amazon-sagemaker-community/encoder_decoder_es

1{2  "architectures": [3    "EncoderDecoderModel"4  ],5  "decoder": {6    "_name_or_path": "bertin-project/bertin-roberta-base-spanish",7    "add_cross_attention": true,8    "architectures": [9      "RobertaForMaskedLM"10    ],11    "attention_probs_dropout_prob": 0.1,12    "bad_words_ids": null,13    "bos_token_id": 0,14    "chunk_size_feed_forward": 0,15    "classifier_dropout": null,16    "cross_attention_hidden_size": null,17    "decoder_start_token_id": null,18    "diversity_penalty": 0.0,19    "do_sample": false,20    "early_stopping": false,21    "encoder_no_repeat_ngram_size": 0,22    "eos_token_id": 2,23    "finetuning_task": null,24    "forced_bos_token_id": null,25    "forced_eos_token_id": null,26    "gradient_checkpointing": false,27    "hidden_act": "gelu",28    "hidden_dropout_prob": 0.1,29    "hidden_size": 768,30    "id2label": {31      "0": "LABEL_0",32      "1": "LABEL_1"33    },34    "initializer_range": 0.02,35    "intermediate_size": 3072,36    "is_decoder": true,37    "is_encoder_decoder": false,38    "label2id": {39      "LABEL_0": 0,40      "LABEL_1": 141    },42    "layer_norm_eps": 1e-05,43    "length_penalty": 1.0,44    "max_length": 20,45    "max_position_embeddings": 514,46    "min_length": 0,47    "model_type": "roberta",48    "no_repeat_ngram_size": 0,49    "num_attention_heads": 12,50    "num_beam_groups": 1,51    "num_beams": 1,52    "num_hidden_layers": 12,53    "num_return_sequences": 1,54    "output_attentions": false,55    "output_hidden_states": false,56    "output_scores": false,57    "pad_token_id": 1,58    "position_embedding_type": "absolute",59    "prefix": null,60    "problem_type": null,61    "pruned_heads": {},62    "remove_invalid_values": false,63    "repetition_penalty": 1.0,64    "return_dict": true,65    "return_dict_in_generate": false,66    "sep_token_id": null,67    "task_specific_params": null,68    "temperature": 1.0,69    "tie_encoder_decoder": false,70    "tie_word_embeddings": true,71    "tokenizer_class": null,72    "top_k": 50,73    "top_p": 1.0,74    "torch_dtype": null,75    "torchscript": false,76    "transformers_version": "4.12.3",77    "type_vocab_size": 1,78    "use_bfloat16": false,79    "use_cache": true,80    "vocab_size": 5026581  },82  "decoder_start_token_id": 0,83  "early_stopping": true,84  "encoder": {85    "_name_or_path": "bertin-project/bertin-roberta-base-spanish",86    "add_cross_attention": false,87    "architectures": [88      "RobertaForMaskedLM"89    ],90    "attention_probs_dropout_prob": 0.1,91    "bad_words_ids": null,92    "bos_token_id": 0,93    "chunk_size_feed_forward": 0,94    "classifier_dropout": null,95    "cross_attention_hidden_size": null,96    "decoder_start_token_id": null,97    "diversity_penalty": 0.0,98    "do_sample": false,99    "early_stopping": false,100    "encoder_no_repeat_ngram_size": 0,101    "eos_token_id": 2,102    "finetuning_task": null,103    "forced_bos_token_id": null,104    "forced_eos_token_id": null,105    "gradient_checkpointing": false,106    "hidden_act": "gelu",107    "hidden_dropout_prob": 0.1,108    "hidden_size": 768,109    "id2label": {110      "0": "LABEL_0",111      "1": "LABEL_1"112    },113    "initializer_range": 0.02,114    "intermediate_size": 3072,115    "is_decoder": false,116    "is_encoder_decoder": false,117    "label2id": {118      "LABEL_0": 0,119      "LABEL_1": 1120    },121    "layer_norm_eps": 1e-05,122    "length_penalty": 1.0,123    "max_length": 20,124    "max_position_embeddings": 514,125    "min_length": 0,126    "model_type": "roberta",127    "no_repeat_ngram_size": 0,128    "num_attention_heads": 12,129    "num_beam_groups": 1,130    "num_beams": 1,131    "num_hidden_layers": 12,132    "num_return_sequences": 1,133    "output_attentions": false,134    "output_hidden_states": false,135    "output_scores": false,136    "pad_token_id": 1,137    "position_embedding_type": "absolute",138    "prefix": null,139    "problem_type": null,140    "pruned_heads": {},141    "remove_invalid_values": false,142    "repetition_penalty": 1.0,143    "return_dict": true,144    "return_dict_in_generate": false,145    "sep_token_id": null,146    "task_specific_params": null,147    "temperature": 1.0,148    "tie_encoder_decoder": false,149    "tie_word_embeddings": true,150    "tokenizer_class": null,151    "top_k": 50,152    "top_p": 1.0,153    "torch_dtype": null,154    "torchscript": false,155    "transformers_version": "4.12.3",156    "type_vocab_size": 1,157    "use_bfloat16": false,158    "use_cache": true,159    "vocab_size": 50265160  },161  "eos_token_id": 2,162  "is_encoder_decoder": true,163  "length_penalty": 2.0,164  "max_length": 128,165  "min_length": 16,166  "model_type": "encoder-decoder",167  "no_repeat_ngram_size": 3,168  "num_beams": 4,169  "pad_token_id": 1,170  "repetition_penalty": 3.0,171  "tie_encoder_decoder": true,172  "torch_dtype": "float32",173  "transformers_version": null,174  "vocab_size": 50265175}176