explosion-testing/gpt-neox-float16

1{2  "_name_or_path": "trl-internal-testing/tiny-random-GPTNeoXForCausalLM-safetensors-sharded",3  "architectures": [4    "GPTNeoXForCausalLM"5  ],6  "attention_bias": true,7  "attention_dropout": 0.0,8  "attention_probs_dropout_prob": 0.1,9  "bos_token_id": 0,10  "classifier_dropout": 0.1,11  "eos_token_id": 0,12  "hidden_act": "gelu",13  "hidden_dropout": 0.0,14  "hidden_dropout_prob": 0.1,15  "hidden_size": 32,16  "initializer_range": 0.02,17  "intermediate_size": 37,18  "is_decoder": true,19  "layer_norm_eps": 1e-05,20  "max_position_embeddings": 512,21  "model_type": "gpt_neox",22  "num_attention_heads": 4,23  "num_hidden_layers": 5,24  "rope_scaling": null,25  "rotary_emb_base": 10000,26  "rotary_pct": 0.25,27  "tie_word_embeddings": false,28  "torch_dtype": "float16",29  "transformers_version": "4.38.2",30  "type_vocab_size": 16,31  "use_cache": true,32  "use_parallel_residual": true,33  "vocab_size": 102434}35