raxtemur/SONAR_200_text_decoder
11k
1{2 "_name_or_path": "facebook/nllb-200-distilled-1.3B",3 "activation_dropout": 0.0,4 "activation_function": "relu",5 "architectures": [6 "M2M100ForConditionalGeneration"7 ],8 "attention_dropout": 0.1,9 "bos_token_id": 0,10 "d_model": 1024,11 "decoder_attention_heads": 16,12 "decoder_ffn_dim": 8192,13 "decoder_layerdrop": 0,14 "decoder_layers": 24,15 "decoder_start_token_id": 2,16 "dropout": 0.1,17 "encoder_attention_heads": 16,18 "encoder_ffn_dim": 8192,19 "encoder_layerdrop": 0,20 "encoder_layers": 24,21 "eos_token_id": 2,22 "init_std": 0.02,23 "is_encoder_decoder": true,24 "max_length": null,25 "max_position_embeddings": 1024,26 "model_type": "m2m_100",27 "num_hidden_layers": 24,28 "pad_token_id": 1,29 "scale_embedding": true,30 "torch_dtype": "float32",31 "transformers_version": "4.46.3",32 "use_cache": true,33 "vocab_size": 25620634}35 