initial commit
Browse files- added_tokens.json +1 -0
- config.json +114 -0
- entity_vocab.json +1 -0
- pytorch_model.bin +3 -0
- sentencepiece.bpe.model +3 -0
- special_tokens_map.json +1 -0
- tokenizer_config.json +1 -0
added_tokens.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"<ent>": 250002, "<ent2>": 250003}
|
config.json
ADDED
@@ -0,0 +1,114 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "studio-ousia/mluke-large-lite",
|
3 |
+
"architectures": [
|
4 |
+
"LukeForEntityPairClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"bert_model_name": "xlm-roberta-large",
|
8 |
+
"bos_token_id": 0,
|
9 |
+
"classifier_bias": true,
|
10 |
+
"cls_entity_prediction": false,
|
11 |
+
"deepspeed_transformer_layer_args": {},
|
12 |
+
"entity_emb_size": 256,
|
13 |
+
"entity_vocab_size": 4,
|
14 |
+
"eos_token_id": 2,
|
15 |
+
"gradient_checkpointing": false,
|
16 |
+
"hidden_act": "gelu",
|
17 |
+
"hidden_dropout_prob": 0.1,
|
18 |
+
"hidden_size": 1024,
|
19 |
+
"id2label": {
|
20 |
+
"0": "per:employee_of(e1,e2)",
|
21 |
+
"1": "no_relation",
|
22 |
+
"2": "per:employee_of(e2,e1)",
|
23 |
+
"3": "org:city_of_headquarters(e1,e2)",
|
24 |
+
"4": "per:countries_of_residence(e1,e2)",
|
25 |
+
"5": "per:countries_of_residence(e2,e1)",
|
26 |
+
"6": "org:country_of_headquarters(e1,e2)",
|
27 |
+
"7": "org:members(e2,e1)",
|
28 |
+
"8": "per:stateorprovinces_of_residence(e1,e2)",
|
29 |
+
"9": "org:subsidiaries(e2,e1)",
|
30 |
+
"10": "per:cities_of_residence(e1,e2)",
|
31 |
+
"11": "per:title(e2,e1)",
|
32 |
+
"12": "org:subsidiaries(e1,e2)",
|
33 |
+
"13": "org:stateorprovince_of_headquarters(e1,e2)",
|
34 |
+
"14": "org:country_of_headquarters(e2,e1)",
|
35 |
+
"15": "org:top_members/employees(e2,e1)",
|
36 |
+
"16": "org:city_of_headquarters(e2,e1)",
|
37 |
+
"17": "org:alternate_names(e1,e2)",
|
38 |
+
"18": "org:top_members/employees(e1,e2)",
|
39 |
+
"19": "org:founded(e1,e2)",
|
40 |
+
"20": "per:country_of_birth(e1,e2)",
|
41 |
+
"21": "per:cities_of_residence(e2,e1)",
|
42 |
+
"22": "org:alternate_names(e2,e1)",
|
43 |
+
"23": "per:title(e1,e2)",
|
44 |
+
"24": "per:stateorprovinces_of_residence(e2,e1)",
|
45 |
+
"25": "org:members(e1,e2)",
|
46 |
+
"26": "org:founded_by(e1,e2)",
|
47 |
+
"27": "org:founded_by(e2,e1)",
|
48 |
+
"28": "org:founded(e2,e1)",
|
49 |
+
"29": "per:spouse(e1,e2)",
|
50 |
+
"30": "per:origin(e2,e1)",
|
51 |
+
"31": "per:origin(e1,e2)",
|
52 |
+
"32": "org:stateorprovince_of_headquarters(e2,e1)",
|
53 |
+
"33": "per:spouse(e2,e1)",
|
54 |
+
"34": "per:country_of_birth(e2,e1)",
|
55 |
+
"35": "per:alternate_names(e1,e2)",
|
56 |
+
"36": "per:alternate_names(e2,e1)"
|
57 |
+
},
|
58 |
+
"initializer_range": 0.02,
|
59 |
+
"intermediate_size": 4096,
|
60 |
+
"label2id": {
|
61 |
+
"per:employee_of(e1,e2)": 0,
|
62 |
+
"no_relation": 1,
|
63 |
+
"per:employee_of(e2,e1)": 2,
|
64 |
+
"org:city_of_headquarters(e1,e2)": 3,
|
65 |
+
"per:countries_of_residence(e1,e2)": 4,
|
66 |
+
"per:countries_of_residence(e2,e1)": 5,
|
67 |
+
"org:country_of_headquarters(e1,e2)": 6,
|
68 |
+
"org:members(e2,e1)": 7,
|
69 |
+
"per:stateorprovinces_of_residence(e1,e2)": 8,
|
70 |
+
"org:subsidiaries(e2,e1)": 9,
|
71 |
+
"per:cities_of_residence(e1,e2)": 10,
|
72 |
+
"per:title(e2,e1)": 11,
|
73 |
+
"org:subsidiaries(e1,e2)": 12,
|
74 |
+
"org:stateorprovince_of_headquarters(e1,e2)": 13,
|
75 |
+
"org:country_of_headquarters(e2,e1)": 14,
|
76 |
+
"org:top_members/employees(e2,e1)": 15,
|
77 |
+
"org:city_of_headquarters(e2,e1)": 16,
|
78 |
+
"org:alternate_names(e1,e2)": 17,
|
79 |
+
"org:top_members/employees(e1,e2)": 18,
|
80 |
+
"org:founded(e1,e2)": 19,
|
81 |
+
"per:country_of_birth(e1,e2)": 20,
|
82 |
+
"per:cities_of_residence(e2,e1)": 21,
|
83 |
+
"org:alternate_names(e2,e1)": 22,
|
84 |
+
"per:title(e1,e2)": 23,
|
85 |
+
"per:stateorprovinces_of_residence(e2,e1)": 24,
|
86 |
+
"org:members(e1,e2)": 25,
|
87 |
+
"org:founded_by(e1,e2)": 26,
|
88 |
+
"org:founded_by(e2,e1)": 27,
|
89 |
+
"org:founded(e2,e1)": 28,
|
90 |
+
"per:spouse(e1,e2)": 29,
|
91 |
+
"per:origin(e2,e1)": 30,
|
92 |
+
"per:origin(e1,e2)": 31,
|
93 |
+
"org:stateorprovince_of_headquarters(e2,e1)": 32,
|
94 |
+
"per:spouse(e2,e1)": 33,
|
95 |
+
"per:country_of_birth(e2,e1)": 34,
|
96 |
+
"per:alternate_names(e1,e2)": 35,
|
97 |
+
"per:alternate_names(e2,e1)": 36
|
98 |
+
},
|
99 |
+
"layer_norm_eps": 1e-05,
|
100 |
+
"max_position_embeddings": 514,
|
101 |
+
"model_type": "luke",
|
102 |
+
"num_attention_heads": 16,
|
103 |
+
"num_hidden_layers": 24,
|
104 |
+
"output_past": true,
|
105 |
+
"pad_token_id": 1,
|
106 |
+
"position_embedding_type": "absolute",
|
107 |
+
"torch_dtype": "float32",
|
108 |
+
"transformers_version": "4.15.0",
|
109 |
+
"type_vocab_size": 1,
|
110 |
+
"use_cache": true,
|
111 |
+
"use_deepspeed_transformer_layer": false,
|
112 |
+
"use_entity_aware_attention": false,
|
113 |
+
"vocab_size": 250004
|
114 |
+
}
|
entity_vocab.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"[MASK]": 0, "[UNK]": 1, "[PAD]": 2, "[MASK2]": 3}
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:088bc5673593da08359cbe30f34e44b95e198d61ad174406b7843bbb308f27be
|
3 |
+
size 2243197603
|
sentencepiece.bpe.model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
|
3 |
+
size 5069051
|
special_tokens_map.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true}, "additional_special_tokens": [{"content": "<ent>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}, {"content": "<ent2>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true}]}
|
tokenizer_config.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "cls_token": "<s>", "pad_token": "<pad>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "sp_model_kwargs": {}, "task": null, "max_entity_length": 32, "max_mention_length": 30, "entity_token_1": {"content": "<ent>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "entity_token_2": {"content": "<ent2>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "entity_unk_token": "[UNK]", "entity_pad_token": "[PAD]", "entity_mask_token": "[MASK]", "entity_mask2_token": "[MASK2]", "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "studio-ousia/mluke-large-lite", "additional_special_tokens": [{"content": "<ent>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, {"content": "<ent2>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, {"content": "<ent>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, {"content": "<ent2>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}], "tokenizer_file": null, "tokenizer_class": "MLukeTokenizer"}
|