2021-08-05 17:56:49 +00:00
|
|
|
{
|
|
|
|
|
"activation_function": "gelu_new",
|
|
|
|
|
"architectures": [
|
|
|
|
|
"GPTJForCausalLM"
|
|
|
|
|
],
|
|
|
|
|
"attention_dropout": 0,
|
|
|
|
|
"bos_token_id": 50256,
|
|
|
|
|
"embed_dropout": 0,
|
|
|
|
|
"eos_token_id": 50256,
|
2021-08-05 20:06:46 +00:00
|
|
|
"jax": true,
|
|
|
|
|
"rotary": true,
|
|
|
|
|
"rotary_dim": 64,
|
2021-08-05 17:56:49 +00:00
|
|
|
"gradient_checkpointing": false,
|
2021-08-05 20:05:38 +00:00
|
|
|
"n_embd": 4096,
|
2021-08-05 17:56:49 +00:00
|
|
|
"initializer_range": 0.02,
|
|
|
|
|
"intermediate_size": null,
|
|
|
|
|
"layer_norm_epsilon": 1e-05,
|
2021-08-05 20:05:38 +00:00
|
|
|
"n_positions": 2048,
|
2021-08-05 17:56:49 +00:00
|
|
|
"model_type": "gptj",
|
2021-08-05 20:05:38 +00:00
|
|
|
"n_head": 16,
|
|
|
|
|
"n_layer": 28,
|
2021-08-05 17:56:49 +00:00
|
|
|
"rotary_dim": 64,
|
|
|
|
|
"summary_activation": null,
|
|
|
|
|
"summary_first_dropout": 0.1,
|
|
|
|
|
"summary_proj_to_labels": true,
|
|
|
|
|
"summary_type": "cls_index",
|
|
|
|
|
"summary_use_proj": true,
|
|
|
|
|
"transformers_version": "4.10.0.dev0",
|
|
|
|
|
"use_cache": true,
|
|
|
|
|
"vocab_size": 50400,
|
|
|
|
|
"window_size": 256,
|
|
|
|
|
"tokenizer_class": "GPT2Tokenizer",
|
|
|
|
|
"task_specific_params": {
|
|
|
|
|
"text-generation": {
|
|
|
|
|
"do_sample": true,
|
|
|
|
|
"temperature": 1.0,
|
|
|
|
|
"max_length": 50
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
