{ "_name_or_path": "/raid/rpa020/D1", "apply_residual_connection_post_layernorm": false, "architectures": [ "BloomForCausalLM" ], "attention_dropout": 0.0, "bos_token_id": 1, "d_model": 1024, "eos_token_id": 2, "hidden_dropout": 0.0, "hidden_size": 64, "initializer_range": 0.02, "layer_norm_epsilon": 1e-05, "max_position_embeddings": 512, "model_type": "bloom", "n_head": 8, "n_layer": 2, "num_decoder_layers": 16, "num_encoder_layers": 16, "pretraining_tp": 1, "slow_but_exact": false, "torch_dtype": "float32", "transformers_version": "4.41.2", "use_cache": true, "vocab_size": 96103 }