It seems that the get_latest_training_loss function in fasttext returns only 0. Both gensim 4.1.0 and 4.0.0 do not work.
from gensim.models.callbacks import CallbackAny2Vec
from pprint import pprint as print
from gensim.models.fasttext import FastText
from gensim.test.utils import datapath
class callback(CallbackAny2Vec):
'''Callback to print loss after each epoch.'''
def __init__(self):
self.epoch = 0
def on_epoch_end(self, model):
loss = model.get_latest_training_loss()
print('Loss after epoch {}: {}'.format(self.epoch, loss))
self.epoch += 1
# Set file names for train and test data
corpus_file = datapath('lee_background.cor')
model = FastText(vector_size=100, callbacks=[callback()])
# build the vocabulary
model.build_vocab(corpus_file=corpus_file)
# train the model
model.train(
corpus_file=corpus_file, epochs=model.epochs,
total_examples=model.corpus_count, total_words=model.corpus_total_words,
callbacks=model.callbacks, compute_loss=True,
)
print(model)
'Loss after epoch 0: 0.0'
'Loss after epoch 1: 0.0'
'Loss after epoch 2: 0.0'
'Loss after epoch 3: 0.0'
'Loss after epoch 4: 0.0'
If currently FastText does not support get_latest_training_loss, the documentation here needs to be removed:
I have tried this in three different environments and neither of them works.
First environment:
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import platform; print(platform.platform())
Linux-3.10.0-1160.36.2.el7.x86_64-x86_64-with-glibc2.17
>>> import sys; print("Python", sys.version)
Python 3.9.6 | packaged by conda-forge | (default, Jul 11 2021, 03:39:48)
[GCC 9.3.0]
>>> import struct; print("Bits", 8 * struct.calcsize("P"))
Bits 64
>>> import numpy; print("NumPy", numpy.__version__)
NumPy 1.21.2
>>> import scipy; print("SciPy", scipy.__version__)
SciPy 1.7.1
>>> import gensim; print("gensim", gensim.__version__)
gensim 4.1.0
>>> from gensim.models import word2vec;print("FAST_VERSION", word2vec.FAST_VERSION)
FAST_VERSION 0
Second environment:
Python 3.9.5 (default, May 18 2021, 12:31:01)
[Clang 10.0.0 ] :: Anaconda, Inc. on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import platform; print(platform.platform())
macOS-10.16-x86_64-i386-64bit
>>> import sys; print("Python", sys.version)
Python 3.9.5 (default, May 18 2021, 12:31:01)
[Clang 10.0.0 ]
>>> import struct; print("Bits", 8 * struct.calcsize("P"))
Bits 64
>>> import numpy; print("NumPy", numpy.__version__)
NumPy 1.20.3
>>> import scipy; print("SciPy", scipy.__version__)
SciPy 1.7.1
>>> import gensim; print("gensim", gensim.__version__)
gensim 4.1.0
>>> from gensim.models import word2vec;print("FAST_VERSION", word2vec.FAST_VERSION)
FAST_VERSION 0
Third environment:
Python 3.9.5 (default, May 18 2021, 12:31:01)
[Clang 10.0.0 ] :: Anaconda, Inc. on darwin
Type "help", "copyright", "credits" or "license" for more information.
>>> import platform; print(platform.platform())
macOS-10.16-x86_64-i386-64bit
>>> import sys; print("Python", sys.version)
Python 3.9.5 (default, May 18 2021, 12:31:01)
[Clang 10.0.0 ]
>>> import struct; print("Bits", 8 * struct.calcsize("P"))
Bits 64
>>> import numpy; print("NumPy", numpy.__version__)
NumPy 1.20.3
>>> import scipy; print("SciPy", scipy.__version__)
SciPy 1.7.1
>>> import gensim; print("gensim", gensim.__version__)
/Users/jinhuawang/miniconda3/lib/python3.9/site-packages/gensim/similarities/__init__.py:15: UserWarning: The gensim.similarities.levenshtein submodule is disabled, because the optional Levenshtein package <https://pypi.org/project/python-Levenshtein/> is unavailable. Install Levenhstein (e.g. `pip install python-Levenshtein`) to suppress this warning.
warnings.warn(msg)
gensim 4.0.0
>>> from gensim.models import word2vec;print("FAST_VERSION", word2vec.FAST_VERSION)
FAST_VERSION 0
Indeed, loss-tracking hasn't ever been implemented in Gensim's FastText model, at least through release 4.1.0 (August 2021).
The docs for that method appear in error, due to the inherited method from the Word2Vec superclass not being overriden to prevent the default assumption that superclass methods work.
There is a long-open issue to fill the gaps & fix the problems in Gensim's loss-tracking (which is also somewhat buggy & incomplete for Word2Vec). But, at the moment I don't think any contributor is working on it, & it hasn't been prioritized for any upcoming release. It may require someone to volunteer to step forward & fix things.