| { | |
| "architectures": [ | |
| "FairseqT5ForConditionalGeneration" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_fairseq_t5.FairseqT5Config", | |
| "AutoModelForSeq2SeqLM": "modeling_fairseq_t5.FairseqT5ForConditionalGeneration" | |
| }, | |
| "d_ff": 3072, | |
| "d_kv": 64, | |
| "d_model": 768, | |
| "decoder_start_token_id": 2, | |
| "dropout_rate": 0.1, | |
| "eos_token_id": 2, | |
| "feed_forward_proj": "relu", | |
| "initializer_factor": 1.0, | |
| "is_encoder_decoder": true, | |
| "layer_norm_epsilon": 1e-05, | |
| "max_positions": 1280, | |
| "model_type": "fairseq_t5", | |
| "num_decoder_layers": 12, | |
| "num_heads": 12, | |
| "num_layers": 12, | |
| "pad_token_id": 1, | |
| "relative_attention_max_distance": 256, | |
| "relative_attention_num_buckets": 128, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.37.2", | |
| "use_cache": true, | |
| "vocab_size": 101265 | |
| } | |