Update spaCy pipeline
Browse files- .gitattributes +3 -0
- README.md +48 -0
- config.cfg +157 -0
- meta.json +84 -0
- textcat_multilabel/cfg +10 -0
- textcat_multilabel/model +3 -0
- tokenizer/cfg +3 -0
- transformer/cfg +3 -0
- transformer/model +3 -0
- vocab/key2row +1 -0
- vocab/lookups.bin +3 -0
- vocab/strings.json +0 -0
- vocab/vectors +0 -0
- vocab/vectors.cfg +3 -0
- zh_game_layer_classifier-any-py3-none-any.whl +3 -0
.gitattributes
CHANGED
@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
|
|
33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
36 |
+
textcat_multilabel/model filter=lfs diff=lfs merge=lfs -text
|
37 |
+
transformer/model filter=lfs diff=lfs merge=lfs -text
|
38 |
+
zh_game_layer_classifier-any-py3-none-any.whl filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
@@ -0,0 +1,48 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- spacy
|
4 |
+
- text-classification
|
5 |
+
language:
|
6 |
+
- zh
|
7 |
+
model-index:
|
8 |
+
- name: zh_game_layer_classifier
|
9 |
+
results: []
|
10 |
+
---
|
11 |
+
| Feature | Description |
|
12 |
+
| --- | --- |
|
13 |
+
| **Name** | `zh_game_layer_classifier` |
|
14 |
+
| **Version** | `0.0.0` |
|
15 |
+
| **spaCy** | `>=3.6.1,<3.7.0` |
|
16 |
+
| **Default Pipeline** | `transformer`, `textcat_multilabel` |
|
17 |
+
| **Components** | `transformer`, `textcat_multilabel` |
|
18 |
+
| **Vectors** | 0 keys, 0 unique vectors (0 dimensions) |
|
19 |
+
| **Sources** | n/a |
|
20 |
+
| **License** | n/a |
|
21 |
+
| **Author** | [n/a]() |
|
22 |
+
|
23 |
+
### Label Scheme
|
24 |
+
|
25 |
+
<details>
|
26 |
+
|
27 |
+
<summary>View label scheme (5 labels for 1 components)</summary>
|
28 |
+
|
29 |
+
| Component | Labels |
|
30 |
+
| --- | --- |
|
31 |
+
| **`textcat_multilabel`** | `球員`, `教練`, `裁判`, `戰術`, `球隊` |
|
32 |
+
|
33 |
+
</details>
|
34 |
+
|
35 |
+
### Accuracy
|
36 |
+
|
37 |
+
| Type | Score |
|
38 |
+
| --- | --- |
|
39 |
+
| `CATS_SCORE` | 99.73 |
|
40 |
+
| `CATS_MICRO_P` | 98.94 |
|
41 |
+
| `CATS_MICRO_R` | 98.88 |
|
42 |
+
| `CATS_MICRO_F` | 98.91 |
|
43 |
+
| `CATS_MACRO_P` | 98.97 |
|
44 |
+
| `CATS_MACRO_R` | 98.79 |
|
45 |
+
| `CATS_MACRO_F` | 98.88 |
|
46 |
+
| `CATS_MACRO_AUC` | 99.73 |
|
47 |
+
| `TRANSFORMER_LOSS` | 3256.99 |
|
48 |
+
| `TEXTCAT_MULTILABEL_LOSS` | 859.81 |
|
config.cfg
ADDED
@@ -0,0 +1,157 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
[paths]
|
2 |
+
train = "./data/train.spacy"
|
3 |
+
dev = "./data/valid.spacy"
|
4 |
+
vectors = null
|
5 |
+
init_tok2vec = null
|
6 |
+
|
7 |
+
[system]
|
8 |
+
gpu_allocator = "pytorch"
|
9 |
+
seed = 0
|
10 |
+
|
11 |
+
[nlp]
|
12 |
+
lang = "zh"
|
13 |
+
pipeline = ["transformer","textcat_multilabel"]
|
14 |
+
batch_size = 128
|
15 |
+
disabled = []
|
16 |
+
before_creation = null
|
17 |
+
after_creation = null
|
18 |
+
after_pipeline_creation = null
|
19 |
+
|
20 |
+
[nlp.tokenizer]
|
21 |
+
@tokenizers = "spacy.zh.ChineseTokenizer"
|
22 |
+
segmenter = "char"
|
23 |
+
|
24 |
+
[components]
|
25 |
+
|
26 |
+
[components.textcat_multilabel]
|
27 |
+
factory = "textcat_multilabel"
|
28 |
+
scorer = {"@scorers":"spacy.textcat_multilabel_scorer.v2"}
|
29 |
+
threshold = 0.5
|
30 |
+
|
31 |
+
[components.textcat_multilabel.model]
|
32 |
+
@architectures = "spacy.TextCatEnsemble.v2"
|
33 |
+
nO = null
|
34 |
+
|
35 |
+
[components.textcat_multilabel.model.linear_model]
|
36 |
+
@architectures = "spacy.TextCatBOW.v2"
|
37 |
+
exclusive_classes = false
|
38 |
+
ngram_size = 1
|
39 |
+
no_output_layer = false
|
40 |
+
nO = null
|
41 |
+
|
42 |
+
[components.textcat_multilabel.model.tok2vec]
|
43 |
+
@architectures = "spacy-transformers.TransformerListener.v1"
|
44 |
+
grad_factor = 1.0
|
45 |
+
pooling = {"@layers":"reduce_mean.v1"}
|
46 |
+
upstream = "*"
|
47 |
+
|
48 |
+
[components.transformer]
|
49 |
+
factory = "transformer"
|
50 |
+
max_batch_items = 4096
|
51 |
+
set_extra_annotations = {"@annotation_setters":"spacy-transformers.null_annotation_setter.v1"}
|
52 |
+
|
53 |
+
[components.transformer.model]
|
54 |
+
@architectures = "spacy-transformers.TransformerModel.v3"
|
55 |
+
name = "bert-base-chinese"
|
56 |
+
mixed_precision = false
|
57 |
+
|
58 |
+
[components.transformer.model.get_spans]
|
59 |
+
@span_getters = "spacy-transformers.strided_spans.v1"
|
60 |
+
window = 128
|
61 |
+
stride = 96
|
62 |
+
|
63 |
+
[components.transformer.model.grad_scaler_config]
|
64 |
+
|
65 |
+
[components.transformer.model.tokenizer_config]
|
66 |
+
use_fast = true
|
67 |
+
|
68 |
+
[components.transformer.model.transformer_config]
|
69 |
+
|
70 |
+
[corpora]
|
71 |
+
|
72 |
+
[corpora.dev]
|
73 |
+
@readers = "spacy.Corpus.v1"
|
74 |
+
path = ${paths.dev}
|
75 |
+
max_length = 0
|
76 |
+
gold_preproc = false
|
77 |
+
limit = 0
|
78 |
+
augmenter = null
|
79 |
+
|
80 |
+
[corpora.train]
|
81 |
+
@readers = "spacy.Corpus.v1"
|
82 |
+
path = ${paths.train}
|
83 |
+
max_length = 0
|
84 |
+
gold_preproc = false
|
85 |
+
limit = 0
|
86 |
+
augmenter = null
|
87 |
+
|
88 |
+
[training]
|
89 |
+
accumulate_gradient = 3
|
90 |
+
dev_corpus = "corpora.dev"
|
91 |
+
train_corpus = "corpora.train"
|
92 |
+
seed = ${system.seed}
|
93 |
+
gpu_allocator = ${system.gpu_allocator}
|
94 |
+
dropout = 0.1
|
95 |
+
patience = 1600
|
96 |
+
max_epochs = 0
|
97 |
+
max_steps = 20000
|
98 |
+
eval_frequency = 200
|
99 |
+
frozen_components = []
|
100 |
+
annotating_components = []
|
101 |
+
before_to_disk = null
|
102 |
+
before_update = null
|
103 |
+
|
104 |
+
[training.batcher]
|
105 |
+
@batchers = "spacy.batch_by_padded.v1"
|
106 |
+
discard_oversize = true
|
107 |
+
size = 2000
|
108 |
+
buffer = 256
|
109 |
+
get_length = null
|
110 |
+
|
111 |
+
[training.logger]
|
112 |
+
@loggers = "spacy.ConsoleLogger.v1"
|
113 |
+
progress_bar = false
|
114 |
+
|
115 |
+
[training.optimizer]
|
116 |
+
@optimizers = "Adam.v1"
|
117 |
+
beta1 = 0.9
|
118 |
+
beta2 = 0.999
|
119 |
+
L2_is_weight_decay = true
|
120 |
+
L2 = 0.01
|
121 |
+
grad_clip = 1.0
|
122 |
+
use_averages = false
|
123 |
+
eps = 0.00000001
|
124 |
+
|
125 |
+
[training.optimizer.learn_rate]
|
126 |
+
@schedules = "warmup_linear.v1"
|
127 |
+
warmup_steps = 250
|
128 |
+
total_steps = 20000
|
129 |
+
initial_rate = 0.00005
|
130 |
+
|
131 |
+
[training.score_weights]
|
132 |
+
cats_score = 1.0
|
133 |
+
cats_score_desc = null
|
134 |
+
cats_micro_p = null
|
135 |
+
cats_micro_r = null
|
136 |
+
cats_micro_f = null
|
137 |
+
cats_macro_p = null
|
138 |
+
cats_macro_r = null
|
139 |
+
cats_macro_f = null
|
140 |
+
cats_macro_auc = null
|
141 |
+
cats_f_per_type = null
|
142 |
+
|
143 |
+
[pretraining]
|
144 |
+
|
145 |
+
[initialize]
|
146 |
+
vectors = ${paths.vectors}
|
147 |
+
init_tok2vec = ${paths.init_tok2vec}
|
148 |
+
vocab_data = null
|
149 |
+
lookups = null
|
150 |
+
before_init = null
|
151 |
+
after_init = null
|
152 |
+
|
153 |
+
[initialize.components]
|
154 |
+
|
155 |
+
[initialize.tokenizer]
|
156 |
+
pkuseg_model = null
|
157 |
+
pkuseg_user_dict = "default"
|
meta.json
ADDED
@@ -0,0 +1,84 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"lang":"zh",
|
3 |
+
"name":"game_layer_classifier",
|
4 |
+
"version":"0.0.0",
|
5 |
+
"description":"",
|
6 |
+
"author":"",
|
7 |
+
"email":"",
|
8 |
+
"url":"",
|
9 |
+
"license":"",
|
10 |
+
"spacy_version":">=3.6.1,<3.7.0",
|
11 |
+
"spacy_git_version":"458bc5f45",
|
12 |
+
"vectors":{
|
13 |
+
"width":0,
|
14 |
+
"vectors":0,
|
15 |
+
"keys":0,
|
16 |
+
"name":null
|
17 |
+
},
|
18 |
+
"labels":{
|
19 |
+
"transformer":[
|
20 |
+
|
21 |
+
],
|
22 |
+
"textcat_multilabel":[
|
23 |
+
"\u7403\u54e1",
|
24 |
+
"\u6559\u7df4",
|
25 |
+
"\u88c1\u5224",
|
26 |
+
"\u6230\u8853",
|
27 |
+
"\u7403\u968a"
|
28 |
+
]
|
29 |
+
},
|
30 |
+
"pipeline":[
|
31 |
+
"transformer",
|
32 |
+
"textcat_multilabel"
|
33 |
+
],
|
34 |
+
"components":[
|
35 |
+
"transformer",
|
36 |
+
"textcat_multilabel"
|
37 |
+
],
|
38 |
+
"disabled":[
|
39 |
+
|
40 |
+
],
|
41 |
+
"performance":{
|
42 |
+
"cats_score":0.9973097822,
|
43 |
+
"cats_score_desc":"macro AUC",
|
44 |
+
"cats_micro_p":0.9894019349,
|
45 |
+
"cats_micro_r":0.9888366369,
|
46 |
+
"cats_micro_f":0.9891192052,
|
47 |
+
"cats_macro_p":0.9897095913,
|
48 |
+
"cats_macro_r":0.9879037181,
|
49 |
+
"cats_macro_f":0.9887754518,
|
50 |
+
"cats_macro_auc":0.9973097822,
|
51 |
+
"cats_f_per_type":{
|
52 |
+
"\u7403\u54e1":{
|
53 |
+
"p":0.9866971524,
|
54 |
+
"r":0.9894736842,
|
55 |
+
"f":0.9880834678
|
56 |
+
},
|
57 |
+
"\u6559\u7df4":{
|
58 |
+
"p":0.9918283963,
|
59 |
+
"r":0.9928425358,
|
60 |
+
"f":0.9923352069
|
61 |
+
},
|
62 |
+
"\u88c1\u5224":{
|
63 |
+
"p":0.9929378531,
|
64 |
+
"r":0.9723374827,
|
65 |
+
"f":0.9825296995
|
66 |
+
},
|
67 |
+
"\u6230\u8853":{
|
68 |
+
"p":0.9823627288,
|
69 |
+
"r":0.9949443883,
|
70 |
+
"f":0.9886135298
|
71 |
+
},
|
72 |
+
"\u7403\u968a":{
|
73 |
+
"p":0.994721826,
|
74 |
+
"r":0.9899204997,
|
75 |
+
"f":0.9923153551
|
76 |
+
}
|
77 |
+
},
|
78 |
+
"transformer_loss":32.5699071415,
|
79 |
+
"textcat_multilabel_loss":8.5980506992
|
80 |
+
},
|
81 |
+
"requirements":[
|
82 |
+
"spacy-transformers>=1.2.5,<1.3.0"
|
83 |
+
]
|
84 |
+
}
|
textcat_multilabel/cfg
ADDED
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"labels":[
|
3 |
+
"\u7403\u54e1",
|
4 |
+
"\u6559\u7df4",
|
5 |
+
"\u88c1\u5224",
|
6 |
+
"\u6230\u8853",
|
7 |
+
"\u7403\u968a"
|
8 |
+
],
|
9 |
+
"threshold":0.5
|
10 |
+
}
|
textcat_multilabel/model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8ed1b6c463eb47a5ff3f398d1f8dead98d740133389c273bf326a285d2abf4ca
|
3 |
+
size 12357563
|
tokenizer/cfg
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"segmenter":"char"
|
3 |
+
}
|
transformer/cfg
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"max_batch_items":4096
|
3 |
+
}
|
transformer/model
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:2ab52306c5249d878b7ee923a149236b9eba282805d3058e1cbe8ae62c9af305
|
3 |
+
size 409704459
|
vocab/key2row
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
�
|
vocab/lookups.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:76be8b528d0075f7aae98d6fa57a6d3c83ae480a8469e668d7b0af968995ac71
|
3 |
+
size 1
|
vocab/strings.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
vocab/vectors
ADDED
Binary file (128 Bytes). View file
|
|
vocab/vectors.cfg
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"mode":"default"
|
3 |
+
}
|
zh_game_layer_classifier-any-py3-none-any.whl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:984d57643bb402b485b427b5596b7bbb307fc37f7aa581820572af5ac835ed37
|
3 |
+
size 386815234
|