korca's picture
First version of model and tokenizer
0d3049a
{
"activation_function": "gelu_new",
"architectures": [
"GPT2LMHeadModel"
],
"attn_pdrop": 0.1,
"bos_token_id": 0,
"embd_pdrop": 0.1,
"eos_token_id": 8,
"gradient_checkpointing": false,
"initializer_range": 0.02,
"layer_norm_epsilon": 1e-05,
"model_type": "gpt2",
"n_ctx": 1024,
"n_embd": 1920,
"n_head": 16,
"n_inner": 7680,
"n_layer": 24,
"n_positions": 1024,
"pad_token_id": 8,
"resid_pdrop": 0.1,
"scale_attn_weights": true,
"summary_activation": null,
"summary_first_dropout": 0.1,
"summary_proj_to_labels": true,
"summary_type": "cls_index",
"summary_use_proj": true,
"torch_dtype": null,
"transformers_version": "4.7.0",
"use_cache": true,
"vocab_size": 51200
}