aragpt2-mega / config.json
wissamantoun's picture
Update README.md and config.json, and add aragpt2-large model to ARAGPT2_PRETRAINED_MODEL_ARCHIVE_LIST
18d0924
{
"activation_function": "gelu_new",
"architectures": [
"AraGPT2LMHeadModel"
],
"auto_map": {
"AutoConfig": "configuration_aragpt2.AraGPT2Config",
"AutoModelForCausalLM": "modeling_aragpt2.AraGPT2LMHeadModel",
"AutoModel": "modeling_aragpt2.AraGPT2Model"
},
"attention_probs_dropout_prob": 0.1,
"attn_pdrop": 0.1,
"bos_token_id": 0,
"embd_pdrop": 0.1,
"eos_token_id": 0,
"hidden_act": "gelu",
"hidden_dropout_prob": 0.1,
"initializer_range": 0.014142135623731,
"intermediate_size": 6144,
"layer_norm_epsilon": 1e-05,
"model_type": "aragpt2",
"n_ctx": 1024,
"n_embd": 1536,
"n_head": 24,
"n_inner": null,
"n_layer": 48,
"n_positions": 1024,
"resid_pdrop": 0.1,
"summary_activation": null,
"summary_first_dropout": 0.1,
"summary_proj_to_labels": true,
"summary_type": "cls_index",
"summary_use_proj": true,
"task_specific_params": {
"text-generation": {
"do_sample": true,
"max_length": 50,
"num_beams": 5,
"top_p": 0.95,
"repetition_penalty": 3.0,
"no_repeat_ngram_size": 3
}
},
"vocab_size": 64000,
"tokenizer_class": "GPT2Tokenizer"
}