CoolFace
Modelpublic

RichardErkhov/microsoft_-_phi-2-4bits

sourceHugging Faceupdated 2y agoView on Hugging Face
0likes22downloads
config.json46 linesDownload Raw Back to root
1{2  "_name_or_path": "phi-2",3  "architectures": [4    "PhiForCausalLM"5  ],6  "attention_dropout": 0.0,7  "bos_token_id": 50256,8  "embd_pdrop": 0.0,9  "eos_token_id": 50256,10  "hidden_act": "gelu_new",11  "hidden_size": 2560,12  "initializer_range": 0.02,13  "intermediate_size": 10240,14  "layer_norm_eps": 1e-05,15  "max_position_embeddings": 2048,16  "model_type": "phi",17  "num_attention_heads": 32,18  "num_hidden_layers": 32,19  "num_key_value_heads": 32,20  "partial_rotary_factor": 0.4,21  "qk_layernorm": false,22  "quantization_config": {23    "_load_in_4bit": true,24    "_load_in_8bit": false,25    "bnb_4bit_compute_dtype": "float32",26    "bnb_4bit_quant_storage": "uint8",27    "bnb_4bit_quant_type": "fp4",28    "bnb_4bit_use_double_quant": false,29    "llm_int8_enable_fp32_cpu_offload": false,30    "llm_int8_has_fp16_weight": false,31    "llm_int8_skip_modules": null,32    "llm_int8_threshold": 6.0,33    "load_in_4bit": true,34    "load_in_8bit": false,35    "quant_method": "bitsandbytes"36  },37  "resid_pdrop": 0.1,38  "rope_scaling": null,39  "rope_theta": 10000.0,40  "tie_word_embeddings": false,41  "torch_dtype": "float16",42  "transformers_version": "4.39.3",43  "use_cache": true,44  "vocab_size": 5120045}46