|
{ |
|
"command": "/usr/local/bin/ludwig experiment --config /src/intent-classifier-lstm.yaml --dataset /data/data.json --output_directory /results", |
|
"compute": { |
|
"num_nodes": 1 |
|
}, |
|
"config": { |
|
"backend": null, |
|
"combiner": { |
|
"activation": "relu", |
|
"bias_initializer": "zeros", |
|
"dropout": 0.0, |
|
"fc_layers": null, |
|
"flatten_inputs": false, |
|
"norm": null, |
|
"norm_params": null, |
|
"num_fc_layers": 0, |
|
"output_size": 256, |
|
"residual": false, |
|
"type": "concat", |
|
"use_bias": true, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"defaults": { |
|
"audio": { |
|
"encoder": { |
|
"activation": "relu", |
|
"bias_initializer": "zeros", |
|
"conv_layers": null, |
|
"dropout": 0.0, |
|
"embedding_size": 256, |
|
"embeddings_on_cpu": false, |
|
"embeddings_trainable": true, |
|
"fc_layers": null, |
|
"filter_size": 3, |
|
"max_sequence_length": null, |
|
"norm": null, |
|
"norm_params": null, |
|
"num_conv_layers": null, |
|
"num_fc_layers": null, |
|
"num_filters": 256, |
|
"output_size": 256, |
|
"pool_function": "max", |
|
"pool_size": null, |
|
"pretrained_embeddings": null, |
|
"reduce_output": "sum", |
|
"representation": "dense", |
|
"should_embed": true, |
|
"skip": false, |
|
"type": "parallel_cnn", |
|
"use_bias": true, |
|
"vocab": null, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"preprocessing": { |
|
"audio_file_length_limit_in_s": 7.5, |
|
"computed_fill_value": null, |
|
"fill_value": null, |
|
"in_memory": true, |
|
"missing_value_strategy": "bfill", |
|
"norm": null, |
|
"num_fft_points": null, |
|
"num_filter_bands": 80, |
|
"padding_value": 0.0, |
|
"type": "fbank", |
|
"window_length_in_s": 0.04, |
|
"window_shift_in_s": 0.02, |
|
"window_type": "hamming" |
|
} |
|
}, |
|
"bag": { |
|
"encoder": { |
|
"activation": "relu", |
|
"bias_initializer": "zeros", |
|
"dropout": 0.0, |
|
"embedding_size": 50, |
|
"embeddings_on_cpu": false, |
|
"embeddings_trainable": true, |
|
"fc_layers": null, |
|
"force_embedding_size": false, |
|
"norm": null, |
|
"norm_params": null, |
|
"num_fc_layers": 0, |
|
"output_size": 10, |
|
"pretrained_embeddings": null, |
|
"representation": "dense", |
|
"skip": false, |
|
"type": "embed", |
|
"use_bias": true, |
|
"vocab": null, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"preprocessing": { |
|
"computed_fill_value": "<UNK>", |
|
"fill_value": "<UNK>", |
|
"lowercase": false, |
|
"missing_value_strategy": "fill_with_const", |
|
"most_common": 10000, |
|
"tokenizer": "space" |
|
} |
|
}, |
|
"binary": { |
|
"decoder": { |
|
"bias_initializer": "zeros", |
|
"fc_activation": "relu", |
|
"fc_bias_initializer": "zeros", |
|
"fc_dropout": 0.0, |
|
"fc_layers": null, |
|
"fc_norm": null, |
|
"fc_norm_params": null, |
|
"fc_output_size": 256, |
|
"fc_use_bias": true, |
|
"fc_weights_initializer": "xavier_uniform", |
|
"input_size": null, |
|
"num_fc_layers": 0, |
|
"type": "regressor", |
|
"use_bias": true, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"encoder": { |
|
"skip": false, |
|
"type": "passthrough" |
|
}, |
|
"loss": { |
|
"confidence_penalty": 0, |
|
"positive_class_weight": null, |
|
"robust_lambda": 0, |
|
"type": "binary_weighted_cross_entropy", |
|
"weight": 1.0 |
|
}, |
|
"preprocessing": { |
|
"computed_fill_value": null, |
|
"fallback_true_label": null, |
|
"fill_value": null, |
|
"missing_value_strategy": "fill_with_false" |
|
} |
|
}, |
|
"category": { |
|
"decoder": { |
|
"bias_initializer": "zeros", |
|
"fc_activation": "relu", |
|
"fc_bias_initializer": "zeros", |
|
"fc_dropout": 0.0, |
|
"fc_layers": null, |
|
"fc_norm": null, |
|
"fc_norm_params": null, |
|
"fc_output_size": 256, |
|
"fc_use_bias": true, |
|
"fc_weights_initializer": "xavier_uniform", |
|
"input_size": null, |
|
"num_classes": null, |
|
"num_fc_layers": 0, |
|
"type": "classifier", |
|
"use_bias": true, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"encoder": { |
|
"dropout": 0.0, |
|
"embedding_initializer": null, |
|
"embedding_size": 50, |
|
"embeddings_on_cpu": false, |
|
"embeddings_trainable": true, |
|
"pretrained_embeddings": null, |
|
"skip": false, |
|
"type": "dense", |
|
"vocab": null |
|
}, |
|
"loss": { |
|
"class_similarities": null, |
|
"class_similarities_temperature": 0, |
|
"class_weights": null, |
|
"confidence_penalty": 0, |
|
"robust_lambda": 0, |
|
"type": "softmax_cross_entropy", |
|
"weight": 1.0 |
|
}, |
|
"preprocessing": { |
|
"cache_encoder_embeddings": false, |
|
"computed_fill_value": "<UNK>", |
|
"fill_value": "<UNK>", |
|
"lowercase": false, |
|
"missing_value_strategy": "fill_with_const", |
|
"most_common": 10000 |
|
} |
|
}, |
|
"date": { |
|
"encoder": { |
|
"activation": "relu", |
|
"bias_initializer": "zeros", |
|
"dropout": 0.0, |
|
"embedding_size": 10, |
|
"embeddings_on_cpu": false, |
|
"fc_layers": null, |
|
"norm": null, |
|
"norm_params": null, |
|
"num_fc_layers": 0, |
|
"output_size": 10, |
|
"skip": false, |
|
"type": "embed", |
|
"use_bias": true, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"preprocessing": { |
|
"computed_fill_value": "", |
|
"datetime_format": null, |
|
"fill_value": "", |
|
"missing_value_strategy": "fill_with_const" |
|
} |
|
}, |
|
"h3": { |
|
"encoder": { |
|
"activation": "relu", |
|
"bias_initializer": "zeros", |
|
"dropout": 0.0, |
|
"embedding_size": 10, |
|
"embeddings_on_cpu": false, |
|
"fc_layers": null, |
|
"norm": null, |
|
"norm_params": null, |
|
"num_fc_layers": 0, |
|
"output_size": 10, |
|
"reduce_output": "sum", |
|
"skip": false, |
|
"type": "embed", |
|
"use_bias": true, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"preprocessing": { |
|
"computed_fill_value": 576495936675512319, |
|
"fill_value": 576495936675512319, |
|
"missing_value_strategy": "fill_with_const" |
|
} |
|
}, |
|
"image": { |
|
"augmentation": [], |
|
"decoder": { |
|
"conv_norm": "batch", |
|
"fc_activation": "relu", |
|
"fc_bias_initializer": "zeros", |
|
"fc_dropout": 0.0, |
|
"fc_layers": null, |
|
"fc_norm": null, |
|
"fc_norm_params": null, |
|
"fc_output_size": 256, |
|
"fc_use_bias": true, |
|
"fc_weights_initializer": "xavier_uniform", |
|
"height": null, |
|
"input_size": 1024, |
|
"num_channels": null, |
|
"num_classes": null, |
|
"num_fc_layers": 0, |
|
"type": "unet", |
|
"width": null |
|
}, |
|
"encoder": { |
|
"conv_activation": "relu", |
|
"conv_dropout": 0.0, |
|
"conv_layers": null, |
|
"conv_norm": null, |
|
"conv_norm_params": null, |
|
"conv_use_bias": true, |
|
"dilation": 1, |
|
"fc_activation": "relu", |
|
"fc_bias_initializer": "zeros", |
|
"fc_dropout": 0.0, |
|
"fc_layers": null, |
|
"fc_norm": null, |
|
"fc_norm_params": null, |
|
"fc_use_bias": true, |
|
"fc_weights_initializer": "xavier_uniform", |
|
"groups": 1, |
|
"height": null, |
|
"kernel_size": 3, |
|
"num_channels": null, |
|
"num_conv_layers": null, |
|
"num_fc_layers": 1, |
|
"out_channels": 32, |
|
"output_size": 128, |
|
"padding": "valid", |
|
"padding_mode": "zeros", |
|
"pool_dilation": 1, |
|
"pool_function": "max", |
|
"pool_kernel_size": 2, |
|
"pool_padding": 0, |
|
"pool_stride": null, |
|
"skip": false, |
|
"stride": 1, |
|
"type": "stacked_cnn", |
|
"width": null |
|
}, |
|
"loss": { |
|
"class_similarities": null, |
|
"class_similarities_temperature": 0, |
|
"class_weights": null, |
|
"confidence_penalty": 0, |
|
"robust_lambda": 0, |
|
"type": "softmax_cross_entropy", |
|
"weight": 1.0 |
|
}, |
|
"preprocessing": { |
|
"computed_fill_value": null, |
|
"fill_value": null, |
|
"height": null, |
|
"in_memory": true, |
|
"infer_image_dimensions": true, |
|
"infer_image_max_height": 256, |
|
"infer_image_max_width": 256, |
|
"infer_image_num_channels": true, |
|
"infer_image_num_classes": false, |
|
"infer_image_sample_size": 100, |
|
"missing_value_strategy": "bfill", |
|
"num_channels": null, |
|
"num_classes": null, |
|
"num_processes": 1, |
|
"requires_equal_dimensions": false, |
|
"resize_method": "interpolate", |
|
"standardize_image": null, |
|
"width": null |
|
} |
|
}, |
|
"number": { |
|
"decoder": { |
|
"bias_initializer": "zeros", |
|
"fc_activation": "relu", |
|
"fc_bias_initializer": "zeros", |
|
"fc_dropout": 0.0, |
|
"fc_layers": null, |
|
"fc_norm": null, |
|
"fc_norm_params": null, |
|
"fc_output_size": 256, |
|
"fc_use_bias": true, |
|
"fc_weights_initializer": "xavier_uniform", |
|
"input_size": null, |
|
"num_fc_layers": 0, |
|
"type": "regressor", |
|
"use_bias": true, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"encoder": { |
|
"skip": false, |
|
"type": "passthrough" |
|
}, |
|
"loss": { |
|
"type": "mean_squared_error", |
|
"weight": 1.0 |
|
}, |
|
"preprocessing": { |
|
"computed_fill_value": 0.0, |
|
"computed_outlier_fill_value": 0.0, |
|
"fill_value": 0.0, |
|
"missing_value_strategy": "fill_with_const", |
|
"normalization": "zscore", |
|
"outlier_strategy": null, |
|
"outlier_threshold": 3.0 |
|
} |
|
}, |
|
"sequence": { |
|
"decoder": { |
|
"cell_type": "gru", |
|
"fc_activation": "relu", |
|
"fc_bias_initializer": "zeros", |
|
"fc_dropout": 0.0, |
|
"fc_layers": null, |
|
"fc_norm": null, |
|
"fc_norm_params": null, |
|
"fc_output_size": 256, |
|
"fc_use_bias": true, |
|
"fc_weights_initializer": "xavier_uniform", |
|
"input_size": 256, |
|
"max_sequence_length": null, |
|
"num_fc_layers": 0, |
|
"num_layers": 1, |
|
"reduce_input": "sum", |
|
"type": "generator", |
|
"vocab_size": null |
|
}, |
|
"encoder": { |
|
"dropout": 0.0, |
|
"embedding_size": 256, |
|
"embeddings_on_cpu": false, |
|
"embeddings_trainable": true, |
|
"max_sequence_length": null, |
|
"pretrained_embeddings": null, |
|
"reduce_output": "sum", |
|
"representation": "dense", |
|
"skip": false, |
|
"type": "embed", |
|
"vocab": null, |
|
"weights_initializer": "uniform" |
|
}, |
|
"loss": { |
|
"class_similarities": null, |
|
"class_similarities_temperature": 0, |
|
"class_weights": null, |
|
"confidence_penalty": 0, |
|
"robust_lambda": 0, |
|
"type": "sequence_softmax_cross_entropy", |
|
"unique": false, |
|
"weight": 1.0 |
|
}, |
|
"preprocessing": { |
|
"cache_encoder_embeddings": false, |
|
"computed_fill_value": "<UNK>", |
|
"fill_value": "<UNK>", |
|
"lowercase": false, |
|
"max_sequence_length": 256, |
|
"missing_value_strategy": "fill_with_const", |
|
"most_common": 20000, |
|
"ngram_size": 2, |
|
"padding": "right", |
|
"padding_symbol": "<PAD>", |
|
"sequence_length": null, |
|
"tokenizer": "space", |
|
"unknown_symbol": "<UNK>", |
|
"vocab_file": null |
|
} |
|
}, |
|
"set": { |
|
"decoder": { |
|
"bias_initializer": "zeros", |
|
"fc_activation": "relu", |
|
"fc_bias_initializer": "zeros", |
|
"fc_dropout": 0.0, |
|
"fc_layers": null, |
|
"fc_norm": null, |
|
"fc_norm_params": null, |
|
"fc_output_size": 256, |
|
"fc_use_bias": true, |
|
"fc_weights_initializer": "xavier_uniform", |
|
"input_size": null, |
|
"num_classes": null, |
|
"num_fc_layers": 0, |
|
"type": "classifier", |
|
"use_bias": true, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"encoder": { |
|
"activation": "relu", |
|
"bias_initializer": "zeros", |
|
"dropout": 0.0, |
|
"embedding_size": 50, |
|
"embeddings_on_cpu": false, |
|
"embeddings_trainable": true, |
|
"fc_layers": null, |
|
"norm": null, |
|
"norm_params": null, |
|
"num_fc_layers": 0, |
|
"output_size": 10, |
|
"pretrained_embeddings": null, |
|
"representation": "dense", |
|
"skip": false, |
|
"type": "embed", |
|
"use_bias": true, |
|
"vocab": null, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"loss": { |
|
"class_weights": null, |
|
"type": "sigmoid_cross_entropy", |
|
"weight": 1.0 |
|
}, |
|
"preprocessing": { |
|
"computed_fill_value": "<UNK>", |
|
"fill_value": "<UNK>", |
|
"lowercase": false, |
|
"missing_value_strategy": "fill_with_const", |
|
"most_common": 10000, |
|
"tokenizer": "space" |
|
} |
|
}, |
|
"text": { |
|
"decoder": { |
|
"cell_type": "gru", |
|
"fc_activation": "relu", |
|
"fc_bias_initializer": "zeros", |
|
"fc_dropout": 0.0, |
|
"fc_layers": null, |
|
"fc_norm": null, |
|
"fc_norm_params": null, |
|
"fc_output_size": 256, |
|
"fc_use_bias": true, |
|
"fc_weights_initializer": "xavier_uniform", |
|
"input_size": 256, |
|
"max_sequence_length": null, |
|
"num_fc_layers": 0, |
|
"num_layers": 1, |
|
"reduce_input": "sum", |
|
"type": "generator", |
|
"vocab_size": null |
|
}, |
|
"encoder": { |
|
"activation": "relu", |
|
"bias_initializer": "zeros", |
|
"conv_layers": null, |
|
"dropout": 0.0, |
|
"embedding_size": 256, |
|
"embeddings_on_cpu": false, |
|
"embeddings_trainable": true, |
|
"fc_layers": null, |
|
"filter_size": 3, |
|
"max_sequence_length": null, |
|
"norm": null, |
|
"norm_params": null, |
|
"num_conv_layers": null, |
|
"num_fc_layers": null, |
|
"num_filters": 256, |
|
"output_size": 256, |
|
"pool_function": "max", |
|
"pool_size": null, |
|
"pretrained_embeddings": null, |
|
"reduce_output": "sum", |
|
"representation": "dense", |
|
"should_embed": true, |
|
"skip": false, |
|
"type": "parallel_cnn", |
|
"use_bias": true, |
|
"vocab": null, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"loss": { |
|
"class_similarities": null, |
|
"class_similarities_temperature": 0, |
|
"class_weights": null, |
|
"confidence_penalty": 0, |
|
"robust_lambda": 0, |
|
"type": "sequence_softmax_cross_entropy", |
|
"unique": false, |
|
"weight": 1.0 |
|
}, |
|
"preprocessing": { |
|
"cache_encoder_embeddings": false, |
|
"compute_idf": false, |
|
"computed_fill_value": "<UNK>", |
|
"fill_value": "<UNK>", |
|
"lowercase": false, |
|
"max_sequence_length": 256, |
|
"missing_value_strategy": "fill_with_const", |
|
"most_common": 20000, |
|
"ngram_size": 2, |
|
"padding": "right", |
|
"padding_symbol": "<PAD>", |
|
"pretrained_model_name_or_path": null, |
|
"prompt": { |
|
"retrieval": { |
|
"index_name": null, |
|
"k": 0, |
|
"model_name": null, |
|
"type": null |
|
}, |
|
"task": null, |
|
"template": null |
|
}, |
|
"sequence_length": null, |
|
"tokenizer": "space_punct", |
|
"unknown_symbol": "<UNK>", |
|
"vocab_file": null |
|
} |
|
}, |
|
"timeseries": { |
|
"decoder": { |
|
"activation": null, |
|
"bias_initializer": "zeros", |
|
"clip": null, |
|
"fc_activation": "relu", |
|
"fc_bias_initializer": "zeros", |
|
"fc_dropout": 0.0, |
|
"fc_layers": null, |
|
"fc_norm": null, |
|
"fc_norm_params": null, |
|
"fc_output_size": 256, |
|
"fc_use_bias": true, |
|
"fc_weights_initializer": "xavier_uniform", |
|
"input_size": null, |
|
"multiplier": 1.0, |
|
"num_fc_layers": 0, |
|
"output_size": null, |
|
"type": "projector", |
|
"use_bias": true, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"encoder": { |
|
"activation": "relu", |
|
"bias_initializer": "zeros", |
|
"conv_layers": null, |
|
"dropout": 0.0, |
|
"embedding_size": 256, |
|
"embeddings_on_cpu": false, |
|
"embeddings_trainable": true, |
|
"fc_layers": null, |
|
"filter_size": 3, |
|
"max_sequence_length": null, |
|
"norm": null, |
|
"norm_params": null, |
|
"num_conv_layers": null, |
|
"num_fc_layers": null, |
|
"num_filters": 256, |
|
"output_size": 256, |
|
"pool_function": "max", |
|
"pool_size": null, |
|
"pretrained_embeddings": null, |
|
"reduce_output": "sum", |
|
"representation": "dense", |
|
"should_embed": true, |
|
"skip": false, |
|
"type": "parallel_cnn", |
|
"use_bias": true, |
|
"vocab": null, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"loss": { |
|
"delta": 1.0, |
|
"type": "huber", |
|
"weight": 1.0 |
|
}, |
|
"preprocessing": { |
|
"computed_fill_value": "", |
|
"fill_value": "", |
|
"missing_value_strategy": "fill_with_const", |
|
"padding": "right", |
|
"padding_value": 0.0, |
|
"timeseries_length_limit": 256, |
|
"tokenizer": "space", |
|
"window_size": 0 |
|
} |
|
}, |
|
"vector": { |
|
"decoder": { |
|
"activation": null, |
|
"bias_initializer": "zeros", |
|
"clip": null, |
|
"fc_activation": "relu", |
|
"fc_bias_initializer": "zeros", |
|
"fc_dropout": 0.0, |
|
"fc_layers": null, |
|
"fc_norm": null, |
|
"fc_norm_params": null, |
|
"fc_output_size": 256, |
|
"fc_use_bias": true, |
|
"fc_weights_initializer": "xavier_uniform", |
|
"input_size": null, |
|
"multiplier": 1.0, |
|
"num_fc_layers": 0, |
|
"output_size": null, |
|
"type": "projector", |
|
"use_bias": true, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"encoder": { |
|
"activation": "relu", |
|
"bias_initializer": "zeros", |
|
"dropout": 0.0, |
|
"fc_layers": null, |
|
"input_size": null, |
|
"norm": null, |
|
"norm_params": null, |
|
"num_layers": 1, |
|
"output_size": 256, |
|
"skip": false, |
|
"type": "dense", |
|
"use_bias": true, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"loss": { |
|
"type": "mean_squared_error", |
|
"weight": 1.0 |
|
}, |
|
"preprocessing": { |
|
"computed_fill_value": "", |
|
"fill_value": "", |
|
"missing_value_strategy": "fill_with_const", |
|
"vector_size": null |
|
} |
|
} |
|
}, |
|
"hyperopt": null, |
|
"input_features": [ |
|
{ |
|
"active": true, |
|
"column": "text", |
|
"encoder": { |
|
"activation": "tanh", |
|
"bias_initializer": "zeros", |
|
"bidirectional": true, |
|
"cell_type": "lstm", |
|
"conv_activation": "relu", |
|
"conv_dropout": 0.0, |
|
"conv_layers": null, |
|
"dilation_rate": 1, |
|
"dropout": 0.2, |
|
"embedding_size": 256, |
|
"embeddings_on_cpu": false, |
|
"embeddings_trainable": true, |
|
"fc_activation": "relu", |
|
"fc_dropout": 0.0, |
|
"fc_layers": null, |
|
"filter_size": 5, |
|
"max_sequence_length": null, |
|
"norm": null, |
|
"norm_params": null, |
|
"num_conv_layers": null, |
|
"num_fc_layers": 0, |
|
"num_filters": 256, |
|
"num_rec_layers": 1, |
|
"output_size": 256, |
|
"padding": "same", |
|
"pool_function": "max", |
|
"pool_padding": "same", |
|
"pool_size": null, |
|
"pool_strides": null, |
|
"pretrained_embeddings": null, |
|
"recurrent_activation": "sigmoid", |
|
"recurrent_dropout": 0.0, |
|
"recurrent_initializer": "orthogonal", |
|
"reduce_output": null, |
|
"representation": "dense", |
|
"should_embed": true, |
|
"skip": false, |
|
"state_size": 256, |
|
"strides": 1, |
|
"type": "cnnrnn", |
|
"unit_forget_bias": true, |
|
"use_bias": true, |
|
"vocab": null, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"name": "text", |
|
"preprocessing": { |
|
"cache_encoder_embeddings": false, |
|
"computed_fill_value": "<UNK>", |
|
"fill_value": "<UNK>", |
|
"lowercase": true, |
|
"max_sequence_length": 256, |
|
"missing_value_strategy": "fill_with_const", |
|
"most_common": 20000, |
|
"ngram_size": 3, |
|
"padding": "right", |
|
"padding_symbol": "<PAD>", |
|
"sequence_length": null, |
|
"tokenizer": "ngram", |
|
"unknown_symbol": "<UNK>", |
|
"vocab_file": null |
|
}, |
|
"proc_column": "text_llKKzv", |
|
"tied": null, |
|
"type": "sequence" |
|
} |
|
], |
|
"ludwig_version": "0.10.2.dev", |
|
"model_type": "ecd", |
|
"output_features": [ |
|
{ |
|
"active": true, |
|
"calibration": false, |
|
"column": "intent", |
|
"decoder": { |
|
"bias_initializer": "zeros", |
|
"fc_activation": "relu", |
|
"fc_bias_initializer": "zeros", |
|
"fc_dropout": 0.2, |
|
"fc_layers": null, |
|
"fc_norm": null, |
|
"fc_norm_params": null, |
|
"fc_output_size": 256, |
|
"fc_use_bias": true, |
|
"fc_weights_initializer": "xavier_uniform", |
|
"input_size": null, |
|
"num_classes": null, |
|
"num_fc_layers": 0, |
|
"type": "classifier", |
|
"use_bias": true, |
|
"weights_initializer": "xavier_uniform" |
|
}, |
|
"default_validation_metric": "accuracy", |
|
"dependencies": [], |
|
"input_size": null, |
|
"loss": { |
|
"class_similarities": null, |
|
"class_similarities_temperature": 0, |
|
"class_weights": null, |
|
"confidence_penalty": 0, |
|
"robust_lambda": 0, |
|
"type": "softmax_cross_entropy", |
|
"weight": 1.0 |
|
}, |
|
"name": "intent", |
|
"num_classes": null, |
|
"preprocessing": { |
|
"cache_encoder_embeddings": false, |
|
"computed_fill_value": "<UNK>", |
|
"fill_value": "<UNK>", |
|
"lowercase": false, |
|
"missing_value_strategy": "drop_row", |
|
"most_common": 10000 |
|
}, |
|
"proc_column": "intent_YwJjWN", |
|
"reduce_dependencies": "sum", |
|
"reduce_input": "sum", |
|
"top_k": 3, |
|
"type": "category" |
|
} |
|
], |
|
"preprocessing": { |
|
"global_max_sequence_length": null, |
|
"oversample_minority": null, |
|
"sample_ratio": 1.0, |
|
"sample_size": null, |
|
"split": { |
|
"column": "intent", |
|
"probabilities": [ |
|
0.85, |
|
0.1, |
|
0.05 |
|
], |
|
"type": "stratify" |
|
}, |
|
"undersample_majority": null |
|
}, |
|
"trainer": { |
|
"batch_size": "auto", |
|
"bucketing_field": null, |
|
"checkpoints_per_epoch": 0, |
|
"compile": false, |
|
"early_stop": 0, |
|
"effective_batch_size": "auto", |
|
"enable_gradient_checkpointing": false, |
|
"enable_profiling": false, |
|
"epochs": 10, |
|
"eval_batch_size": null, |
|
"eval_steps": null, |
|
"evaluate_training_set": false, |
|
"gradient_accumulation_steps": "auto", |
|
"gradient_clipping": { |
|
"clipglobalnorm": 0.5, |
|
"clipnorm": null, |
|
"clipvalue": null |
|
}, |
|
"increase_batch_size_eval_metric": "loss", |
|
"increase_batch_size_eval_split": "training", |
|
"increase_batch_size_on_plateau": 0, |
|
"increase_batch_size_on_plateau_patience": 5, |
|
"increase_batch_size_on_plateau_rate": 2.0, |
|
"learning_rate": 0.001, |
|
"learning_rate_scaling": "linear", |
|
"learning_rate_scheduler": { |
|
"decay": "exponential", |
|
"decay_rate": 0.96, |
|
"decay_steps": 10000, |
|
"eta_min": 0, |
|
"reduce_eval_metric": "loss", |
|
"reduce_eval_split": "training", |
|
"reduce_on_plateau": 0, |
|
"reduce_on_plateau_patience": 10, |
|
"reduce_on_plateau_rate": 0.1, |
|
"staircase": false, |
|
"t_0": null, |
|
"t_mult": 1, |
|
"warmup_evaluations": 0, |
|
"warmup_fraction": 0.0 |
|
}, |
|
"max_batch_size": 1099511627776, |
|
"optimizer": { |
|
"amsgrad": false, |
|
"betas": [ |
|
0.9, |
|
0.999 |
|
], |
|
"eps": 1e-08, |
|
"type": "adam", |
|
"weight_decay": 0.0 |
|
}, |
|
"profiler": { |
|
"active": 3, |
|
"repeat": 5, |
|
"skip_first": 0, |
|
"wait": 1, |
|
"warmup": 1 |
|
}, |
|
"regularization_lambda": 0.0, |
|
"regularization_type": "l2", |
|
"should_shuffle": true, |
|
"skip_all_evaluation": false, |
|
"steps_per_checkpoint": 0, |
|
"train_steps": null, |
|
"use_mixed_precision": false, |
|
"validation_field": "intent", |
|
"validation_metric": "accuracy" |
|
} |
|
}, |
|
"data_format": "json", |
|
"dataset": "/data/data.json", |
|
"ludwig_version": "0.10.2.dev", |
|
"random_seed": 42, |
|
"torch_version": "2.0.0+cpu" |
|
} |