athrvk commited on
Commit
0695d62
·
verified ·
1 Parent(s): e1ef46e

Upload folder using huggingface_hub

Browse files
charset.json CHANGED
@@ -69,6 +69,7 @@
69
  "ए",
70
  "ऑ",
71
  "ओ",
 
72
  "क",
73
  "ख",
74
  "ग",
 
69
  "ए",
70
  "ऑ",
71
  "ओ",
72
+ "औ",
73
  "क",
74
  "ख",
75
  "ग",
gold_metrics_json_float32.json CHANGED
@@ -1,27 +1,27 @@
1
  {
2
  "per_file": {
3
  "tests/gold.jsonl": {
4
- "examples": 184,
5
- "spans": 165,
6
- "precision": 0.8971428571428571,
7
- "recall": 0.9515151515151515,
8
- "f1": 0.9235294117647059,
9
- "value_acc": 0.9515151515151515,
10
  "categories": {
11
  "digits": {
12
- "spans": 51,
13
- "value_acc": 0.9607843137254902,
14
- "exact_match_rate": 0.9607843137254902
15
  },
16
  "words": {
17
- "spans": 114,
18
- "value_acc": 0.9473684210526315,
19
- "exact_match_rate": 0.9473684210526315
20
  },
21
  "prefix": {
22
- "spans": 50,
23
- "value_acc": 0.92,
24
- "exact_match_rate": 0.92
25
  },
26
  "range": {
27
  "spans": 14,
@@ -30,38 +30,38 @@
30
  },
31
  "currency": {
32
  "spans": 27,
33
- "value_acc": 0.9629629629629629,
34
- "exact_match_rate": 0.9629629629629629
35
  },
36
  "multi_unit": {
37
- "spans": 11,
38
- "value_acc": 1.0,
39
  "exact_match_rate": 1.0
40
  },
41
  "symbol_unit": {
42
- "spans": 17,
43
- "value_acc": 1.0,
44
- "exact_match_rate": 1.0
45
  },
46
  "long": {
47
- "spans": 8,
48
- "value_acc": 0.75,
49
- "exact_match_rate": 0.75
50
  }
51
  },
52
  "negatives": {
53
- "examples": 25,
54
- "false_positives": 3,
55
- "fp_rate": 0.12
56
  }
57
  },
58
  "tests/gold_deva.jsonl": {
59
- "examples": 169,
60
- "spans": 144,
61
- "precision": 0.9662162162162162,
62
- "recall": 0.9930555555555556,
63
- "f1": 0.9794520547945206,
64
- "value_acc": 0.9861111111111112,
65
  "categories": {
66
  "digits": {
67
  "spans": 37,
@@ -69,12 +69,12 @@
69
  "exact_match_rate": 1.0
70
  },
71
  "words": {
72
- "spans": 107,
73
- "value_acc": 0.9813084112149533,
74
- "exact_match_rate": 0.9906542056074766
75
  },
76
  "prefix": {
77
- "spans": 27,
78
  "value_acc": 1.0,
79
  "exact_match_rate": 1.0
80
  },
@@ -89,7 +89,7 @@
89
  "exact_match_rate": 1.0
90
  },
91
  "multi_unit": {
92
- "spans": 6,
93
  "value_acc": 1.0,
94
  "exact_match_rate": 1.0
95
  },
@@ -105,41 +105,41 @@
105
  }
106
  },
107
  "negatives": {
108
- "examples": 30,
109
- "false_positives": 1,
110
- "fp_rate": 0.03333333333333333
111
  }
112
  }
113
  },
114
  "combined": {
115
- "examples": 353,
116
- "spans": 309,
117
- "precision": 0.9287925696594427,
118
- "recall": 0.970873786407767,
119
- "f1": 0.9493670886075949,
120
- "value_acc": 0.9676375404530745,
121
  "miss_summary": {
122
- "missed_span": 0,
123
- "spurious_span": 13,
124
- "wrong_value": 1,
125
- "wrong_boundary": 9,
126
- "total": 23
127
  },
128
  "categories": {
129
  "digits": {
130
- "spans": 88,
131
- "value_acc": 0.9772727272727273,
132
- "exact_match_rate": 0.9772727272727273
133
  },
134
  "words": {
135
- "spans": 221,
136
- "value_acc": 0.9638009049773756,
137
- "exact_match_rate": 0.9683257918552036
138
  },
139
  "prefix": {
140
- "spans": 77,
141
- "value_acc": 0.948051948051948,
142
- "exact_match_rate": 0.948051948051948
143
  },
144
  "range": {
145
  "spans": 24,
@@ -148,18 +148,18 @@
148
  },
149
  "currency": {
150
  "spans": 47,
151
- "value_acc": 0.9787234042553191,
152
- "exact_match_rate": 0.9787234042553191
153
  },
154
  "multi_unit": {
155
- "spans": 17,
156
- "value_acc": 1.0,
157
  "exact_match_rate": 1.0
158
  },
159
  "symbol_unit": {
160
- "spans": 31,
161
- "value_acc": 1.0,
162
- "exact_match_rate": 1.0
163
  },
164
  "mixed_script": {
165
  "spans": 2,
@@ -167,15 +167,15 @@
167
  "exact_match_rate": 1.0
168
  },
169
  "long": {
170
- "spans": 8,
171
- "value_acc": 0.75,
172
- "exact_match_rate": 0.75
173
  }
174
  },
175
  "negatives": {
176
- "examples": 55,
177
- "false_positives": 4,
178
- "fp_rate": 0.07272727272727272
179
  }
180
  }
181
  }
 
1
  {
2
  "per_file": {
3
  "tests/gold.jsonl": {
4
+ "examples": 214,
5
+ "spans": 188,
6
+ "precision": 0.9166666666666666,
7
+ "recall": 0.9361702127659575,
8
+ "f1": 0.9263157894736843,
9
+ "value_acc": 0.925531914893617,
10
  "categories": {
11
  "digits": {
12
+ "spans": 54,
13
+ "value_acc": 0.9444444444444444,
14
+ "exact_match_rate": 0.9444444444444444
15
  },
16
  "words": {
17
+ "spans": 134,
18
+ "value_acc": 0.917910447761194,
19
+ "exact_match_rate": 0.9328358208955224
20
  },
21
  "prefix": {
22
+ "spans": 52,
23
+ "value_acc": 0.8846153846153846,
24
+ "exact_match_rate": 0.9038461538461539
25
  },
26
  "range": {
27
  "spans": 14,
 
30
  },
31
  "currency": {
32
  "spans": 27,
33
+ "value_acc": 1.0,
34
+ "exact_match_rate": 1.0
35
  },
36
  "multi_unit": {
37
+ "spans": 14,
38
+ "value_acc": 0.9285714285714286,
39
  "exact_match_rate": 1.0
40
  },
41
  "symbol_unit": {
42
+ "spans": 19,
43
+ "value_acc": 0.9473684210526315,
44
+ "exact_match_rate": 0.9473684210526315
45
  },
46
  "long": {
47
+ "spans": 10,
48
+ "value_acc": 0.8,
49
+ "exact_match_rate": 0.8
50
  }
51
  },
52
  "negatives": {
53
+ "examples": 35,
54
+ "false_positives": 0,
55
+ "fp_rate": 0.0
56
  }
57
  },
58
  "tests/gold_deva.jsonl": {
59
+ "examples": 182,
60
+ "spans": 151,
61
+ "precision": 0.9736842105263158,
62
+ "recall": 0.9801324503311258,
63
+ "f1": 0.976897689768977,
64
+ "value_acc": 0.9668874172185431,
65
  "categories": {
66
  "digits": {
67
  "spans": 37,
 
69
  "exact_match_rate": 1.0
70
  },
71
  "words": {
72
+ "spans": 114,
73
+ "value_acc": 0.956140350877193,
74
+ "exact_match_rate": 0.9736842105263158
75
  },
76
  "prefix": {
77
+ "spans": 29,
78
  "value_acc": 1.0,
79
  "exact_match_rate": 1.0
80
  },
 
89
  "exact_match_rate": 1.0
90
  },
91
  "multi_unit": {
92
+ "spans": 7,
93
  "value_acc": 1.0,
94
  "exact_match_rate": 1.0
95
  },
 
105
  }
106
  },
107
  "negatives": {
108
+ "examples": 38,
109
+ "false_positives": 0,
110
+ "fp_rate": 0.0
111
  }
112
  }
113
  },
114
  "combined": {
115
+ "examples": 396,
116
+ "spans": 339,
117
+ "precision": 0.9418604651162791,
118
+ "recall": 0.9557522123893806,
119
+ "f1": 0.9487554904831625,
120
+ "value_acc": 0.943952802359882,
121
  "miss_summary": {
122
+ "missed_span": 2,
123
+ "spurious_span": 7,
124
+ "wrong_value": 4,
125
+ "wrong_boundary": 13,
126
+ "total": 26
127
  },
128
  "categories": {
129
  "digits": {
130
+ "spans": 91,
131
+ "value_acc": 0.967032967032967,
132
+ "exact_match_rate": 0.967032967032967
133
  },
134
  "words": {
135
+ "spans": 248,
136
+ "value_acc": 0.9354838709677419,
137
+ "exact_match_rate": 0.9516129032258065
138
  },
139
  "prefix": {
140
+ "spans": 81,
141
+ "value_acc": 0.9259259259259259,
142
+ "exact_match_rate": 0.9382716049382716
143
  },
144
  "range": {
145
  "spans": 24,
 
148
  },
149
  "currency": {
150
  "spans": 47,
151
+ "value_acc": 1.0,
152
+ "exact_match_rate": 1.0
153
  },
154
  "multi_unit": {
155
+ "spans": 21,
156
+ "value_acc": 0.9523809523809523,
157
  "exact_match_rate": 1.0
158
  },
159
  "symbol_unit": {
160
+ "spans": 33,
161
+ "value_acc": 0.9696969696969697,
162
+ "exact_match_rate": 0.9696969696969697
163
  },
164
  "mixed_script": {
165
  "spans": 2,
 
167
  "exact_match_rate": 1.0
168
  },
169
  "long": {
170
+ "spans": 10,
171
+ "value_acc": 0.8,
172
+ "exact_match_rate": 0.8
173
  }
174
  },
175
  "negatives": {
176
+ "examples": 73,
177
+ "false_positives": 0,
178
+ "fp_rate": 0.0
179
  }
180
  }
181
  }
gold_metrics_json_int8.json CHANGED
@@ -1,27 +1,27 @@
1
  {
2
  "per_file": {
3
  "tests/gold.jsonl": {
4
- "examples": 184,
5
- "spans": 165,
6
- "precision": 0.8971428571428571,
7
- "recall": 0.9515151515151515,
8
- "f1": 0.9235294117647059,
9
- "value_acc": 0.9515151515151515,
10
  "categories": {
11
  "digits": {
12
- "spans": 51,
13
- "value_acc": 0.9607843137254902,
14
- "exact_match_rate": 0.9607843137254902
15
  },
16
  "words": {
17
- "spans": 114,
18
- "value_acc": 0.9473684210526315,
19
- "exact_match_rate": 0.9473684210526315
20
  },
21
  "prefix": {
22
- "spans": 50,
23
- "value_acc": 0.92,
24
- "exact_match_rate": 0.92
25
  },
26
  "range": {
27
  "spans": 14,
@@ -30,38 +30,38 @@
30
  },
31
  "currency": {
32
  "spans": 27,
33
- "value_acc": 0.9629629629629629,
34
- "exact_match_rate": 0.9629629629629629
35
  },
36
  "multi_unit": {
37
- "spans": 11,
38
- "value_acc": 1.0,
39
  "exact_match_rate": 1.0
40
  },
41
  "symbol_unit": {
42
- "spans": 17,
43
- "value_acc": 1.0,
44
- "exact_match_rate": 1.0
45
  },
46
  "long": {
47
- "spans": 8,
48
- "value_acc": 0.75,
49
- "exact_match_rate": 0.75
50
  }
51
  },
52
  "negatives": {
53
- "examples": 25,
54
- "false_positives": 3,
55
- "fp_rate": 0.12
56
  }
57
  },
58
  "tests/gold_deva.jsonl": {
59
- "examples": 169,
60
- "spans": 144,
61
- "precision": 0.9662162162162162,
62
- "recall": 0.9930555555555556,
63
- "f1": 0.9794520547945206,
64
- "value_acc": 0.9861111111111112,
65
  "categories": {
66
  "digits": {
67
  "spans": 37,
@@ -69,12 +69,12 @@
69
  "exact_match_rate": 1.0
70
  },
71
  "words": {
72
- "spans": 107,
73
- "value_acc": 0.9813084112149533,
74
- "exact_match_rate": 0.9906542056074766
75
  },
76
  "prefix": {
77
- "spans": 27,
78
  "value_acc": 1.0,
79
  "exact_match_rate": 1.0
80
  },
@@ -89,7 +89,7 @@
89
  "exact_match_rate": 1.0
90
  },
91
  "multi_unit": {
92
- "spans": 6,
93
  "value_acc": 1.0,
94
  "exact_match_rate": 1.0
95
  },
@@ -105,41 +105,41 @@
105
  }
106
  },
107
  "negatives": {
108
- "examples": 30,
109
- "false_positives": 1,
110
- "fp_rate": 0.03333333333333333
111
  }
112
  }
113
  },
114
  "combined": {
115
- "examples": 353,
116
- "spans": 309,
117
- "precision": 0.9287925696594427,
118
- "recall": 0.970873786407767,
119
- "f1": 0.9493670886075949,
120
- "value_acc": 0.9676375404530745,
121
  "miss_summary": {
122
- "missed_span": 0,
123
- "spurious_span": 13,
124
- "wrong_value": 1,
125
- "wrong_boundary": 9,
126
- "total": 23
127
  },
128
  "categories": {
129
  "digits": {
130
- "spans": 88,
131
- "value_acc": 0.9772727272727273,
132
- "exact_match_rate": 0.9772727272727273
133
  },
134
  "words": {
135
- "spans": 221,
136
- "value_acc": 0.9638009049773756,
137
- "exact_match_rate": 0.9683257918552036
138
  },
139
  "prefix": {
140
- "spans": 77,
141
- "value_acc": 0.948051948051948,
142
- "exact_match_rate": 0.948051948051948
143
  },
144
  "range": {
145
  "spans": 24,
@@ -148,18 +148,18 @@
148
  },
149
  "currency": {
150
  "spans": 47,
151
- "value_acc": 0.9787234042553191,
152
- "exact_match_rate": 0.9787234042553191
153
  },
154
  "multi_unit": {
155
- "spans": 17,
156
- "value_acc": 1.0,
157
  "exact_match_rate": 1.0
158
  },
159
  "symbol_unit": {
160
- "spans": 31,
161
- "value_acc": 1.0,
162
- "exact_match_rate": 1.0
163
  },
164
  "mixed_script": {
165
  "spans": 2,
@@ -167,15 +167,15 @@
167
  "exact_match_rate": 1.0
168
  },
169
  "long": {
170
- "spans": 8,
171
- "value_acc": 0.75,
172
- "exact_match_rate": 0.75
173
  }
174
  },
175
  "negatives": {
176
- "examples": 55,
177
- "false_positives": 4,
178
- "fp_rate": 0.07272727272727272
179
  }
180
  }
181
  }
 
1
  {
2
  "per_file": {
3
  "tests/gold.jsonl": {
4
+ "examples": 214,
5
+ "spans": 188,
6
+ "precision": 0.9166666666666666,
7
+ "recall": 0.9361702127659575,
8
+ "f1": 0.9263157894736843,
9
+ "value_acc": 0.925531914893617,
10
  "categories": {
11
  "digits": {
12
+ "spans": 54,
13
+ "value_acc": 0.9444444444444444,
14
+ "exact_match_rate": 0.9444444444444444
15
  },
16
  "words": {
17
+ "spans": 134,
18
+ "value_acc": 0.917910447761194,
19
+ "exact_match_rate": 0.9328358208955224
20
  },
21
  "prefix": {
22
+ "spans": 52,
23
+ "value_acc": 0.8846153846153846,
24
+ "exact_match_rate": 0.9038461538461539
25
  },
26
  "range": {
27
  "spans": 14,
 
30
  },
31
  "currency": {
32
  "spans": 27,
33
+ "value_acc": 1.0,
34
+ "exact_match_rate": 1.0
35
  },
36
  "multi_unit": {
37
+ "spans": 14,
38
+ "value_acc": 0.9285714285714286,
39
  "exact_match_rate": 1.0
40
  },
41
  "symbol_unit": {
42
+ "spans": 19,
43
+ "value_acc": 0.9473684210526315,
44
+ "exact_match_rate": 0.9473684210526315
45
  },
46
  "long": {
47
+ "spans": 10,
48
+ "value_acc": 0.8,
49
+ "exact_match_rate": 0.8
50
  }
51
  },
52
  "negatives": {
53
+ "examples": 35,
54
+ "false_positives": 0,
55
+ "fp_rate": 0.0
56
  }
57
  },
58
  "tests/gold_deva.jsonl": {
59
+ "examples": 182,
60
+ "spans": 151,
61
+ "precision": 0.9736842105263158,
62
+ "recall": 0.9801324503311258,
63
+ "f1": 0.976897689768977,
64
+ "value_acc": 0.9735099337748344,
65
  "categories": {
66
  "digits": {
67
  "spans": 37,
 
69
  "exact_match_rate": 1.0
70
  },
71
  "words": {
72
+ "spans": 114,
73
+ "value_acc": 0.9649122807017544,
74
+ "exact_match_rate": 0.9736842105263158
75
  },
76
  "prefix": {
77
+ "spans": 29,
78
  "value_acc": 1.0,
79
  "exact_match_rate": 1.0
80
  },
 
89
  "exact_match_rate": 1.0
90
  },
91
  "multi_unit": {
92
+ "spans": 7,
93
  "value_acc": 1.0,
94
  "exact_match_rate": 1.0
95
  },
 
105
  }
106
  },
107
  "negatives": {
108
+ "examples": 38,
109
+ "false_positives": 0,
110
+ "fp_rate": 0.0
111
  }
112
  }
113
  },
114
  "combined": {
115
+ "examples": 396,
116
+ "spans": 339,
117
+ "precision": 0.9418604651162791,
118
+ "recall": 0.9557522123893806,
119
+ "f1": 0.9487554904831625,
120
+ "value_acc": 0.9469026548672567,
121
  "miss_summary": {
122
+ "missed_span": 2,
123
+ "spurious_span": 7,
124
+ "wrong_value": 3,
125
+ "wrong_boundary": 13,
126
+ "total": 25
127
  },
128
  "categories": {
129
  "digits": {
130
+ "spans": 91,
131
+ "value_acc": 0.967032967032967,
132
+ "exact_match_rate": 0.967032967032967
133
  },
134
  "words": {
135
+ "spans": 248,
136
+ "value_acc": 0.9395161290322581,
137
+ "exact_match_rate": 0.9516129032258065
138
  },
139
  "prefix": {
140
+ "spans": 81,
141
+ "value_acc": 0.9259259259259259,
142
+ "exact_match_rate": 0.9382716049382716
143
  },
144
  "range": {
145
  "spans": 24,
 
148
  },
149
  "currency": {
150
  "spans": 47,
151
+ "value_acc": 1.0,
152
+ "exact_match_rate": 1.0
153
  },
154
  "multi_unit": {
155
+ "spans": 21,
156
+ "value_acc": 0.9523809523809523,
157
  "exact_match_rate": 1.0
158
  },
159
  "symbol_unit": {
160
+ "spans": 33,
161
+ "value_acc": 0.9696969696969697,
162
+ "exact_match_rate": 0.9696969696969697
163
  },
164
  "mixed_script": {
165
  "spans": 2,
 
167
  "exact_match_rate": 1.0
168
  },
169
  "long": {
170
+ "spans": 10,
171
+ "value_acc": 0.8,
172
+ "exact_match_rate": 0.8
173
  }
174
  },
175
  "negatives": {
176
+ "examples": 73,
177
+ "false_positives": 0,
178
+ "fp_rate": 0.0
179
  }
180
  }
181
  }
gold_metrics_torch.json CHANGED
@@ -1,27 +1,27 @@
1
  {
2
  "per_file": {
3
  "tests/gold.jsonl": {
4
- "examples": 184,
5
- "spans": 165,
6
- "precision": 0.8971428571428571,
7
- "recall": 0.9515151515151515,
8
- "f1": 0.9235294117647059,
9
- "value_acc": 0.9515151515151515,
10
  "categories": {
11
  "digits": {
12
- "spans": 51,
13
- "value_acc": 0.9607843137254902,
14
- "exact_match_rate": 0.9607843137254902
15
  },
16
  "words": {
17
- "spans": 114,
18
- "value_acc": 0.9473684210526315,
19
- "exact_match_rate": 0.9473684210526315
20
  },
21
  "prefix": {
22
- "spans": 50,
23
- "value_acc": 0.92,
24
- "exact_match_rate": 0.92
25
  },
26
  "range": {
27
  "spans": 14,
@@ -30,38 +30,38 @@
30
  },
31
  "currency": {
32
  "spans": 27,
33
- "value_acc": 0.9629629629629629,
34
- "exact_match_rate": 0.9629629629629629
35
  },
36
  "multi_unit": {
37
- "spans": 11,
38
- "value_acc": 1.0,
39
  "exact_match_rate": 1.0
40
  },
41
  "symbol_unit": {
42
- "spans": 17,
43
- "value_acc": 1.0,
44
- "exact_match_rate": 1.0
45
  },
46
  "long": {
47
- "spans": 8,
48
- "value_acc": 0.75,
49
- "exact_match_rate": 0.75
50
  }
51
  },
52
  "negatives": {
53
- "examples": 25,
54
- "false_positives": 3,
55
- "fp_rate": 0.12
56
  }
57
  },
58
  "tests/gold_deva.jsonl": {
59
- "examples": 169,
60
- "spans": 144,
61
- "precision": 0.9662162162162162,
62
- "recall": 0.9930555555555556,
63
- "f1": 0.9794520547945206,
64
- "value_acc": 0.9861111111111112,
65
  "categories": {
66
  "digits": {
67
  "spans": 37,
@@ -69,12 +69,12 @@
69
  "exact_match_rate": 1.0
70
  },
71
  "words": {
72
- "spans": 107,
73
- "value_acc": 0.9813084112149533,
74
- "exact_match_rate": 0.9906542056074766
75
  },
76
  "prefix": {
77
- "spans": 27,
78
  "value_acc": 1.0,
79
  "exact_match_rate": 1.0
80
  },
@@ -89,7 +89,7 @@
89
  "exact_match_rate": 1.0
90
  },
91
  "multi_unit": {
92
- "spans": 6,
93
  "value_acc": 1.0,
94
  "exact_match_rate": 1.0
95
  },
@@ -105,41 +105,41 @@
105
  }
106
  },
107
  "negatives": {
108
- "examples": 30,
109
- "false_positives": 1,
110
- "fp_rate": 0.03333333333333333
111
  }
112
  }
113
  },
114
  "combined": {
115
- "examples": 353,
116
- "spans": 309,
117
- "precision": 0.9287925696594427,
118
- "recall": 0.970873786407767,
119
- "f1": 0.9493670886075949,
120
- "value_acc": 0.9676375404530745,
121
  "miss_summary": {
122
- "missed_span": 0,
123
- "spurious_span": 13,
124
- "wrong_value": 1,
125
- "wrong_boundary": 9,
126
- "total": 23
127
  },
128
  "categories": {
129
  "digits": {
130
- "spans": 88,
131
- "value_acc": 0.9772727272727273,
132
- "exact_match_rate": 0.9772727272727273
133
  },
134
  "words": {
135
- "spans": 221,
136
- "value_acc": 0.9638009049773756,
137
- "exact_match_rate": 0.9683257918552036
138
  },
139
  "prefix": {
140
- "spans": 77,
141
- "value_acc": 0.948051948051948,
142
- "exact_match_rate": 0.948051948051948
143
  },
144
  "range": {
145
  "spans": 24,
@@ -148,18 +148,18 @@
148
  },
149
  "currency": {
150
  "spans": 47,
151
- "value_acc": 0.9787234042553191,
152
- "exact_match_rate": 0.9787234042553191
153
  },
154
  "multi_unit": {
155
- "spans": 17,
156
- "value_acc": 1.0,
157
  "exact_match_rate": 1.0
158
  },
159
  "symbol_unit": {
160
- "spans": 31,
161
- "value_acc": 1.0,
162
- "exact_match_rate": 1.0
163
  },
164
  "mixed_script": {
165
  "spans": 2,
@@ -167,15 +167,15 @@
167
  "exact_match_rate": 1.0
168
  },
169
  "long": {
170
- "spans": 8,
171
- "value_acc": 0.75,
172
- "exact_match_rate": 0.75
173
  }
174
  },
175
  "negatives": {
176
- "examples": 55,
177
- "false_positives": 4,
178
- "fp_rate": 0.07272727272727272
179
  }
180
  }
181
  }
 
1
  {
2
  "per_file": {
3
  "tests/gold.jsonl": {
4
+ "examples": 214,
5
+ "spans": 188,
6
+ "precision": 0.9166666666666666,
7
+ "recall": 0.9361702127659575,
8
+ "f1": 0.9263157894736843,
9
+ "value_acc": 0.925531914893617,
10
  "categories": {
11
  "digits": {
12
+ "spans": 54,
13
+ "value_acc": 0.9444444444444444,
14
+ "exact_match_rate": 0.9444444444444444
15
  },
16
  "words": {
17
+ "spans": 134,
18
+ "value_acc": 0.917910447761194,
19
+ "exact_match_rate": 0.9328358208955224
20
  },
21
  "prefix": {
22
+ "spans": 52,
23
+ "value_acc": 0.8846153846153846,
24
+ "exact_match_rate": 0.9038461538461539
25
  },
26
  "range": {
27
  "spans": 14,
 
30
  },
31
  "currency": {
32
  "spans": 27,
33
+ "value_acc": 1.0,
34
+ "exact_match_rate": 1.0
35
  },
36
  "multi_unit": {
37
+ "spans": 14,
38
+ "value_acc": 0.9285714285714286,
39
  "exact_match_rate": 1.0
40
  },
41
  "symbol_unit": {
42
+ "spans": 19,
43
+ "value_acc": 0.9473684210526315,
44
+ "exact_match_rate": 0.9473684210526315
45
  },
46
  "long": {
47
+ "spans": 10,
48
+ "value_acc": 0.8,
49
+ "exact_match_rate": 0.8
50
  }
51
  },
52
  "negatives": {
53
+ "examples": 35,
54
+ "false_positives": 0,
55
+ "fp_rate": 0.0
56
  }
57
  },
58
  "tests/gold_deva.jsonl": {
59
+ "examples": 182,
60
+ "spans": 151,
61
+ "precision": 0.9736842105263158,
62
+ "recall": 0.9801324503311258,
63
+ "f1": 0.976897689768977,
64
+ "value_acc": 0.9668874172185431,
65
  "categories": {
66
  "digits": {
67
  "spans": 37,
 
69
  "exact_match_rate": 1.0
70
  },
71
  "words": {
72
+ "spans": 114,
73
+ "value_acc": 0.956140350877193,
74
+ "exact_match_rate": 0.9736842105263158
75
  },
76
  "prefix": {
77
+ "spans": 29,
78
  "value_acc": 1.0,
79
  "exact_match_rate": 1.0
80
  },
 
89
  "exact_match_rate": 1.0
90
  },
91
  "multi_unit": {
92
+ "spans": 7,
93
  "value_acc": 1.0,
94
  "exact_match_rate": 1.0
95
  },
 
105
  }
106
  },
107
  "negatives": {
108
+ "examples": 38,
109
+ "false_positives": 0,
110
+ "fp_rate": 0.0
111
  }
112
  }
113
  },
114
  "combined": {
115
+ "examples": 396,
116
+ "spans": 339,
117
+ "precision": 0.9418604651162791,
118
+ "recall": 0.9557522123893806,
119
+ "f1": 0.9487554904831625,
120
+ "value_acc": 0.943952802359882,
121
  "miss_summary": {
122
+ "missed_span": 2,
123
+ "spurious_span": 7,
124
+ "wrong_value": 4,
125
+ "wrong_boundary": 13,
126
+ "total": 26
127
  },
128
  "categories": {
129
  "digits": {
130
+ "spans": 91,
131
+ "value_acc": 0.967032967032967,
132
+ "exact_match_rate": 0.967032967032967
133
  },
134
  "words": {
135
+ "spans": 248,
136
+ "value_acc": 0.9354838709677419,
137
+ "exact_match_rate": 0.9516129032258065
138
  },
139
  "prefix": {
140
+ "spans": 81,
141
+ "value_acc": 0.9259259259259259,
142
+ "exact_match_rate": 0.9382716049382716
143
  },
144
  "range": {
145
  "spans": 24,
 
148
  },
149
  "currency": {
150
  "spans": 47,
151
+ "value_acc": 1.0,
152
+ "exact_match_rate": 1.0
153
  },
154
  "multi_unit": {
155
+ "spans": 21,
156
+ "value_acc": 0.9523809523809523,
157
  "exact_match_rate": 1.0
158
  },
159
  "symbol_unit": {
160
+ "spans": 33,
161
+ "value_acc": 0.9696969696969697,
162
+ "exact_match_rate": 0.9696969696969697
163
  },
164
  "mixed_script": {
165
  "spans": 2,
 
167
  "exact_match_rate": 1.0
168
  },
169
  "long": {
170
+ "spans": 10,
171
+ "value_acc": 0.8,
172
+ "exact_match_rate": 0.8
173
  }
174
  },
175
  "negatives": {
176
+ "examples": 73,
177
+ "false_positives": 0,
178
+ "fp_rate": 0.0
179
  }
180
  }
181
  }
kaggle_metrics.json CHANGED
@@ -10,7 +10,7 @@
10
  "epochs": 20,
11
  "arch": "v2",
12
  "channels": 48,
13
- "seed": 0,
14
  "matrix": "v2:48:0,v2:48:1,v2:48:2",
15
  "smoke": false
16
  },
@@ -18,27 +18,27 @@
18
  "torch": {
19
  "per_file": {
20
  "tests/gold.jsonl": {
21
- "examples": 184,
22
- "spans": 165,
23
- "precision": 0.8971428571428571,
24
- "recall": 0.9515151515151515,
25
- "f1": 0.9235294117647059,
26
- "value_acc": 0.9515151515151515,
27
  "categories": {
28
  "digits": {
29
- "spans": 51,
30
- "value_acc": 0.9607843137254902,
31
- "exact_match_rate": 0.9607843137254902
32
  },
33
  "words": {
34
- "spans": 114,
35
- "value_acc": 0.9473684210526315,
36
- "exact_match_rate": 0.9473684210526315
37
  },
38
  "prefix": {
39
- "spans": 50,
40
- "value_acc": 0.92,
41
- "exact_match_rate": 0.92
42
  },
43
  "range": {
44
  "spans": 14,
@@ -47,38 +47,38 @@
47
  },
48
  "currency": {
49
  "spans": 27,
50
- "value_acc": 0.9629629629629629,
51
- "exact_match_rate": 0.9629629629629629
52
  },
53
  "multi_unit": {
54
- "spans": 11,
55
- "value_acc": 1.0,
56
  "exact_match_rate": 1.0
57
  },
58
  "symbol_unit": {
59
- "spans": 17,
60
- "value_acc": 1.0,
61
- "exact_match_rate": 1.0
62
  },
63
  "long": {
64
- "spans": 8,
65
- "value_acc": 0.75,
66
- "exact_match_rate": 0.75
67
  }
68
  },
69
  "negatives": {
70
- "examples": 25,
71
- "false_positives": 3,
72
- "fp_rate": 0.12
73
  }
74
  },
75
  "tests/gold_deva.jsonl": {
76
- "examples": 169,
77
- "spans": 144,
78
- "precision": 0.9662162162162162,
79
- "recall": 0.9930555555555556,
80
- "f1": 0.9794520547945206,
81
- "value_acc": 0.9861111111111112,
82
  "categories": {
83
  "digits": {
84
  "spans": 37,
@@ -86,12 +86,12 @@
86
  "exact_match_rate": 1.0
87
  },
88
  "words": {
89
- "spans": 107,
90
- "value_acc": 0.9813084112149533,
91
- "exact_match_rate": 0.9906542056074766
92
  },
93
  "prefix": {
94
- "spans": 27,
95
  "value_acc": 1.0,
96
  "exact_match_rate": 1.0
97
  },
@@ -106,7 +106,7 @@
106
  "exact_match_rate": 1.0
107
  },
108
  "multi_unit": {
109
- "spans": 6,
110
  "value_acc": 1.0,
111
  "exact_match_rate": 1.0
112
  },
@@ -122,41 +122,41 @@
122
  }
123
  },
124
  "negatives": {
125
- "examples": 30,
126
- "false_positives": 1,
127
- "fp_rate": 0.03333333333333333
128
  }
129
  }
130
  },
131
  "combined": {
132
- "examples": 353,
133
- "spans": 309,
134
- "precision": 0.9287925696594427,
135
- "recall": 0.970873786407767,
136
- "f1": 0.9493670886075949,
137
- "value_acc": 0.9676375404530745,
138
  "miss_summary": {
139
- "missed_span": 0,
140
- "spurious_span": 13,
141
- "wrong_value": 1,
142
- "wrong_boundary": 9,
143
- "total": 23
144
  },
145
  "categories": {
146
  "digits": {
147
- "spans": 88,
148
- "value_acc": 0.9772727272727273,
149
- "exact_match_rate": 0.9772727272727273
150
  },
151
  "words": {
152
- "spans": 221,
153
- "value_acc": 0.9638009049773756,
154
- "exact_match_rate": 0.9683257918552036
155
  },
156
  "prefix": {
157
- "spans": 77,
158
- "value_acc": 0.948051948051948,
159
- "exact_match_rate": 0.948051948051948
160
  },
161
  "range": {
162
  "spans": 24,
@@ -165,18 +165,18 @@
165
  },
166
  "currency": {
167
  "spans": 47,
168
- "value_acc": 0.9787234042553191,
169
- "exact_match_rate": 0.9787234042553191
170
  },
171
  "multi_unit": {
172
- "spans": 17,
173
- "value_acc": 1.0,
174
  "exact_match_rate": 1.0
175
  },
176
  "symbol_unit": {
177
- "spans": 31,
178
- "value_acc": 1.0,
179
- "exact_match_rate": 1.0
180
  },
181
  "mixed_script": {
182
  "spans": 2,
@@ -184,42 +184,42 @@
184
  "exact_match_rate": 1.0
185
  },
186
  "long": {
187
- "spans": 8,
188
- "value_acc": 0.75,
189
- "exact_match_rate": 0.75
190
  }
191
  },
192
  "negatives": {
193
- "examples": 55,
194
- "false_positives": 4,
195
- "fp_rate": 0.07272727272727272
196
  }
197
  }
198
  },
199
  "json_float32": {
200
  "per_file": {
201
  "tests/gold.jsonl": {
202
- "examples": 184,
203
- "spans": 165,
204
- "precision": 0.8971428571428571,
205
- "recall": 0.9515151515151515,
206
- "f1": 0.9235294117647059,
207
- "value_acc": 0.9515151515151515,
208
  "categories": {
209
  "digits": {
210
- "spans": 51,
211
- "value_acc": 0.9607843137254902,
212
- "exact_match_rate": 0.9607843137254902
213
  },
214
  "words": {
215
- "spans": 114,
216
- "value_acc": 0.9473684210526315,
217
- "exact_match_rate": 0.9473684210526315
218
  },
219
  "prefix": {
220
- "spans": 50,
221
- "value_acc": 0.92,
222
- "exact_match_rate": 0.92
223
  },
224
  "range": {
225
  "spans": 14,
@@ -228,38 +228,38 @@
228
  },
229
  "currency": {
230
  "spans": 27,
231
- "value_acc": 0.9629629629629629,
232
- "exact_match_rate": 0.9629629629629629
233
  },
234
  "multi_unit": {
235
- "spans": 11,
236
- "value_acc": 1.0,
237
  "exact_match_rate": 1.0
238
  },
239
  "symbol_unit": {
240
- "spans": 17,
241
- "value_acc": 1.0,
242
- "exact_match_rate": 1.0
243
  },
244
  "long": {
245
- "spans": 8,
246
- "value_acc": 0.75,
247
- "exact_match_rate": 0.75
248
  }
249
  },
250
  "negatives": {
251
- "examples": 25,
252
- "false_positives": 3,
253
- "fp_rate": 0.12
254
  }
255
  },
256
  "tests/gold_deva.jsonl": {
257
- "examples": 169,
258
- "spans": 144,
259
- "precision": 0.9662162162162162,
260
- "recall": 0.9930555555555556,
261
- "f1": 0.9794520547945206,
262
- "value_acc": 0.9861111111111112,
263
  "categories": {
264
  "digits": {
265
  "spans": 37,
@@ -267,12 +267,12 @@
267
  "exact_match_rate": 1.0
268
  },
269
  "words": {
270
- "spans": 107,
271
- "value_acc": 0.9813084112149533,
272
- "exact_match_rate": 0.9906542056074766
273
  },
274
  "prefix": {
275
- "spans": 27,
276
  "value_acc": 1.0,
277
  "exact_match_rate": 1.0
278
  },
@@ -287,7 +287,7 @@
287
  "exact_match_rate": 1.0
288
  },
289
  "multi_unit": {
290
- "spans": 6,
291
  "value_acc": 1.0,
292
  "exact_match_rate": 1.0
293
  },
@@ -303,41 +303,41 @@
303
  }
304
  },
305
  "negatives": {
306
- "examples": 30,
307
- "false_positives": 1,
308
- "fp_rate": 0.03333333333333333
309
  }
310
  }
311
  },
312
  "combined": {
313
- "examples": 353,
314
- "spans": 309,
315
- "precision": 0.9287925696594427,
316
- "recall": 0.970873786407767,
317
- "f1": 0.9493670886075949,
318
- "value_acc": 0.9676375404530745,
319
  "miss_summary": {
320
- "missed_span": 0,
321
- "spurious_span": 13,
322
- "wrong_value": 1,
323
- "wrong_boundary": 9,
324
- "total": 23
325
  },
326
  "categories": {
327
  "digits": {
328
- "spans": 88,
329
- "value_acc": 0.9772727272727273,
330
- "exact_match_rate": 0.9772727272727273
331
  },
332
  "words": {
333
- "spans": 221,
334
- "value_acc": 0.9638009049773756,
335
- "exact_match_rate": 0.9683257918552036
336
  },
337
  "prefix": {
338
- "spans": 77,
339
- "value_acc": 0.948051948051948,
340
- "exact_match_rate": 0.948051948051948
341
  },
342
  "range": {
343
  "spans": 24,
@@ -346,18 +346,18 @@
346
  },
347
  "currency": {
348
  "spans": 47,
349
- "value_acc": 0.9787234042553191,
350
- "exact_match_rate": 0.9787234042553191
351
  },
352
  "multi_unit": {
353
- "spans": 17,
354
- "value_acc": 1.0,
355
  "exact_match_rate": 1.0
356
  },
357
  "symbol_unit": {
358
- "spans": 31,
359
- "value_acc": 1.0,
360
- "exact_match_rate": 1.0
361
  },
362
  "mixed_script": {
363
  "spans": 2,
@@ -365,42 +365,42 @@
365
  "exact_match_rate": 1.0
366
  },
367
  "long": {
368
- "spans": 8,
369
- "value_acc": 0.75,
370
- "exact_match_rate": 0.75
371
  }
372
  },
373
  "negatives": {
374
- "examples": 55,
375
- "false_positives": 4,
376
- "fp_rate": 0.07272727272727272
377
  }
378
  }
379
  },
380
  "json_int8": {
381
  "per_file": {
382
  "tests/gold.jsonl": {
383
- "examples": 184,
384
- "spans": 165,
385
- "precision": 0.8971428571428571,
386
- "recall": 0.9515151515151515,
387
- "f1": 0.9235294117647059,
388
- "value_acc": 0.9515151515151515,
389
  "categories": {
390
  "digits": {
391
- "spans": 51,
392
- "value_acc": 0.9607843137254902,
393
- "exact_match_rate": 0.9607843137254902
394
  },
395
  "words": {
396
- "spans": 114,
397
- "value_acc": 0.9473684210526315,
398
- "exact_match_rate": 0.9473684210526315
399
  },
400
  "prefix": {
401
- "spans": 50,
402
- "value_acc": 0.92,
403
- "exact_match_rate": 0.92
404
  },
405
  "range": {
406
  "spans": 14,
@@ -409,38 +409,38 @@
409
  },
410
  "currency": {
411
  "spans": 27,
412
- "value_acc": 0.9629629629629629,
413
- "exact_match_rate": 0.9629629629629629
414
  },
415
  "multi_unit": {
416
- "spans": 11,
417
- "value_acc": 1.0,
418
  "exact_match_rate": 1.0
419
  },
420
  "symbol_unit": {
421
- "spans": 17,
422
- "value_acc": 1.0,
423
- "exact_match_rate": 1.0
424
  },
425
  "long": {
426
- "spans": 8,
427
- "value_acc": 0.75,
428
- "exact_match_rate": 0.75
429
  }
430
  },
431
  "negatives": {
432
- "examples": 25,
433
- "false_positives": 3,
434
- "fp_rate": 0.12
435
  }
436
  },
437
  "tests/gold_deva.jsonl": {
438
- "examples": 169,
439
- "spans": 144,
440
- "precision": 0.9662162162162162,
441
- "recall": 0.9930555555555556,
442
- "f1": 0.9794520547945206,
443
- "value_acc": 0.9861111111111112,
444
  "categories": {
445
  "digits": {
446
  "spans": 37,
@@ -448,12 +448,12 @@
448
  "exact_match_rate": 1.0
449
  },
450
  "words": {
451
- "spans": 107,
452
- "value_acc": 0.9813084112149533,
453
- "exact_match_rate": 0.9906542056074766
454
  },
455
  "prefix": {
456
- "spans": 27,
457
  "value_acc": 1.0,
458
  "exact_match_rate": 1.0
459
  },
@@ -468,7 +468,7 @@
468
  "exact_match_rate": 1.0
469
  },
470
  "multi_unit": {
471
- "spans": 6,
472
  "value_acc": 1.0,
473
  "exact_match_rate": 1.0
474
  },
@@ -484,41 +484,41 @@
484
  }
485
  },
486
  "negatives": {
487
- "examples": 30,
488
- "false_positives": 1,
489
- "fp_rate": 0.03333333333333333
490
  }
491
  }
492
  },
493
  "combined": {
494
- "examples": 353,
495
- "spans": 309,
496
- "precision": 0.9287925696594427,
497
- "recall": 0.970873786407767,
498
- "f1": 0.9493670886075949,
499
- "value_acc": 0.9676375404530745,
500
  "miss_summary": {
501
- "missed_span": 0,
502
- "spurious_span": 13,
503
- "wrong_value": 1,
504
- "wrong_boundary": 9,
505
- "total": 23
506
  },
507
  "categories": {
508
  "digits": {
509
- "spans": 88,
510
- "value_acc": 0.9772727272727273,
511
- "exact_match_rate": 0.9772727272727273
512
  },
513
  "words": {
514
- "spans": 221,
515
- "value_acc": 0.9638009049773756,
516
- "exact_match_rate": 0.9683257918552036
517
  },
518
  "prefix": {
519
- "spans": 77,
520
- "value_acc": 0.948051948051948,
521
- "exact_match_rate": 0.948051948051948
522
  },
523
  "range": {
524
  "spans": 24,
@@ -527,18 +527,18 @@
527
  },
528
  "currency": {
529
  "spans": 47,
530
- "value_acc": 0.9787234042553191,
531
- "exact_match_rate": 0.9787234042553191
532
  },
533
  "multi_unit": {
534
- "spans": 17,
535
- "value_acc": 1.0,
536
  "exact_match_rate": 1.0
537
  },
538
  "symbol_unit": {
539
- "spans": 31,
540
- "value_acc": 1.0,
541
- "exact_match_rate": 1.0
542
  },
543
  "mixed_script": {
544
  "spans": 2,
@@ -546,15 +546,15 @@
546
  "exact_match_rate": 1.0
547
  },
548
  "long": {
549
- "spans": 8,
550
- "value_acc": 0.75,
551
- "exact_match_rate": 0.75
552
  }
553
  },
554
  "negatives": {
555
- "examples": 55,
556
- "false_positives": 4,
557
- "fp_rate": 0.07272727272727272
558
  }
559
  }
560
  }
 
10
  "epochs": 20,
11
  "arch": "v2",
12
  "channels": 48,
13
+ "seed": 2,
14
  "matrix": "v2:48:0,v2:48:1,v2:48:2",
15
  "smoke": false
16
  },
 
18
  "torch": {
19
  "per_file": {
20
  "tests/gold.jsonl": {
21
+ "examples": 214,
22
+ "spans": 188,
23
+ "precision": 0.9166666666666666,
24
+ "recall": 0.9361702127659575,
25
+ "f1": 0.9263157894736843,
26
+ "value_acc": 0.925531914893617,
27
  "categories": {
28
  "digits": {
29
+ "spans": 54,
30
+ "value_acc": 0.9444444444444444,
31
+ "exact_match_rate": 0.9444444444444444
32
  },
33
  "words": {
34
+ "spans": 134,
35
+ "value_acc": 0.917910447761194,
36
+ "exact_match_rate": 0.9328358208955224
37
  },
38
  "prefix": {
39
+ "spans": 52,
40
+ "value_acc": 0.8846153846153846,
41
+ "exact_match_rate": 0.9038461538461539
42
  },
43
  "range": {
44
  "spans": 14,
 
47
  },
48
  "currency": {
49
  "spans": 27,
50
+ "value_acc": 1.0,
51
+ "exact_match_rate": 1.0
52
  },
53
  "multi_unit": {
54
+ "spans": 14,
55
+ "value_acc": 0.9285714285714286,
56
  "exact_match_rate": 1.0
57
  },
58
  "symbol_unit": {
59
+ "spans": 19,
60
+ "value_acc": 0.9473684210526315,
61
+ "exact_match_rate": 0.9473684210526315
62
  },
63
  "long": {
64
+ "spans": 10,
65
+ "value_acc": 0.8,
66
+ "exact_match_rate": 0.8
67
  }
68
  },
69
  "negatives": {
70
+ "examples": 35,
71
+ "false_positives": 0,
72
+ "fp_rate": 0.0
73
  }
74
  },
75
  "tests/gold_deva.jsonl": {
76
+ "examples": 182,
77
+ "spans": 151,
78
+ "precision": 0.9736842105263158,
79
+ "recall": 0.9801324503311258,
80
+ "f1": 0.976897689768977,
81
+ "value_acc": 0.9668874172185431,
82
  "categories": {
83
  "digits": {
84
  "spans": 37,
 
86
  "exact_match_rate": 1.0
87
  },
88
  "words": {
89
+ "spans": 114,
90
+ "value_acc": 0.956140350877193,
91
+ "exact_match_rate": 0.9736842105263158
92
  },
93
  "prefix": {
94
+ "spans": 29,
95
  "value_acc": 1.0,
96
  "exact_match_rate": 1.0
97
  },
 
106
  "exact_match_rate": 1.0
107
  },
108
  "multi_unit": {
109
+ "spans": 7,
110
  "value_acc": 1.0,
111
  "exact_match_rate": 1.0
112
  },
 
122
  }
123
  },
124
  "negatives": {
125
+ "examples": 38,
126
+ "false_positives": 0,
127
+ "fp_rate": 0.0
128
  }
129
  }
130
  },
131
  "combined": {
132
+ "examples": 396,
133
+ "spans": 339,
134
+ "precision": 0.9418604651162791,
135
+ "recall": 0.9557522123893806,
136
+ "f1": 0.9487554904831625,
137
+ "value_acc": 0.943952802359882,
138
  "miss_summary": {
139
+ "missed_span": 2,
140
+ "spurious_span": 7,
141
+ "wrong_value": 4,
142
+ "wrong_boundary": 13,
143
+ "total": 26
144
  },
145
  "categories": {
146
  "digits": {
147
+ "spans": 91,
148
+ "value_acc": 0.967032967032967,
149
+ "exact_match_rate": 0.967032967032967
150
  },
151
  "words": {
152
+ "spans": 248,
153
+ "value_acc": 0.9354838709677419,
154
+ "exact_match_rate": 0.9516129032258065
155
  },
156
  "prefix": {
157
+ "spans": 81,
158
+ "value_acc": 0.9259259259259259,
159
+ "exact_match_rate": 0.9382716049382716
160
  },
161
  "range": {
162
  "spans": 24,
 
165
  },
166
  "currency": {
167
  "spans": 47,
168
+ "value_acc": 1.0,
169
+ "exact_match_rate": 1.0
170
  },
171
  "multi_unit": {
172
+ "spans": 21,
173
+ "value_acc": 0.9523809523809523,
174
  "exact_match_rate": 1.0
175
  },
176
  "symbol_unit": {
177
+ "spans": 33,
178
+ "value_acc": 0.9696969696969697,
179
+ "exact_match_rate": 0.9696969696969697
180
  },
181
  "mixed_script": {
182
  "spans": 2,
 
184
  "exact_match_rate": 1.0
185
  },
186
  "long": {
187
+ "spans": 10,
188
+ "value_acc": 0.8,
189
+ "exact_match_rate": 0.8
190
  }
191
  },
192
  "negatives": {
193
+ "examples": 73,
194
+ "false_positives": 0,
195
+ "fp_rate": 0.0
196
  }
197
  }
198
  },
199
  "json_float32": {
200
  "per_file": {
201
  "tests/gold.jsonl": {
202
+ "examples": 214,
203
+ "spans": 188,
204
+ "precision": 0.9166666666666666,
205
+ "recall": 0.9361702127659575,
206
+ "f1": 0.9263157894736843,
207
+ "value_acc": 0.925531914893617,
208
  "categories": {
209
  "digits": {
210
+ "spans": 54,
211
+ "value_acc": 0.9444444444444444,
212
+ "exact_match_rate": 0.9444444444444444
213
  },
214
  "words": {
215
+ "spans": 134,
216
+ "value_acc": 0.917910447761194,
217
+ "exact_match_rate": 0.9328358208955224
218
  },
219
  "prefix": {
220
+ "spans": 52,
221
+ "value_acc": 0.8846153846153846,
222
+ "exact_match_rate": 0.9038461538461539
223
  },
224
  "range": {
225
  "spans": 14,
 
228
  },
229
  "currency": {
230
  "spans": 27,
231
+ "value_acc": 1.0,
232
+ "exact_match_rate": 1.0
233
  },
234
  "multi_unit": {
235
+ "spans": 14,
236
+ "value_acc": 0.9285714285714286,
237
  "exact_match_rate": 1.0
238
  },
239
  "symbol_unit": {
240
+ "spans": 19,
241
+ "value_acc": 0.9473684210526315,
242
+ "exact_match_rate": 0.9473684210526315
243
  },
244
  "long": {
245
+ "spans": 10,
246
+ "value_acc": 0.8,
247
+ "exact_match_rate": 0.8
248
  }
249
  },
250
  "negatives": {
251
+ "examples": 35,
252
+ "false_positives": 0,
253
+ "fp_rate": 0.0
254
  }
255
  },
256
  "tests/gold_deva.jsonl": {
257
+ "examples": 182,
258
+ "spans": 151,
259
+ "precision": 0.9736842105263158,
260
+ "recall": 0.9801324503311258,
261
+ "f1": 0.976897689768977,
262
+ "value_acc": 0.9668874172185431,
263
  "categories": {
264
  "digits": {
265
  "spans": 37,
 
267
  "exact_match_rate": 1.0
268
  },
269
  "words": {
270
+ "spans": 114,
271
+ "value_acc": 0.956140350877193,
272
+ "exact_match_rate": 0.9736842105263158
273
  },
274
  "prefix": {
275
+ "spans": 29,
276
  "value_acc": 1.0,
277
  "exact_match_rate": 1.0
278
  },
 
287
  "exact_match_rate": 1.0
288
  },
289
  "multi_unit": {
290
+ "spans": 7,
291
  "value_acc": 1.0,
292
  "exact_match_rate": 1.0
293
  },
 
303
  }
304
  },
305
  "negatives": {
306
+ "examples": 38,
307
+ "false_positives": 0,
308
+ "fp_rate": 0.0
309
  }
310
  }
311
  },
312
  "combined": {
313
+ "examples": 396,
314
+ "spans": 339,
315
+ "precision": 0.9418604651162791,
316
+ "recall": 0.9557522123893806,
317
+ "f1": 0.9487554904831625,
318
+ "value_acc": 0.943952802359882,
319
  "miss_summary": {
320
+ "missed_span": 2,
321
+ "spurious_span": 7,
322
+ "wrong_value": 4,
323
+ "wrong_boundary": 13,
324
+ "total": 26
325
  },
326
  "categories": {
327
  "digits": {
328
+ "spans": 91,
329
+ "value_acc": 0.967032967032967,
330
+ "exact_match_rate": 0.967032967032967
331
  },
332
  "words": {
333
+ "spans": 248,
334
+ "value_acc": 0.9354838709677419,
335
+ "exact_match_rate": 0.9516129032258065
336
  },
337
  "prefix": {
338
+ "spans": 81,
339
+ "value_acc": 0.9259259259259259,
340
+ "exact_match_rate": 0.9382716049382716
341
  },
342
  "range": {
343
  "spans": 24,
 
346
  },
347
  "currency": {
348
  "spans": 47,
349
+ "value_acc": 1.0,
350
+ "exact_match_rate": 1.0
351
  },
352
  "multi_unit": {
353
+ "spans": 21,
354
+ "value_acc": 0.9523809523809523,
355
  "exact_match_rate": 1.0
356
  },
357
  "symbol_unit": {
358
+ "spans": 33,
359
+ "value_acc": 0.9696969696969697,
360
+ "exact_match_rate": 0.9696969696969697
361
  },
362
  "mixed_script": {
363
  "spans": 2,
 
365
  "exact_match_rate": 1.0
366
  },
367
  "long": {
368
+ "spans": 10,
369
+ "value_acc": 0.8,
370
+ "exact_match_rate": 0.8
371
  }
372
  },
373
  "negatives": {
374
+ "examples": 73,
375
+ "false_positives": 0,
376
+ "fp_rate": 0.0
377
  }
378
  }
379
  },
380
  "json_int8": {
381
  "per_file": {
382
  "tests/gold.jsonl": {
383
+ "examples": 214,
384
+ "spans": 188,
385
+ "precision": 0.9166666666666666,
386
+ "recall": 0.9361702127659575,
387
+ "f1": 0.9263157894736843,
388
+ "value_acc": 0.925531914893617,
389
  "categories": {
390
  "digits": {
391
+ "spans": 54,
392
+ "value_acc": 0.9444444444444444,
393
+ "exact_match_rate": 0.9444444444444444
394
  },
395
  "words": {
396
+ "spans": 134,
397
+ "value_acc": 0.917910447761194,
398
+ "exact_match_rate": 0.9328358208955224
399
  },
400
  "prefix": {
401
+ "spans": 52,
402
+ "value_acc": 0.8846153846153846,
403
+ "exact_match_rate": 0.9038461538461539
404
  },
405
  "range": {
406
  "spans": 14,
 
409
  },
410
  "currency": {
411
  "spans": 27,
412
+ "value_acc": 1.0,
413
+ "exact_match_rate": 1.0
414
  },
415
  "multi_unit": {
416
+ "spans": 14,
417
+ "value_acc": 0.9285714285714286,
418
  "exact_match_rate": 1.0
419
  },
420
  "symbol_unit": {
421
+ "spans": 19,
422
+ "value_acc": 0.9473684210526315,
423
+ "exact_match_rate": 0.9473684210526315
424
  },
425
  "long": {
426
+ "spans": 10,
427
+ "value_acc": 0.8,
428
+ "exact_match_rate": 0.8
429
  }
430
  },
431
  "negatives": {
432
+ "examples": 35,
433
+ "false_positives": 0,
434
+ "fp_rate": 0.0
435
  }
436
  },
437
  "tests/gold_deva.jsonl": {
438
+ "examples": 182,
439
+ "spans": 151,
440
+ "precision": 0.9736842105263158,
441
+ "recall": 0.9801324503311258,
442
+ "f1": 0.976897689768977,
443
+ "value_acc": 0.9735099337748344,
444
  "categories": {
445
  "digits": {
446
  "spans": 37,
 
448
  "exact_match_rate": 1.0
449
  },
450
  "words": {
451
+ "spans": 114,
452
+ "value_acc": 0.9649122807017544,
453
+ "exact_match_rate": 0.9736842105263158
454
  },
455
  "prefix": {
456
+ "spans": 29,
457
  "value_acc": 1.0,
458
  "exact_match_rate": 1.0
459
  },
 
468
  "exact_match_rate": 1.0
469
  },
470
  "multi_unit": {
471
+ "spans": 7,
472
  "value_acc": 1.0,
473
  "exact_match_rate": 1.0
474
  },
 
484
  }
485
  },
486
  "negatives": {
487
+ "examples": 38,
488
+ "false_positives": 0,
489
+ "fp_rate": 0.0
490
  }
491
  }
492
  },
493
  "combined": {
494
+ "examples": 396,
495
+ "spans": 339,
496
+ "precision": 0.9418604651162791,
497
+ "recall": 0.9557522123893806,
498
+ "f1": 0.9487554904831625,
499
+ "value_acc": 0.9469026548672567,
500
  "miss_summary": {
501
+ "missed_span": 2,
502
+ "spurious_span": 7,
503
+ "wrong_value": 3,
504
+ "wrong_boundary": 13,
505
+ "total": 25
506
  },
507
  "categories": {
508
  "digits": {
509
+ "spans": 91,
510
+ "value_acc": 0.967032967032967,
511
+ "exact_match_rate": 0.967032967032967
512
  },
513
  "words": {
514
+ "spans": 248,
515
+ "value_acc": 0.9395161290322581,
516
+ "exact_match_rate": 0.9516129032258065
517
  },
518
  "prefix": {
519
+ "spans": 81,
520
+ "value_acc": 0.9259259259259259,
521
+ "exact_match_rate": 0.9382716049382716
522
  },
523
  "range": {
524
  "spans": 24,
 
527
  },
528
  "currency": {
529
  "spans": 47,
530
+ "value_acc": 1.0,
531
+ "exact_match_rate": 1.0
532
  },
533
  "multi_unit": {
534
+ "spans": 21,
535
+ "value_acc": 0.9523809523809523,
536
  "exact_match_rate": 1.0
537
  },
538
  "symbol_unit": {
539
+ "spans": 33,
540
+ "value_acc": 0.9696969696969697,
541
+ "exact_match_rate": 0.9696969696969697
542
  },
543
  "mixed_script": {
544
  "spans": 2,
 
546
  "exact_match_rate": 1.0
547
  },
548
  "long": {
549
+ "spans": 10,
550
+ "value_acc": 0.8,
551
+ "exact_match_rate": 0.8
552
  }
553
  },
554
  "negatives": {
555
+ "examples": 73,
556
+ "false_positives": 0,
557
+ "fp_rate": 0.0
558
  }
559
  }
560
  }
matrix.md CHANGED
@@ -1,7 +1,7 @@
1
  | arch | channels | seed | val_value_acc | gold_torch_value_acc | gold_json_float32_value_acc | gold_int8_value_acc | gold_int8_f1 | winner |
2
  |---|---|---|---|---|---|---|---|---|
3
- | v2 | 48 | 0 | 0.9316 | 0.9676 | 0.9676 | 0.9676 | 0.9494 | <-- winner |
4
- | v2 | 48 | 1 | 0.9327 | 0.9417 | 0.9417 | 0.9482 | 0.9342 | |
5
- | v2 | 48 | 2 | 0.9338 | 0.9547 | 0.9547 | 0.9515 | 0.9354 | |
6
 
7
  winning config: **v2:48**
 
1
  | arch | channels | seed | val_value_acc | gold_torch_value_acc | gold_json_float32_value_acc | gold_int8_value_acc | gold_int8_f1 | winner |
2
  |---|---|---|---|---|---|---|---|---|
3
+ | v2 | 48 | 0 | 0.9252 | 0.9528 | 0.9528 | 0.9499 | 0.9431 | |
4
+ | v2 | 48 | 1 | 0.9262 | 0.9469 | 0.9469 | 0.9499 | 0.9475 | |
5
+ | v2 | 48 | 2 | 0.9252 | 0.9440 | 0.9440 | 0.9469 | 0.9488 | <-- winner |
6
 
7
  winning config: **v2:48**
sankhya.onnx CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a5f9289d74641398828bb412f0fca5a34fa966794cf0568fc5404368a9311bbe
3
- size 160960
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:082d373fc9b8324d4ab9b0634076bb614c959536290a152bc17f0c05bb9dbdba
3
+ size 161024
sankhya.pt CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:6b4d8108be14780da2a17fd2d0b36537ff76422158a8570f5d4652c3185010a7
3
- size 168246
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:75156b533c2e5baf13d0bd6037a7e106dffd0a3add0cc5a9ee9fb4d09baebfa1
3
+ size 168310
sankhya.weights.int8.json CHANGED
The diff for this file is too large to render. See raw diff
 
sankhya.weights.json CHANGED
The diff for this file is too large to render. See raw diff