Initial commit
Browse files- config.json +37 -0
- pytorch_model.bin +3 -0
- vocab.txt +0 -0
config.json
ADDED
|
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_dropout": 0.0,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"FunnelModel"
|
| 5 |
+
],
|
| 6 |
+
"attention_dropout": 0.1,
|
| 7 |
+
"attention_type": "relative_shift",
|
| 8 |
+
"block_repeats": [
|
| 9 |
+
1,
|
| 10 |
+
1,
|
| 11 |
+
1
|
| 12 |
+
],
|
| 13 |
+
"block_sizes": [
|
| 14 |
+
4,
|
| 15 |
+
4,
|
| 16 |
+
4
|
| 17 |
+
],
|
| 18 |
+
"d_head": 64,
|
| 19 |
+
"d_inner": 3072,
|
| 20 |
+
"d_model": 768,
|
| 21 |
+
"hidden_act": "gelu_new",
|
| 22 |
+
"hidden_dropout": 0.1,
|
| 23 |
+
"initializer_range": 0.1,
|
| 24 |
+
"initializer_std": null,
|
| 25 |
+
"layer_norm_eps": 1e-09,
|
| 26 |
+
"max_position_embeddings": 512,
|
| 27 |
+
"model_type": "funnel",
|
| 28 |
+
"n_head": 12,
|
| 29 |
+
"num_decoder_layers": 2,
|
| 30 |
+
"pool_q_only": true,
|
| 31 |
+
"pooling_type": "mean",
|
| 32 |
+
"rel_attn_type": "factorized",
|
| 33 |
+
"separate_cls": true,
|
| 34 |
+
"truncate_seq": true,
|
| 35 |
+
"type_vocab_size": 3,
|
| 36 |
+
"vocab_size": 50000
|
| 37 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:166293638511f21ecd8d7fe5f1d32b7a2fc52c933e2890b817520ebd2b409cbc
|
| 3 |
+
size 522373790
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|