gpt-j-random-tinier / config.json
kirisame's picture
add model
40a9146
{
"activation_function": "gelu_new",
"architectures": [
"GPTJForCausalLM"
],
"attn_pdrop": 0.0,
"bos_token_id": 50256,
"embd_pdrop": 0.0,
"eos_token_id": 50256,
"initializer_range": 0.02,
"layer_norm_epsilon": 1e-05,
"model_type": "gptj",
"n_ctx": 2048,
"n_embd": 128,
"n_head": 2,
"n_inner": null,
"n_layer": 2,
"n_positions": 2048,
"resid_pdrop": 0.0,
"rotary_dim": 64,
"scale_attn_weights": true,
"torch_dtype": "float32",
"transformers_version": "4.11.0.dev0",
"use_cache": true,
"vocab_size": 50400
}