Divyasreepat
commited on
Commit
•
61ad469
1
Parent(s):
6bd500a
Upload folder using huggingface_hub
Browse files- README.md +18 -0
- assets/tokenizer/vocabulary.txt +0 -0
- config.json +20 -0
- metadata.json +6 -0
- model.weights.h5 +3 -0
- tokenizer.json +21 -0
README.md
ADDED
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: keras-hub
|
3 |
+
---
|
4 |
+
This is a [`Electra` model](https://keras.io/api/keras_hub/models/electra) uploaded using the KerasHub library and can be used with JAX, TensorFlow, and PyTorch backends.
|
5 |
+
Model config:
|
6 |
+
* **name:** electra_backbone
|
7 |
+
* **trainable:** True
|
8 |
+
* **vocab_size:** 30522
|
9 |
+
* **num_layers:** 24
|
10 |
+
* **num_heads:** 4
|
11 |
+
* **hidden_dim:** 256
|
12 |
+
* **embedding_dim:** 1024
|
13 |
+
* **intermediate_dim:** 1024
|
14 |
+
* **dropout:** 0.1
|
15 |
+
* **max_sequence_length:** 512
|
16 |
+
* **num_segments:** 2
|
17 |
+
|
18 |
+
This model card has been generated automatically and should be completed by the model author. See [Model Cards documentation](https://huggingface.co/docs/hub/model-cards) for more information.
|
assets/tokenizer/vocabulary.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|
config.json
ADDED
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"module": "keras_nlp.models.electra.electra_backbone",
|
3 |
+
"class_name": "ElectraBackbone",
|
4 |
+
"config": {
|
5 |
+
"name": "electra_backbone",
|
6 |
+
"trainable": true,
|
7 |
+
"vocab_size": 30522,
|
8 |
+
"num_layers": 24,
|
9 |
+
"num_heads": 4,
|
10 |
+
"hidden_dim": 256,
|
11 |
+
"embedding_dim": 1024,
|
12 |
+
"intermediate_dim": 1024,
|
13 |
+
"dropout": 0.1,
|
14 |
+
"max_sequence_length": 512,
|
15 |
+
"num_segments": 2
|
16 |
+
},
|
17 |
+
"registered_name": "keras_nlp>ElectraBackbone",
|
18 |
+
"assets": [],
|
19 |
+
"weights": "model.weights.h5"
|
20 |
+
}
|
metadata.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"keras_version": "3.0.2",
|
3 |
+
"keras_nlp_version": "0.9.0",
|
4 |
+
"parameter_count": 51065344,
|
5 |
+
"date_saved": "2024-03-24@02:18:32"
|
6 |
+
}
|
model.weights.h5
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:5276b180931b8c0ce4f267ba41f4f498837195b6f339d84aa52abeca56249886
|
3 |
+
size 205053960
|
tokenizer.json
ADDED
@@ -0,0 +1,21 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"module": "keras_nlp.models.electra.electra_tokenizer",
|
3 |
+
"class_name": "ElectraTokenizer",
|
4 |
+
"config": {
|
5 |
+
"name": "electra_tokenizer",
|
6 |
+
"trainable": true,
|
7 |
+
"dtype": "int32",
|
8 |
+
"vocabulary": null,
|
9 |
+
"sequence_length": null,
|
10 |
+
"lowercase": true,
|
11 |
+
"strip_accents": false,
|
12 |
+
"split": true,
|
13 |
+
"suffix_indicator": "##",
|
14 |
+
"oov_token": "[UNK]"
|
15 |
+
},
|
16 |
+
"registered_name": "keras_nlp>ElectraTokenizer",
|
17 |
+
"assets": [
|
18 |
+
"assets/tokenizer/vocabulary.txt"
|
19 |
+
],
|
20 |
+
"weights": null
|
21 |
+
}
|