config.json

{
  "architectures": [
    "MegatronGPTForCausalLM"
  ],
  "attention_dropout": 0.0,
  "auto_map": {
    "AutoConfig": "configuration_megatron_gpt.MegatronGPTConfig",
    "AutoModelForCausalLM": "modeling_megatron_gpt.MegatronGPTForCausalLM"
  },
  "bias": true,
  "bos_token_id": 2,
  "classifier_dropout": 0.0,
  "eos_token_id": 2,
  "hidden_act": "gelu",
  "hidden_dropout": 0.0,
  "hidden_size": 4096,
  "initializer_range": 0.01,
  "intermediate_size": 10880,
  "layer_norm_eps": 1e-05,
  "max_position_embeddings": 2048,
  "model_type": "megatron_gpt",
  "normalization": "layernorm1p",
  "normalize_attention_scores": true,
  "num_attention_heads": 32,
  "num_hidden_layers": 32,
  "pad_token_id": 0,
  "rope_scaling": null,
  "rotary_emb_base": 10000,
  "rotary_pct": 0.5,
  "self_attention_relative_position_bias": true,
  "tie_word_embeddings": false,
  "tokenizer_class": "GPT2Tokenizer",
  "torch_dtype": "float16",
  "transformers_version": "4.31.0",
  "use_cache": true,
  "vocab_size": 56064
}