| { | |
| "architectures": [ | |
| "NeedleForCausalLM" | |
| ], | |
| "auto_map": { | |
| "AutoConfig": "configuration_needle.NeedleConfig", | |
| "AutoModel": "modeling_needle.NeedleModel", | |
| "AutoModelForCausalLM": "modeling_needle.NeedleForCausalLM", | |
| "AutoModelForSeq2SeqLM": "modeling_needle.NeedleForCausalLM" | |
| }, | |
| "bos_token_id": 2, | |
| "d_model": 512, | |
| "decoder_start_token_id": 1, | |
| "eos_token_id": 1, | |
| "hidden_size": 512, | |
| "is_encoder_decoder": true, | |
| "model_type": "needle", | |
| "num_attention_heads": 8, | |
| "num_decoder_layers": 8, | |
| "num_encoder_layers": 12, | |
| "num_heads": 8, | |
| "num_hidden_layers": 8, | |
| "num_key_value_heads": 4, | |
| "num_kv_heads": 4, | |
| "pad_token_id": 0, | |
| "rms_norm_eps": 1e-06, | |
| "rope_theta": 10000.0, | |
| "tie_word_embeddings": true, | |
| "torch_dtype": "bfloat16", | |
| "transformers_version": "5.5.4", | |
| "unk_token_id": 3, | |
| "vocab_size": 8192 | |
| } | |