Update README.md
Browse files
README.md
CHANGED
@@ -7,15 +7,18 @@ model-index:
|
|
7 |
results: []
|
8 |
---
|
9 |
|
10 |
-
# bart-base-spelling-nl
|
11 |
|
12 |
This model is a Dutch fine-tuned version of
|
13 |
[facebook/bart-base](https://huggingface.co/facebook/bart-base).
|
14 |
|
15 |
-
It achieves the following results on
|
|
|
|
|
|
|
|
|
|
|
16 |
|
17 |
-
- Loss: 0.0053
|
18 |
-
- Cer: 0.0117
|
19 |
|
20 |
## Model description
|
21 |
|
@@ -35,12 +38,12 @@ checker.
|
|
35 |
|
36 |
## Training and evaluation data
|
37 |
|
38 |
-
The model was trained on a Dutch dataset composed of
|
39 |
of text from three public Dutch sources, downloaded from the
|
40 |
[Opus corpus](https://opus.nlpl.eu/):
|
41 |
|
42 |
-
- nl-europarlv7.
|
43 |
-
- nl-opensubtitles2016.
|
44 |
- nl-wikipedia.txt (964,203 lines)
|
45 |
|
46 |
|
@@ -59,196 +62,6 @@ The following hyperparameters were used during training:
|
|
59 |
- lr_scheduler_type: linear
|
60 |
- num_epochs: 2.0
|
61 |
|
62 |
-
### Training results
|
63 |
-
|
64 |
-
| Training Loss | Epoch | Step | Validation Loss | Cer |
|
65 |
-
|:-------------:|:-----:|:------:|:---------------:|:------:|
|
66 |
-
| 0.0834 | 0.01 | 1000 | 0.0603 | 0.9216 |
|
67 |
-
| 0.0566 | 0.02 | 2000 | 0.0467 | 0.9217 |
|
68 |
-
| 0.0534 | 0.03 | 3000 | 0.0436 | 0.9216 |
|
69 |
-
| 0.0461 | 0.04 | 4000 | 0.0392 | 0.9216 |
|
70 |
-
| 0.0416 | 0.05 | 5000 | 0.0354 | 0.9216 |
|
71 |
-
| 0.0433 | 0.06 | 6000 | 0.0336 | 0.9216 |
|
72 |
-
| 0.045 | 0.08 | 7000 | 0.0315 | 0.9216 |
|
73 |
-
| 0.0452 | 0.09 | 8000 | 0.0305 | 0.9216 |
|
74 |
-
| 0.04 | 0.1 | 9000 | 0.0281 | 0.9216 |
|
75 |
-
| 0.0307 | 0.11 | 10000 | 0.0273 | 0.9216 |
|
76 |
-
| 0.0382 | 0.12 | 11000 | 0.0269 | 0.9216 |
|
77 |
-
| 0.036 | 0.13 | 12000 | 0.0254 | 0.9216 |
|
78 |
-
| 0.0412 | 0.14 | 13000 | 0.0258 | 0.9216 |
|
79 |
-
| 0.0404 | 0.15 | 14000 | 0.0238 | 0.9216 |
|
80 |
-
| 0.0265 | 0.16 | 15000 | 0.0239 | 0.9216 |
|
81 |
-
| 0.029 | 0.17 | 16000 | 0.0235 | 0.9216 |
|
82 |
-
| 0.0295 | 0.18 | 17000 | 0.0218 | 0.9216 |
|
83 |
-
| 0.0262 | 0.19 | 18000 | 0.0214 | 0.9216 |
|
84 |
-
| 0.0274 | 0.21 | 19000 | 0.0222 | 0.9216 |
|
85 |
-
| 0.0317 | 0.22 | 20000 | 0.0204 | 0.9216 |
|
86 |
-
| 0.0248 | 0.23 | 21000 | 0.0204 | 0.9216 |
|
87 |
-
| 0.0258 | 0.24 | 22000 | 0.0195 | 0.9216 |
|
88 |
-
| 0.0247 | 0.25 | 23000 | 0.0188 | 0.9216 |
|
89 |
-
| 0.0285 | 0.26 | 24000 | 0.0191 | 0.9215 |
|
90 |
-
| 0.031 | 0.27 | 25000 | 0.0192 | 0.9216 |
|
91 |
-
| 0.0267 | 0.28 | 26000 | 0.0188 | 0.9216 |
|
92 |
-
| 0.0245 | 0.29 | 27000 | 0.0177 | 0.9216 |
|
93 |
-
| 0.0258 | 0.3 | 28000 | 0.0177 | 0.9216 |
|
94 |
-
| 0.0235 | 0.31 | 29000 | 0.0169 | 0.9216 |
|
95 |
-
| 0.0235 | 0.32 | 30000 | 0.0176 | 0.9216 |
|
96 |
-
| 0.0223 | 0.34 | 31000 | 0.0165 | 0.9216 |
|
97 |
-
| 0.0219 | 0.35 | 32000 | 0.0167 | 0.9216 |
|
98 |
-
| 0.0214 | 0.36 | 33000 | 0.0165 | 0.9216 |
|
99 |
-
| 0.0232 | 0.37 | 34000 | 0.0163 | 0.9216 |
|
100 |
-
| 0.0192 | 0.38 | 35000 | 0.0162 | 0.9216 |
|
101 |
-
| 0.0159 | 0.39 | 36000 | 0.0160 | 0.9216 |
|
102 |
-
| 0.0205 | 0.4 | 37000 | 0.0150 | 0.9216 |
|
103 |
-
| 0.0197 | 0.41 | 38000 | 0.0152 | 0.9216 |
|
104 |
-
| 0.0205 | 0.42 | 39000 | 0.0150 | 0.9216 |
|
105 |
-
| 0.0182 | 0.43 | 40000 | 0.0145 | 0.9216 |
|
106 |
-
| 0.0204 | 0.44 | 41000 | 0.0139 | 0.9216 |
|
107 |
-
| 0.0201 | 0.45 | 42000 | 0.0146 | 0.9216 |
|
108 |
-
| 0.0202 | 0.46 | 43000 | 0.0132 | 0.9215 |
|
109 |
-
| 0.0219 | 0.48 | 44000 | 0.0146 | 0.9216 |
|
110 |
-
| 0.0161 | 0.49 | 45000 | 0.0134 | 0.9215 |
|
111 |
-
| 0.0172 | 0.5 | 46000 | 0.0137 | 0.9216 |
|
112 |
-
| 0.0199 | 0.51 | 47000 | 0.0133 | 0.9215 |
|
113 |
-
| 0.0211 | 0.52 | 48000 | 0.0132 | 0.9215 |
|
114 |
-
| 0.0184 | 0.53 | 49000 | 0.0136 | 0.9216 |
|
115 |
-
| 0.0191 | 0.54 | 50000 | 0.0129 | 0.9216 |
|
116 |
-
| 0.017 | 0.55 | 51000 | 0.0127 | 0.9216 |
|
117 |
-
| 0.0188 | 0.56 | 52000 | 0.0127 | 0.9215 |
|
118 |
-
| 0.0157 | 0.57 | 53000 | 0.0128 | 0.9216 |
|
119 |
-
| 0.0158 | 0.58 | 54000 | 0.0127 | 0.9216 |
|
120 |
-
| 0.0145 | 0.59 | 55000 | 0.0117 | 0.9216 |
|
121 |
-
| 0.0148 | 0.61 | 56000 | 0.0123 | 0.9216 |
|
122 |
-
| 0.0153 | 0.62 | 57000 | 0.0115 | 0.9216 |
|
123 |
-
| 0.0182 | 0.63 | 58000 | 0.0115 | 0.9216 |
|
124 |
-
| 0.0178 | 0.64 | 59000 | 0.0112 | 0.9215 |
|
125 |
-
| 0.0187 | 0.65 | 60000 | 0.0113 | 0.9215 |
|
126 |
-
| 0.0174 | 0.66 | 61000 | 0.0119 | 0.9216 |
|
127 |
-
| 0.0135 | 0.67 | 62000 | 0.0115 | 0.9215 |
|
128 |
-
| 0.0167 | 0.68 | 63000 | 0.0112 | 0.9216 |
|
129 |
-
| 0.0163 | 0.69 | 64000 | 0.0111 | 0.9215 |
|
130 |
-
| 0.0128 | 0.7 | 65000 | 0.0110 | 0.9215 |
|
131 |
-
| 0.0178 | 0.71 | 66000 | 0.0113 | 0.9215 |
|
132 |
-
| 0.0142 | 0.72 | 67000 | 0.0110 | 0.9215 |
|
133 |
-
| 0.0143 | 0.74 | 68000 | 0.0110 | 0.9215 |
|
134 |
-
| 0.0168 | 0.75 | 69000 | 0.0106 | 0.9216 |
|
135 |
-
| 0.0136 | 0.76 | 70000 | 0.0107 | 0.9215 |
|
136 |
-
| 0.0141 | 0.77 | 71000 | 0.0104 | 0.9215 |
|
137 |
-
| 0.0217 | 0.78 | 72000 | 0.0115 | 0.9216 |
|
138 |
-
| 0.012 | 0.79 | 73000 | 0.0105 | 0.9215 |
|
139 |
-
| 0.0141 | 0.8 | 74000 | 0.0100 | 0.9215 |
|
140 |
-
| 0.0136 | 0.81 | 75000 | 0.0096 | 0.9215 |
|
141 |
-
| 0.0106 | 0.82 | 76000 | 0.0104 | 0.9216 |
|
142 |
-
| 0.0176 | 0.83 | 77000 | 0.0102 | 0.9216 |
|
143 |
-
| 0.0169 | 0.84 | 78000 | 0.0099 | 0.9215 |
|
144 |
-
| 0.0118 | 0.85 | 79000 | 0.0102 | 0.9215 |
|
145 |
-
| 0.0178 | 0.86 | 80000 | 0.0095 | 0.9215 |
|
146 |
-
| 0.0145 | 0.88 | 81000 | 0.0097 | 0.9216 |
|
147 |
-
| 0.0154 | 0.89 | 82000 | 0.0099 | 0.9215 |
|
148 |
-
| 0.0129 | 0.9 | 83000 | 0.0094 | 0.9215 |
|
149 |
-
| 0.0125 | 0.91 | 84000 | 0.0097 | 0.9215 |
|
150 |
-
| 0.0147 | 0.92 | 85000 | 0.0093 | 0.9215 |
|
151 |
-
| 0.0145 | 0.93 | 86000 | 0.0091 | 0.9215 |
|
152 |
-
| 0.0121 | 0.94 | 87000 | 0.0089 | 0.9215 |
|
153 |
-
| 0.0125 | 0.95 | 88000 | 0.0094 | 0.9215 |
|
154 |
-
| 0.0113 | 0.96 | 89000 | 0.0088 | 0.9216 |
|
155 |
-
| 0.0098 | 0.97 | 90000 | 0.0094 | 0.9216 |
|
156 |
-
| 0.0137 | 0.98 | 91000 | 0.0089 | 0.9215 |
|
157 |
-
| 0.0105 | 0.99 | 92000 | 0.0091 | 0.9215 |
|
158 |
-
| 0.01 | 1.01 | 93000 | 0.0090 | 0.9216 |
|
159 |
-
| 0.0103 | 1.02 | 94000 | 0.0087 | 0.9216 |
|
160 |
-
| 0.0103 | 1.03 | 95000 | 0.0091 | 0.9215 |
|
161 |
-
| 0.0107 | 1.04 | 96000 | 0.0088 | 0.9216 |
|
162 |
-
| 0.0109 | 1.05 | 97000 | 0.0087 | 0.9215 |
|
163 |
-
| 0.0102 | 1.06 | 98000 | 0.0090 | 0.9216 |
|
164 |
-
| 0.0109 | 1.07 | 99000 | 0.0087 | 0.9215 |
|
165 |
-
| 0.0094 | 1.08 | 100000 | 0.0084 | 0.9215 |
|
166 |
-
| 0.009 | 1.09 | 101000 | 0.0085 | 0.9215 |
|
167 |
-
| 0.0085 | 1.1 | 102000 | 0.0084 | 0.9216 |
|
168 |
-
| 0.0123 | 1.11 | 103000 | 0.0085 | 0.9215 |
|
169 |
-
| 0.0094 | 1.12 | 104000 | 0.0084 | 0.9215 |
|
170 |
-
| 0.0076 | 1.14 | 105000 | 0.0081 | 0.9215 |
|
171 |
-
| 0.0119 | 1.15 | 106000 | 0.0079 | 0.9216 |
|
172 |
-
| 0.0079 | 1.16 | 107000 | 0.0081 | 0.9216 |
|
173 |
-
| 0.0108 | 1.17 | 108000 | 0.0080 | 0.9216 |
|
174 |
-
| 0.01 | 1.18 | 109000 | 0.0077 | 0.9216 |
|
175 |
-
| 0.0112 | 1.19 | 110000 | 0.0077 | 0.9216 |
|
176 |
-
| 0.0092 | 1.2 | 111000 | 0.0076 | 0.9215 |
|
177 |
-
| 0.0097 | 1.21 | 112000 | 0.0077 | 0.9215 |
|
178 |
-
| 0.0093 | 1.22 | 113000 | 0.0078 | 0.9215 |
|
179 |
-
| 0.0106 | 1.23 | 114000 | 0.0077 | 0.9215 |
|
180 |
-
| 0.0107 | 1.24 | 115000 | 0.0076 | 0.9215 |
|
181 |
-
| 0.0111 | 1.25 | 116000 | 0.0077 | 0.9215 |
|
182 |
-
| 0.0118 | 1.26 | 117000 | 0.0076 | 0.9215 |
|
183 |
-
| 0.0088 | 1.28 | 118000 | 0.0076 | 0.9215 |
|
184 |
-
| 0.01 | 1.29 | 119000 | 0.0076 | 0.9215 |
|
185 |
-
| 0.0102 | 1.3 | 120000 | 0.0076 | 0.9215 |
|
186 |
-
| 0.0106 | 1.31 | 121000 | 0.0076 | 0.9215 |
|
187 |
-
| 0.0099 | 1.32 | 122000 | 0.0077 | 0.9215 |
|
188 |
-
| 0.0099 | 1.33 | 123000 | 0.0077 | 0.9216 |
|
189 |
-
| 0.0105 | 1.34 | 124000 | 0.0075 | 0.9216 |
|
190 |
-
| 0.0082 | 1.35 | 125000 | 0.0074 | 0.9216 |
|
191 |
-
| 0.0088 | 1.36 | 126000 | 0.0072 | 0.9215 |
|
192 |
-
| 0.0077 | 1.37 | 127000 | 0.0070 | 0.9215 |
|
193 |
-
| 0.0063 | 1.38 | 128000 | 0.0074 | 0.9216 |
|
194 |
-
| 0.0084 | 1.39 | 129000 | 0.0069 | 0.9215 |
|
195 |
-
| 0.0085 | 1.41 | 130000 | 0.0071 | 0.9215 |
|
196 |
-
| 0.0107 | 1.42 | 131000 | 0.0067 | 0.9215 |
|
197 |
-
| 0.0064 | 1.43 | 132000 | 0.0068 | 0.9215 |
|
198 |
-
| 0.0064 | 1.44 | 133000 | 0.0069 | 0.9215 |
|
199 |
-
| 0.0139 | 1.45 | 134000 | 0.0067 | 0.9216 |
|
200 |
-
| 0.0093 | 1.46 | 135000 | 0.0068 | 0.9216 |
|
201 |
-
| 0.009 | 1.47 | 136000 | 0.0067 | 0.9215 |
|
202 |
-
| 0.0083 | 1.48 | 137000 | 0.0065 | 0.9216 |
|
203 |
-
| 0.0108 | 1.49 | 138000 | 0.0064 | 0.9215 |
|
204 |
-
| 0.0074 | 1.5 | 139000 | 0.0066 | 0.9215 |
|
205 |
-
| 0.009 | 1.51 | 140000 | 0.0064 | 0.9216 |
|
206 |
-
| 0.0062 | 1.52 | 141000 | 0.0064 | 0.9215 |
|
207 |
-
| 0.007 | 1.54 | 142000 | 0.0063 | 0.9215 |
|
208 |
-
| 0.0082 | 1.55 | 143000 | 0.0062 | 0.9215 |
|
209 |
-
| 0.0077 | 1.56 | 144000 | 0.0064 | 0.9215 |
|
210 |
-
| 0.0094 | 1.57 | 145000 | 0.0062 | 0.9215 |
|
211 |
-
| 0.0085 | 1.58 | 146000 | 0.0063 | 0.9215 |
|
212 |
-
| 0.0091 | 1.59 | 147000 | 0.0062 | 0.9215 |
|
213 |
-
| 0.0087 | 1.6 | 148000 | 0.0061 | 0.9215 |
|
214 |
-
| 0.0066 | 1.61 | 149000 | 0.0062 | 0.9215 |
|
215 |
-
| 0.0087 | 1.62 | 150000 | 0.0061 | 0.9215 |
|
216 |
-
| 0.0059 | 1.63 | 151000 | 0.0059 | 0.9215 |
|
217 |
-
| 0.0086 | 1.64 | 152000 | 0.0059 | 0.9215 |
|
218 |
-
| 0.0066 | 1.65 | 153000 | 0.0059 | 0.9215 |
|
219 |
-
| 0.0076 | 1.66 | 154000 | 0.0058 | 0.9215 |
|
220 |
-
| 0.0073 | 1.68 | 155000 | 0.0060 | 0.9215 |
|
221 |
-
| 0.0118 | 1.69 | 156000 | 0.0060 | 0.9215 |
|
222 |
-
| 0.0058 | 1.7 | 157000 | 0.0059 | 0.9215 |
|
223 |
-
| 0.0093 | 1.71 | 158000 | 0.0058 | 0.9215 |
|
224 |
-
| 0.0079 | 1.72 | 159000 | 0.0058 | 0.9215 |
|
225 |
-
| 0.0063 | 1.73 | 160000 | 0.0059 | 0.9215 |
|
226 |
-
| 0.0065 | 1.74 | 161000 | 0.0056 | 0.9215 |
|
227 |
-
| 0.0105 | 1.75 | 162000 | 0.0057 | 0.9215 |
|
228 |
-
| 0.0075 | 1.76 | 163000 | 0.0055 | 0.9215 |
|
229 |
-
| 0.0069 | 1.77 | 164000 | 0.0056 | 0.9215 |
|
230 |
-
| 0.0075 | 1.78 | 165000 | 0.0056 | 0.9215 |
|
231 |
-
| 0.0067 | 1.79 | 166000 | 0.0055 | 0.9215 |
|
232 |
-
| 0.0069 | 1.81 | 167000 | 0.0056 | 0.9215 |
|
233 |
-
| 0.0063 | 1.82 | 168000 | 0.0056 | 0.9215 |
|
234 |
-
| 0.0058 | 1.83 | 169000 | 0.0055 | 0.9215 |
|
235 |
-
| 0.0058 | 1.84 | 170000 | 0.0054 | 0.9215 |
|
236 |
-
| 0.0081 | 1.85 | 171000 | 0.0055 | 0.9215 |
|
237 |
-
| 0.0071 | 1.86 | 172000 | 0.0054 | 0.9215 |
|
238 |
-
| 0.0077 | 1.87 | 173000 | 0.0054 | 0.9215 |
|
239 |
-
| 0.0053 | 1.88 | 174000 | 0.0053 | 0.9215 |
|
240 |
-
| 0.0067 | 1.89 | 175000 | 0.0053 | 0.9215 |
|
241 |
-
| 0.0066 | 1.9 | 176000 | 0.0053 | 0.9215 |
|
242 |
-
| 0.0084 | 1.91 | 177000 | 0.0053 | 0.9215 |
|
243 |
-
| 0.0066 | 1.92 | 178000 | 0.0052 | 0.9215 |
|
244 |
-
| 0.0057 | 1.94 | 179000 | 0.0053 | 0.9215 |
|
245 |
-
| 0.0059 | 1.95 | 180000 | 0.0052 | 0.9215 |
|
246 |
-
| 0.0053 | 1.96 | 181000 | 0.0053 | 0.9215 |
|
247 |
-
| 0.0056 | 1.97 | 182000 | 0.0052 | 0.9215 |
|
248 |
-
| 0.0054 | 1.98 | 183000 | 0.0052 | 0.9215 |
|
249 |
-
| 0.0053 | 1.99 | 184000 | 0.0052 | 0.9215 |
|
250 |
-
| 0.0066 | 2.0 | 185000 | 0.0052 | 0.9215 |
|
251 |
-
|
252 |
|
253 |
### Framework versions
|
254 |
|
|
|
7 |
results: []
|
8 |
---
|
9 |
|
10 |
+
# bart-base-spelling-nl
|
11 |
|
12 |
This model is a Dutch fine-tuned version of
|
13 |
[facebook/bart-base](https://huggingface.co/facebook/bart-base).
|
14 |
|
15 |
+
It achieves the following results on an external evaluation set:
|
16 |
+
|
17 |
+
CER - 0.025
|
18 |
+
WER - 0.090
|
19 |
+
BLEU - 0.837
|
20 |
+
METEOR - 0.932
|
21 |
|
|
|
|
|
22 |
|
23 |
## Model description
|
24 |
|
|
|
38 |
|
39 |
## Training and evaluation data
|
40 |
|
41 |
+
The model was trained on a Dutch dataset composed of 6,351,203 lines
|
42 |
of text from three public Dutch sources, downloaded from the
|
43 |
[Opus corpus](https://opus.nlpl.eu/):
|
44 |
|
45 |
+
- nl-europarlv7.txt (2,387,000 lines)
|
46 |
+
- nl-opensubtitles2016.3m.txt (3,000,000 lines)
|
47 |
- nl-wikipedia.txt (964,203 lines)
|
48 |
|
49 |
|
|
|
62 |
- lr_scheduler_type: linear
|
63 |
- num_epochs: 2.0
|
64 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
65 |
|
66 |
### Framework versions
|
67 |
|