acmc's picture
Pushing model (#1)
e8e0235
{
"_name_or_path": "acmc/summarizer_google_bigbird-pegasus-large-pubmed_base_faceted",
"activation_dropout": 0.0,
"activation_function": "gelu_new",
"architectures": [
"BigBirdPegasusForConditionalGeneration"
],
"attention_dropout": 0.0,
"attention_type": "block_sparse",
"block_size": 64,
"bos_token_id": 2,
"classifier_dropout": 0.0,
"d_model": 1024,
"decoder_attention_heads": 16,
"decoder_ffn_dim": 4096,
"decoder_layerdrop": 0.0,
"decoder_layers": 16,
"decoder_start_token_id": 2,
"dropout": 0.1,
"encoder_attention_heads": 16,
"encoder_ffn_dim": 4096,
"encoder_layerdrop": 0.0,
"encoder_layers": 16,
"eos_token_id": 1,
"gradient_checkpointing": false,
"init_std": 0.02,
"is_encoder_decoder": true,
"length_penalty": 0.8,
"max_length": 512,
"max_position_embeddings": 4096,
"model_type": "bigbird_pegasus",
"num_beams": 5,
"num_hidden_layers": 16,
"num_random_blocks": 3,
"pad_token_id": 0,
"scale_embedding": true,
"tokenizer_class": "PegasusTokenizer",
"torch_dtype": "float32",
"transformers_version": "4.30.2",
"use_bias": false,
"use_cache": true,
"vocab_size": 96111
}