FloatingDuck/zoom_model

1{2  "architectures": [3    "LlamaForCausalLM"4  ],5  "attention_bias": false,6  "attention_dropout": 0.0,7  "bos_token_id": 128000,8  "eos_token_id": 128009,9  "head_dim": 128,10  "hidden_act": "silu",11  "hidden_size": 3072,12  "initializer_range": 0.02,13  "intermediate_size": 8192,14  "max_position_embeddings": 131072,15  "mlp_bias": false,16  "model_type": "llama",17  "num_attention_heads": 24,18  "num_hidden_layers": 28,19  "num_key_value_heads": 8,20  "pad_token_id": 128004,21  "pretraining_tp": 1,22  "quantization_config": {23    "_load_in_4bit": true,24    "_load_in_8bit": false,25    "bnb_4bit_compute_dtype": "bfloat16",26    "bnb_4bit_quant_storage": "uint8",27    "bnb_4bit_quant_type": "nf4",28    "bnb_4bit_use_double_quant": true,29    "llm_int8_enable_fp32_cpu_offload": false,30    "llm_int8_has_fp16_weight": false,31    "llm_int8_skip_modules": [32      "lm_head",33      "multi_modal_projector",34      "merger",35      "modality_projection"36    ],37    "llm_int8_threshold": 6.0,38    "load_in_4bit": true,39    "load_in_8bit": false,40    "quant_method": "bitsandbytes"41  },42  "rms_norm_eps": 1e-05,43  "rope_scaling": {44    "factor": 32.0,45    "high_freq_factor": 4.0,46    "low_freq_factor": 1.0,47    "original_max_position_embeddings": 8192,48    "rope_type": "llama3"49  },50  "rope_theta": 500000.0,51  "tie_word_embeddings": true,52  "torch_dtype": "bfloat16",53  "transformers_version": "4.52.4",54  "unsloth_fixed": true,55  "use_cache": true,56  "vocab_size": 12825657}58