I have some sentences for which I am creating an embedding and it works great for similarity searching unless there are some truly unusual words in the sentence.
In that case, the truly unusual words in fact contain the very most similarity information of any words in the sentence BUT all of that information is lost during embedding due to the fact that the word is apparently not in the vocabulary of the model.
I'd like to get a list of all of the words known by the GUSE embedding model so that I can mask those known words out of my sentence, leaving only the "novel" words.
I can then do an exact word search for those novel words in my target corpus and achieve usability for my similar sentence searching.
e.g. "I love to use Xapian!" gets embedded as "I love to use UNK".
If I just do a keyword search for "Xapian" instead of a semantic similarity search, I'll get much more relevant results than I would using GUSE and vector KNN.
Any ideas on how I can extract the vocabulary known/used by GUSE?
I combine the earlier answer from #Roee Shenberg and the solution provided here to come up with solution, which is applicable for USE v4:
import importlib
loader_impl = importlib.import_module('tensorflow.python.saved_model.loader_impl')
saved_model = loader_impl.parse_saved_model("/tmp/tfhub_modules/063d866c06683311b44b4992fd46003be952409c/")
graph = saved_model.meta_graphs[0].graph_def
fns = [f for f in saved_model.meta_graphs[0].graph_def.library.function if "ptb" in str(f).lower()];
print(len(fns)) # should be 1
nodes_with_sp = [n for n in fns[0].node_def if n.name == "Embeddings_words"]
print(len(nodes_with_sp)) # should be 1
words_tensor = nodes_with_sp[0].attr.get("value").tensor
word_list = [i.decode('utf-8') for i in words_tensor.string_val]
print(len(word_list)) # should be 400004
If you are just curious about the words I upload them here.
I'm assuming you have tensorflow & tensorflow_hub installed, and youhave already downloaded the model.
IMPORTANT: I'm assuming you're looking at https://tfhub.dev/google/universal-sentence-encoder/4! There's no guarantee the object graph looks the same for different versions, it's likely that modifications will be needed.
Find it's location on disk - it's somewhere at /tmp/tfhub_modules unless you set the TFHUB_CACHE_DIR environment variable (Windows/Mac have different locations). The path should contain a file called saved_model.pb, which is the model, serialized using Protocol Buffers.
Unfortunately, the dictionary is serialized inside the model's Protocol Buffers file and not as an external asset, so we'll have to load the model and get the variable from it.
The strategy is to use tensorflow's code to deserialize the file, and then travel down the serialized object tree all the way to the dictionary.
import importlib
MODEL_PATH = 'path/to/model/dir' # e.g. '/tmp/tfhub_modules/063d866c06683311b44b4992fd46003be952409c/'
# Use the tensorflow internal Protobuf loader. A regular import statement will fail.
loader_impl = importlib.import_module('tensorflow.python.saved_model.loader_impl')
saved_model = loader_impl.parse_saved_model(MODEL_PATH)
# reach into the object graph to get the tensor
graph = saved_model.meta_graphs[0].graph_def
function = graph.library.function
node_type, node_value = function[5].node_def
# if you print(node_type) you'll see it's called "text_preprocessor/hash_table"
# as well as get insight into this branch of the object graph we're looking at
words_tensor = node_value.attr.get("value").tensor
word_list = [i.decode('utf-8') for i in words_tensor.string_val]
print(len(word_list)) # -> 400004
Some resources that helped:
A GitHub issue relating to changing the vocabulary
A Tensorflow Google-group thread linked from the issue
Extra Notes
Despite what the GitHub issue may lead you to think, the 400k words here are not the GloVe 400k vocabulary. You can verify this by downloading the GloVe 6B embeddings (file link), extracting glove.6B.50d.txt, and then using the following code to compare the two dictionaries:
with open('/path/to/glove.6B.50d.txt') as f:
glove_vocabulary = set(line.strip().split(maxsplit=1)[0] for line in f)
USE_vocabulary = set(word_list) # from above
print(len(USE_vocabulary - glove_vocabulary)) # -> 281150
Inspecting the different vocabularies is interesting in and of itself, e.g. why does GloVe have an entry for '287.9'?
Related
I am currently doing the following tf tutorial : https://www.tensorflow.org/tutorials/text/solve_glue_tasks_using_bert_on_tpu
Testing the outputs of the tokenize function on different sentences, I wonder what happens when tokenizing unknown words.
Loading model:
bert_model_name = 'bert_en_uncased_L-12_H-768_A-12'
tfhub_handle_encoder = 'https://tfhub.dev/tensorflow/bert_en_uncased_L-12_H-768_A-12/3'
tfhub_handle_preprocess = 'https://tfhub.dev/tensorflow/bert_en_uncased_preprocess/3'
bert_preprocess = hub.load(tfhub_handle_preprocess)
Tokenizing sentence/word:
tok = bert_preprocess.tokenize(tf.constant(['Tensorsss bla']))
print(tok)
# Output:
<tf.RaggedTensor [[[23435, 4757, 2015], [1038, 2721]]]>
Shouldnt it be so every word is tokenized to a single token ? Those are obviously made up words, but I am wondering what happens when you encode those words to fixed length vectors.
Also, how does the tokenizer transform the made up words in 3 different tokens ? Does it split the unknown words into different known parts ?
The default cache location for the tensorflow/bert_en_uncased_preprocess/3 model is /tmp/tfhub_modules/602d30248ff7929470db09f7385fc895e9ceb4c0 (more on caching). In the assets directory, you'll find vocab.txt, which is the used vocabulary. You can use the file to look up what token the token-id i corresponds to by looking at line i+1 of the file i.e.
sed '23436q;d' /tmp/tfhub_modules/602d30248ff7929470db09f7385fc895e9ceb4c0/assets/vocab.txt
> tensor
Doing that for all token-ids returns
[tensor, ##ss, ##s], [b, ##la]
As you can see, this confirms your theory that words are split into different known parts. More details on the exact algorithm can be found in Subword tokenizers.
Problem
I'm working on training and deploying an instance segmentation model using TF's object detection API. I'm able to successfully train the model, package it into a TF Serving Docker image (latest tag as of Oct 2020), and process inference requests via the REST interface. However, the amount of data returned from an inference request is very large (hundreds of Mb). This is a big problem when the inference request and processing don't happen on the same machine because all that returned data has to go over the network.
Is there a way to trim down the number of outputs (either during model export or within the TF Serving image) so allow faster round trip times during inference?
Details
I'm using TF OD API (with TF2) to train a Mask RCNN model, which is a modified version of this config. I believe the full list of outputs is described in code here. The list of items I get during inference is also pasted below. For a model with 100 object proposals, that information is ~270 Mb if I just write the returned inference as json to disk.
inference_payload['outputs'].keys()
dict_keys(['detection_masks', 'rpn_features_to_crop', 'detection_anchor_indices', 'refined_box_encodings', 'final_anchors', 'mask_predictions', 'detection_classes', 'num_detections', 'rpn_box_predictor_features', 'class_predictions_with_background', 'proposal_boxes', 'raw_detection_boxes', 'rpn_box_encodings', 'box_classifier_features', 'raw_detection_scores', 'proposal_boxes_normalized', 'detection_multiclass_scores', 'anchors', 'num_proposals', 'detection_boxes', 'image_shape', 'rpn_objectness_predictions_with_background', 'detection_scores'])
I already encode the images within my inference requests as base64, so the request payload is not too large when going over the network. It's just that the inference response is gigantic in comparison. I only need 4 or 5 of the items out of this response, so it'd be great to exclude the rest and avoid passing such a large package of bits over the network.
Things I've tried
I've tried setting the score_threshold to a higher value during the export (code example here) to reduce the number of outputs. However, this seems to just threshold the detection_scores. All the extraneous inference information is still returned.
I also tried just manually excluding some of these inference outputs by adding the names of keys to remove here. That also didn't seem to have any effect, and I'm worried this is a bad idea because some of those keys might be needed during scoring/evaluation.
I also searched here and on tensorflow/models repo, but I wasn't able to find anything.
I was able to find a hacky workaround. In the export process (here), some of the components of the prediction dict are deleted. I added additional items to the non_tensor_predictions list, which contains all keys that will get removed during the postprocess step. Augmenting this list cut down my inference outputs from ~200MB to ~12MB.
Full code for the if self._number_of_stages == 3 block:
if self._number_of_stages == 3:
non_tensor_predictions = [
k for k, v in prediction_dict.items() if not isinstance(v, tf.Tensor)]
# Add additional keys to delete during postprocessing
non_tensor_predictions = non_tensor_predictions + ['raw_detection_scores', 'detection_multiclass_scores', 'anchors', 'rpn_objectness_predictions_with_background', 'detection_anchor_indices', 'refined_box_encodings', 'class_predictions_with_background', 'raw_detection_boxes', 'final_anchors', 'rpn_box_encodings', 'box_classifier_features']
for k in non_tensor_predictions:
tf.logging.info('Removing {0} from prediction_dict'.format(k))
prediction_dict.pop(k)
return prediction_dict
I think there's a more "proper" way to deal with this using signature definitions during the creation of the TF Serving image, but this worked for a quick and dirty fix.
I've ran into the same problem. In the exporter_main_v2 code there is stated that the outputs should be:
and the following output nodes returned by the model.postprocess(..):
* `num_detections`: Outputs float32 tensors of the form [batch]
that specifies the number of valid boxes per image in the batch.
* `detection_boxes`: Outputs float32 tensors of the form
[batch, num_boxes, 4] containing detected boxes.
* `detection_scores`: Outputs float32 tensors of the form
[batch, num_boxes] containing class scores for the detections.
* `detection_classes`: Outputs float32 tensors of the form
[batch, num_boxes] containing classes for the detections.
I've submitted an issue on the tensorflow object detection github repo, I hope we will get feedback from the tensorflow dev team.
The github issue can be found here
If you are using exporter_main_v2.py file to export your model, you can try this hack way to solve this problem.
Just add following codes in the function _run_inference_on_images of exporter_lib_v2.py file:
detections[classes_field] = (
tf.cast(detections[classes_field], tf.float32) + label_id_offset)
############# START ##########
ignored_model_output_names = ["raw_detection_boxes", "raw_detection_scores"]
for key in ignored_model_output_names:
if key in detections.keys(): del detections[key]
############# END ##########
for key, val in detections.items():
detections[key] = tf.cast(val, tf.float32)
Therefore, the generated model will not output the values of ignored_model_output_names.
Please let me know if this can solve your problem.
Another approach would be to alter the signatures of the saved model:
model = tf.saved_model.load(path.join("models", "efficientdet_d7_coco17_tpu-32", "saved_model"))
infer = model.signatures["serving_default"]
outputs = infer.structured_outputs
for o in ["raw_detection_boxes", "raw_detection_scores"]:
outputs.pop(o)
tf.saved_model.save(
model,
export_dir="export",
signatures={"serving_default" : infer},
options=None
)
I am using this code to train a word2vec model. I am trying to train it incrementally, with using saver.restore(). I am using new data after restoring the model. Since vocabulary size for the old data and new data are not the same, I got an exception like this:
InvalidArgumentError (see above for traceback): Assign requires shapes of both tensors to match. lhs shape= [28908,200] rhs shape= [71291,200]
Here 71291 is vocabulary size for the old data and 28908 is for new data.
It gets the vocabulary words from the train_data file here, and constructs the network model using size of the vocabulary. I thought that if I could set vocabulary size the same for my old data and new data, I can solve this problem.
So, my question is: Can I do that in this code? As far as I understand, I cannot reach skipgram_word2vec() function.
Or, is there any other way of solving this issue in this code beside what I thought? If it is not possible using this code, I will try other ways for my purpose.
Any help is appreciated.
Having taken a look at the source of word2vec_optimized.py I'd say you will need to change the code there. It operates by opening a text file right up front as "training data". For your purposes, you have to change the build_graph method and allow it to get an option to set all that data ( words, counts, words_per_epoch, current_epoch, total_words_processed, examples, labels, opts.vocab_words, opts.vocab_counts, opts.words_per_epoch ) when initializing, and not from a text file.
Then you need to merge the two text files, and load them once, to produce the vocabulary. Then save all the data above, and use that to restore the network at each subsequent run.
If you use more than 2 texts, you need to include all the text you plan to use in the first data to produce the vocabulary, however.
I am training a simple model for text classification (currently with scikit-learn). To transform my document samples into word count vectors using a vocabulary I use
CountVectorizer(vocabulary=myDictionaryWords).fit_transform(myDocumentsAsArrays)
from sklearn.feature_extraction.text.
This works great and I can subsequently train my classifier on this word count vectors as feature vectors. But what I don't know is how to inverse transform these word count vectors to the original documents. CountVectorizer indeed has a function inverse_transform(X) but this only gives you back the unique non-zero tokens.
As far as I know CountVectorizer doesn't have any implementation of a mapping back to the original documents.
Anyone know how I can restore the original sequences of tokens from their count-vectorized representation? Is there maybe a Tensorflow or any other module for this?
CountVectorizer is "lossy", i.e. for a document :
This is the amazing string in amazing program , it will only store counts of words in the document (i.e. string -> 1, amazing ->2 etc), but loses the position information.
So by reversing it, you can create a document having same words repeated same number of times, but their sequence in the document cannot be retraced.
Reading the tensorflow word2vec model output how can I output the words related to a specific word ?
Reading the src : https://github.com/tensorflow/tensorflow/blob/r0.11/tensorflow/examples/tutorials/word2vec/word2vec_basic.py can view how the image is plotted.
But is there a data structure (e.g dictionary) created as part of training the model that allows to access nearest n words closest to given word ?
For example if word2vec generated image :
image src: https://www.tensorflow.org/versions/r0.11/tutorials/word2vec/index.html
In this image the words 'to , he , it' are contained in same cluster, is there a function which takes as input 'to' and outputs 'he , it' (in this case n=2) ?
This approach apply to word2vec in general. If you can save the word2vec in text/binary file like google/GloVe word vector. Then what you need is just the gensim.
To install:
Via github
Python code:
from gensim.models import Word2Vec
gmodel=Word2Vec.load_word2vec_format(fname)
ms=gmodel.most_similar('good',10)
for x in ms:
print x[0],x[1]
However this will search all the words to give the results, there are approximate nearest neighbor (ANN) which will give you the result faster but with a trade off in accuracy.
In the latest gensim, annoy is used to perform the ANN, see this notebooks for more information.
Flann is another library for Approximate Nearest Neighbors.
I will assume that you don't want to use gensim, and would prefer to stick with tensorflow. In that case, I'll offer two options
Option 1 - Tensorboard:
If you are just trying to do this from an exploratory standpoint, I would suggest using Tensorboard's embedding visualizer to search for the closest embeddings. It provides a cool interface and you can use both cosine and euclidian distances with a set number of neighbors.
Link to Tensorflow documentation
Option 2 - Direct Calculation
Within the word2vec_basic.py file, there is an example of how they are calculating closest words, and you could go ahead and use that if you mess with the function a little bit. The following is found in the graph itself:
# Compute the cosine similarity between minibatch examples and all embeddings.
norm = tf.sqrt(tf.reduce_sum(tf.square(embeddings), 1, keep_dims=True))
normalized_embeddings = embeddings / norm
valid_embeddings = tf.nn.embedding_lookup(
normalized_embeddings, valid_dataset)
similarity = tf.matmul(
valid_embeddings, normalized_embeddings, transpose_b=True)
Then, during training (every 10000 steps) they run this next bit of code (while the session is active). When they call similarity.eval() it is getting the literal numpy array evaluation of the similarity tensor in the graph.
# Note that this is expensive (~20% slowdown if computed every 500 steps)
if step % 10000 == 0:
sim = similarity.eval()
for i in xrange(valid_size):
valid_word = reverse_dictionary[valid_examples[i]]
top_k = 8 # number of nearest neighbors
nearest = (-sim[i, :]).argsort()[1:top_k+1]
log_str = "Nearest to %s:" % valid_word
for k in xrange(top_k):
close_word = reverse_dictionary[nearest[k]]
log_str = "%s %s," % (log_str, close_word)
print(log_str)
If you want to adapt this for yourself, you will have to do some finessing with changing reverse_dictionary[valid_examples[i]] to be the word/words idxs that you want to get the k-closest words for.
Get gensim and use similar_by_word method on gensim.models.Word2Vec model.
similar_by_word takes 3 parameters,
The input word
n - for top n similar words (optional, default=10)
restrict_vocab (optional, default=None)
Example
import gensim, nltk
class FileToSent(object):
"""A class to load a text file efficiently """
def __init__(self, filename):
self.filename = filename
# To remove stop words (optional)
self.stop = set(nltk.corpus.stopwords.words('english'))
def __iter__(self):
for line in open(self.filename, 'r'):
ll = [i for i in unicode(line, 'utf-8').lower().split() if i not in self.stop]
yield ll
Then depending on your input sentences (sentence_file.txt),
sentences = FileToSent('sentence_file.txt')
model = gensim.models.Word2Vec(sentences=sentences, min_count=2, hs=1)
print model.similar_by_word('hack', 2) # Get two most similar words to 'hack'
# [(u'debug', 0.967338502407074), (u'patch', 0.952264130115509)] (Output specific to my dataset)