| { | |
| "_name_or_path": null, | |
| "architectures": [ | |
| "EvaByteForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_class": "eva", | |
| "attention_dropout": 0.0, | |
| "auto_map": { | |
| "AutoConfig": "configuration_evabyte.EvaByteConfig", | |
| "AutoModelForCausalLM": "modeling_evabyte.EvaByteForCausalLM" | |
| }, | |
| "bos_token_id": 1, | |
| "chunk_size": 16, | |
| "eos_token_id": 2, | |
| "fp32_ln": true, | |
| "fp32_logits": true, | |
| "fp32_skip_add": false, | |
| "hidden_act": "silu", | |
| "hidden_size": 5120, | |
| "init_cutoff_factor": null, | |
| "init_fn": "v2", | |
| "init_std": 0.01275, | |
| "initializer_range": 0.01275, | |
| "intermediate_size": 16384, | |
| "lazy_init": true, | |
| "max_position_embeddings": 8192, | |
| "max_seq_length": 8192, | |
| "mixedp_attn": true, | |
| "model_type": "evabyte", | |
| "norm_add_unit_offset": true, | |
| "num_attention_heads": 40, | |
| "num_chunks": null, | |
| "num_hidden_layers": 40, | |
| "num_key_value_heads": 40, | |
| "num_pred_heads": 1, | |
| "pad_token_id": 0, | |
| "return_dict": false, | |
| "rms_norm_eps": 1e-06, | |
| "rope_scaling": null, | |
| "rope_theta": 100000.0, | |
| "tie_word_embeddings": false, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "4.47.1", | |
| "use_cache": true, | |
| "vocab_size": 65536, | |
| "window_size": 2048 | |
| } |