| { | |
| "architectures": [ | |
| "FlashHeadLlamaForCausalLM" | |
| ], | |
| "attention_bias": false, | |
| "attention_dropout": 0.0, | |
| "auto_map": { | |
| "AutoConfig": "configuration_flash_head_llama.FlashHeadLlamaConfig", | |
| "AutoModelForCausalLM": "modeling_flash_head_llama.FlashHeadLlamaForCausalLM" | |
| }, | |
| "bos_token_id": 128000, | |
| "creation_time": 1753785304.4597948, | |
| "dtype": "bfloat16", | |
| "enforce_equal_cluster_sizes": true, | |
| "eos_token_id": [ | |
| 128001, | |
| 128008, | |
| 128009 | |
| ], | |
| "flash_head_cache_dir": "flash_head_assets", | |
| "flash_head_special_token_ids": [], | |
| "head_dim": 64, | |
| "hidden_act": "silu", | |
| "hidden_size": 2048, | |
| "initializer_range": 0.02, | |
| "intermediate_size": 8192, | |
| "max_position_embeddings": 131072, | |
| "mlp_bias": false, | |
| "model_type": "flash_head_llama", | |
| "n_clusters": 8016, | |
| "num_attention_heads": 32, | |
| "num_hidden_layers": 16, | |
| "num_key_value_heads": 8, | |
| "pretraining_tp": 1, | |
| "rms_norm_eps": 1e-05, | |
| "rope_scaling": { | |
| "factor": 32.0, | |
| "high_freq_factor": 4.0, | |
| "low_freq_factor": 1.0, | |
| "original_max_position_embeddings": 8192, | |
| "rope_type": "llama3" | |
| }, | |
| "rope_theta": 500000.0, | |
| "tie_word_embeddings": true, | |
| "transformers_version": "4.57.0", | |
| "use_cache": true, | |
| "vocab_size": 128256 | |
| } |