| { |
| "attention_bias": false, |
| "attention_dropout": 0.0, |
| "bos_token_id": 128000, |
| "eos_token_id": 128001, |
| "head_dim": 64, |
| "hidden_act": "silu", |
| "hidden_size": 256, |
| "initializer_range": 0.02, |
| "intermediate_size": 768, |
| "max_position_embeddings": 8192, |
| "mlp_bias": false, |
| "model_type": "llama", |
| "num_attention_heads": 4, |
| "num_hidden_layers": 4, |
| "num_key_value_heads": 4, |
| "pretraining_tp": 1, |
| "rms_norm_eps": 1e-05, |
| "rope_scaling": null, |
| "rope_theta": 10000.0, |
| "tie_word_embeddings": false, |
| "transformers_version": "4.57.3", |
| "use_cache": true, |
| "vocab_size": 100007 |
| } |
|
|
