_WER_(Argmax): 0.4332590763851437
2025-10-14 06:52:09,497 - train - INFO - val_WER : 0.4332590763851437
2025-10-14 06:52:09,497 - train - INFO - val_CER : 0.1497014184247999
2025-10-14 06:52:09,497 - train - INFO - test_loss : 0.49085995009759575
2025-10-14 06:52:09,497 - train - INFO - test_CER_(Argmax): 0.1494535796558252
2025-10-14 06:52:09,497 - train - INFO - test_WER_(Argmax): 0.4310647728310606
2025-10-14 06:52:09,497 - train - INFO - test_WER : 0.4310647728310606
2025-10-14 06:52:09,497 - train - INFO - test_CER : 0.14945445213192635
2025-10-14 06:52:10,157 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 06:57:44,296 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 07:11:49,105 - train - INFO - epoch : 14
2025-10-14 07:11:49,105 - train - INFO - loss : 0.32730326041579244
2025-10-14 07:11:49,105 - train - INFO - grad_norm : 0.37786272332072257
2025-10-14 07:11:49,105 - train - INFO - val_loss : 0.5277743637561798
2025-10-14 07:11:49,106 - train - INFO - val_CER_(Argmax): 0.15338718155863898
2025-10-14 07:11:49,106 - train - INFO - val_WER_(Argmax): 0.44184830517379275
2025-10-14 07:11:49,106 - train - INFO - val_WER : 0.44184830517379275
2025-10-14 07:11:49,106 - train - INFO - val_CER : 0.15328479401941472
2025-10-14 07:11:49,106 - train - INFO - test_loss : 0.5241630139147363
2025-10-14 07:11:49,106 - train - INFO - test_CER_(Argmax): 0.14972207345168487
2025-10-14 07:11:49,106 - train - INFO - test_WER_(Argmax): 0.43158279188478127
2025-10-14 07:11:49,106 - train - INFO - test_WER : 0.43158279188478127
2025-10-14 07:11:49,106 - train - INFO - test_CER : 0.14960003782834835
2025-10-14 07:31:25,217 - train - INFO - epoch : 15
2025-10-14 07:31:25,218 - train - INFO - loss : 0.3190571042895317
2025-10-14 07:31:25,219 - train - INFO - grad_norm : 0.3733240906894207
2025-10-14 07:31:25,219 - train - INFO - val_loss : 0.4872164645615746
2025-10-14 07:31:25,219 - train - INFO - val_CER_(Argmax): 0.14572484758853044
2025-10-14 07:31:25,219 - train - INFO - val_WER_(Argmax): 0.4191253071785101
2025-10-14 07:31:25,219 - train - INFO - val_WER : 0.4191253071785101
2025-10-14 07:31:25,219 - train - INFO - val_CER : 0.1452825863232472
2025-10-14 07:31:25,219 - train - INFO - test_loss : 0.48767780676120664
2025-10-14 07:31:25,219 - train - INFO - test_CER_(Argmax): 0.1442184158052019
2025-10-14 07:31:25,219 - train - INFO - test_WER_(Argmax): 0.4156251226875968
2025-10-14 07:31:25,219 - train - INFO - test_WER : 0.4156251226875968
2025-10-14 07:31:25,220 - train - INFO - test_CER : 0.14397894549224005
2025-10-14 07:31:25,873 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 07:32:07,397 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 07:38:29,332 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 07:51:09,085 - train - INFO - epoch : 16
2025-10-14 07:51:09,085 - train - INFO - loss : 0.32288431376218796
2025-10-14 07:51:09,085 - train - INFO - grad_norm : 0.3875794377923012
2025-10-14 07:51:09,085 - train - INFO - val_loss : 0.4824349988909329
2025-10-14 07:51:09,086 - train - INFO - val_CER_(Argmax): 0.1440481174169628
2025-10-14 07:51:09,086 - train - INFO - val_WER_(Argmax): 0.41937533251399106
2025-10-14 07:51:09,086 - train - INFO - val_WER : 0.41937533251399106
2025-10-14 07:51:09,086 - train - INFO - val_CER : 0.14402863422008916
2025-10-14 07:51:09,086 - train - INFO - test_loss : 0.4875093171509301
2025-10-14 07:51:09,086 - train - INFO - test_CER_(Argmax): 0.1435495187572233
2025-10-14 07:51:09,086 - train - INFO - test_WER_(Argmax): 0.41857261206475077
2025-10-14 07:51:09,086 - train - INFO - test_WER : 0.41857261206475077
2025-10-14 07:51:09,086 - train - INFO - test_CER : 0.14345600543528397
2025-10-14 07:53:04,340 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 08:11:00,249 - train - INFO - epoch : 17
2025-10-14 08:11:00,250 - train - INFO - loss : 0.3117878006398678
2025-10-14 08:11:00,250 - train - INFO - grad_norm : 0.37129794418811796
2025-10-14 08:11:00,250 - train - INFO - val_loss : 0.46727428611587074
2025-10-14 08:11:00,251 - train - INFO - val_CER_(Argmax): 0.1398674099185118
2025-10-14 08:11:00,251 - train - INFO - val_WER_(Argmax): 0.4029957532757304
2025-10-14 08:11:00,251 - train - INFO - val_WER : 0.4029957532757304
2025-10-14 08:11:00,251 - train - INFO - val_CER : 0.138428207923737
2025-10-14 08:11:00,251 - train - INFO - test_loss : 0.4605345202655327
2025-10-14 08:11:00,251 - train - INFO - test_CER_(Argmax): 0.13763281168297406
2025-10-14 08:11:00,251 - train - INFO - test_WER_(Argmax): 0.40026093361768017
2025-10-14 08:11:00,251 - train - INFO - test_WER : 0.40026093361768017
2025-10-14 08:11:00,251 - train - INFO - test_CER : 0.13695235893427762
2025-10-14 08:11:00,897 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 08:20:41,969 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 08:31:03,757 - train - INFO - epoch : 18
2025-10-14 08:31:03,757 - train - INFO - loss : 0.3081890671846256
2025-10-14 08:31:03,758 - train - INFO - grad_norm : 0.37016375235576726
2025-10-14 08:31:03,758 - train - INFO - val_loss : 0.4678590574685265
2025-10-14 08:31:03,758 - train - INFO - val_CER_(Argmax): 0.13379779068102082
2025-10-14 08:31:03,758 - train - INFO - val_WER_(Argmax): 0.3922874175956862
2025-10-14 08:31:03,758 - train - INFO - val_WER : 0.3922874175956862
2025-10-14 08:31:03,758 - train - INFO - val_CER : 0.13369347277137167
2025-10-14 08:31:03,758 - train - INFO - test_loss : 0.4725317922307224
2025-10-14 08:31:03,758 - train - INFO - test_CER_(Argmax): 0.13567544579262084
2025-10-14 08:31:03,758 - train - INFO - test_WER_(Argmax): 0.3943974565677792
2025-10-14 08:31:03,758 - train - INFO - test_WER : 0.3943974565677792
2025-10-14 08:31:03,759 - train - INFO - test_CER : 0.13559223799657805
2025-10-14 08:31:04,419 - train - INFO - Saving current best: model_best.pth ...
2025-10-14 08:36:17,211 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 08:50:46,825 - train - INFO - epoch : 19
2025-10-14 08:50:46,825 - train - INFO - loss : 0.2682218397408724
2025-10-14 08:50:46,825 - train - INFO - grad_norm : 0.32187166772782805
2025-10-14 08:50:46,826 - train - INFO - val_loss : 0.47410393462461586
2025-10-14 08:50:46,826 - train - INFO - val_CER_(Argmax): 0.14253732992734036
2025-10-14 08:50:46,826 - train - INFO - val_WER_(Argmax): 0.409711592751532
2025-10-14 08:50:46,826 - train - INFO - val_WER : 0.409711592751532
2025-10-14 08:50:46,826 - train - INFO - val_CER : 0.14251998090573448
2025-10-14 08:50:46,826 - train - INFO - test_loss : 0.470459710170583
2025-10-14 08:50:46,826 - train - INFO - test_CER_(Argmax): 0.14044063030058693
2025-10-14 08:50:46,826 - train - INFO - test_WER_(Argmax): 0.40506228117177334
2025-10-14 08:50:46,826 - train - INFO - test_WER : 0.40506228117177334
2025-10-14 08:50:46,827 - train - INFO - test_CER : 0.1404145806112171
2025-10-14 08:53:19,357 - urllib3.connectionpool - WARNING - Retrying (Retry(total=2, connect=None, read=None, redirect=None, status=None)) after connection broken by 'ReadTimeoutError("HTTPSConnectionPool(host='www.comet.com', port=443): Read timed out. (read timeout=10)")': /clientlib/status-report/update
2025-10-14 08:53:21,310 - urllib3.connectionpool - WARNING - Retrying (Retry(total=2, connect=None, read=None, redirect=None, status=None)) after connection broken by 'ReadTimeoutError("HTTPSConnectionPool(host='www.comet.com', port=443): Read timed out. (read timeout=10)")': /clientlib/rest/v2/write/experiment/output
2025-10-14 08:57:22,238 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 09:02:29,246 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 09:10:42,467 - train - INFO - epoch : 20
2025-10-14 09:10:42,468 - train - INFO - loss : 0.26297851376235487
2025-10-14 09:10:42,468 - train - INFO - grad_norm : 0.36333887211978433
2025-10-14 09:10:42,468 - train - INFO - val_loss : 0.498136385048137
2025-10-14 09:10:42,468 - train - INFO - val_CER_(Argmax): 0.1409829398308886
2025-10-14 09:10:42,468 - train - INFO - val_WER_(Argmax): 0.40756040553729267
2025-10-14 09:10:42,468 - train - INFO - val_WER : 0.40756040553729267
2025-10-14 09:10:42,468 - train - INFO - val_CER : 0.14095097487519742
2025-10-14 09:10:42,468 - train - INFO - test_loss : 0.48370018637761836
2025-10-14 09:10:42,469 - train - INFO - test_CER_(Argmax): 0.1360583083333615
2025-10-14 09:10:42,469 - train - INFO - test_WER_(Argmax): 0.39381178720670307
2025-10-14 09:10:42,469 - train - INFO - test_WER : 0.39381178720670307
2025-10-14 09:10:42,469 - train - INFO - test_CER : 0.1360182842490938
2025-10-14 09:10:42,992 - train - INFO - Saving checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch20.pth ...
2025-10-14 09:11:48,738 - train - INFO - Saving model on keyboard interrupt
2025-10-14 09:11:49,259 - train - INFO - Saving checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch21.pth ...
2025-10-14 23:15:37,958 - train - INFO - ConformerModel(
(linear_layer_1): Linear(in_features=128, out_features=256, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
(conformer_blocks): Sequential(
(0): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(1): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(2): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(3): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(4): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(5): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(6): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(7): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(8): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(9): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
)
(linear_layer_2): Linear(in_features=256, out_features=28, bias=True)
)
2025-10-14 23:15:37,974 - train - INFO - Loading checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch21.pth ...
2025-10-14 23:15:38,389 - train - INFO - Checkpoint loaded. Resume training from epoch 22
2025-10-14 23:23:52,739 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 23:26:06,300 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 23:36:17,410 - train - INFO - epoch : 22
2025-10-14 23:36:17,411 - train - INFO - loss : 0.2934816901129906
2025-10-14 23:36:17,411 - train - INFO - grad_norm : 0.3821600427982783
2025-10-14 23:36:17,411 - train - INFO - val_loss : 0.4874178209725548
2025-10-14 23:36:17,411 - train - INFO - val_CER_(Argmax): 0.13945358821010811
2025-10-14 23:36:17,411 - train - INFO - val_WER_(Argmax): 0.40280195242601624
2025-10-14 23:36:17,412 - train - INFO - val_WER : 0.40280195242601624
2025-10-14 23:36:17,412 - train - INFO - val_CER : 0.13942245845447518
2025-10-14 23:36:17,412 - train - INFO - test_loss : 0.48479962312593694
2025-10-14 23:36:17,412 - train - INFO - test_CER_(Argmax): 0.13631144181611937
2025-10-14 23:36:17,412 - train - INFO - test_WER_(Argmax): 0.3958590163824199
2025-10-14 23:36:17,412 - train - INFO - test_WER : 0.3958590163824199
2025-10-14 23:36:17,412 - train - INFO - test_CER : 0.13629110738251396
2025-10-14 23:46:20,954 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 23:48:42,744 - train - WARNING - OOM on batch. Skipping batch.
2025-10-14 23:56:39,785 - train - INFO - epoch : 23
2025-10-14 23:56:39,786 - train - INFO - loss : 0.2551273109775093
2025-10-14 23:56:39,786 - train - INFO - grad_norm : 0.3137171796368594
2025-10-14 23:56:39,786 - train - INFO - val_loss : 0.46369018204071943
2025-10-14 23:56:39,786 - train - INFO - val_CER_(Argmax): 0.13684675847070465
2025-10-14 23:56:39,786 - train - INFO - val_WER_(Argmax): 0.39398508692774187
2025-10-14 23:56:39,787 - train - INFO - val_WER : 0.39398508692774187
2025-10-14 23:56:39,787 - train - INFO - val_CER : 0.13685407007272402
2025-10-14 23:56:39,787 - train - INFO - test_loss : 0.4581556451029894
2025-10-14 23:56:39,787 - train - INFO - test_CER_(Argmax): 0.1350574156452121
2025-10-14 23:56:39,787 - train - INFO - test_WER_(Argmax): 0.3875261612050869
2025-10-14 23:56:39,787 - train - INFO - test_WER : 0.3875261612050869
2025-10-14 23:56:39,787 - train - INFO - test_CER : 0.13511500315245614
2025-10-15 00:08:52,697 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 00:17:07,562 - train - INFO - epoch : 24
2025-10-15 00:17:07,562 - train - INFO - loss : 0.245378298163414
2025-10-15 00:17:07,563 - train - INFO - grad_norm : 0.3387820851057768
2025-10-15 00:17:07,563 - train - INFO - val_loss : 0.5287014912156498
2025-10-15 00:17:07,563 - train - INFO - val_CER_(Argmax): 0.1539106275040438
2025-10-15 00:17:07,563 - train - INFO - val_WER_(Argmax): 0.42460000884399807
2025-10-15 00:17:07,563 - train - INFO - val_WER : 0.42460000884399807
2025-10-15 00:17:07,563 - train - INFO - val_CER : 0.15297842974722475
2025-10-15 00:17:07,563 - train - INFO - test_loss : 0.5029914589916787
2025-10-15 00:17:07,563 - train - INFO - test_CER_(Argmax): 0.14657635076122771
2025-10-15 00:17:07,564 - train - INFO - test_WER_(Argmax): 0.41203503318541046
2025-10-15 00:17:07,564 - train - INFO - test_WER : 0.41203503318541046
2025-10-15 00:17:07,564 - train - INFO - test_CER : 0.1457842404666151
2025-10-15 00:37:36,383 - train - INFO - epoch : 25
2025-10-15 00:37:36,384 - train - INFO - loss : 0.23428807348012926
2025-10-15 00:37:36,384 - train - INFO - grad_norm : 0.3239304776489735
2025-10-15 00:37:36,384 - train - INFO - val_loss : 0.44715122440282035
2025-10-15 00:37:36,384 - train - INFO - val_CER_(Argmax): 0.12522702466393065
2025-10-15 00:37:36,384 - train - INFO - val_WER_(Argmax): 0.36798136361847184
2025-10-15 00:37:36,385 - train - INFO - val_WER : 0.36798136361847184
2025-10-15 00:37:36,385 - train - INFO - val_CER : 0.12516017792671708
2025-10-15 00:37:36,385 - train - INFO - test_loss : 0.4525063423121848
2025-10-15 00:37:36,385 - train - INFO - test_CER_(Argmax): 0.12637926517100662
2025-10-15 00:37:36,385 - train - INFO - test_WER_(Argmax): 0.3720632078294957
2025-10-15 00:37:36,385 - train - INFO - test_WER : 0.3720632078294957
2025-10-15 00:37:36,385 - train - INFO - test_CER : 0.1263555695987811
2025-10-15 00:37:36,962 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 00:42:33,247 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 00:57:31,683 - train - INFO - epoch : 26
2025-10-15 00:57:31,684 - train - INFO - loss : 0.21956679798662662
2025-10-15 00:57:31,684 - train - INFO - grad_norm : 0.30377400361001494
2025-10-15 00:57:31,684 - train - INFO - val_loss : 0.4440635372610653
2025-10-15 00:57:31,684 - train - INFO - val_CER_(Argmax): 0.12553994948531128
2025-10-15 00:57:31,684 - train - INFO - val_WER_(Argmax): 0.3683140692106527
2025-10-15 00:57:31,684 - train - INFO - val_WER : 0.3683140692106527
2025-10-15 00:57:31,685 - train - INFO - val_CER : 0.12547953415857194
2025-10-15 00:57:31,685 - train - INFO - test_loss : 0.43620326751615945
2025-10-15 00:57:31,685 - train - INFO - test_CER_(Argmax): 0.12184782350276652
2025-10-15 00:57:31,685 - train - INFO - test_WER_(Argmax): 0.36093858955137864
2025-10-15 00:57:31,685 - train - INFO - test_WER : 0.36093858955137864
2025-10-15 00:57:31,685 - train - INFO - test_CER : 0.12182641046616761
2025-10-15 00:58:46,849 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 01:08:54,287 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 01:18:58,969 - train - INFO - epoch : 27
2025-10-15 01:18:58,970 - train - INFO - loss : 0.23763787642121315
2025-10-15 01:18:58,970 - train - INFO - grad_norm : 0.33571442432701587
2025-10-15 01:18:58,970 - train - INFO - val_loss : 0.43435747272828046
2025-10-15 01:18:58,970 - train - INFO - val_CER_(Argmax): 0.12460465956934305
2025-10-15 01:18:58,970 - train - INFO - val_WER_(Argmax): 0.3689147427735292
2025-10-15 01:18:58,970 - train - INFO - val_WER : 0.3689147427735292
2025-10-15 01:18:58,970 - train - INFO - val_CER : 0.12435221723652463
2025-10-15 01:18:58,970 - train - INFO - test_loss : 0.42551243159829116
2025-10-15 01:18:58,971 - train - INFO - test_CER_(Argmax): 0.12051501557557807
2025-10-15 01:18:58,971 - train - INFO - test_WER_(Argmax): 0.35871069551154516
2025-10-15 01:18:58,971 - train - INFO - test_WER : 0.35871069551154516
2025-10-15 01:18:58,971 - train - INFO - test_CER : 0.12040524447063847
2025-10-15 01:28:06,616 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 01:40:43,152 - train - INFO - epoch : 28
2025-10-15 01:40:43,152 - train - INFO - loss : 0.2343284837414871
2025-10-15 01:40:43,152 - train - INFO - grad_norm : 0.32623336228293986
2025-10-15 01:40:43,152 - train - INFO - val_loss : 0.45126108702491313
2025-10-15 01:40:43,153 - train - INFO - val_CER_(Argmax): 0.12340088072525852
2025-10-15 01:40:43,153 - train - INFO - val_WER_(Argmax): 0.36056492495645576
2025-10-15 01:40:43,153 - train - INFO - val_WER : 0.36056492495645576
2025-10-15 01:40:43,153 - train - INFO - val_CER : 0.12344191243753086
2025-10-15 01:40:43,153 - train - INFO - test_loss : 0.44489473068132634
2025-10-15 01:40:43,153 - train - INFO - test_CER_(Argmax): 0.12100421883695911
2025-10-15 01:40:43,153 - train - INFO - test_WER_(Argmax): 0.3557473436617356
2025-10-15 01:40:43,153 - train - INFO - test_WER : 0.3557473436617356
2025-10-15 01:40:43,153 - train - INFO - test_CER : 0.1209957554068205
2025-10-15 01:40:43,788 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 01:45:52,826 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 02:02:15,203 - train - INFO - epoch : 29
2025-10-15 02:02:15,204 - train - INFO - loss : 0.2276789116859436
2025-10-15 02:02:15,204 - train - INFO - grad_norm : 0.31875048592686656
2025-10-15 02:02:15,204 - train - INFO - val_loss : 0.4515169536366182
2025-10-15 02:02:15,204 - train - INFO - val_CER_(Argmax): 0.12343411279442047
2025-10-15 02:02:15,205 - train - INFO - val_WER_(Argmax): 0.35693396607175665
2025-10-15 02:02:15,205 - train - INFO - val_WER : 0.35693396607175665
2025-10-15 02:02:15,205 - train - INFO - val_CER : 0.12296479435758914
2025-10-15 02:02:15,205 - train - INFO - test_loss : 0.4444568338917523
2025-10-15 02:02:15,205 - train - INFO - test_CER_(Argmax): 0.12155681713388337
2025-10-15 02:02:15,205 - train - INFO - test_WER_(Argmax): 0.355160702184518
2025-10-15 02:02:15,206 - train - INFO - test_WER : 0.355160702184518
2025-10-15 02:02:15,206 - train - INFO - test_CER : 0.1213823591345638
2025-10-15 02:02:16,139 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 02:10:06,354 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 02:23:50,140 - train - INFO - epoch : 30
2025-10-15 02:23:50,141 - train - INFO - loss : 0.23757737919912866
2025-10-15 02:23:50,141 - train - INFO - grad_norm : 0.3481074444313145
2025-10-15 02:23:50,141 - train - INFO - val_loss : 0.4355435743051417
2025-10-15 02:23:50,141 - train - INFO - val_CER_(Argmax): 0.12227846859888573
2025-10-15 02:23:50,141 - train - INFO - val_WER_(Argmax): 0.3600137657491089
2025-10-15 02:23:50,141 - train - INFO - val_WER : 0.3600137657491089
2025-10-15 02:23:50,142 - train - INFO - val_CER : 0.12224294816692473
2025-10-15 02:23:50,142 - train - INFO - test_loss : 0.4234782213844904
2025-10-15 02:23:50,142 - train - INFO - test_CER_(Argmax): 0.11799567883835432
2025-10-15 02:23:50,142 - train - INFO - test_WER_(Argmax): 0.3509967284700314
2025-10-15 02:23:50,142 - train - INFO - test_WER : 0.3509967284700314
2025-10-15 02:23:50,142 - train - INFO - test_CER : 0.11798016735938993
2025-10-15 02:23:50,653 - train - INFO - Saving checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch30.pth ...
2025-10-15 02:24:02,264 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 02:34:18,495 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 02:45:15,288 - train - INFO - epoch : 31
2025-10-15 02:45:15,289 - train - INFO - loss : 0.2335624401271343
2025-10-15 02:45:15,289 - train - INFO - grad_norm : 0.33499817200005055
2025-10-15 02:45:15,289 - train - INFO - val_loss : 0.42463123097139244
2025-10-15 02:45:15,289 - train - INFO - val_CER_(Argmax): 0.1214074309645935
2025-10-15 02:45:15,289 - train - INFO - val_WER_(Argmax): 0.36004170998968105
2025-10-15 02:45:15,289 - train - INFO - val_WER : 0.36004170998968105
2025-10-15 02:45:15,289 - train - INFO - val_CER : 0.12133784581497396
2025-10-15 02:45:15,289 - train - INFO - test_loss : 0.4196070029241283
2025-10-15 02:45:15,289 - train - INFO - test_CER_(Argmax): 0.1190913597673928
2025-10-15 02:45:15,289 - train - INFO - test_WER_(Argmax): 0.3543684771257035
2025-10-15 02:45:15,289 - train - INFO - test_WER : 0.3543684771257035
2025-10-15 02:45:15,290 - train - INFO - test_CER : 0.11909188803428304
2025-10-15 03:06:34,594 - train - INFO - epoch : 32
2025-10-15 03:06:34,595 - train - INFO - loss : 0.19390573374927045
2025-10-15 03:06:34,595 - train - INFO - grad_norm : 0.2884246703982353
2025-10-15 03:06:34,595 - train - INFO - val_loss : 0.4251282506129321
2025-10-15 03:06:34,595 - train - INFO - val_CER_(Argmax): 0.11977977920474925
2025-10-15 03:06:34,596 - train - INFO - val_WER_(Argmax): 0.35403802314681054
2025-10-15 03:06:34,596 - train - INFO - val_WER : 0.35403802314681054
2025-10-15 03:06:34,596 - train - INFO - val_CER : 0.11974593748907238
2025-10-15 03:06:34,596 - train - INFO - test_loss : 0.4268707185983658
2025-10-15 03:06:34,596 - train - INFO - test_CER_(Argmax): 0.11750315678556099
2025-10-15 03:06:34,596 - train - INFO - test_WER_(Argmax): 0.34796513829846193
2025-10-15 03:06:34,596 - train - INFO - test_WER : 0.34796513829846193
2025-10-15 03:06:34,596 - train - INFO - test_CER : 0.11745897923149805
2025-10-15 03:06:35,206 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 03:07:47,566 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 03:17:49,399 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 03:27:58,262 - train - INFO - epoch : 33
2025-10-15 03:27:58,263 - train - INFO - loss : 0.18811449602246286
2025-10-15 03:27:58,263 - train - INFO - grad_norm : 0.3027683352679014
2025-10-15 03:27:58,263 - train - INFO - val_loss : 0.4447146780350629
2025-10-15 03:27:58,263 - train - INFO - val_CER_(Argmax): 0.1167911485471282
2025-10-15 03:27:58,263 - train - INFO - val_WER_(Argmax): 0.34777324203356125
2025-10-15 03:27:58,263 - train - INFO - val_WER : 0.34777324203356125
2025-10-15 03:27:58,264 - train - INFO - val_CER : 0.11672586478730562
2025-10-15 03:27:58,264 - train - INFO - test_loss : 0.44641188295876105
2025-10-15 03:27:58,264 - train - INFO - test_CER_(Argmax): 0.11520872571090347
2025-10-15 03:27:58,264 - train - INFO - test_WER_(Argmax): 0.34454126380885003
2025-10-15 03:27:58,264 - train - INFO - test_WER : 0.34454126380885003
2025-10-15 03:27:58,264 - train - INFO - test_CER : 0.11515496969675332
2025-10-15 03:27:58,876 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 03:39:06,057 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 03:49:07,853 - train - INFO - epoch : 34
2025-10-15 03:49:07,854 - train - INFO - loss : 0.19895173616707326
2025-10-15 03:49:07,854 - train - INFO - grad_norm : 0.3446432762593031
2025-10-15 03:49:07,854 - train - INFO - val_loss : 0.438978297920788
2025-10-15 03:49:07,854 - train - INFO - val_CER_(Argmax): 0.12138526768120507
2025-10-15 03:49:07,854 - train - INFO - val_WER_(Argmax): 0.3567880885491568
2025-10-15 03:49:07,854 - train - INFO - val_WER : 0.3567880885491568
2025-10-15 03:49:07,855 - train - INFO - val_CER : 0.12133834485438902
2025-10-15 03:49:07,855 - train - INFO - test_loss : 0.4366752952337265
2025-10-15 03:49:07,855 - train - INFO - test_CER_(Argmax): 0.11889285769529587
2025-10-15 03:49:07,855 - train - INFO - test_WER_(Argmax): 0.3499597095170695
2025-10-15 03:49:07,855 - train - INFO - test_WER : 0.3499597095170695
2025-10-15 03:49:07,855 - train - INFO - test_CER : 0.11886283480129935
2025-10-15 03:54:43,121 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 04:10:10,751 - train - INFO - epoch : 35
2025-10-15 04:10:10,752 - train - INFO - loss : 0.18583452805876732
2025-10-15 04:10:10,752 - train - INFO - grad_norm : 0.2995685636997223
2025-10-15 04:10:10,752 - train - INFO - val_loss : 0.4410898643381455
2025-10-15 04:10:10,752 - train - INFO - val_CER_(Argmax): 0.11951493903695873
2025-10-15 04:10:10,752 - train - INFO - val_WER_(Argmax): 0.3514972872290408
2025-10-15 04:10:10,752 - train - INFO - val_WER : 0.3514972872290408
2025-10-15 04:10:10,752 - train - INFO - val_CER : 0.11950992517253951
2025-10-15 04:10:10,752 - train - INFO - test_loss : 0.43057214050758175
2025-10-15 04:10:10,753 - train - INFO - test_CER_(Argmax): 0.11632696628536299
2025-10-15 04:10:10,753 - train - INFO - test_WER_(Argmax): 0.34279318429294053
2025-10-15 04:10:10,753 - train - INFO - test_WER : 0.34279318429294053
2025-10-15 04:10:10,753 - train - INFO - test_CER : 0.11630289332594408
2025-10-15 04:10:11,240 - train - INFO - Saving checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch35.pth ...
2025-10-15 04:31:14,040 - train - INFO - epoch : 36
2025-10-15 04:31:14,041 - train - INFO - loss : 0.20375656172633172
2025-10-15 04:31:14,041 - train - INFO - grad_norm : 0.3344234121590853
2025-10-15 04:31:14,041 - train - INFO - val_loss : 0.4425490912269143
2025-10-15 04:31:14,042 - train - INFO - val_CER_(Argmax): 0.12107380933012289
2025-10-15 04:31:14,042 - train - INFO - val_WER_(Argmax): 0.35624895795851813
2025-10-15 04:31:14,042 - train - INFO - val_WER : 0.35624895795851813
2025-10-15 04:31:14,042 - train - INFO - val_CER : 0.12094349863299866
2025-10-15 04:31:14,042 - train - INFO - test_loss : 0.4421439832303582
2025-10-15 04:31:14,042 - train - INFO - test_CER_(Argmax): 0.1203417035792212
2025-10-15 04:31:14,042 - train - INFO - test_WER_(Argmax): 0.3539471761721452
2025-10-15 04:31:14,042 - train - INFO - test_WER : 0.3539471761721452
2025-10-15 04:31:14,042 - train - INFO - test_CER : 0.12014604748975138
2025-10-15 04:31:55,800 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 04:38:23,938 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 04:52:26,034 - train - INFO - epoch : 37
2025-10-15 04:52:26,035 - train - INFO - loss : 0.19795498616993426
2025-10-15 04:52:26,035 - train - INFO - grad_norm : 0.3208508171141148
2025-10-15 04:52:26,036 - train - INFO - val_loss : 0.4592762873453252
2025-10-15 04:52:26,036 - train - INFO - val_CER_(Argmax): 0.1294481027382875
2025-10-15 04:52:26,036 - train - INFO - val_WER_(Argmax): 0.3774834616766213
2025-10-15 04:52:26,036 - train - INFO - val_WER : 0.3774834616766213
2025-10-15 04:52:26,037 - train - INFO - val_CER : 0.1291189030153579
2025-10-15 04:52:26,037 - train - INFO - test_loss : 0.45307019171191426
2025-10-15 04:52:26,037 - train - INFO - test_CER_(Argmax): 0.12595557229485574
2025-10-15 04:52:26,037 - train - INFO - test_WER_(Argmax): 0.36986993718068867
2025-10-15 04:52:26,038 - train - INFO - test_WER : 0.36986993718068867
2025-10-15 04:52:26,038 - train - INFO - test_CER : 0.12586377840829696
2025-10-15 04:54:23,286 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 05:13:56,065 - train - INFO - epoch : 38
2025-10-15 05:13:56,065 - train - INFO - loss : 0.19663494504988194
2025-10-15 05:13:56,065 - train - INFO - grad_norm : 0.3321287203580141
2025-10-15 05:13:56,066 - train - INFO - val_loss : 0.4284120608778561
2025-10-15 05:13:56,066 - train - INFO - val_CER_(Argmax): 0.11491743825574693
2025-10-15 05:13:56,066 - train - INFO - val_WER_(Argmax): 0.34405708317994527
2025-10-15 05:13:56,066 - train - INFO - val_WER : 0.34405708317994527
2025-10-15 05:13:56,066 - train - INFO - val_CER : 0.11487393368251399
2025-10-15 05:13:56,066 - train - INFO - test_loss : 0.42923609730674
2025-10-15 05:13:56,066 - train - INFO - test_CER_(Argmax): 0.11365153750471271
2025-10-15 05:13:56,066 - train - INFO - test_WER_(Argmax): 0.34081102085255915
2025-10-15 05:13:56,066 - train - INFO - test_WER : 0.34081102085255915
2025-10-15 05:13:56,066 - train - INFO - test_CER : 0.11363074827669829
2025-10-15 05:13:56,668 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 05:23:38,950 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 05:35:12,536 - train - INFO - epoch : 39
2025-10-15 05:35:12,538 - train - INFO - loss : 0.18985670647728983
2025-10-15 05:35:12,538 - train - INFO - grad_norm : 0.3164519183150488
2025-10-15 05:35:12,538 - train - INFO - val_loss : 0.4361394615734325
2025-10-15 05:35:12,538 - train - INFO - val_CER_(Argmax): 0.11303924020618653
2025-10-15 05:35:12,538 - train - INFO - val_WER_(Argmax): 0.336099117293402
2025-10-15 05:35:12,539 - train - INFO - val_WER : 0.336099117293402
2025-10-15 05:35:12,539 - train - INFO - val_CER : 0.11297259763035278
2025-10-15 05:35:12,539 - train - INFO - test_loss : 0.4445284605026245
2025-10-15 05:35:12,539 - train - INFO - test_CER_(Argmax): 0.11357265394174307
2025-10-15 05:35:12,540 - train - INFO - test_WER_(Argmax): 0.33712349274851
2025-10-15 05:35:12,540 - train - INFO - test_WER : 0.33712349274851
2025-10-15 05:35:12,540 - train - INFO - test_CER : 0.11356115257526862
2025-10-15 05:35:13,651 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 05:40:26,534 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 05:56:08,877 - train - INFO - epoch : 40
2025-10-15 05:56:08,878 - train - INFO - loss : 0.1758589457720518
2025-10-15 05:56:08,878 - train - INFO - grad_norm : 0.3102634911984205
2025-10-15 05:56:08,878 - train - INFO - val_loss : 0.43493274906102347
2025-10-15 05:56:08,878 - train - INFO - val_CER_(Argmax): 0.1127735913870602
2025-10-15 05:56:08,878 - train - INFO - val_WER_(Argmax): 0.3333324824481823
2025-10-15 05:56:08,878 - train - INFO - val_WER : 0.3333324824481823
2025-10-15 05:56:08,879 - train - INFO - val_CER : 0.11270196248096376
2025-10-15 05:56:08,879 - train - INFO - test_loss : 0.4304133802652359
2025-10-15 05:56:08,879 - train - INFO - test_CER_(Argmax): 0.10964191141102644
2025-10-15 05:56:08,879 - train - INFO - test_WER_(Argmax): 0.3288502833506312
2025-10-15 05:56:08,879 - train - INFO - test_WER : 0.3288502833506312
2025-10-15 05:56:08,879 - train - INFO - test_CER : 0.10959007331535237
2025-10-15 05:56:09,478 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 06:02:47,603 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 06:07:57,641 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 06:17:32,447 - train - INFO - epoch : 41
2025-10-15 06:17:32,448 - train - INFO - loss : 0.1584758448600769
2025-10-15 06:17:32,448 - train - INFO - grad_norm : 0.31132353775203225
2025-10-15 06:17:32,448 - train - INFO - val_loss : 0.4486117271816029
2025-10-15 06:17:32,449 - train - INFO - val_CER_(Argmax): 0.11394301416811013
2025-10-15 06:17:32,449 - train - INFO - val_WER_(Argmax): 0.3352175680667114
2025-10-15 06:17:32,449 - train - INFO - val_WER : 0.3352175680667114
2025-10-15 06:17:32,449 - train - INFO - val_CER : 0.11388101864339778
2025-10-15 06:17:32,449 - train - INFO - test_loss : 0.4439057616925821
2025-10-15 06:17:32,450 - train - INFO - test_CER_(Argmax): 0.11181283918602594
2025-10-15 06:17:32,450 - train - INFO - test_WER_(Argmax): 0.3330420680106056
2025-10-15 06:17:32,450 - train - INFO - test_WER : 0.3330420680106056
2025-10-15 06:17:32,450 - train - INFO - test_CER : 0.11175747260237462
2025-10-15 06:38:32,900 - train - INFO - epoch : 42
2025-10-15 06:38:32,901 - train - INFO - loss : 0.15154085498303174
2025-10-15 06:38:32,901 - train - INFO - grad_norm : 0.31321163214743136
2025-10-15 06:38:32,901 - train - INFO - val_loss : 0.4170303176431095
2025-10-15 06:38:32,901 - train - INFO - val_CER_(Argmax): 0.11403384671386754
2025-10-15 06:38:32,901 - train - INFO - val_WER_(Argmax): 0.339910982840436
2025-10-15 06:38:32,901 - train - INFO - val_WER : 0.339910982840436
2025-10-15 06:38:32,901 - train - INFO - val_CER : 0.11396475986393424
2025-10-15 06:38:32,902 - train - INFO - test_loss : 0.4154184039046125
2025-10-15 06:38:32,902 - train - INFO - test_CER_(Argmax): 0.11247534045841451
2025-10-15 06:38:32,902 - train - INFO - test_WER_(Argmax): 0.3328717059031543
2025-10-15 06:38:32,902 - train - INFO - test_WER : 0.3328717059031543
2025-10-15 06:38:32,902 - train - INFO - test_CER : 0.1123930726890102
2025-10-15 06:42:35,159 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 06:49:28,757 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 06:59:48,120 - train - INFO - epoch : 43
2025-10-15 06:59:48,120 - train - INFO - loss : 0.15414640687406064
2025-10-15 06:59:48,121 - train - INFO - grad_norm : 0.2877813006937504
2025-10-15 06:59:48,121 - train - INFO - val_loss : 0.4414702657391043
2025-10-15 06:59:48,121 - train - INFO - val_CER_(Argmax): 0.12070044864926999
2025-10-15 06:59:48,121 - train - INFO - val_WER_(Argmax): 0.3537302879447558
2025-10-15 06:59:48,121 - train - INFO - val_WER : 0.3537302879447558
2025-10-15 06:59:48,121 - train - INFO - val_CER : 0.12068197627972231
2025-10-15 06:59:48,121 - train - INFO - test_loss : 0.4438759840843154
2025-10-15 06:59:48,121 - train - INFO - test_CER_(Argmax): 0.11862870450615648
2025-10-15 06:59:48,121 - train - INFO - test_WER_(Argmax): 0.34733577061982585
2025-10-15 06:59:48,121 - train - INFO - test_WER : 0.34733577061982585
2025-10-15 06:59:48,122 - train - INFO - test_CER : 0.11862613781757472
2025-10-15 07:20:43,415 - train - INFO - epoch : 44
2025-10-15 07:20:43,415 - train - INFO - loss : 0.1363305367529392
2025-10-15 07:20:43,416 - train - INFO - grad_norm : 0.2479817882925272
2025-10-15 07:20:43,416 - train - INFO - val_loss : 0.4520348629530738
2025-10-15 07:20:43,416 - train - INFO - val_CER_(Argmax): 0.11507168537958189
2025-10-15 07:20:43,416 - train - INFO - val_WER_(Argmax): 0.3395088421236314
2025-10-15 07:20:43,416 - train - INFO - val_WER : 0.3395088421236314
2025-10-15 07:20:43,417 - train - INFO - val_CER : 0.11507427832449559
2025-10-15 07:20:43,417 - train - INFO - test_loss : 0.4543668962106472
2025-10-15 07:20:43,417 - train - INFO - test_CER_(Argmax): 0.11321039333306016
2025-10-15 07:20:43,417 - train - INFO - test_WER_(Argmax): 0.3319809926281677
2025-10-15 07:20:43,417 - train - INFO - test_WER : 0.3319809926281677
2025-10-15 07:20:43,417 - train - INFO - test_CER : 0.1131822929806515
2025-10-15 07:20:51,021 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 07:23:10,474 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 07:41:51,854 - train - INFO - epoch : 45
2025-10-15 07:41:51,855 - train - INFO - loss : 0.13987333193421364
2025-10-15 07:41:51,855 - train - INFO - grad_norm : 0.26614324234426023
2025-10-15 07:41:51,855 - train - INFO - val_loss : 0.4230976984781377
2025-10-15 07:41:51,855 - train - INFO - val_CER_(Argmax): 0.11541515342609154
2025-10-15 07:41:51,855 - train - INFO - val_WER_(Argmax): 0.34157605694474263
2025-10-15 07:41:51,855 - train - INFO - val_WER : 0.34157605694474263
2025-10-15 07:41:51,856 - train - INFO - val_CER : 0.11539626185146624
2025-10-15 07:41:51,856 - train - INFO - test_loss : 0.4274728712512226
2025-10-15 07:41:51,856 - train - INFO - test_CER_(Argmax): 0.11464870047179306
2025-10-15 07:41:51,856 - train - INFO - test_WER_(Argmax): 0.3363408865261514
2025-10-15 07:41:51,856 - train - INFO - test_WER : 0.3363408865261514
2025-10-15 07:41:51,856 - train - INFO - test_CER : 0.11463597320095606
2025-10-15 07:41:52,350 - train - INFO - Saving checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch45.pth ...
2025-10-15 07:49:37,815 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 08:03:12,872 - train - INFO - epoch : 46
2025-10-15 08:03:12,873 - train - INFO - loss : 0.15541664376180975
2025-10-15 08:03:12,873 - train - INFO - grad_norm : 0.30704583459763074
2025-10-15 08:03:12,873 - train - INFO - val_loss : 0.4262906141140882
2025-10-15 08:03:12,873 - train - INFO - val_CER_(Argmax): 0.10697360241698037
2025-10-15 08:03:12,873 - train - INFO - val_WER_(Argmax): 0.31894312216255877
2025-10-15 08:03:12,873 - train - INFO - val_WER : 0.31894312216255877
2025-10-15 08:03:12,873 - train - INFO - val_CER : 0.10697432792093566
2025-10-15 08:03:12,874 - train - INFO - test_loss : 0.4236972459205767
2025-10-15 08:03:12,874 - train - INFO - test_CER_(Argmax): 0.1036966429887784
2025-10-15 08:03:12,874 - train - INFO - test_WER_(Argmax): 0.31260322305849975
2025-10-15 08:03:12,874 - train - INFO - test_WER : 0.31260322305849975
2025-10-15 08:03:12,874 - train - INFO - test_CER : 0.10368049391363919
2025-10-15 08:03:13,468 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 08:11:04,939 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 08:14:33,659 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 08:24:28,330 - train - INFO - epoch : 47
2025-10-15 08:24:28,331 - train - INFO - loss : 0.1484343280369912
2025-10-15 08:24:28,331 - train - INFO - grad_norm : 0.2978131380212966
2025-10-15 08:24:28,331 - train - INFO - val_loss : 0.41686498592881593
2025-10-15 08:24:28,331 - train - INFO - val_CER_(Argmax): 0.10715906838873794
2025-10-15 08:24:28,331 - train - INFO - val_WER_(Argmax): 0.3189208901687594
2025-10-15 08:24:28,331 - train - INFO - val_WER : 0.3189208901687594
2025-10-15 08:24:28,331 - train - INFO - val_CER : 0.1070721680237797
2025-10-15 08:24:28,331 - train - INFO - test_loss : 0.4248736776956698
2025-10-15 08:24:28,332 - train - INFO - test_CER_(Argmax): 0.10606248870942041
2025-10-15 08:24:28,332 - train - INFO - test_WER_(Argmax): 0.31392775456406125
2025-10-15 08:24:28,332 - train - INFO - test_WER : 0.31392775456406125
2025-10-15 08:24:28,332 - train - INFO - test_CER : 0.10608059527809617
2025-10-15 08:24:28,964 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 08:45:42,227 - train - INFO - epoch : 48
2025-10-15 08:45:42,227 - train - INFO - loss : 0.1307941133156419
2025-10-15 08:45:42,228 - train - INFO - grad_norm : 0.27077246375381947
2025-10-15 08:45:42,228 - train - INFO - val_loss : 0.42694155994583577
2025-10-15 08:45:42,228 - train - INFO - val_CER_(Argmax): 0.10582719959543849
2025-10-15 08:45:42,228 - train - INFO - val_WER_(Argmax): 0.31559900742699704
2025-10-15 08:45:42,228 - train - INFO - val_WER : 0.31559900742699704
2025-10-15 08:45:42,228 - train - INFO - val_CER : 0.10572084591343127
2025-10-15 08:45:42,228 - train - INFO - test_loss : 0.43384504790713146
2025-10-15 08:45:42,228 - train - INFO - test_CER_(Argmax): 0.10381538597811517
2025-10-15 08:45:42,228 - train - INFO - test_WER_(Argmax): 0.311420565217544
2025-10-15 08:45:42,229 - train - INFO - test_WER : 0.311420565217544
2025-10-15 08:45:42,229 - train - INFO - test_CER : 0.10370990247742001
2025-10-15 08:45:42,872 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 08:50:42,467 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 08:53:44,773 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 09:07:09,369 - train - INFO - epoch : 49
2025-10-15 09:07:09,371 - train - INFO - loss : 0.10918797556329612
2025-10-15 09:07:09,371 - train - INFO - grad_norm : 0.25354656966487366
2025-10-15 09:07:09,371 - train - INFO - val_loss : 0.45092080375727484
2025-10-15 09:07:09,371 - train - INFO - val_CER_(Argmax): 0.10998150896043028
2025-10-15 09:07:09,371 - train - INFO - val_WER_(Argmax): 0.3251663764785473
2025-10-15 09:07:09,372 - train - INFO - val_WER : 0.3251663764785473
2025-10-15 09:07:09,372 - train - INFO - val_CER : 0.10883912407170318
2025-10-15 09:07:09,372 - train - INFO - test_loss : 0.45026707503853775
2025-10-15 09:07:09,372 - train - INFO - test_CER_(Argmax): 0.1083685399307039
2025-10-15 09:07:09,372 - train - INFO - test_WER_(Argmax): 0.31903264381702373
2025-10-15 09:07:09,373 - train - INFO - test_WER : 0.31903264381702373
2025-10-15 09:07:09,373 - train - INFO - test_CER : 0.10773803375770985
2025-10-15 09:28:10,598 - train - INFO - epoch : 50
2025-10-15 09:28:10,598 - train - INFO - loss : 0.10687003966420888
2025-10-15 09:28:10,598 - train - INFO - grad_norm : 0.2543030245602131
2025-10-15 09:28:10,599 - train - INFO - val_loss : 0.45245280335931215
2025-10-15 09:28:10,599 - train - INFO - val_CER_(Argmax): 0.10794419929359747
2025-10-15 09:28:10,599 - train - INFO - val_WER_(Argmax): 0.319912166248811
2025-10-15 09:28:10,599 - train - INFO - val_WER : 0.319912166248811
2025-10-15 09:28:10,599 - train - INFO - val_CER : 0.10779546715487702
2025-10-15 09:28:10,599 - train - INFO - test_loss : 0.44923570715799566
2025-10-15 09:28:10,599 - train - INFO - test_CER_(Argmax): 0.10577866073952201
2025-10-15 09:28:10,599 - train - INFO - test_WER_(Argmax): 0.3148882348303448
2025-10-15 09:28:10,599 - train - INFO - test_WER : 0.3148882348303448
2025-10-15 09:28:10,599 - train - INFO - test_CER : 0.10566427550858948
2025-10-15 09:28:11,086 - train - INFO - Saving checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch50.pth ...
2025-10-15 09:31:04,615 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 09:33:16,245 - train - INFO - Saving model on keyboard interrupt
2025-10-15 09:33:16,720 - train - INFO - Saving checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch51.pth ...
2025-10-15 21:28:51,805 - train - INFO - ConformerModel(
(linear_layer_1): Linear(in_features=128, out_features=256, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
(conformer_blocks): Sequential(
(0): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(1): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(2): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(3): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(4): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(5): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(6): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(7): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(8): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(9): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
)
(linear_layer_2): Linear(in_features=256, out_features=28, bias=True)
)
2025-10-15 21:28:51,826 - train - INFO - Loading checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/conformer_30m/checkpoint-epoch51.pth ...
2025-10-15 21:29:48,755 - train - INFO - ConformerModel(
(linear_layer_1): Linear(in_features=128, out_features=256, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
(conformer_blocks): Sequential(
(0): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(1): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(2): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(3): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(4): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(5): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(6): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(7): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(8): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(9): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
)
(linear_layer_2): Linear(in_features=256, out_features=28, bias=True)
)
2025-10-15 21:29:48,772 - train - INFO - Loading checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch51.pth ...
2025-10-15 21:29:49,382 - train - WARNING - Warning: Optimizer or lr_scheduler given in the config file is different from that of the checkpoint. Optimizer and scheduler parameters are not resumed.
2025-10-15 21:29:49,383 - train - INFO - Checkpoint loaded. Resume training from epoch 52
2025-10-15 21:32:46,047 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 21:36:14,907 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 21:36:28,019 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 21:37:11,355 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 21:37:28,040 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 21:38:05,327 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 21:38:37,856 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 21:59:34,376 - train - INFO - ConformerModel(
(linear_layer_1): Linear(in_features=128, out_features=256, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
(conformer_blocks): Sequential(
(0): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(1): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(2): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(3): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(4): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(5): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(6): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(7): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(8): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(9): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
)
(linear_layer_2): Linear(in_features=256, out_features=28, bias=True)
)
2025-10-15 21:59:34,392 - train - INFO - Loading checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch51.pth ...
2025-10-15 21:59:34,755 - train - WARNING - Warning: Optimizer or lr_scheduler given in the config file is different from that of the checkpoint. Optimizer and scheduler parameters are not resumed.
2025-10-15 21:59:34,756 - train - INFO - Checkpoint loaded. Resume training from epoch 52
2025-10-15 22:02:34,318 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:06:06,199 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:06:18,792 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:07:20,904 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:07:57,912 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:08:30,148 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:09:56,709 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:10:56,816 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:18:26,896 - train - INFO - epoch : 52
2025-10-15 22:18:26,897 - train - INFO - loss : 0.5125516548677145
2025-10-15 22:18:26,897 - train - INFO - grad_norm : 0.5027066829240867
2025-10-15 22:18:26,897 - train - INFO - val_loss : 0.3037505878524466
2025-10-15 22:18:26,897 - train - INFO - val_CER_(Argmax): 0.09129654682855147
2025-10-15 22:18:26,897 - train - INFO - val_WER_(Argmax): 0.28161343918875303
2025-10-15 22:18:26,897 - train - INFO - val_WER : 0.28161343918875303
2025-10-15 22:18:26,897 - train - INFO - val_CER : 0.09130862404528663
2025-10-15 22:18:26,897 - train - INFO - test_loss : 0.30807594836435537
2025-10-15 22:18:26,897 - train - INFO - test_CER_(Argmax): 0.09054438299605952
2025-10-15 22:18:26,898 - train - INFO - test_WER_(Argmax): 0.2782349555610264
2025-10-15 22:18:26,898 - train - INFO - test_WER : 0.2782349555610264
2025-10-15 22:18:26,898 - train - INFO - test_CER : 0.09054140648346386
2025-10-15 22:18:27,452 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 22:18:34,798 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:19:35,681 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:20:45,068 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:21:17,097 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:28:13,972 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:28:43,629 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:37:01,553 - train - INFO - epoch : 53
2025-10-15 22:37:01,554 - train - INFO - loss : 0.4767994123697281
2025-10-15 22:37:01,554 - train - INFO - grad_norm : 0.5589224849641323
2025-10-15 22:37:01,554 - train - INFO - val_loss : 0.287550592979232
2025-10-15 22:37:01,555 - train - INFO - val_CER_(Argmax): 0.08663647521630864
2025-10-15 22:37:01,555 - train - INFO - val_WER_(Argmax): 0.26653980578276465
2025-10-15 22:37:01,555 - train - INFO - val_WER : 0.26653980578276465
2025-10-15 22:37:01,555 - train - INFO - val_CER : 0.08663647998507391
2025-10-15 22:37:01,555 - train - INFO - test_loss : 0.29402880726212804
2025-10-15 22:37:01,555 - train - INFO - test_CER_(Argmax): 0.08634445495520625
2025-10-15 22:37:01,555 - train - INFO - test_WER_(Argmax): 0.26760766956081206
2025-10-15 22:37:01,555 - train - INFO - test_WER : 0.26760766956081206
2025-10-15 22:37:01,555 - train - INFO - test_CER : 0.0863314313346999
2025-10-15 22:37:02,160 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 22:38:36,055 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:41:57,959 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:42:05,218 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:47:32,830 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:55:34,476 - train - INFO - epoch : 54
2025-10-15 22:55:34,476 - train - INFO - loss : 0.4517164458334446
2025-10-15 22:55:34,476 - train - INFO - grad_norm : 0.6237075063586235
2025-10-15 22:55:34,476 - train - INFO - val_loss : 0.28015144562328254
2025-10-15 22:55:34,476 - train - INFO - val_CER_(Argmax): 0.08501164964665332
2025-10-15 22:55:34,477 - train - INFO - val_WER_(Argmax): 0.26355578235185695
2025-10-15 22:55:34,477 - train - INFO - val_WER : 0.26355578235185695
2025-10-15 22:55:34,477 - train - INFO - val_CER : 0.08501273757687683
2025-10-15 22:55:34,477 - train - INFO - test_loss : 0.28677379729395563
2025-10-15 22:55:34,477 - train - INFO - test_CER_(Argmax): 0.08556362216950926
2025-10-15 22:55:34,477 - train - INFO - test_WER_(Argmax): 0.26548957249503236
2025-10-15 22:55:34,477 - train - INFO - test_WER : 0.26548957249503236
2025-10-15 22:55:34,477 - train - INFO - test_CER : 0.08552682603878249
2025-10-15 22:55:35,084 - train - INFO - Saving current best: model_best.pth ...
2025-10-15 22:58:30,490 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 22:59:07,816 - urllib3.connectionpool - WARNING - Retrying (Retry(total=2, connect=None, read=None, redirect=None, status=None)) after connection broken by 'ReadTimeoutError("HTTPSConnectionPool(host='www.comet.com', port=443): Read timed out. (read timeout=10)")': /clientlib/rest/v2/write/experiment/output
2025-10-15 23:03:27,906 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:14:15,775 - train - INFO - epoch : 55
2025-10-15 23:14:15,775 - train - INFO - loss : 0.43951718381900884
2025-10-15 23:14:15,775 - train - INFO - grad_norm : 0.610138273718369
2025-10-15 23:14:15,775 - train - INFO - val_loss : 0.28494494465681225
2025-10-15 23:14:15,775 - train - INFO - val_CER_(Argmax): 0.08762042370466316
2025-10-15 23:14:15,775 - train - INFO - val_WER_(Argmax): 0.26894677637977005
2025-10-15 23:14:15,776 - train - INFO - val_WER : 0.26894677637977005
2025-10-15 23:14:15,776 - train - INFO - val_CER : 0.08761403647985697
2025-10-15 23:14:15,776 - train - INFO - test_loss : 0.284894541583278
2025-10-15 23:14:15,776 - train - INFO - test_CER_(Argmax): 0.08618727251740227
2025-10-15 23:14:15,776 - train - INFO - test_WER_(Argmax): 0.2651266555833503
2025-10-15 23:14:15,776 - train - INFO - test_WER : 0.2651266555833503
2025-10-15 23:14:15,776 - train - INFO - test_CER : 0.08613433873893521
2025-10-15 23:14:16,264 - train - INFO - Saving checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch55.pth ...
2025-10-15 23:17:14,837 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:24:01,730 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:32:58,682 - train - INFO - epoch : 56
2025-10-15 23:32:58,682 - train - INFO - loss : 0.4343975530564785
2025-10-15 23:32:58,682 - train - INFO - grad_norm : 0.6229039686918258
2025-10-15 23:32:58,682 - train - INFO - val_loss : 0.29983231277911215
2025-10-15 23:32:58,682 - train - INFO - val_CER_(Argmax): 0.08955679162254075
2025-10-15 23:32:58,683 - train - INFO - val_WER_(Argmax): 0.27534571803696956
2025-10-15 23:32:58,683 - train - INFO - val_WER : 0.27534571803696956
2025-10-15 23:32:58,683 - train - INFO - val_CER : 0.08951839197980865
2025-10-15 23:32:58,683 - train - INFO - test_loss : 0.29415979016233573
2025-10-15 23:32:58,683 - train - INFO - test_CER_(Argmax): 0.08593879618718768
2025-10-15 23:32:58,683 - train - INFO - test_WER_(Argmax): 0.27041946687163093
2025-10-15 23:32:58,683 - train - INFO - test_WER : 0.27041946687163093
2025-10-15 23:32:58,683 - train - INFO - test_CER : 0.0859091788688324
2025-10-15 23:34:07,974 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:34:55,062 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:37:46,718 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:51:41,538 - train - INFO - epoch : 57
2025-10-15 23:51:41,539 - train - INFO - loss : 0.44212916776537897
2025-10-15 23:51:41,539 - train - INFO - grad_norm : 0.6302911515533924
2025-10-15 23:51:41,539 - train - INFO - val_loss : 0.3111286821601155
2025-10-15 23:51:41,539 - train - INFO - val_CER_(Argmax): 0.09671470586458569
2025-10-15 23:51:41,539 - train - INFO - val_WER_(Argmax): 0.2970807443308442
2025-10-15 23:51:41,539 - train - INFO - val_WER : 0.2970807443308442
2025-10-15 23:51:41,539 - train - INFO - val_CER : 0.09666179815724249
2025-10-15 23:51:41,540 - train - INFO - test_loss : 0.3062010456553914
2025-10-15 23:51:41,540 - train - INFO - test_CER_(Argmax): 0.09478064453852167
2025-10-15 23:51:41,540 - train - INFO - test_WER_(Argmax): 0.2917665250029616
2025-10-15 23:51:41,540 - train - INFO - test_WER : 0.2917665250029616
2025-10-15 23:51:41,540 - train - INFO - test_CER : 0.09473352732799264
2025-10-15 23:52:00,536 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:52:02,784 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:52:37,963 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:53:55,377 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:54:35,227 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:55:49,514 - train - WARNING - OOM on batch. Skipping batch.
2025-10-15 23:59:59,392 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:01:41,820 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:11:38,086 - train - INFO - epoch : 58
2025-10-16 00:11:38,087 - train - INFO - loss : 0.4711415856028322
2025-10-16 00:11:38,088 - train - INFO - grad_norm : 0.5886639985606898
2025-10-16 00:11:38,088 - train - INFO - val_loss : 0.3353017005291614
2025-10-16 00:11:38,088 - train - INFO - val_CER_(Argmax): 0.10235041302713359
2025-10-16 00:11:38,088 - train - INFO - val_WER_(Argmax): 0.3116658465074636
2025-10-16 00:11:38,088 - train - INFO - val_WER : 0.3116658465074636
2025-10-16 00:11:38,088 - train - INFO - val_CER : 0.1023209039028333
2025-10-16 00:11:38,088 - train - INFO - test_loss : 0.3331021143292839
2025-10-16 00:11:38,089 - train - INFO - test_CER_(Argmax): 0.09998714099883693
2025-10-16 00:11:38,089 - train - INFO - test_WER_(Argmax): 0.30518309281193207
2025-10-16 00:11:38,089 - train - INFO - test_WER : 0.30518309281193207
2025-10-16 00:11:38,089 - train - INFO - test_CER : 0.10000657190144936
2025-10-16 00:15:31,484 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:17:40,029 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:21:20,102 - train - INFO - Saving model on keyboard interrupt
2025-10-16 00:21:20,725 - train - INFO - Saving checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch59.pth ...
2025-10-16 00:25:35,941 - train - INFO - ConformerModel(
(linear_layer_1): Linear(in_features=128, out_features=256, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
(conformer_blocks): Sequential(
(0): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(1): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(2): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(3): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(4): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(5): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(6): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(7): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(8): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(9): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
)
(linear_layer_2): Linear(in_features=256, out_features=28, bias=True)
)
2025-10-16 00:25:35,957 - train - INFO - Loading checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch55.pth ...
2025-10-16 00:25:36,354 - train - WARNING - Warning: Optimizer or lr_scheduler given in the config file is different from that of the checkpoint. Optimizer and scheduler parameters are not resumed.
2025-10-16 00:25:36,354 - train - INFO - Checkpoint loaded. Resume training from epoch 56
2025-10-16 00:28:38,563 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:32:12,325 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:32:25,198 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:33:27,019 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:34:03,944 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:34:36,076 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:36:03,448 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:37:05,549 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:45:33,019 - train - INFO - epoch : 56
2025-10-16 00:45:33,019 - train - INFO - loss : 0.3975452154420959
2025-10-16 00:45:33,020 - train - INFO - grad_norm : 0.5853364494851399
2025-10-16 00:45:33,020 - train - INFO - val_loss : 0.26601492441617525
2025-10-16 00:45:33,020 - train - INFO - val_CER_(Argmax): 0.08095286654242909
2025-10-16 00:45:33,020 - train - INFO - val_WER_(Argmax): 0.25238181992836795
2025-10-16 00:45:33,020 - train - INFO - val_WER : 0.25238181992836795
2025-10-16 00:45:33,020 - train - INFO - val_CER : 0.08094647931762287
2025-10-16 00:45:33,020 - train - INFO - test_loss : 0.2680981419980526
2025-10-16 00:45:33,020 - train - INFO - test_CER_(Argmax): 0.08014114209410332
2025-10-16 00:45:33,020 - train - INFO - test_WER_(Argmax): 0.2506958176100273
2025-10-16 00:45:33,021 - train - INFO - test_WER : 0.2506958176100273
2025-10-16 00:45:33,021 - train - INFO - test_CER : 0.08013405791041706
2025-10-16 00:45:33,634 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 00:45:40,678 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:46:41,544 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:47:51,335 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:48:23,005 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 00:55:25,594 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 01:05:45,591 - train - INFO - epoch : 57
2025-10-16 01:05:45,591 - train - INFO - loss : 0.3919293682277203
2025-10-16 01:05:45,591 - train - INFO - grad_norm : 0.5571760645508766
2025-10-16 01:05:45,592 - train - INFO - val_loss : 0.26085271825502204
2025-10-16 01:05:45,592 - train - INFO - val_CER_(Argmax): 0.07941454723878524
2025-10-16 01:05:45,592 - train - INFO - val_WER_(Argmax): 0.24847200659684884
2025-10-16 01:05:45,592 - train - INFO - val_WER : 0.24847200659684884
2025-10-16 01:05:45,592 - train - INFO - val_CER : 0.07940334027231717
2025-10-16 01:05:45,592 - train - INFO - test_loss : 0.2639189895919778
2025-10-16 01:05:45,592 - train - INFO - test_CER_(Argmax): 0.07906222839036091
2025-10-16 01:05:45,592 - train - INFO - test_WER_(Argmax): 0.2467830935703038
2025-10-16 01:05:45,592 - train - INFO - test_WER : 0.2467830935703038
2025-10-16 01:05:45,592 - train - INFO - test_CER : 0.07904224464832055
2025-10-16 01:05:46,206 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 01:07:22,051 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 01:08:02,388 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 01:10:48,748 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 01:10:56,086 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 01:16:24,729 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 01:25:47,709 - train - INFO - epoch : 58
2025-10-16 01:25:47,709 - train - INFO - loss : 0.36956372365355494
2025-10-16 01:25:47,710 - train - INFO - grad_norm : 0.59020700648427
2025-10-16 01:25:47,710 - train - INFO - val_loss : 0.2571919283696583
2025-10-16 01:25:47,710 - train - INFO - val_CER_(Argmax): 0.07819992281237179
2025-10-16 01:25:47,710 - train - INFO - val_WER_(Argmax): 0.24590658841056448
2025-10-16 01:25:47,710 - train - INFO - val_WER : 0.24590658841056448
2025-10-16 01:25:47,710 - train - INFO - val_CER : 0.07818871584590371
2025-10-16 01:25:47,710 - train - INFO - test_loss : 0.26166995834897866
2025-10-16 01:25:47,710 - train - INFO - test_CER_(Argmax): 0.07749860080767827
2025-10-16 01:25:47,710 - train - INFO - test_WER_(Argmax): 0.24307131531532164
2025-10-16 01:25:47,710 - train - INFO - test_WER : 0.24307131531532164
2025-10-16 01:25:47,711 - train - INFO - test_CER : 0.0774786170656379
2025-10-16 01:25:48,325 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 01:28:44,667 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 01:31:32,043 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 01:33:48,273 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 01:46:04,253 - train - INFO - epoch : 59
2025-10-16 01:46:04,253 - train - INFO - loss : 0.33486752002382997
2025-10-16 01:46:04,254 - train - INFO - grad_norm : 0.5633255804902945
2025-10-16 01:46:04,254 - train - INFO - val_loss : 0.25902849910678444
2025-10-16 01:46:04,254 - train - INFO - val_CER_(Argmax): 0.07788373600524053
2025-10-16 01:46:04,254 - train - INFO - val_WER_(Argmax): 0.24356013677092545
2025-10-16 01:46:04,254 - train - INFO - val_WER : 0.24356013677092545
2025-10-16 01:46:04,254 - train - INFO - val_CER : 0.07787252903877245
2025-10-16 01:46:04,254 - train - INFO - test_loss : 0.26355371505699376
2025-10-16 01:46:04,254 - train - INFO - test_CER_(Argmax): 0.07704687640540066
2025-10-16 01:46:04,254 - train - INFO - test_WER_(Argmax): 0.24105306456868927
2025-10-16 01:46:04,255 - train - INFO - test_WER : 0.24105306456868927
2025-10-16 01:46:04,255 - train - INFO - test_CER : 0.07702689266336028
2025-10-16 01:46:04,866 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 01:49:04,404 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 01:55:55,381 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 01:55:55,714 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:06:10,617 - train - INFO - epoch : 60
2025-10-16 02:06:10,617 - train - INFO - loss : 0.38520306527614595
2025-10-16 02:06:10,617 - train - INFO - grad_norm : 0.6394448788464069
2025-10-16 02:06:10,618 - train - INFO - val_loss : 0.2523955891093055
2025-10-16 02:06:10,618 - train - INFO - val_CER_(Argmax): 0.07642980123388085
2025-10-16 02:06:10,618 - train - INFO - val_WER_(Argmax): 0.24021242375268156
2025-10-16 02:06:10,618 - train - INFO - val_WER : 0.24021242375268156
2025-10-16 02:06:10,618 - train - INFO - val_CER : 0.07641859426741279
2025-10-16 02:06:10,618 - train - INFO - test_loss : 0.2567285626110705
2025-10-16 02:06:10,618 - train - INFO - test_CER_(Argmax): 0.07615406577241463
2025-10-16 02:06:10,618 - train - INFO - test_WER_(Argmax): 0.23853556294885694
2025-10-16 02:06:10,619 - train - INFO - test_WER : 0.23853556294885694
2025-10-16 02:06:10,619 - train - INFO - test_CER : 0.07614019151229019
2025-10-16 02:06:11,228 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 02:06:43,709 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:07:21,268 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:08:09,949 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:11:04,696 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:26:31,842 - train - INFO - epoch : 61
2025-10-16 02:26:31,844 - train - INFO - loss : 0.36128015503287314
2025-10-16 02:26:31,844 - train - INFO - grad_norm : 0.6772882167994976
2025-10-16 02:26:31,844 - train - INFO - val_loss : 0.25020998798228883
2025-10-16 02:26:31,844 - train - INFO - val_CER_(Argmax): 0.07584409640850495
2025-10-16 02:26:31,844 - train - INFO - val_WER_(Argmax): 0.23786885936936988
2025-10-16 02:26:31,845 - train - INFO - val_WER : 0.23786885936936988
2025-10-16 02:26:31,845 - train - INFO - val_CER : 0.075829165821292
2025-10-16 02:26:31,845 - train - INFO - test_loss : 0.25460747815668583
2025-10-16 02:26:31,845 - train - INFO - test_CER_(Argmax): 0.07530382475604616
2025-10-16 02:26:31,845 - train - INFO - test_WER_(Argmax): 0.23598896298484998
2025-10-16 02:26:31,845 - train - INFO - test_WER : 0.23598896298484998
2025-10-16 02:26:31,845 - train - INFO - test_CER : 0.07528315230877163
2025-10-16 02:26:32,993 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 02:26:52,328 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:26:54,618 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:27:29,413 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:28:46,742 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:29:27,005 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:30:43,684 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:34:56,804 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:36:39,409 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:43:32,699 - train - INFO - Saving model on keyboard interrupt
2025-10-16 02:43:33,623 - train - INFO - Saving checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch62.pth ...
2025-10-16 02:49:17,502 - train - INFO - ConformerModel(
(linear_layer_1): Linear(in_features=128, out_features=256, bias=True)
(dropout): Dropout(p=0.1, inplace=False)
(conformer_blocks): Sequential(
(0): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(1): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(2): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(3): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(4): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(5): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(6): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(7): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(8): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
(9): ConformerBlock(
(feedforward): FeedForward(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(linear_1): Linear(in_features=256, out_features=1024, bias=True)
(swish): SiLU()
(dropout_1): Dropout(p=0.1, inplace=False)
(linear_2): Linear(in_features=1024, out_features=256, bias=True)
(dropout_2): Dropout(p=0.1, inplace=False)
)
(multiheadselfattn): MultiHeadSelfAttn(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(multiheadattn): MultiheadAttention(
(out_proj): NonDynamicallyQuantizableLinear(in_features=256, out_features=256, bias=True)
)
(dropout): Dropout(p=0.1, inplace=False)
)
(convmod): ConvolutionModule(
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
(conv_1): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
(GLu): GLU(dim=1)
(conv_2): Conv1d(256, 256, kernel_size=(31,), stride=(1,), padding=(15,))
(bn): BatchNorm1d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
(swish): SiLU()
(conv_3): Conv1d(256, 256, kernel_size=(1,), stride=(1,))
(dp): Dropout(p=0.1, inplace=False)
)
(layer_norm): LayerNorm((256,), eps=1e-05, elementwise_affine=True)
)
)
(linear_layer_2): Linear(in_features=256, out_features=28, bias=True)
)
2025-10-16 02:49:17,523 - train - INFO - Loading checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch62.pth ...
2025-10-16 02:49:17,913 - train - WARNING - Warning: Optimizer or lr_scheduler given in the config file is different from that of the checkpoint. Optimizer and scheduler parameters are not resumed.
2025-10-16 02:49:17,914 - train - INFO - Checkpoint loaded. Resume training from epoch 63
2025-10-16 02:52:15,957 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:55:47,500 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:56:00,240 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:57:01,514 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:57:38,947 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:58:11,042 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 02:59:37,819 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:00:38,610 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:01:03,954 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:02:05,006 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:03:14,868 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:03:46,604 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:12:15,075 - train - INFO - epoch : 63
2025-10-16 03:12:15,075 - train - INFO - loss : 0.3650146044299082
2025-10-16 03:12:15,076 - train - INFO - grad_norm : 0.732061750695185
2025-10-16 03:12:15,076 - train - INFO - val_loss : 0.25124785574999725
2025-10-16 03:12:15,076 - train - INFO - val_CER_(Argmax): 0.0746809727347122
2025-10-16 03:12:15,076 - train - INFO - val_WER_(Argmax): 0.2340455252925752
2025-10-16 03:12:15,077 - train - INFO - val_WER : 0.2340455252925752
2025-10-16 03:12:15,077 - train - INFO - val_CER : 0.07466787604501343
2025-10-16 03:12:15,077 - train - INFO - test_loss : 0.6272238085464555
2025-10-16 03:12:15,077 - train - INFO - test_CER_(Argmax): 0.18450640920311695
2025-10-16 03:12:15,077 - train - INFO - test_WER_(Argmax): 0.45838625598859056
2025-10-16 03:12:15,078 - train - INFO - test_WER : 0.45838625598859056
2025-10-16 03:12:15,078 - train - INFO - test_CER : 0.1844315619994462
2025-10-16 03:12:16,653 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 03:18:39,991 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:21:29,785 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:21:50,742 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:22:29,530 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:25:13,780 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:25:20,946 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:35:12,338 - train - INFO - epoch : 64
2025-10-16 03:35:12,339 - train - INFO - loss : 0.34100347862701225
2025-10-16 03:35:12,339 - train - INFO - grad_norm : 0.6329810723210826
2025-10-16 03:35:12,339 - train - INFO - val_loss : 0.24926169436763634
2025-10-16 03:35:12,339 - train - INFO - val_CER_(Argmax): 0.07435775967825689
2025-10-16 03:35:12,339 - train - INFO - val_WER_(Argmax): 0.23329625474160692
2025-10-16 03:35:12,339 - train - INFO - val_WER : 0.23329625474160692
2025-10-16 03:35:12,339 - train - INFO - val_CER : 0.07438597296022
2025-10-16 03:35:12,340 - train - INFO - test_loss : 0.6233528654794304
2025-10-16 03:35:12,340 - train - INFO - test_CER_(Argmax): 0.18366725143983528
2025-10-16 03:35:12,340 - train - INFO - test_WER_(Argmax): 0.45719180165705925
2025-10-16 03:35:12,340 - train - INFO - test_WER : 0.45719180165705925
2025-10-16 03:35:12,340 - train - INFO - test_CER : 0.18355957764546682
2025-10-16 03:35:12,971 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 03:38:54,738 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:42:35,097 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:45:20,154 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:47:35,186 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:48:14,719 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:49:30,920 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 03:58:14,580 - train - INFO - epoch : 65
2025-10-16 03:58:14,581 - train - INFO - loss : 0.31584615042113295
2025-10-16 03:58:14,581 - train - INFO - grad_norm : 0.5978777054885421
2025-10-16 03:58:14,581 - train - INFO - val_loss : 0.2521649480543353
2025-10-16 03:58:14,581 - train - INFO - val_CER_(Argmax): 0.07417007913966524
2025-10-16 03:58:14,582 - train - INFO - val_WER_(Argmax): 0.23324526469575566
2025-10-16 03:58:14,582 - train - INFO - val_WER : 0.23324526469575566
2025-10-16 03:58:14,582 - train - INFO - val_CER : 0.07415698244996646
2025-10-16 03:58:14,582 - train - INFO - test_loss : 0.6326038317412747
2025-10-16 03:58:14,583 - train - INFO - test_CER_(Argmax): 0.1823612314736419
2025-10-16 03:58:14,583 - train - INFO - test_WER_(Argmax): 0.4545763313591499
2025-10-16 03:58:14,583 - train - INFO - test_WER : 0.4545763313591499
2025-10-16 03:58:14,583 - train - INFO - test_CER : 0.18233351503535009
2025-10-16 03:58:15,691 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 04:02:20,768 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:04:19,315 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:09:07,248 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:09:07,646 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:09:21,702 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:11:07,133 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:11:42,931 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:21:07,208 - train - INFO - epoch : 66
2025-10-16 04:21:07,210 - train - INFO - loss : 0.35260840369896457
2025-10-16 04:21:07,210 - train - INFO - grad_norm : 0.6941393242038861
2025-10-16 04:21:07,210 - train - INFO - val_loss : 0.24919554303315553
2025-10-16 04:21:07,211 - train - INFO - val_CER_(Argmax): 0.07394414424173303
2025-10-16 04:21:07,211 - train - INFO - val_WER_(Argmax): 0.2325940054693939
2025-10-16 04:21:07,211 - train - INFO - val_WER : 0.2325940054693939
2025-10-16 04:21:07,211 - train - INFO - val_CER : 0.07393104755203425
2025-10-16 04:21:07,211 - train - INFO - test_loss : 0.6230674550241354
2025-10-16 04:21:07,212 - train - INFO - test_CER_(Argmax): 0.18231329163386267
2025-10-16 04:21:07,212 - train - INFO - test_WER_(Argmax): 0.4550589076913371
2025-10-16 04:21:07,212 - train - INFO - test_WER : 0.4550589076913371
2025-10-16 04:21:07,212 - train - INFO - test_CER : 0.18230684880807213
2025-10-16 04:21:08,325 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 04:22:33,557 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:27:24,290 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:27:45,211 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:30:06,656 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:33:36,847 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:34:10,539 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:34:23,483 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:34:37,393 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:43:57,489 - train - INFO - epoch : 67
2025-10-16 04:43:57,489 - train - INFO - loss : 0.3447438172843991
2025-10-16 04:43:57,489 - train - INFO - grad_norm : 0.6474652308590558
2025-10-16 04:43:57,490 - train - INFO - val_loss : 0.24890769882635636
2025-10-16 04:43:57,490 - train - INFO - val_CER_(Argmax): 0.07402464042128627
2025-10-16 04:43:57,490 - train - INFO - val_WER_(Argmax): 0.23276514236104118
2025-10-16 04:43:57,490 - train - INFO - val_WER : 0.23276514236104118
2025-10-16 04:43:57,490 - train - INFO - val_CER : 0.07401007745592768
2025-10-16 04:43:57,490 - train - INFO - test_loss : 0.6215816164503292
2025-10-16 04:43:57,490 - train - INFO - test_CER_(Argmax): 0.18244981545202055
2025-10-16 04:43:57,490 - train - INFO - test_WER_(Argmax): 0.4545860848775349
2025-10-16 04:43:57,490 - train - INFO - test_WER : 0.4545860848775349
2025-10-16 04:43:57,490 - train - INFO - test_CER : 0.18234336381594599
2025-10-16 04:48:59,867 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:49:49,544 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:52:50,756 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:54:56,214 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:56:32,796 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 04:56:41,395 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:06:56,800 - train - INFO - epoch : 68
2025-10-16 05:06:56,801 - train - INFO - loss : 0.34905940896332865
2025-10-16 05:06:56,801 - train - INFO - grad_norm : 0.6677498564575658
2025-10-16 05:06:56,801 - train - INFO - val_loss : 0.24816725386137312
2025-10-16 05:06:56,801 - train - INFO - val_CER_(Argmax): 0.07392974273734511
2025-10-16 05:06:56,801 - train - INFO - val_WER_(Argmax): 0.23159585286071424
2025-10-16 05:06:56,802 - train - INFO - val_WER : 0.23159585286071424
2025-10-16 05:06:56,802 - train - INFO - val_CER : 0.07391517977198651
2025-10-16 05:06:56,802 - train - INFO - test_loss : 0.6195810704815145
2025-10-16 05:06:56,802 - train - INFO - test_CER_(Argmax): 0.18233763775457032
2025-10-16 05:06:56,802 - train - INFO - test_WER_(Argmax): 0.4545479979074868
2025-10-16 05:06:56,802 - train - INFO - test_WER : 0.4545479979074868
2025-10-16 05:06:56,802 - train - INFO - test_CER : 0.1823116928582287
2025-10-16 05:06:57,368 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 05:07:50,812 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:09:15,900 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:09:28,324 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:12:10,737 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:12:15,674 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:15:57,493 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:17:11,071 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:17:28,170 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:18:48,935 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:19:16,909 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:20:11,772 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:20:34,940 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:29:55,155 - train - INFO - epoch : 69
2025-10-16 05:29:55,156 - train - INFO - loss : 0.34450486673013814
2025-10-16 05:29:55,156 - train - INFO - grad_norm : 0.6989231766177918
2025-10-16 05:29:55,157 - train - INFO - val_loss : 0.24752564846792005
2025-10-16 05:29:55,157 - train - INFO - val_CER_(Argmax): 0.07375343535866713
2025-10-16 05:29:55,157 - train - INFO - val_WER_(Argmax): 0.23090194026229824
2025-10-16 05:29:55,157 - train - INFO - val_WER : 0.23090194026229824
2025-10-16 05:29:55,157 - train - INFO - val_CER : 0.07373887239330854
2025-10-16 05:29:55,157 - train - INFO - test_loss : 0.6168282333077216
2025-10-16 05:29:55,157 - train - INFO - test_CER_(Argmax): 0.18170050435403806
2025-10-16 05:29:55,157 - train - INFO - test_WER_(Argmax): 0.45372425677781314
2025-10-16 05:29:55,157 - train - INFO - test_WER : 0.45372425677781314
2025-10-16 05:29:55,157 - train - INFO - test_CER : 0.1816881820827073
2025-10-16 05:29:55,750 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 05:31:10,506 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:32:32,422 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:37:27,413 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:38:39,694 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:39:33,806 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:39:38,992 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:40:03,551 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:43:34,425 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:52:58,744 - train - INFO - epoch : 70
2025-10-16 05:52:58,745 - train - INFO - loss : 0.3450617952644825
2025-10-16 05:52:58,745 - train - INFO - grad_norm : 0.6964107640087605
2025-10-16 05:52:58,745 - train - INFO - val_loss : 0.2472096932205287
2025-10-16 05:52:58,745 - train - INFO - val_CER_(Argmax): 0.07284972328502712
2025-10-16 05:52:58,745 - train - INFO - val_WER_(Argmax): 0.22901276614441682
2025-10-16 05:52:58,745 - train - INFO - val_WER : 0.22901276614441682
2025-10-16 05:52:58,745 - train - INFO - val_CER : 0.07284273607724429
2025-10-16 05:52:58,746 - train - INFO - test_loss : 0.6180139366461306
2025-10-16 05:52:58,746 - train - INFO - test_CER_(Argmax): 0.1804781446442257
2025-10-16 05:52:58,746 - train - INFO - test_WER_(Argmax): 0.45284049314400276
2025-10-16 05:52:58,746 - train - INFO - test_WER : 0.45284049314400276
2025-10-16 05:52:58,746 - train - INFO - test_CER : 0.1804666251609065
2025-10-16 05:52:59,372 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 05:56:08,152 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:59:33,037 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 05:59:56,321 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:05:25,117 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:07:33,143 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:08:00,984 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:16:10,784 - train - INFO - epoch : 71
2025-10-16 06:16:10,784 - train - INFO - loss : 0.34330349164692003
2025-10-16 06:16:10,784 - train - INFO - grad_norm : 0.7429921634532698
2025-10-16 06:16:10,785 - train - INFO - val_loss : 0.2456764829429713
2025-10-16 06:16:10,785 - train - INFO - val_CER_(Argmax): 0.07272585291214713
2025-10-16 06:16:10,785 - train - INFO - val_WER_(Argmax): 0.2281453037552794
2025-10-16 06:16:10,785 - train - INFO - val_WER : 0.2281453037552794
2025-10-16 06:16:10,785 - train - INFO - val_CER : 0.07276095324645183
2025-10-16 06:16:10,785 - train - INFO - test_loss : 0.6153945196039823
2025-10-16 06:16:10,785 - train - INFO - test_CER_(Argmax): 0.18028797138951858
2025-10-16 06:16:10,785 - train - INFO - test_WER_(Argmax): 0.45096762978426175
2025-10-16 06:16:10,785 - train - INFO - test_WER : 0.45096762978426175
2025-10-16 06:16:10,785 - train - INFO - test_CER : 0.18027404470283673
2025-10-16 06:16:11,370 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 06:22:36,692 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:22:50,342 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:24:49,221 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:25:27,796 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:27:48,359 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:29:38,242 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:39:03,837 - train - INFO - epoch : 72
2025-10-16 06:39:03,838 - train - INFO - loss : 0.339138802103322
2025-10-16 06:39:03,838 - train - INFO - grad_norm : 0.7133045122779981
2025-10-16 06:39:03,838 - train - INFO - val_loss : 0.2443471114066514
2025-10-16 06:39:03,839 - train - INFO - val_CER_(Argmax): 0.07240100881127012
2025-10-16 06:39:03,839 - train - INFO - val_WER_(Argmax): 0.22696484163654596
2025-10-16 06:39:03,839 - train - INFO - val_WER : 0.22696484163654596
2025-10-16 06:39:03,839 - train - INFO - val_CER : 0.07238644584591152
2025-10-16 06:39:03,839 - train - INFO - test_loss : 0.612513683888377
2025-10-16 06:39:03,840 - train - INFO - test_CER_(Argmax): 0.1791687001827859
2025-10-16 06:39:03,840 - train - INFO - test_WER_(Argmax): 0.4488865730597953
2025-10-16 06:39:03,840 - train - INFO - test_WER : 0.4488865730597953
2025-10-16 06:39:03,840 - train - INFO - test_CER : 0.17915718069946662
2025-10-16 06:39:04,972 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 06:41:02,810 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:45:51,176 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:47:05,013 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:49:14,158 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 06:49:39,422 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:01:47,234 - train - INFO - epoch : 73
2025-10-16 07:01:47,235 - train - INFO - loss : 0.3470729561895132
2025-10-16 07:01:47,235 - train - INFO - grad_norm : 0.7514531427621841
2025-10-16 07:01:47,235 - train - INFO - val_loss : 0.24378775940700012
2025-10-16 07:01:47,235 - train - INFO - val_CER_(Argmax): 0.07183435190371569
2025-10-16 07:01:47,235 - train - INFO - val_WER_(Argmax): 0.2256147255339385
2025-10-16 07:01:47,235 - train - INFO - val_WER : 0.2256147255339385
2025-10-16 07:01:47,235 - train - INFO - val_CER : 0.0718197889383571
2025-10-16 07:01:47,235 - train - INFO - test_loss : 0.6091715346793739
2025-10-16 07:01:47,236 - train - INFO - test_CER_(Argmax): 0.17911457671301328
2025-10-16 07:01:47,236 - train - INFO - test_WER_(Argmax): 0.44794237444369694
2025-10-16 07:01:47,236 - train - INFO - test_WER : 0.44794237444369694
2025-10-16 07:01:47,236 - train - INFO - test_CER : 0.17910649294741543
2025-10-16 07:01:47,826 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 07:04:43,210 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:05:38,351 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:07:33,361 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:10:39,941 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:16:21,980 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:24:38,614 - train - INFO - epoch : 74
2025-10-16 07:24:38,615 - train - INFO - loss : 0.33683984458446503
2025-10-16 07:24:38,615 - train - INFO - grad_norm : 0.7467496731877327
2025-10-16 07:24:38,615 - train - INFO - val_loss : 0.2443155725909905
2025-10-16 07:24:38,616 - train - INFO - val_CER_(Argmax): 0.07227646308735522
2025-10-16 07:24:38,616 - train - INFO - val_WER_(Argmax): 0.22735330118624314
2025-10-16 07:24:38,616 - train - INFO - val_WER : 0.22735330118624314
2025-10-16 07:24:38,616 - train - INFO - val_CER : 0.07227911775285062
2025-10-16 07:24:38,616 - train - INFO - test_loss : 0.6112303989274162
2025-10-16 07:24:38,616 - train - INFO - test_CER_(Argmax): 0.17907364584589097
2025-10-16 07:24:38,616 - train - INFO - test_WER_(Argmax): 0.44814157881320876
2025-10-16 07:24:38,616 - train - INFO - test_WER : 0.44814157881320876
2025-10-16 07:24:38,616 - train - INFO - test_CER : 0.17904936512532063
2025-10-16 07:25:13,751 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:25:37,831 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:28:57,741 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:29:38,177 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:34:16,623 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:37:09,990 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:47:25,461 - train - INFO - epoch : 75
2025-10-16 07:47:25,461 - train - INFO - loss : 0.33465448658370495
2025-10-16 07:47:25,461 - train - INFO - grad_norm : 0.7519346656811298
2025-10-16 07:47:25,462 - train - INFO - val_loss : 0.242401727729223
2025-10-16 07:47:25,462 - train - INFO - val_CER_(Argmax): 0.07174316270698235
2025-10-16 07:47:25,462 - train - INFO - val_WER_(Argmax): 0.2253097904291825
2025-10-16 07:47:25,462 - train - INFO - val_WER : 0.2253097904291825
2025-10-16 07:47:25,462 - train - INFO - val_CER : 0.07172859974162374
2025-10-16 07:47:25,462 - train - INFO - test_loss : 0.6091889444054389
2025-10-16 07:47:25,462 - train - INFO - test_CER_(Argmax): 0.17898326704627793
2025-10-16 07:47:25,462 - train - INFO - test_WER_(Argmax): 0.4477697155209368
2025-10-16 07:47:25,462 - train - INFO - test_WER : 0.4477697155209368
2025-10-16 07:47:25,462 - train - INFO - test_CER : 0.17895496802095612
2025-10-16 07:47:26,060 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 07:50:39,738 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:50:41,382 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:51:44,271 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:52:08,305 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:57:55,955 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 07:58:00,203 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:00:15,606 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:10:09,966 - train - INFO - epoch : 76
2025-10-16 08:10:09,967 - train - INFO - loss : 0.32825977680970675
2025-10-16 08:10:09,967 - train - INFO - grad_norm : 0.7922337020761404
2025-10-16 08:10:09,967 - train - INFO - val_loss : 0.24259263547984036
2025-10-16 08:10:09,967 - train - INFO - val_CER_(Argmax): 0.07152227206748903
2025-10-16 08:10:09,967 - train - INFO - val_WER_(Argmax): 0.22444476221779855
2025-10-16 08:10:09,967 - train - INFO - val_WER : 0.22444476221779855
2025-10-16 08:10:09,968 - train - INFO - val_CER : 0.0715126284252316
2025-10-16 08:10:09,968 - train - INFO - test_loss : 0.60749830001471
2025-10-16 08:10:09,968 - train - INFO - test_CER_(Argmax): 0.1776300184520875
2025-10-16 08:10:09,968 - train - INFO - test_WER_(Argmax): 0.4444251261265295
2025-10-16 08:10:09,968 - train - INFO - test_WER : 0.4444251261265295
2025-10-16 08:10:09,968 - train - INFO - test_CER : 0.177609227132328
2025-10-16 08:10:10,533 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 08:11:59,900 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:12:54,045 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:14:34,964 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:21:55,106 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:21:55,451 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:22:06,483 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:33:19,822 - train - INFO - epoch : 77
2025-10-16 08:33:19,822 - train - INFO - loss : 0.33436640658354394
2025-10-16 08:33:19,822 - train - INFO - grad_norm : 0.7152383050640222
2025-10-16 08:33:19,822 - train - INFO - val_loss : 0.24181210639124567
2025-10-16 08:33:19,823 - train - INFO - val_CER_(Argmax): 0.07168098424058468
2025-10-16 08:33:19,823 - train - INFO - val_WER_(Argmax): 0.22612596969878818
2025-10-16 08:33:19,823 - train - INFO - val_WER : 0.22612596969878818
2025-10-16 08:33:19,823 - train - INFO - val_CER : 0.07166642127522609
2025-10-16 08:33:19,823 - train - INFO - test_loss : 0.6055028849110311
2025-10-16 08:33:19,823 - train - INFO - test_CER_(Argmax): 0.17730527841163873
2025-10-16 08:33:19,823 - train - INFO - test_WER_(Argmax): 0.444182600467874
2025-10-16 08:33:19,823 - train - INFO - test_WER : 0.444182600467874
2025-10-16 08:33:19,823 - train - INFO - test_CER : 0.1772793335152971
2025-10-16 08:35:13,806 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:36:08,863 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:36:27,911 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:37:11,720 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:42:38,767 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:44:59,770 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:56:30,978 - train - INFO - epoch : 78
2025-10-16 08:56:30,979 - train - INFO - loss : 0.3246919604102571
2025-10-16 08:56:30,979 - train - INFO - grad_norm : 0.7791670113053154
2025-10-16 08:56:30,979 - train - INFO - val_loss : 0.24172794463282282
2025-10-16 08:56:30,979 - train - INFO - val_CER_(Argmax): 0.07153011484291474
2025-10-16 08:56:30,979 - train - INFO - val_WER_(Argmax): 0.22511833258022118
2025-10-16 08:56:30,979 - train - INFO - val_WER : 0.22511833258022118
2025-10-16 08:56:30,979 - train - INFO - val_CER : 0.071512884357283
2025-10-16 08:56:30,979 - train - INFO - test_loss : 0.6050408068968325
2025-10-16 08:56:30,980 - train - INFO - test_CER_(Argmax): 0.17660421668923618
2025-10-16 08:56:30,980 - train - INFO - test_WER_(Argmax): 0.44288891060599644
2025-10-16 08:56:30,980 - train - INFO - test_WER : 0.44288891060599644
2025-10-16 08:56:30,980 - train - INFO - test_CER : 0.1765579884047018
2025-10-16 08:59:27,293 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 08:59:27,869 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:02:30,628 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:07:50,034 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:09:45,540 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:10:48,602 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:19:14,555 - train - INFO - epoch : 79
2025-10-16 09:19:14,556 - train - INFO - loss : 0.32903695121558024
2025-10-16 09:19:14,556 - train - INFO - grad_norm : 0.7922589413144372
2025-10-16 09:19:14,556 - train - INFO - val_loss : 0.24085671488534321
2025-10-16 09:19:14,556 - train - INFO - val_CER_(Argmax): 0.07120119172424885
2025-10-16 09:19:14,556 - train - INFO - val_WER_(Argmax): 0.2238828798942847
2025-10-16 09:19:14,556 - train - INFO - val_WER : 0.2238828798942847
2025-10-16 09:19:14,556 - train - INFO - val_CER : 0.07118662875889026
2025-10-16 09:19:14,557 - train - INFO - test_loss : 0.6060074610369546
2025-10-16 09:19:14,557 - train - INFO - test_CER_(Argmax): 0.17613761186333135
2025-10-16 09:19:14,557 - train - INFO - test_WER_(Argmax): 0.4415207936896337
2025-10-16 09:19:14,557 - train - INFO - test_WER : 0.4415207936896337
2025-10-16 09:19:14,557 - train - INFO - test_CER : 0.1760987778408496
2025-10-16 09:19:15,154 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 09:21:32,328 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:25:32,846 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:26:24,808 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:27:11,142 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:31:43,693 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:33:07,799 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:42:14,548 - train - INFO - epoch : 80
2025-10-16 09:42:14,549 - train - INFO - loss : 0.32665394328946445
2025-10-16 09:42:14,549 - train - INFO - grad_norm : 0.785485486439125
2025-10-16 09:42:14,549 - train - INFO - val_loss : 0.23916942947967487
2025-10-16 09:42:14,549 - train - INFO - val_CER_(Argmax): 0.07079503833489652
2025-10-16 09:42:14,549 - train - INFO - val_WER_(Argmax): 0.2233928609416103
2025-10-16 09:42:14,549 - train - INFO - val_WER : 0.2233928609416103
2025-10-16 09:42:14,549 - train - INFO - val_CER : 0.07078805112711367
2025-10-16 09:42:14,549 - train - INFO - test_loss : 0.5997372880882147
2025-10-16 09:42:14,549 - train - INFO - test_CER_(Argmax): 0.17529654974260578
2025-10-16 09:42:14,550 - train - INFO - test_WER_(Argmax): 0.4405526126548696
2025-10-16 09:42:14,550 - train - INFO - test_WER : 0.4405526126548696
2025-10-16 09:42:14,550 - train - INFO - test_CER : 0.1752632917210163
2025-10-16 09:42:15,152 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 09:42:15,591 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:42:48,136 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:43:41,631 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:44:09,545 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:45:05,149 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:49:31,128 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:55:41,850 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 09:55:42,189 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:05:05,706 - train - INFO - epoch : 81
2025-10-16 10:05:05,707 - train - INFO - loss : 0.32321790718671045
2025-10-16 10:05:05,707 - train - INFO - grad_norm : 0.8345202321958061
2025-10-16 10:05:05,707 - train - INFO - val_loss : 0.2388677727431059
2025-10-16 10:05:05,707 - train - INFO - val_CER_(Argmax): 0.07033447157901968
2025-10-16 10:05:05,707 - train - INFO - val_WER_(Argmax): 0.22307846231609646
2025-10-16 10:05:05,707 - train - INFO - val_WER : 0.22307846231609646
2025-10-16 10:05:05,707 - train - INFO - val_CER : 0.0703199086136611
2025-10-16 10:05:05,707 - train - INFO - test_loss : 0.599141996429891
2025-10-16 10:05:05,708 - train - INFO - test_CER_(Argmax): 0.174724708652118
2025-10-16 10:05:05,708 - train - INFO - test_WER_(Argmax): 0.43903859863372224
2025-10-16 10:05:05,708 - train - INFO - test_WER : 0.43903859863372224
2025-10-16 10:05:05,708 - train - INFO - test_CER : 0.1747035251574909
2025-10-16 10:05:06,299 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 10:09:04,114 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:10:02,193 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:10:55,240 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:14:42,223 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:15:31,459 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:18:16,351 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:19:34,572 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:28:09,798 - train - INFO - epoch : 82
2025-10-16 10:28:09,799 - train - INFO - loss : 0.3183249425348924
2025-10-16 10:28:09,799 - train - INFO - grad_norm : 0.8161137418531293
2025-10-16 10:28:09,799 - train - INFO - val_loss : 0.2384590067985383
2025-10-16 10:28:09,799 - train - INFO - val_CER_(Argmax): 0.07017255953778744
2025-10-16 10:28:09,799 - train - INFO - val_WER_(Argmax): 0.22131364034164988
2025-10-16 10:28:09,799 - train - INFO - val_WER : 0.22131364034164988
2025-10-16 10:28:09,800 - train - INFO - val_CER : 0.07015799657242883
2025-10-16 10:28:09,800 - train - INFO - test_loss : 0.5989029520020193
2025-10-16 10:28:09,800 - train - INFO - test_CER_(Argmax): 0.1743205254276692
2025-10-16 10:28:09,800 - train - INFO - test_WER_(Argmax): 0.43704721422289666
2025-10-16 10:28:09,800 - train - INFO - test_WER : 0.43704721422289666
2025-10-16 10:28:09,800 - train - INFO - test_CER : 0.1743163534871922
2025-10-16 10:28:10,392 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 10:30:51,295 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:32:11,943 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:33:27,960 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:36:33,747 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:37:19,556 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:40:31,804 - train - WARNING - OOM on batch. Skipping batch.
2025-10-16 10:51:18,473 - train - INFO - epoch : 83
2025-10-16 10:51:18,473 - train - INFO - loss : 0.3176834620722574
2025-10-16 10:51:18,473 - train - INFO - grad_norm : 0.8232347652241213
2025-10-16 10:51:18,474 - train - INFO - val_loss : 0.23727865584871985
2025-10-16 10:51:18,474 - train - INFO - val_CER_(Argmax): 0.07039769349766453
2025-10-16 10:51:18,474 - train - INFO - val_WER_(Argmax): 0.2212025348655912
2025-10-16 10:51:18,474 - train - INFO - val_WER : 0.2212025348655912
2025-10-16 10:51:18,474 - train - INFO - val_CER : 0.07038313053230592
2025-10-16 10:51:18,474 - train - INFO - test_loss : 0.5976807195313123
2025-10-16 10:51:18,474 - train - INFO - test_CER_(Argmax): 0.17417757347516163
2025-10-16 10:51:18,474 - train - INFO - test_WER_(Argmax): 0.4383814222677989
2025-10-16 10:51:18,474 - train - INFO - test_WER : 0.4383814222677989
2025-10-16 10:51:18,474 - train - INFO - test_CER : 0.1741524139310583
2025-10-16 10:51:19,085 - train - INFO - Saving current best: model_best.pth ...
2025-10-16 10:51:40,238 - train - INFO - Saving model on keyboard interrupt
2025-10-16 10:51:40,720 - train - INFO - Saving checkpoint: /home/nyakovchuk/wave_rover/rover-Conformer-ASR/saved/conformer_30m/checkpoint-epoch84.pth ...
|