mlong-t5-base-sumstew / config.json
Joemgu's picture
Training in progress, step 200
b9a9be5
raw
history blame
1.25 kB
{
"_name_or_path": "Joemgu/mlong-t5-base-sumstew",
"architectures": [
"LongT5ForConditionalGeneration"
],
"d_ff": 2048,
"d_kv": 64,
"d_model": 768,
"decoder_start_token_id": 0,
"dense_act_fn": "gelu_new",
"dropout_rate": 0.1,
"encoder_attention_type": "transient-global",
"eos_token_id": 1,
"feed_forward_proj": "gated-gelu",
"global_block_size": 16,
"initializer_factor": 1.0,
"is_encoder_decoder": true,
"is_gated_act": true,
"layer_norm_epsilon": 1e-06,
"local_radius": 127,
"model_type": "longt5",
"n_positions": 4096,
"num_decoder_layers": 12,
"num_heads": 12,
"num_layers": 12,
"output_past": true,
"pad_token_id": 0,
"relative_attention_max_distance": 128,
"relative_attention_num_buckets": 32,
"task_specific_params": {
"summarization": {
"early_stopping": true,
"encoder_no_repeat_ngram_size": 5,
"length_penalty": 2.0,
"max_length": 256,
"min_length": 16,
"no_repeat_ngram_size": 4,
"num_beams": 4,
"prefix": "Write a title and summarize: ",
"repetition_penalty": 3.5
}
},
"tie_word_embeddings": false,
"torch_dtype": "float32",
"transformers_version": "4.30.2",
"use_cache": false,
"vocab_size": 256384
}