Upload config.json
Browse files- config.json +45 -0
config.json
ADDED
|
@@ -0,0 +1,45 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "None",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"NetFoundLanguageModelling"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"classifier_dropout": null,
|
| 8 |
+
"embedding_size": 1024,
|
| 9 |
+
"encoder_layout": null,
|
| 10 |
+
"flat": false,
|
| 11 |
+
"hidden_act": "gelu",
|
| 12 |
+
"hidden_dropout_prob": 0.1,
|
| 13 |
+
"hidden_size": 1024,
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 3072,
|
| 16 |
+
"layer_norm_eps": 1e-12,
|
| 17 |
+
"limit_bursts": false,
|
| 18 |
+
"max_burst_length": 109,
|
| 19 |
+
"max_bursts": 12,
|
| 20 |
+
"max_position_embeddings": 109,
|
| 21 |
+
"metaFeatures": 4,
|
| 22 |
+
"model_max_length": 1308,
|
| 23 |
+
"model_type": "NetFound",
|
| 24 |
+
"no_direction_loss": false,
|
| 25 |
+
"no_meta": false,
|
| 26 |
+
"no_metadata_loss": false,
|
| 27 |
+
"no_mlm": false,
|
| 28 |
+
"no_swapped_bursts": false,
|
| 29 |
+
"num_attention_heads": 16,
|
| 30 |
+
"num_hidden_layers": 24,
|
| 31 |
+
"p": 0,
|
| 32 |
+
"pad_token_id": 0,
|
| 33 |
+
"position_embedding_type": "absolute",
|
| 34 |
+
"pretraining": true,
|
| 35 |
+
"rep_output_path": null,
|
| 36 |
+
"roformer": false,
|
| 37 |
+
"rotary_value": false,
|
| 38 |
+
"subflow_bursts": 3,
|
| 39 |
+
"subflow_len": -1,
|
| 40 |
+
"torch_dtype": "float32",
|
| 41 |
+
"transformers_version": "4.47.0.dev0",
|
| 42 |
+
"type_vocab_size": 2,
|
| 43 |
+
"use_cache": true,
|
| 44 |
+
"vocab_size": 65539
|
| 45 |
+
}
|