Abc7347/clone4bit

1{2  "architectures": [3    "Gemma3ForCausalLM"4  ],5  "attention_bias": false,6  "attention_dropout": 0.0,7  "attn_logit_softcapping": null,8  "bos_token_id": 2,9  "cache_implementation": "hybrid",10  "eos_token_id": 106,11  "final_logit_softcapping": null,12  "head_dim": 128,13  "hidden_activation": "gelu_pytorch_tanh",14  "hidden_size": 5376,15  "initializer_range": 0.02,16  "intermediate_size": 21504,17  "max_position_embeddings": 131072,18  "model_type": "gemma3_text",19  "num_attention_heads": 32,20  "num_hidden_layers": 62,21  "num_key_value_heads": 16,22  "pad_token_id": 0,23  "quantization_config": {24    "_load_in_4bit": true,25    "_load_in_8bit": false,26    "bnb_4bit_compute_dtype": "bfloat16",27    "bnb_4bit_quant_storage": "uint8",28    "bnb_4bit_quant_type": "nf4",29    "bnb_4bit_use_double_quant": true,30    "llm_int8_enable_fp32_cpu_offload": false,31    "llm_int8_has_fp16_weight": false,32    "llm_int8_skip_modules": [33      "lm_head",34      "multi_modal_projector",35      "merger",36      "modality_projection"37    ],38    "llm_int8_threshold": 6.0,39    "load_in_4bit": true,40    "load_in_8bit": false,41    "quant_method": "bitsandbytes"42  },43  "query_pre_attn_scalar": 168,44  "rms_norm_eps": 1e-06,45  "rope_local_base_freq": 10000,46  "rope_scaling": {47    "factor": 8.0,48    "rope_type": "linear"49  },50  "rope_theta": 1000000,51  "sliding_window": 1024,52  "sliding_window_pattern": 6,53  "torch_dtype": "bfloat16",54  "transformers_version": "4.52.1",55  "unsloth_fixed": true,56  "use_cache": true,57  "vocab_size": 26214458}59