Can not retrive encodings about tweet2vec

[josh-gree]

Hi, thanks for making this available! I really would like to make use of the tweet embeddings but I am getting the same error as above. I will try to fix myself but would be great if you could help since I believe its probably just an error in the config variables...My error logs are below;

2017-03-02 20:20:47,445 - utils - INFO - Loading Data from: /res/tweets.json
2017-03-02 20:20:47,445 - utils - ERROR - Could not read data:[Errno 2] No such file or directory: '/res/tweets.json'
2017-03-02 20:20:47,445 - utils - INFO - Loading Data from path: /home/josh/Tweet2Vec/all_data/models/tweet2vec/new_model//encoding_new_model_3.npy

[jagsileo]

I did not specify any output file, i.e., I removed the part -o /res/outputfilename.npy, it created a folder inside all_data called outputs

[dupsys]

hi,
In your code specifically in Tweet2vec-master/model/t2v.py after the second LSTM layer, please can you to explain future why the used of Reshape upon the output of the LSTM layer? I have this in my code:

Second LSTM for encoding and decoding
l_decoder_1 = LSTM(256, return_sequences=True)(l_in_rep)
l_decoder_2 = LSTM(256, return_sequences=True)(l_decoder_1)
fc_layer = Dense(vocab_size,activation='relu')(l_decoder_2)
drop1_out = Dropout(0.5)(fc_layer)
fc_layer = Dense(vocab_size,activation='softmax')(drop1_out)

Thank you.