Closed rohitrango closed 3 years ago
Hi,
I'm trying to learn a neural network using a custom data set. I was able to convert the pgn files into the text file using the `pgn_to_plain.py' file.
Now, I'm trying to convert it into bin. However, it doesn't work. Here is my output:
(chessvenv) rohit@mypc:~/chessbot/Stockfish/src$ ./stockfish Stockfish 221220 by the Stockfish developers (see AUTHORS file) info string Loaded eval file nn-c3ca321c51c9.nnue learn convert_bin converted_pgn.txt output_file_name pgn_bin.bin INFO: Unknown option: convert_bin. Ignoring. INFO: Unknown option: converted_pgn.txt. Ignoring. INFO: Unknown option: output_file_name. Ignoring. INFO: Unknown option: pgn_bin.bin. Ignoring. INFO: Executing learn command INFO: Input files: INFO: Parameters: - validation count : 2000 - epochs : 18446744073709551615 - epochs * minibatch size : 18446744073708551616 - eval_limit : 32000 - save_only_once : false - shuffle on read : true - Loss Function : ELMO_METHOD(WCSC27) - minibatch size : 1000000 - nn_batch_size : 1000 - nn_options : - learning rate : 1 - max_grad : 1 - use draws in training : 1 - use draws in validation : 1 - skip repeated positions : 1 - winning prob coeff : 0.00276753 - use_wdl : 0 - src_score_min_value : 0 - src_score_max_value : 1 - dest_score_min_value : 0 - dest_score_max_value : 1 - reduction_gameply : 1 - elmo_lambda_low : 1 - elmo_lambda_high : 1 - elmo_lambda_limit : 32000 - eval_save_interval : 100000000 sfens - loss_output_interval : 1000000 sfens - sfen_read_size : 10000000 - thread_buffer_size : 10000 - seed : - verbose : false - learning rate scheduling : newbob with decay - newbob_decay : 0.5 - newbob_num_trials : 4 INFO: Started initialization. INFO (initialize_training): Initializing NN training for Features=HalfKP(Friend)[41024->256x2],Network=AffineTransform[1<-32](ClippedReLU[32](AffineTransform[32<-32](ClippedReLU[32](AffineTransform[32<-512](InputSlice[512(0:512)]))))) Layers: - 0 - HalfKP(Friend)[41024->256x2] - 1 - InputSlice[512(0:512)] - 2 - AffineTransform[32<-512] - 3 - ClippedReLU[32] - 4 - AffineTransform[32<-32] - 5 - ClippedReLU[32] - 6 - AffineTransform[1<-32] Factorizers: - Factorizer<HalfKP(Friend)> -> HalfK, P, HalfRelativeKP Finished initialization. INFO (save_eval): Saving current evaluation file in evalsave/original INFO (save_eval): Finished saving evaluation file in evalsave/original INFO (sfen_reader): End of files. INFO (sfen_reader): End of files. info string NNUE evaluation using nn-c3ca321c51c9.nnue enabled ERROR (sfen_reader): Reading failed. INFO (learn): Error reading validation data. Read 0 out of 2000 ^C
Where do I go from here? Seems like there is no command convert_bin in the learn command.
convert_bin
There is no such thing as convert_bin
See here for conversion docs: https://github.com/nodchip/Stockfish/blob/master/docs/convert.md. You'll need to change the extension from .txt to .plain
Hi,
I'm trying to learn a neural network using a custom data set. I was able to convert the pgn files into the text file using the `pgn_to_plain.py' file.
Now, I'm trying to convert it into bin. However, it doesn't work. Here is my output:
Where do I go from here? Seems like there is no command
convert_bin
in the learn command.