config.json

{
  "activation_function": "gelu_fast",
  "apply_residual_connection_post_layernorm": false,
  "architectures": [
    "MegLMModelForCausalLM"
  ],
  "attention_dropout": 0.0,
  "attn_pdrop": 0.1,
  "auto_map": {
    "AutoConfig": "configuration_meglm.MegLMConfig",
    "AutoModelForCausalLM": "modeling_meglm.MegLMModelForCausalLM",
    "AutoTokenizer": [
      "hf_tokenizer.HFTokenizer",
      null
    ]
  },
  "bos_token_id": 32767,
  "embd_pdrop": 0.1,
  "embed_layernorm": false,
  "eos_token_id": 32767,
  "hidden_dropout": 0.0,
  "hidden_size": 2560,
  "initializer_range": 0.02,
  "layer_norm_epsilon": 1e-05,
  "model_type": "meglm",
  "n_embd": 2560,
  "n_head": 32,
  "n_inner": 10240,
  "n_layer": 32,
  "n_positions": 2048,
  "pos_emb_type": "rope",
  "pretraining_tp": 1,
  "resid_pdrop": 0.1,
  "scale_attn_weights": true,
  "slow_but_exact": false,
  "summary_activation": null,
  "summary_first_dropout": 0.1,
  "summary_proj_to_labels": true,
  "summary_type": "cls_index",
  "summary_use_proj": true,
  "tokenizer_class": "HFTokenizer",
  "transformers_version": "4.31.0",
  "use_cache": true,
  "vocab_size": 32768
}