mt5 config
Browse files- config.json +4 -12
config.json
CHANGED
|
@@ -1,32 +1,24 @@
|
|
| 1 |
{
|
| 2 |
"architectures": [
|
| 3 |
-
"
|
| 4 |
],
|
| 5 |
-
"classifier_dropout": 0.0,
|
| 6 |
"d_ff": 8192,
|
| 7 |
"d_kv": 128,
|
| 8 |
"d_model": 1024,
|
| 9 |
"decoder_start_token_id": 1,
|
| 10 |
-
"dense_act_fn": "gelu_new",
|
| 11 |
"dropout_rate": 0.1,
|
| 12 |
"eos_token_id": 2,
|
| 13 |
"feed_forward_proj": "gated-gelu",
|
| 14 |
"initializer_factor": 1.0,
|
| 15 |
"is_encoder_decoder": true,
|
| 16 |
-
"is_gated_act": true,
|
| 17 |
"layer_norm_epsilon": 1e-06,
|
| 18 |
-
"model_type": "
|
| 19 |
-
"n_positions": 512,
|
| 20 |
"num_decoder_layers": 32,
|
| 21 |
"num_heads": 16,
|
| 22 |
"num_layers": 32,
|
| 23 |
-
"output_past": true,
|
| 24 |
"pad_token_id": 1,
|
| 25 |
-
"relative_attention_max_distance": 128,
|
| 26 |
"relative_attention_num_buckets": 32,
|
| 27 |
-
"task_specific_params": {},
|
| 28 |
"tie_word_embeddings": false,
|
| 29 |
-
"
|
| 30 |
-
"use_cache": true,
|
| 31 |
"vocab_size": 256000
|
| 32 |
-
}
|
|
|
|
| 1 |
{
|
| 2 |
"architectures": [
|
| 3 |
+
"MT5ForConditionalGeneration"
|
| 4 |
],
|
|
|
|
| 5 |
"d_ff": 8192,
|
| 6 |
"d_kv": 128,
|
| 7 |
"d_model": 1024,
|
| 8 |
"decoder_start_token_id": 1,
|
|
|
|
| 9 |
"dropout_rate": 0.1,
|
| 10 |
"eos_token_id": 2,
|
| 11 |
"feed_forward_proj": "gated-gelu",
|
| 12 |
"initializer_factor": 1.0,
|
| 13 |
"is_encoder_decoder": true,
|
|
|
|
| 14 |
"layer_norm_epsilon": 1e-06,
|
| 15 |
+
"model_type": "mt5",
|
|
|
|
| 16 |
"num_decoder_layers": 32,
|
| 17 |
"num_heads": 16,
|
| 18 |
"num_layers": 32,
|
|
|
|
| 19 |
"pad_token_id": 1,
|
|
|
|
| 20 |
"relative_attention_num_buckets": 32,
|
|
|
|
| 21 |
"tie_word_embeddings": false,
|
| 22 |
+
"tokenizer_class": "T5Tokenizer",
|
|
|
|
| 23 |
"vocab_size": 256000
|
| 24 |
+
}
|