OpenMatch/AAR-ANCE
09
1{2 "_name_or_path": "checkpoints/t5-ance",3 "architectures": [4 "T5Model"5 ],6 "d_ff": 3072,7 "d_kv": 64,8 "d_model": 768,9 "decoder_start_token_id": 0,10 "dense_act_fn": "relu",11 "dropout_rate": 0.1,12 "eos_token_id": 1,13 "feed_forward_proj": "relu",14 "id2label": {15 "0": "LABEL_0"16 },17 "initializer_factor": 1.0,18 "is_encoder_decoder": true,19 "is_gated_act": false,20 "label2id": {21 "LABEL_0": 022 },23 "layer_norm_epsilon": 1e-06,24 "model_type": "t5",25 "n_positions": 512,26 "num_decoder_layers": 12,27 "num_heads": 12,28 "num_layers": 12,29 "output_past": true,30 "pad_token_id": 0,31 "relative_attention_max_distance": 128,32 "relative_attention_num_buckets": 32,33 "task_specific_params": {34 "summarization": {35 "early_stopping": true,36 "length_penalty": 2.0,37 "max_length": 200,38 "min_length": 30,39 "no_repeat_ngram_size": 3,40 "num_beams": 4,41 "prefix": "summarize: "42 },43 "translation_en_to_de": {44 "early_stopping": true,45 "max_length": 300,46 "num_beams": 4,47 "prefix": "translate English to German: "48 },49 "translation_en_to_fr": {50 "early_stopping": true,51 "max_length": 300,52 "num_beams": 4,53 "prefix": "translate English to French: "54 },55 "translation_en_to_ro": {56 "early_stopping": true,57 "max_length": 300,58 "num_beams": 4,59 "prefix": "translate English to Romanian: "60 }61 },62 "torch_dtype": "float32",63 "transformers_version": "4.23.1",64 "use_cache": true,65 "vocab_size": 3212866}67 