Rooba
/
spaCy
mirror of https://github.com/explosion/spaCy.git
1
1
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Update PTB config

This commit is contained in:
Matthew Honnibal 2020-07-25 13:33:40 +02:00
parent c003d26b94
commit 17f39eebdc
1 changed files with 30 additions and 15 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

45
examples/experiments/ptb-joint-pos-dep/defaults.cfg
View File

@ -1,4 +1,5 @@
[training] [training]
max_steps = 0
patience = 10000 patience = 10000
eval_frequency = 200 eval_frequency = 200
dropout = 0.2 dropout = 0.2
@ -8,13 +9,20 @@ max_epochs = 100
orth_variant_level = 0.0 orth_variant_level = 0.0
gold_preproc = true gold_preproc = true
max_length = 0 max_length = 0
use_gpu = -1
scores = ["tags_acc", "uas", "las"] scores = ["tags_acc", "uas", "las"]
score_weights = {"las": 0.8, "tags_acc": 0.2} score_weights = {"las": 0.8, "tags_acc": 0.2}
limit = 0 limit = 0
seed = 0 seed = 0
accumulate_gradient = 2 accumulate_gradient = 2
discard_oversize = false discard_oversize = false
raw_text = null
tag_map = null
morph_rules = null
base_model = null
eval_batch_size = 128
use_pytorch_for_gpu_memory = false
batch_by = "padded"
[training.batch_size] [training.batch_size]
@schedules = "compounding.v1" @schedules = "compounding.v1"
@ -30,41 +38,48 @@ beta2 = 0.999
[nlp] [nlp]
lang = "en" lang = "en"
vectors = ${training:vectors} pipeline = ["tok2vec", "tagger", "parser"]
load_vocab_data = false
[nlp.pipeline.tok2vec] [nlp.tokenizer]
@tokenizers = "spacy.Tokenizer.v1"
[nlp.lemmatizer]
@lemmatizers = "spacy.Lemmatizer.v1"
[components]
[components.tok2vec]
factory = "tok2vec" factory = "tok2vec"
[nlp.pipeline.tagger] [components.tagger]
factory = "tagger" factory = "tagger"
[nlp.pipeline.parser] [components.parser]
factory = "parser" factory = "parser"
learn_tokens = false learn_tokens = false
min_action_freq = 1 min_action_freq = 1
beam_width = 1
beam_update_prob = 1.0
[nlp.pipeline.tagger.model] [components.tagger.model]
@architectures = "spacy.Tagger.v1" @architectures = "spacy.Tagger.v1"
[nlp.pipeline.tagger.model.tok2vec] [components.tagger.model.tok2vec]
@architectures = "spacy.Tok2VecTensors.v1" @architectures = "spacy.Tok2VecTensors.v1"
width = ${nlp.pipeline.tok2vec.model:width} width = ${components.tok2vec.model:width}
[nlp.pipeline.parser.model] [components.parser.model]
@architectures = "spacy.TransitionBasedParser.v1" @architectures = "spacy.TransitionBasedParser.v1"
nr_feature_tokens = 8 nr_feature_tokens = 8
hidden_width = 64 hidden_width = 64
maxout_pieces = 3 maxout_pieces = 3
[nlp.pipeline.parser.model.tok2vec] [components.parser.model.tok2vec]
@architectures = "spacy.Tok2VecTensors.v1" @architectures = "spacy.Tok2VecTensors.v1"
width = ${nlp.pipeline.tok2vec.model:width} width = ${components.tok2vec.model:width}
[nlp.pipeline.tok2vec.model] [components.tok2vec.model]
@architectures = "spacy.HashEmbedCNN.v1" @architectures = "spacy.HashEmbedCNN.v1"
pretrained_vectors = ${nlp:vectors} pretrained_vectors = ${training:vectors}
width = 96 width = 96
depth = 4 depth = 4
window_size = 1 window_size = 1