RichardErkhov/LeroyDyer_-_SpydazWebAI_VisionEncoderDecoderModel_Mini548m-8bits

1{2  "_name_or_path": "SpydazWebAI_VisionEncoderDecoderModel_Mini548m",3  "architectures": [4    "MistralForCausalLM"5  ],6  "attention_dropout": 0.0,7  "bos_token_id": 1,8  "eos_token_id": 2,9  "head_dim": 32,10  "hidden_act": "silu",11  "hidden_size": 1024,12  "initializer_range": 0.02,13  "intermediate_size": 4096,14  "max_position_embeddings": 32768,15  "model_type": "mistral",16  "num_attention_heads": 32,17  "num_hidden_layers": 12,18  "num_key_value_heads": 8,19  "quantization_config": {20    "_load_in_4bit": false,21    "_load_in_8bit": true,22    "bnb_4bit_compute_dtype": "float32",23    "bnb_4bit_quant_storage": "uint8",24    "bnb_4bit_quant_type": "fp4",25    "bnb_4bit_use_double_quant": false,26    "llm_int8_enable_fp32_cpu_offload": false,27    "llm_int8_has_fp16_weight": false,28    "llm_int8_skip_modules": null,29    "llm_int8_threshold": 6.0,30    "load_in_4bit": false,31    "load_in_8bit": true,32    "quant_method": "bitsandbytes"33  },34  "rms_norm_eps": 1e-06,35  "rope_theta": 10000.0,36  "sliding_window": 32,37  "tie_word_embeddings": false,38  "torch_dtype": "float16",39  "transformers_version": "4.47.1",40  "use_cache": true,41  "vocab_size": 3200942}43