doubledsbv/ALMA-13B-R-AWQ

1{2  "_name_or_path": "/content/ALMA-13B-R",3  "architectures": [4    "LlamaForCausalLM"5  ],6  "attention_bias": false,7  "attention_dropout": 0.0,8  "bos_token_id": 1,9  "eos_token_id": 2,10  "hidden_act": "silu",11  "hidden_size": 5120,12  "initializer_range": 0.02,13  "intermediate_size": 13824,14  "max_length": 512,15  "max_position_embeddings": 4096,16  "model_type": "llama",17  "num_attention_heads": 40,18  "num_hidden_layers": 40,19  "num_key_value_heads": 40,20  "pad_token_id": 0,21  "pretraining_tp": 1,22  "quantization_config": {23    "bits": 4,24    "group_size": 128,25    "modules_to_not_convert": null,26    "quant_method": "awq",27    "version": "gemm",28    "zero_point": true29  },30  "rms_norm_eps": 1e-05,31  "rope_scaling": null,32  "rope_theta": 10000.0,33  "tie_word_embeddings": false,34  "torch_dtype": "float16",35  "transformers_version": "4.38.0.dev0",36  "use_cache": true,37  "vocab_size": 3200038}39