[E143] Labels for component 'tagger' not initialized.

[funghetto]

Hello,

I'm trying to build an Italian model to use for commercial purposes as the ones provided have a restrictive license but get E143 when trying to train, this is the corpus I'm using:

https://github.com/UniversalDependencies/UD_Italian-MarkIT

This is the full output I get:

PS C:\Users\Michele\OneDrive\Documents\C\Programmaraffa2\Trainingmodel\UD_Italian-MarkIT-r2.10> python.exe -m spacy train .\config.cfg --output .\UD_Italian-MarkIT-r2.10\Fine --gpu-id 0 --verbose
ℹ Saving to output directory: UD_Italian-MarkIT-r2.10\Fine
ℹ Using GPU: 0

=========================== Initializing pipeline ===========================
[2022-09-24 07:27:56,369] [INFO] Set up nlp object from config
[2022-09-24 07:27:56,383] [DEBUG] Loading corpus from path: "C:\Users\Michele\OneDrive\Documents\C\Programmaraffa2\Trainingmodel\UD_Italian-MarkIT-r2.10\it_markit-ud-dev.spacy"
[2022-09-24 07:27:56,386] [DEBUG] Loading corpus from path: "C:\Users\Michele\OneDrive\Documents\C\Programmaraffa2\Trainingmodel\UD_Italian-MarkIT-r2.10\it_markit-ud-train.spacy"
[2022-09-24 07:27:56,387] [INFO] Pipeline: ['transformer', 'tagger']
[2022-09-24 07:27:56,391] [INFO] Created vocabulary
[2022-09-24 07:27:56,393] [INFO] Finished initializing nlp object
Some weights of the model checkpoint at bert-base-multilingual-uncased were not used when initializing BertModel: ['cls.seq_relationship.weight', 'cls.seq_relationship.bias', 'cls.predictions.bias', 'cls.predictions.transform.dense.weight', 'cls.predictions.transform.LayerNorm.bias', 'cls.predictions.decoder.weight', 'cls.predictions.transform.LayerNorm.weight', 'cls.predictions.transform.dense.bias']
- This IS expected if you are initializing BertModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).
- This IS NOT expected if you are initializing BertModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).
C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\spacy\training\corpus.py:80: UserWarning: [W090] Could not locate any .spacy files in path '"C:\Users\Michele\OneDrive\Documents\C\Programmaraffa2\Trainingmodel\UD_Italian-MarkIT-r2.10\it_markit-ud-train.spacy"'.
  warnings.warn(Warnings.W090.format(path=orig_path, format=file_type))
Traceback (most recent call last):
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\spacy\__main__.py", line 4, in <module>
    setup_cli()
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\spacy\cli\_util.py", line 71, in setup_cli
    command(prog_name=COMMAND)
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\click\core.py", line 1130, in __call__
    return self.main(*args, **kwargs)
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\click\core.py", line 1055, in main
    rv = self.invoke(ctx)
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\click\core.py", line 1657, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\click\core.py", line 1404, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\click\core.py", line 760, in invoke
    return __callback(*args, **kwargs)
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\typer\main.py", line 532, in wrapper
    return callback(**use_params)  # type: ignore
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\spacy\cli\train.py", line 45, in train_cli
    train(config_path, output_path, use_gpu=use_gpu, overrides=overrides)
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\spacy\cli\train.py", line 72, in train
    nlp = init_nlp(config, use_gpu=use_gpu)
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\spacy\training\initialize.py", line 84, in init_nlp
    nlp.initialize(lambda: train_corpus(nlp), sgd=optimizer)
  File "C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\spacy\language.py", line 1317, in initialize
    proc.initialize(get_examples, nlp=self, **p_settings)
  File "spacy\pipeline\tagger.pyx", line 305, in spacy.pipeline.tagger.Tagger.initialize
  File "spacy\pipeline\pipe.pyx", line 119, in spacy.pipeline.pipe.Pipe._require_labels
ValueError: [E143] Labels for component 'tagger' not initialized. This can be fixed by calling add_label, or by providing a representative batch of examples to the component's `initialize` method.

This is my config:

[paths]
train = "C:\Users\Michele\OneDrive\Documents\C\Programmaraffa2\Trainingmodel\UD_Italian-MarkIT-r2.10\it_markit-ud-train.spacy"
dev = "C:\Users\Michele\OneDrive\Documents\C\Programmaraffa2\Trainingmodel\UD_Italian-MarkIT-r2.10\it_markit-ud-dev.spacy"
vectors = null
init_tok2vec = null

[system]
gpu_allocator = "pytorch"
seed = 0

[nlp]
lang = "it"
pipeline = ["transformer","tagger"]
batch_size = 128
disabled = []
before_creation = null
after_creation = null
after_pipeline_creation = null
tokenizer = {"@tokenizers":"spacy.Tokenizer.v1"}

[components]

[components.tagger]
factory = "tagger"
neg_prefix = "!"
overwrite = false
scorer = {"@scorers":"spacy.tagger_scorer.v1"}

[components.tagger.model]
@architectures = "spacy.Tagger.v2"
nO = null
normalize = false

[components.tagger.model.tok2vec]
@architectures = "spacy-transformers.TransformerListener.v1"
grad_factor = 1.0
pooling = {"@layers":"reduce_mean.v1"}
upstream = "*"

[components.transformer]
factory = "transformer"
max_batch_items = 4096
set_extra_annotations = {"@annotation_setters":"spacy-transformers.null_annotation_setter.v1"}

[components.transformer.model]
@architectures = "spacy-transformers.TransformerModel.v3"
name = "bert-base-multilingual-uncased"
mixed_precision = false

[components.transformer.model.get_spans]
@span_getters = "spacy-transformers.strided_spans.v1"
window = 128
stride = 96

[components.transformer.model.grad_scaler_config]

[components.transformer.model.tokenizer_config]
use_fast = true

[components.transformer.model.transformer_config]

[corpora]

[corpora.dev]
@readers = "spacy.Corpus.v1"
path = ${paths.dev}
max_length = 0
gold_preproc = false
limit = 0
augmenter = null

[corpora.train]
@readers = "spacy.Corpus.v1"
path = ${paths.train}
max_length = 0
gold_preproc = false
limit = 0
augmenter = null

[training]
accumulate_gradient = 3
dev_corpus = "corpora.dev"
train_corpus = "corpora.train"
seed = ${system.seed}
gpu_allocator = ${system.gpu_allocator}
dropout = 0.1
patience = 1600
max_epochs = 0
max_steps = 20000
eval_frequency = 200
frozen_components = []
annotating_components = []
before_to_disk = null

[training.batcher]
@batchers = "spacy.batch_by_padded.v1"
discard_oversize = true
size = 2000
buffer = 256
get_length = null

[training.logger]
@loggers = "spacy.ConsoleLogger.v1"
progress_bar = false

[training.optimizer]
@optimizers = "Adam.v1"
beta1 = 0.9
beta2 = 0.999
L2_is_weight_decay = true
L2 = 0.01
grad_clip = 1.0
use_averages = false
eps = 0.00000001

[training.optimizer.learn_rate]
@schedules = "warmup_linear.v1"
warmup_steps = 250
total_steps = 20000
initial_rate = 0.00005

[training.score_weights]
tag_acc = 1.0

[pretraining]

[initialize]
vectors = ${paths.vectors}
init_tok2vec = ${paths.init_tok2vec}
vocab_data = null
lookups = null
before_init = null
after_init = null

[initialize.components]

[initialize.tokenizer]

spacy info output:

============================== Info about spaCy ==============================

spaCy version    3.4.1
Location         C:\Users\Michele\AppData\Local\Programs\Python\Python310\lib\site-packages\spacy
Platform         Windows-10-10.0.19044-SP0
Python version   3.10.6
Pipelines        en_core_web_sm (3.4.0), it_core_news_sm (3.4.0)

I tested the code provided in a similar github issue showing that there are tags in the document.

import spacy
from spacy.tokens import DocBin

nlp = spacy.blank("en")
db = DocBin().from_disk(r"C:\Users\Michele\OneDrive\Documents\C\Programmaraffa2\Trainingmodel\UD_Italian-MarkIT-r2.10\it_markit-ud-train.spacy")
for doc in db.get_docs(nlp.vocab):
    assert doc.has_annotation("TAG")
    # or just inspect the tags
    for token in doc:
        print(token.text, token.tag_)

Part of output:

essere S
vivente A
è V
l' RD
istinto S
di E
sopravvivenza S
la RD
cosa S
che PR
conta V
più B
di E
ogni DI
altra PI
. FS

Thanks for the help.

[adrianeboyd]

No errors jump out at me in the data or the config and you've checked the right things when inspecting the DocBin. This error does occur if there are no tags in the training data or if there are accidentally no documents at all in the .spacy file, which can happen sometimes if something goes wrong in the conversion.

One additional check to run would be spacy debug data with the exact same config and corpus settings as for spacy train, just in case there's an incorrect .spacy path that might be leading to the initialization error?

A few additional notes, not related to this particular error:

• use spacy convert -T when converting from CoNLL-U to merge the multiword tokens, since spacy's default Italian tokenizer doesn't split these tokens
• this corpus may be too small to train a tagger

[funghetto]

Using spacy debug data I'm getting this error:

UserWarning: [W090] Could not locate any .spacy files 
in path '"C:\Users\Michele\OneDrive\Documents\C\Programmaraffa2\Trainingmodel\UD_Italian-MarkIT-r2.10\it_markit-ud-train.spacy"'.

But the path seems correct:

PS C:\Users\Michele\OneDrive\Documents\C\Programmaraffa2\Trainingmodel\UD_Italian-MarkIT-r2.10> ls .\it_markit-ud-train.spacy


    Directory: C:\Users\Michele\OneDrive\Documents\C\Programmaraffa2\Trainingmodel\UD_Italian-MarkIT-r2.10


Mode                 LastWriteTime         Length Name
----                 -------------         ------ ----
-a---l        10/10/2022     03:07         243093 it_markit-ud-train.spacy

Any clue on what may be going wrong?

EDIT: Ok I can see the issue, the paths need not be surrounded by ".