openwebtext训练的electra small
Browse files- config.json +18 -0
- pytorch_model.bin +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +4 -0
- vocab.txt +0 -0
config.json
ADDED
|
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"attention_probs_dropout_prob": 0.1,
|
| 3 |
+
"embedding_size": 128,
|
| 4 |
+
"hidden_act": "gelu",
|
| 5 |
+
"hidden_dropout_prob": 0.1,
|
| 6 |
+
"hidden_size": 64,
|
| 7 |
+
"initializer_range": 0.02,
|
| 8 |
+
"intermediate_size": 256,
|
| 9 |
+
"layer_norm_eps": 1e-12,
|
| 10 |
+
"max_position_embeddings": 512,
|
| 11 |
+
"model_type": "electra",
|
| 12 |
+
"num_attention_heads": 4,
|
| 13 |
+
"num_hidden_layers": 12,
|
| 14 |
+
"pad_token_id": 0,
|
| 15 |
+
"position_embedding_type": "absolute",
|
| 16 |
+
"type_vocab_size": 2,
|
| 17 |
+
"vocab_size": 30522
|
| 18 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:18acdae4d103a64f0b70d551f0638fdeb846f15519bff584f898508bd4acb532
|
| 3 |
+
size 54273423
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"do_lower_case": true
|
| 3 |
+
}
|
| 4 |
+
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|