| { | |
| "_name_or_path": "/extra/luca.capone/babylmchallange25/ituned_models/llama117M_it_switch_wiki/results_best_switch_wiki", | |
| "architectures": [ | |
| "LlamaForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "attention_probs_dropout_prob": 0.1, | |
| "bos_token_id": 0, | |
| "eos_token_id": 1, | |
| "head_dim": 64, | |
| "hidden_act": "silu", | |
| "hidden_dropout_prob": 0.1, | |
| "hidden_size": 704, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 2816, | |
| "max_position_embeddings": 6144, | |
| "mlp_bias": false, | |
| "model_type": "llama", | |
| "num_attention_heads": 11, | |
| "num_hidden_layers": 12, | |
| "num_key_value_heads": 11, | |
| "pad_token_id": 3, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-06, | |
| "rope_scaling": { | |
| "factor": 2.0, | |
| "rope_type": "linear", | |
| "type": "linear" | |
| }, | |
| "rope_theta": 10000.0, | |
| "tie_word_embeddings": true, | |
| "torch_dtype": "float32", | |
| "transformers_version": "4.49.0", | |
| "unk_token_id": 2, | |
| "use_cache": true, | |
| "vocab_size": 32000 | |
| } | |