kaizerBox/ReFormer-summarization
032
1{2 "architectures": [3 "ReformerModelWithLMHead"4 ],5 "attention_head_size": 64,6 "attn_layers": [7 "local",8 "lsh",9 "local",10 "lsh",11 "local",12 "lsh"13 ],14 "axial_norm_std": 1.0,15 "axial_pos_embds": true,16 "axial_pos_embds_dim": [17 22,18 2219 ],20 "axial_pos_shape": [21 32,22 3223 ],24 "chunk_size_lm_head": 0,25 "classifier_dropout": null,26 "eos_token_id": 50256,27 "feed_forward_size": 64,28 "hash_seed": null,29 "hidden_act": "relu",30 "hidden_dropout_prob": 0.0,31 "hidden_size": 44,32 "initializer_range": 0.02,33 "is_decoder": true,34 "layer_norm_eps": 1e-12,35 "local_attention_probs_dropout_prob": 0.0,36 "local_attn_chunk_length": 64,37 "local_num_chunks_after": 0,38 "local_num_chunks_before": 1,39 "lsh_attention_probs_dropout_prob": 0.0,40 "lsh_attn_chunk_length": 64,41 "lsh_num_chunks_after": 0,42 "lsh_num_chunks_before": 1,43 "max_position_embeddings": 1024,44 "model_type": "reformer",45 "num_attention_heads": 3,46 "num_buckets": 32,47 "num_hashes": 1,48 "num_hidden_layers": 6,49 "pad_token_id": 50257,50 "tie_word_embeddings": false,51 "torch_dtype": "float32",52 "transformers_version": "4.35.2",53 "use_cache": true,54 "vocab_size": 5025955}56 