Commit
·
612bb8d
1
Parent(s):
24f536c
update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,228 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- common_voice
|
| 7 |
+
model-index:
|
| 8 |
+
- name: model
|
| 9 |
+
results: []
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
+
|
| 15 |
+
# model
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the common_voice dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.2220
|
| 20 |
+
- Wer: 0.1301
|
| 21 |
+
|
| 22 |
+
## Model description
|
| 23 |
+
|
| 24 |
+
More information needed
|
| 25 |
+
|
| 26 |
+
## Intended uses & limitations
|
| 27 |
+
|
| 28 |
+
More information needed
|
| 29 |
+
|
| 30 |
+
## Training and evaluation data
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Training procedure
|
| 35 |
+
|
| 36 |
+
### Training hyperparameters
|
| 37 |
+
|
| 38 |
+
The following hyperparameters were used during training:
|
| 39 |
+
- learning_rate: 0.0003
|
| 40 |
+
- train_batch_size: 16
|
| 41 |
+
- eval_batch_size: 8
|
| 42 |
+
- seed: 42
|
| 43 |
+
- gradient_accumulation_steps: 2
|
| 44 |
+
- total_train_batch_size: 32
|
| 45 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
+
- lr_scheduler_type: linear
|
| 47 |
+
- lr_scheduler_warmup_steps: 500
|
| 48 |
+
- num_epochs: 30
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 54 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|
|
| 55 |
+
| 5.9743 | 0.18 | 400 | 2.1457 | 1.0000 |
|
| 56 |
+
| 0.5747 | 0.36 | 800 | 0.3415 | 0.3456 |
|
| 57 |
+
| 0.3383 | 0.54 | 1200 | 0.2797 | 0.3095 |
|
| 58 |
+
| 0.2967 | 0.72 | 1600 | 0.2464 | 0.2568 |
|
| 59 |
+
| 0.2747 | 0.9 | 2000 | 0.2341 | 0.2466 |
|
| 60 |
+
| 0.2501 | 1.08 | 2400 | 0.2299 | 0.2317 |
|
| 61 |
+
| 0.2309 | 1.26 | 2800 | 0.2306 | 0.2328 |
|
| 62 |
+
| 0.2273 | 1.44 | 3200 | 0.2212 | 0.2375 |
|
| 63 |
+
| 0.225 | 1.62 | 3600 | 0.2193 | 0.2267 |
|
| 64 |
+
| 0.2204 | 1.8 | 4000 | 0.2157 | 0.2295 |
|
| 65 |
+
| 0.2256 | 1.98 | 4400 | 0.2165 | 0.2260 |
|
| 66 |
+
| 0.1941 | 2.17 | 4800 | 0.2105 | 0.2163 |
|
| 67 |
+
| 0.1925 | 2.35 | 5200 | 0.2098 | 0.2153 |
|
| 68 |
+
| 0.1925 | 2.53 | 5600 | 0.2120 | 0.2148 |
|
| 69 |
+
| 0.1952 | 2.71 | 6000 | 0.2063 | 0.2178 |
|
| 70 |
+
| 0.1971 | 2.89 | 6400 | 0.2100 | 0.2158 |
|
| 71 |
+
| 0.1888 | 3.07 | 6800 | 0.2131 | 0.2172 |
|
| 72 |
+
| 0.1702 | 3.25 | 7200 | 0.2155 | 0.2203 |
|
| 73 |
+
| 0.173 | 3.43 | 7600 | 0.2141 | 0.2254 |
|
| 74 |
+
| 0.174 | 3.61 | 8000 | 0.2017 | 0.2100 |
|
| 75 |
+
| 0.1802 | 3.79 | 8400 | 0.1998 | 0.2043 |
|
| 76 |
+
| 0.1717 | 3.97 | 8800 | 0.2070 | 0.2110 |
|
| 77 |
+
| 0.162 | 4.15 | 9200 | 0.2082 | 0.2157 |
|
| 78 |
+
| 0.154 | 4.33 | 9600 | 0.2163 | 0.2161 |
|
| 79 |
+
| 0.1598 | 4.51 | 10000 | 0.2070 | 0.2171 |
|
| 80 |
+
| 0.1576 | 4.69 | 10400 | 0.2034 | 0.2116 |
|
| 81 |
+
| 0.1601 | 4.87 | 10800 | 0.1990 | 0.2009 |
|
| 82 |
+
| 0.152 | 5.05 | 11200 | 0.1994 | 0.2039 |
|
| 83 |
+
| 0.1395 | 5.23 | 11600 | 0.2013 | 0.2046 |
|
| 84 |
+
| 0.1407 | 5.41 | 12000 | 0.2009 | 0.2022 |
|
| 85 |
+
| 0.1449 | 5.59 | 12400 | 0.1982 | 0.1961 |
|
| 86 |
+
| 0.1483 | 5.77 | 12800 | 0.2082 | 0.2054 |
|
| 87 |
+
| 0.1514 | 5.95 | 13200 | 0.1953 | 0.1985 |
|
| 88 |
+
| 0.138 | 6.13 | 13600 | 0.2046 | 0.1965 |
|
| 89 |
+
| 0.1322 | 6.31 | 14000 | 0.2076 | 0.1948 |
|
| 90 |
+
| 0.1372 | 6.5 | 14400 | 0.1968 | 0.1944 |
|
| 91 |
+
| 0.136 | 6.68 | 14800 | 0.1971 | 0.1963 |
|
| 92 |
+
| 0.1382 | 6.86 | 15200 | 0.2001 | 0.1990 |
|
| 93 |
+
| 0.1335 | 7.04 | 15600 | 0.2026 | 0.1935 |
|
| 94 |
+
| 0.1206 | 7.22 | 16000 | 0.1986 | 0.1938 |
|
| 95 |
+
| 0.1239 | 7.4 | 16400 | 0.2054 | 0.1919 |
|
| 96 |
+
| 0.1254 | 7.58 | 16800 | 0.1918 | 0.1939 |
|
| 97 |
+
| 0.1262 | 7.76 | 17200 | 0.1960 | 0.1947 |
|
| 98 |
+
| 0.126 | 7.94 | 17600 | 0.1932 | 0.1906 |
|
| 99 |
+
| 0.1169 | 8.12 | 18000 | 0.2037 | 0.1916 |
|
| 100 |
+
| 0.1142 | 8.3 | 18400 | 0.1999 | 0.1900 |
|
| 101 |
+
| 0.1151 | 8.48 | 18800 | 0.1920 | 0.1855 |
|
| 102 |
+
| 0.1121 | 8.66 | 19200 | 0.2007 | 0.1859 |
|
| 103 |
+
| 0.1135 | 8.84 | 19600 | 0.1932 | 0.1879 |
|
| 104 |
+
| 0.1158 | 9.02 | 20000 | 0.1916 | 0.1859 |
|
| 105 |
+
| 0.105 | 9.2 | 20400 | 0.1961 | 0.1831 |
|
| 106 |
+
| 0.1023 | 9.38 | 20800 | 0.1914 | 0.1791 |
|
| 107 |
+
| 0.1004 | 9.56 | 21200 | 0.1881 | 0.1787 |
|
| 108 |
+
| 0.1023 | 9.74 | 21600 | 0.1963 | 0.1817 |
|
| 109 |
+
| 0.1075 | 9.92 | 22000 | 0.1889 | 0.1861 |
|
| 110 |
+
| 0.103 | 10.1 | 22400 | 0.1975 | 0.1791 |
|
| 111 |
+
| 0.0952 | 10.28 | 22800 | 0.1979 | 0.1787 |
|
| 112 |
+
| 0.0957 | 10.46 | 23200 | 0.1922 | 0.1817 |
|
| 113 |
+
| 0.0966 | 10.65 | 23600 | 0.1953 | 0.1857 |
|
| 114 |
+
| 0.0997 | 10.83 | 24000 | 0.1902 | 0.1783 |
|
| 115 |
+
| 0.0981 | 11.01 | 24400 | 0.1959 | 0.1780 |
|
| 116 |
+
| 0.0868 | 11.19 | 24800 | 0.2056 | 0.1783 |
|
| 117 |
+
| 0.0905 | 11.37 | 25200 | 0.1958 | 0.1777 |
|
| 118 |
+
| 0.0892 | 11.55 | 25600 | 0.1935 | 0.1796 |
|
| 119 |
+
| 0.0891 | 11.73 | 26000 | 0.1968 | 0.1763 |
|
| 120 |
+
| 0.0888 | 11.91 | 26400 | 0.2043 | 0.1804 |
|
| 121 |
+
| 0.0842 | 12.09 | 26800 | 0.2043 | 0.1733 |
|
| 122 |
+
| 0.0828 | 12.27 | 27200 | 0.1964 | 0.1715 |
|
| 123 |
+
| 0.0827 | 12.45 | 27600 | 0.1991 | 0.1749 |
|
| 124 |
+
| 0.0844 | 12.63 | 28000 | 0.2014 | 0.1695 |
|
| 125 |
+
| 0.0837 | 12.81 | 28400 | 0.1973 | 0.1759 |
|
| 126 |
+
| 0.0872 | 12.99 | 28800 | 0.1975 | 0.1689 |
|
| 127 |
+
| 0.0778 | 13.17 | 29200 | 0.1979 | 0.1740 |
|
| 128 |
+
| 0.0759 | 13.35 | 29600 | 0.2093 | 0.1753 |
|
| 129 |
+
| 0.076 | 13.53 | 30000 | 0.1990 | 0.1731 |
|
| 130 |
+
| 0.0762 | 13.71 | 30400 | 0.2024 | 0.1690 |
|
| 131 |
+
| 0.0764 | 13.89 | 30800 | 0.2037 | 0.1709 |
|
| 132 |
+
| 0.0756 | 14.07 | 31200 | 0.2007 | 0.1716 |
|
| 133 |
+
| 0.0702 | 14.25 | 31600 | 0.2011 | 0.1680 |
|
| 134 |
+
| 0.0694 | 14.43 | 32000 | 0.2061 | 0.1683 |
|
| 135 |
+
| 0.0713 | 14.61 | 32400 | 0.2014 | 0.1687 |
|
| 136 |
+
| 0.0693 | 14.79 | 32800 | 0.1961 | 0.1658 |
|
| 137 |
+
| 0.071 | 14.98 | 33200 | 0.1921 | 0.1645 |
|
| 138 |
+
| 0.0659 | 15.16 | 33600 | 0.2079 | 0.1682 |
|
| 139 |
+
| 0.0659 | 15.34 | 34000 | 0.2046 | 0.1649 |
|
| 140 |
+
| 0.0685 | 15.52 | 34400 | 0.1994 | 0.1660 |
|
| 141 |
+
| 0.0663 | 15.7 | 34800 | 0.1970 | 0.1652 |
|
| 142 |
+
| 0.0678 | 15.88 | 35200 | 0.1961 | 0.1634 |
|
| 143 |
+
| 0.0644 | 16.06 | 35600 | 0.2141 | 0.1644 |
|
| 144 |
+
| 0.0596 | 16.24 | 36000 | 0.2098 | 0.1628 |
|
| 145 |
+
| 0.0629 | 16.42 | 36400 | 0.1969 | 0.1616 |
|
| 146 |
+
| 0.0598 | 16.6 | 36800 | 0.2026 | 0.1604 |
|
| 147 |
+
| 0.0628 | 16.78 | 37200 | 0.2050 | 0.1620 |
|
| 148 |
+
| 0.0616 | 16.96 | 37600 | 0.1958 | 0.1618 |
|
| 149 |
+
| 0.0538 | 17.14 | 38000 | 0.2093 | 0.1588 |
|
| 150 |
+
| 0.0573 | 17.32 | 38400 | 0.1995 | 0.1588 |
|
| 151 |
+
| 0.0555 | 17.5 | 38800 | 0.2077 | 0.1608 |
|
| 152 |
+
| 0.0555 | 17.68 | 39200 | 0.2036 | 0.1571 |
|
| 153 |
+
| 0.0578 | 17.86 | 39600 | 0.2045 | 0.1572 |
|
| 154 |
+
| 0.056 | 18.04 | 40000 | 0.2065 | 0.1593 |
|
| 155 |
+
| 0.0525 | 18.22 | 40400 | 0.2093 | 0.1580 |
|
| 156 |
+
| 0.0527 | 18.4 | 40800 | 0.2141 | 0.1585 |
|
| 157 |
+
| 0.0529 | 18.58 | 41200 | 0.2137 | 0.1585 |
|
| 158 |
+
| 0.0533 | 18.76 | 41600 | 0.2021 | 0.1558 |
|
| 159 |
+
| 0.0529 | 18.94 | 42000 | 0.2108 | 0.1535 |
|
| 160 |
+
| 0.05 | 19.12 | 42400 | 0.2114 | 0.1555 |
|
| 161 |
+
| 0.0479 | 19.31 | 42800 | 0.2091 | 0.1549 |
|
| 162 |
+
| 0.0509 | 19.49 | 43200 | 0.2145 | 0.1554 |
|
| 163 |
+
| 0.0486 | 19.67 | 43600 | 0.2061 | 0.1536 |
|
| 164 |
+
| 0.049 | 19.85 | 44000 | 0.2132 | 0.1548 |
|
| 165 |
+
| 0.0484 | 20.03 | 44400 | 0.2077 | 0.1523 |
|
| 166 |
+
| 0.0449 | 20.21 | 44800 | 0.2177 | 0.1529 |
|
| 167 |
+
| 0.0452 | 20.39 | 45200 | 0.2204 | 0.1517 |
|
| 168 |
+
| 0.0477 | 20.57 | 45600 | 0.2132 | 0.1517 |
|
| 169 |
+
| 0.048 | 20.75 | 46000 | 0.2119 | 0.1532 |
|
| 170 |
+
| 0.0469 | 20.93 | 46400 | 0.2109 | 0.1524 |
|
| 171 |
+
| 0.0439 | 21.11 | 46800 | 0.2118 | 0.1503 |
|
| 172 |
+
| 0.044 | 21.29 | 47200 | 0.2033 | 0.1474 |
|
| 173 |
+
| 0.0435 | 21.47 | 47600 | 0.2066 | 0.1485 |
|
| 174 |
+
| 0.0418 | 21.65 | 48000 | 0.2125 | 0.1491 |
|
| 175 |
+
| 0.0417 | 21.83 | 48400 | 0.2139 | 0.1487 |
|
| 176 |
+
| 0.0446 | 22.01 | 48800 | 0.2054 | 0.1493 |
|
| 177 |
+
| 0.039 | 22.19 | 49200 | 0.2179 | 0.1459 |
|
| 178 |
+
| 0.0414 | 22.37 | 49600 | 0.2118 | 0.1466 |
|
| 179 |
+
| 0.0394 | 22.55 | 50000 | 0.2104 | 0.1444 |
|
| 180 |
+
| 0.0381 | 22.73 | 50400 | 0.2095 | 0.1458 |
|
| 181 |
+
| 0.0382 | 22.91 | 50800 | 0.2193 | 0.1471 |
|
| 182 |
+
| 0.0391 | 23.09 | 51200 | 0.2143 | 0.1455 |
|
| 183 |
+
| 0.0365 | 23.27 | 51600 | 0.2198 | 0.1445 |
|
| 184 |
+
| 0.0368 | 23.46 | 52000 | 0.2151 | 0.1444 |
|
| 185 |
+
| 0.038 | 23.64 | 52400 | 0.2094 | 0.1439 |
|
| 186 |
+
| 0.038 | 23.82 | 52800 | 0.2137 | 0.1422 |
|
| 187 |
+
| 0.0374 | 24.0 | 53200 | 0.2180 | 0.1425 |
|
| 188 |
+
| 0.0352 | 24.18 | 53600 | 0.2207 | 0.1422 |
|
| 189 |
+
| 0.0343 | 24.36 | 54000 | 0.2269 | 0.1445 |
|
| 190 |
+
| 0.0353 | 24.54 | 54400 | 0.2222 | 0.1438 |
|
| 191 |
+
| 0.0348 | 24.72 | 54800 | 0.2224 | 0.1413 |
|
| 192 |
+
| 0.0342 | 24.9 | 55200 | 0.2146 | 0.1401 |
|
| 193 |
+
| 0.0337 | 25.08 | 55600 | 0.2246 | 0.1408 |
|
| 194 |
+
| 0.0327 | 25.26 | 56000 | 0.2161 | 0.1401 |
|
| 195 |
+
| 0.0339 | 25.44 | 56400 | 0.2212 | 0.1402 |
|
| 196 |
+
| 0.0324 | 25.62 | 56800 | 0.2203 | 0.1394 |
|
| 197 |
+
| 0.0319 | 25.8 | 57200 | 0.2145 | 0.1376 |
|
| 198 |
+
| 0.0317 | 25.98 | 57600 | 0.2147 | 0.1375 |
|
| 199 |
+
| 0.0302 | 26.16 | 58000 | 0.2213 | 0.1362 |
|
| 200 |
+
| 0.0309 | 26.34 | 58400 | 0.2218 | 0.1365 |
|
| 201 |
+
| 0.0308 | 26.52 | 58800 | 0.2167 | 0.1362 |
|
| 202 |
+
| 0.0294 | 26.7 | 59200 | 0.2169 | 0.1368 |
|
| 203 |
+
| 0.0297 | 26.88 | 59600 | 0.2163 | 0.1350 |
|
| 204 |
+
| 0.0289 | 27.06 | 60000 | 0.2188 | 0.1348 |
|
| 205 |
+
| 0.0284 | 27.24 | 60400 | 0.2172 | 0.1338 |
|
| 206 |
+
| 0.0278 | 27.42 | 60800 | 0.2230 | 0.1342 |
|
| 207 |
+
| 0.0283 | 27.6 | 61200 | 0.2233 | 0.1342 |
|
| 208 |
+
| 0.0292 | 27.79 | 61600 | 0.2238 | 0.1335 |
|
| 209 |
+
| 0.0286 | 27.97 | 62000 | 0.2218 | 0.1327 |
|
| 210 |
+
| 0.0262 | 28.15 | 62400 | 0.2220 | 0.1324 |
|
| 211 |
+
| 0.0274 | 28.33 | 62800 | 0.2182 | 0.1323 |
|
| 212 |
+
| 0.0279 | 28.51 | 63200 | 0.2170 | 0.1314 |
|
| 213 |
+
| 0.0269 | 28.69 | 63600 | 0.2228 | 0.1313 |
|
| 214 |
+
| 0.0264 | 28.87 | 64000 | 0.2209 | 0.1313 |
|
| 215 |
+
| 0.0254 | 29.05 | 64400 | 0.2224 | 0.1304 |
|
| 216 |
+
| 0.026 | 29.23 | 64800 | 0.2220 | 0.1302 |
|
| 217 |
+
| 0.0253 | 29.41 | 65200 | 0.2229 | 0.1304 |
|
| 218 |
+
| 0.0244 | 29.59 | 65600 | 0.2217 | 0.1298 |
|
| 219 |
+
| 0.025 | 29.77 | 66000 | 0.2223 | 0.1303 |
|
| 220 |
+
| 0.0255 | 29.95 | 66400 | 0.2220 | 0.1301 |
|
| 221 |
+
|
| 222 |
+
|
| 223 |
+
### Framework versions
|
| 224 |
+
|
| 225 |
+
- Transformers 4.11.3
|
| 226 |
+
- Pytorch 1.9.1+cu102
|
| 227 |
+
- Datasets 1.18.3
|
| 228 |
+
- Tokenizers 0.10.3
|