polish-roberta-base-8k / config.json
sdadas's picture
Upload 7 files
9eea353 verified
{
"architectures": [
"RobertaForMaskedLM"
],
"auto_map": {
"AutoModel": "sdadas/unpad-impl--modeling_roberta.UnpadRobertaModel",
"AutoModelForCausalLM": "sdadas/unpad-impl--modeling_roberta.UnpadRobertaForCausalLM",
"AutoModelForMultipleChoice": "sdadas/unpad-impl--modeling_roberta.UnpadRobertaForMultipleChoice",
"AutoModelForTokenClassification": "sdadas/unpad-impl--modeling_roberta.UnpadRobertaForTokenClassification",
"AutoModelForQuestionAnswering": "sdadas/unpad-impl--modeling_roberta.UnpadRobertaForQuestionAnswering",
"AutoModelForSequenceClassification": "sdadas/unpad-impl--modeling_roberta.UnpadRobertaForSequenceClassification",
"AutoModelForMaskedLM": "sdadas/unpad-impl--modeling_roberta.UnpadRobertaForMaskedLM"
},
"attention_probs_dropout_prob": 0.1,
"bos_token_id": 0,
"classifier_dropout": null,
"eos_token_id": 2,
"gradient_checkpointing": false,
"hidden_act": "gelu",
"hidden_dropout_prob": 0.1,
"hidden_size": 768,
"initializer_range": 0.02,
"intermediate_size": 3072,
"layer_norm_eps": 1e-05,
"max_position_embeddings": 8194,
"model_type": "roberta",
"num_attention_heads": 12,
"num_hidden_layers": 12,
"pad_token_id": 1,
"position_embedding_type": "absolute",
"sep_token_id": 2,
"torch_dtype": "float32",
"transformers_version": "4.55.4",
"type_vocab_size": 1,
"use_cache": true,
"vocab_size": 128064
}