TypeError: Dimension value must be integer or None or have an __index__ method, got TensorShape([None, 1])

user_11625932 0
Viewed 18087

I am trying to convert a bot tutorial I found that uses tflearn to use keras. Right now the output should be data for the models prediction. However, I get the error:

TypeError: Dimension value must be integer or None or have an __index__ method, got TensorShape([None, 1])

my code:

with open("intents.json") as file:
    data = json.load(file)
try:
    with open("data.pickle", "rb") as f:
        words, lables, training, output = pickle.load(f)

except:
    words = []
    labels = []
    docs_x = []
    docs_y = []

    for intent in data["intents"]:
        for pattern in intent["patterns"]:
            wrds = nltk.word_tokenize(pattern)
            words.extend(wrds)
            docs_x.append(wrds)
            docs_y.append(intent["tag"])

        if intent["tag"] not in labels:
            labels.append(intent["tag"])

    words = [stemmer.stem(w.lower()) for w in words if w != "?"]
    words = sorted(list(set(words)))

    labels = sorted(labels)

    training = []
    output = []

    out_empty = [0 for _ in range(len(labels))]

    for x, doc in enumerate(docs_x):
        bag = []

        wrds = [stemmer.stem(w) for w in doc]

        # marks if a word is present in the input (1=yes, 0=no)
        for w in words:
            if w in wrds:
                bag.append(1)
            else:
                bag.append(0)

        output_row = out_empty[:]
        output_row[labels.index(docs_y[x])] = 1

        training.append(bag)
        output.append(output_row)

    training = numpy.array(training)
    output = numpy.array(output)

    with open("data.pickle", "wb") as f:
        pickle.dump((words, labels, training, output), f)

tf.keras.backend.clear_session()

model = tf.keras.Sequential()

model.add(keras.layers.Dense(len(training[0])))
model.add(keras.layers.Dense(8))
model.add(keras.layers.Dense(8))
model.add(keras.layers.Dense(len(output[0]), activation="softmax"))

try:
    keras.models.load_model("model.tflearn")
except:
    model.compile(optimizer="adam", loss="categorical_crossentropy", metrics="accuracy")
    model.fit(training, output, epochs=1000, batch_size=8)
    model.save("model.tflearn")

def bag_of_words(s, words):
    bag = [0 for _ in range(len(words))]

    s_words = nltk.word_tokenize(s)
    s_words = [stemmer.stem(word.lower()) for word in s_words]

    for se in s_words:
        for i, w in enumerate(words):
            if w == se:
                bag[i] = 1
    return numpy.array(bag)

def chat():
    print("Ready to chat! (type quit to stop)")
    while True:
        inp = input("You: ")
        if inp.lower() == "quit":
            break

        results = model.predict([bag_of_words(inp, words)])
        print(results)

chat()

When using the debugging mode it tells me the error is the line(inside of chat()):

results = model.predict([bag_of_words(inp, words)])

keras tensorflow python
4 Answers

Ok, so this is not really 100% an answer to your question, but I came across it when googling the same issue with tf.2.2 and the tf.data.Dataset.from_generator function. Since I found the solution, which is not documented in the tf code, I thought I'll post it here for the community to see in the future. output_shapes simply expects the same number of elements as output_types. I.e. it's not enough to say that all output_types=tf.dtypes.float32, rather one needs a type element per shapes element

user_5960460

Also not an exact answer to the question, but somewhere related on the same lines. I got the error: TypeError: Dimension value must be integer or None or have an __index__ method, got 108.0 which I eventually found that if your generator returns matrices then its shape values are obviously integers e.g. X.shape = (64, 16, 108, 192, 3) where X is a numpy matrix returned from the generator. But you might have mistakenly passed ([64, 16, 108.0, 192.0, 3]) as output_shapes argument in tf.data.Dataset.from_generator function, i.e. a floating point value of 108 instead of an integer. Passing integer values in output_shapes fixed the problem for me. Leaving it for the community.

user_43506170

Not sure if you are still interested in an answer since it has been a while. However, I ran into the same problem when I wrote a custom keras layer. The same error occured for me when my input to the layer is expected to be a Tensor but I provided a list (with a single element). Probably replacing

    results = model.predict([bag_of_words(inp, words)])

by

    results = model.predict(bag_of_words(inp, words))

will to the trick

user_149943110

The issue for me was that my generator was yielding lists:

yield (batch_x_data, batch_y_data)

The solution is having the generator yield np arrays or tf tensors:

yield (np.array(batch_x_data), np.array(batch_y_data))
user_77469610
Related