Hugging Face's logo

drond0174
/
hallucination_detection

Token Classification
Joblib
Safetensors
hallucination-detection
tool-use
Model card Files
xet
 Community
drond0174 commited on
Commit
ee3a8e1
·
verified ·
1 Parent(s): dbb0190

Add tool-aware DeBERTa mixed checkpoint

Browse files
Files changed (8) hide show
  1. deberta_mixed/added_tokens.json +3 -0
  2. deberta_mixed/config.json +43 -0
  3. deberta_mixed/model.safetensors +3 -0
  4. deberta_mixed/run_meta.json +162 -0
  5. deberta_mixed/special_tokens_map.json +15 -0
  6. deberta_mixed/spm.model +3 -0
  7. deberta_mixed/tokenizer.json +0 -0
  8. deberta_mixed/tokenizer_config.json +59 -0
deberta_mixed/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[MASK]": 128000
3
+ }
deberta_mixed/config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "DebertaV2ForTokenClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "dtype": "float32",
7
+ "hidden_act": "gelu",
8
+ "hidden_dropout_prob": 0.1,
9
+ "hidden_size": 768,
10
+ "id2label": {
11
+ "0": "SUPPORTED",
12
+ "1": "HALLUCINATED"
13
+ },
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 3072,
16
+ "label2id": {
17
+ "HALLUCINATED": 1,
18
+ "SUPPORTED": 0
19
+ },
20
+ "layer_norm_eps": 1e-07,
21
+ "legacy": true,
22
+ "max_position_embeddings": 512,
23
+ "max_relative_positions": -1,
24
+ "model_type": "deberta-v2",
25
+ "norm_rel_ebd": "layer_norm",
26
+ "num_attention_heads": 12,
27
+ "num_hidden_layers": 6,
28
+ "pad_token_id": 0,
29
+ "pooler_dropout": 0,
30
+ "pooler_hidden_act": "gelu",
31
+ "pooler_hidden_size": 768,
32
+ "pos_att_type": [
33
+ "p2c",
34
+ "c2p"
35
+ ],
36
+ "position_biased_input": false,
37
+ "position_buckets": 256,
38
+ "relative_attention": true,
39
+ "share_att_key": true,
40
+ "transformers_version": "4.57.3",
41
+ "type_vocab_size": 0,
42
+ "vocab_size": 128100
43
+ }
deberta_mixed/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f59d7ddc11c87d1de8d1ad71e28ca57a7c6861e6d067847c8963fdb5defd7630
3
+ size 565236000
deberta_mixed/run_meta.json ADDED
@@ -0,0 +1,162 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model_name": "microsoft/deberta-v3-small",
3
+ "best_epoch": 2,
4
+ "best_val_loss": 0.14815934832964106,
5
+ "best_threshold": 0.95,
6
+ "pos_weight": 15.933958741715673,
7
+ "training_seconds": 95.76463508605957,
8
+ "n_train": 3617,
9
+ "n_validation": 514,
10
+ "train_log": [
11
+ {
12
+ "epoch": 1,
13
+ "train_loss": 0.246993733105862,
14
+ "val_loss": 0.16851629939267787
15
+ },
16
+ {
17
+ "epoch": 2,
18
+ "train_loss": 0.10282069164992681,
19
+ "val_loss": 0.14815934832964106
20
+ },
21
+ {
22
+ "epoch": 3,
23
+ "train_loss": 0.06773814947488435,
24
+ "val_loss": 0.16870862344072773
25
+ }
26
+ ],
27
+ "threshold_search": [
28
+ {
29
+ "threshold": 0.05,
30
+ "precision": 0.48935965643227997,
31
+ "recall": 0.9732916879143294,
32
+ "f1": 0.6512689272766048,
33
+ "n": 514
34
+ },
35
+ {
36
+ "threshold": 0.1,
37
+ "precision": 0.5677988365547006,
38
+ "recall": 0.9643676695563488,
39
+ "f1": 0.7147615335553822,
40
+ "n": 514
41
+ },
42
+ {
43
+ "threshold": 0.15,
44
+ "precision": 0.6205062664907651,
45
+ "recall": 0.9593957164711882,
46
+ "f1": 0.7536050470658923,
47
+ "n": 514
48
+ },
49
+ {
50
+ "threshold": 0.2,
51
+ "precision": 0.656819775596073,
52
+ "recall": 0.9552524222335543,
53
+ "f1": 0.7784126324537711,
54
+ "n": 514
55
+ },
56
+ {
57
+ "threshold": 0.25,
58
+ "precision": 0.6854035699536548,
59
+ "recall": 0.952129015808261,
60
+ "f1": 0.797043835543342,
61
+ "n": 514
62
+ },
63
+ {
64
+ "threshold": 0.3,
65
+ "precision": 0.7081612305939324,
66
+ "recall": 0.950790413054564,
67
+ "f1": 0.8117330140676444,
68
+ "n": 514
69
+ },
70
+ {
71
+ "threshold": 0.35,
72
+ "precision": 0.7310585574512021,
73
+ "recall": 0.9477944926058134,
74
+ "f1": 0.8254364781969078,
75
+ "n": 514
76
+ },
77
+ {
78
+ "threshold": 0.4,
79
+ "precision": 0.7513157894736842,
80
+ "recall": 0.9463284038755737,
81
+ "f1": 0.8376213044459491,
82
+ "n": 514
83
+ },
84
+ {
85
+ "threshold": 0.45,
86
+ "precision": 0.7724760116812682,
87
+ "recall": 0.9442248852626212,
88
+ "f1": 0.8497590637907297,
89
+ "n": 514
90
+ },
91
+ {
92
+ "threshold": 0.5,
93
+ "precision": 0.7878350955890207,
94
+ "recall": 0.9404003059663437,
95
+ "f1": 0.8573836229441507,
96
+ "n": 514
97
+ },
98
+ {
99
+ "threshold": 0.55,
100
+ "precision": 0.7970857483343264,
101
+ "recall": 0.9379780724120347,
102
+ "f1": 0.8618114732495827,
103
+ "n": 514
104
+ },
105
+ {
106
+ "threshold": 0.6,
107
+ "precision": 0.8108944257223358,
108
+ "recall": 0.9356195818459969,
109
+ "f1": 0.8688034567462785,
110
+ "n": 514
111
+ },
112
+ {
113
+ "threshold": 0.65,
114
+ "precision": 0.822739279491935,
115
+ "recall": 0.9331336053034166,
116
+ "f1": 0.8744661151100623,
117
+ "n": 514
118
+ },
119
+ {
120
+ "threshold": 0.7,
121
+ "precision": 0.8399861759115258,
122
+ "recall": 0.9295639979602244,
123
+ "f1": 0.8825077914611638,
124
+ "n": 514
125
+ },
126
+ {
127
+ "threshold": 0.75,
128
+ "precision": 0.8529963644892694,
129
+ "recall": 0.9272692503824579,
130
+ "f1": 0.8885834707714861,
131
+ "n": 514
132
+ },
133
+ {
134
+ "threshold": 0.8,
135
+ "precision": 0.8671893703615035,
136
+ "recall": 0.9235721570627231,
137
+ "f1": 0.8944931473021361,
138
+ "n": 514
139
+ },
140
+ {
141
+ "threshold": 0.85,
142
+ "precision": 0.8886970463094284,
143
+ "recall": 0.9186639469658338,
144
+ "f1": 0.9034320639398213,
145
+ "n": 514
146
+ },
147
+ {
148
+ "threshold": 0.9,
149
+ "precision": 0.9144846085068336,
150
+ "recall": 0.9127358490566038,
151
+ "f1": 0.9136093919479359,
152
+ "n": 514
153
+ },
154
+ {
155
+ "threshold": 0.95,
156
+ "precision": 0.9486852696538771,
157
+ "recall": 0.9015170831208567,
158
+ "f1": 0.924499934631978,
159
+ "n": 514
160
+ }
161
+ ]
162
+ }
deberta_mixed/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[CLS]",
3
+ "cls_token": "[CLS]",
4
+ "eos_token": "[SEP]",
5
+ "mask_token": "[MASK]",
6
+ "pad_token": "[PAD]",
7
+ "sep_token": "[SEP]",
8
+ "unk_token": {
9
+ "content": "[UNK]",
10
+ "lstrip": false,
11
+ "normalized": true,
12
+ "rstrip": false,
13
+ "single_word": false
14
+ }
15
+ }
deberta_mixed/spm.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
3
+ size 2464616
deberta_mixed/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
deberta_mixed/tokenizer_config.json ADDED
@@ -0,0 +1,59 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "[CLS]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[SEP]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[UNK]",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "128000": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "[CLS]",
45
+ "clean_up_tokenization_spaces": false,
46
+ "cls_token": "[CLS]",
47
+ "do_lower_case": false,
48
+ "eos_token": "[SEP]",
49
+ "extra_special_tokens": {},
50
+ "mask_token": "[MASK]",
51
+ "model_max_length": 1000000000000000019884624838656,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "sp_model_kwargs": {},
55
+ "split_by_punct": false,
56
+ "tokenizer_class": "DebertaV2Tokenizer",
57
+ "unk_token": "[UNK]",
58
+ "vocab_type": "spm"
59
+ }