Deserialize header failed: 1.lstmf

[Coure 2011]

I need to train the default eng data, so that it can also recognize new characters. I created box files and lstm files and when running cmd:

lstmtraining \

  --model_output output/eng_latin \

  --continue_from "/c/Program Files/Tesseract-OCR/ tessdata/eng.lstm" \

  --append_index 5 \

  --net_spec "[Lfx192 O1c129]" \

  --traineddata "/c/Program Files/Tesseract-OCR/tessdata/eng.traineddata" \

  --train_listfile training/training_files.txt \

  --max_iterations 400

getting error

Loaded file C:/Program Files/Tesseract-OCR/tessdata/eng.lstm,

unpacking... Warning: LSTMTrainer deserialized an LSTMRecognizer!

Continuing from C:/Program Files/Tesseract-OCR/tessdata/eng.lstm

Appending a new network to an old one!!Warning: given outputs 129 not equal to unicharset of 111.

Num outputs,weights in Series: Lfx192:192, 221952 Fc111:111, 21423 Total weights = 243375 Built network:[1,36,0,1[C3,3Ft16]Mp3,3TxyLfys64Lfx96RxLrx96Lfx192Fc111] from request [Lfx192 O1c129]

Training parameters: Debug interval = 0, weights = 0.1, learning rate = 0.001, momentum=0.5 null char=110

Deserialize header failed: 1.lstmf

Deserialize header failed: 2.lstmf

Deserialize header failed: 3.lstmf

Deserialize header failed: 4.lstmf

Deserialize header failed: 5.lstmf

Load of page 0 failed! Load of images failed!!

Files data: https://wormhole.app/X6mPda#lT3aG2Jm9u2QquNRyIruMA

Note: I am on windows