Divyasreepat
commited on
Commit
•
784cdc1
1
Parent(s):
f99f279
Upload folder using huggingface_hub
Browse files- README.md +20 -0
- assets/tokenizer/vocabulary.spm +0 -0
- config.json +27 -0
- metadata.json +6 -0
- model.weights.h5 +3 -0
- tokenizer.json +16 -0
README.md
ADDED
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
library_name: keras-hub
|
3 |
+
---
|
4 |
+
This is a [`T5` model](https://keras.io/api/keras_hub/models/t5) uploaded using the KerasHub library and can be used with JAX, TensorFlow, and PyTorch backends.
|
5 |
+
Model config:
|
6 |
+
* **name:** t5_backbone
|
7 |
+
* **trainable:** True
|
8 |
+
* **vocabulary_size:** 32128
|
9 |
+
* **hidden_dim:** 512
|
10 |
+
* **intermediate_dim:** 1024
|
11 |
+
* **num_layers:** 8
|
12 |
+
* **num_heads:** 6
|
13 |
+
* **activation:** {'module': 'builtins', 'class_name': 'function', 'config': 'keras_nlp>gelu_approximate', 'registered_name': 'function'}
|
14 |
+
* **key_value_dim:** 64
|
15 |
+
* **dropout:** 0.1
|
16 |
+
* **use_gated_activation:** True
|
17 |
+
* **layer_norm_epsilon:** 1e-06
|
18 |
+
* **tie_embedding_weights:** False
|
19 |
+
|
20 |
+
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.spm
ADDED
Binary file (792 kB). View file
|
|
config.json
ADDED
@@ -0,0 +1,27 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"module": "keras_nlp.src.models.t5.t5_backbone",
|
3 |
+
"class_name": "T5Backbone",
|
4 |
+
"config": {
|
5 |
+
"name": "t5_backbone",
|
6 |
+
"trainable": true,
|
7 |
+
"vocabulary_size": 32128,
|
8 |
+
"hidden_dim": 512,
|
9 |
+
"intermediate_dim": 1024,
|
10 |
+
"num_layers": 8,
|
11 |
+
"num_heads": 6,
|
12 |
+
"activation": {
|
13 |
+
"module": "builtins",
|
14 |
+
"class_name": "function",
|
15 |
+
"config": "keras_nlp>gelu_approximate",
|
16 |
+
"registered_name": "function"
|
17 |
+
},
|
18 |
+
"key_value_dim": 64,
|
19 |
+
"dropout": 0.1,
|
20 |
+
"use_gated_activation": true,
|
21 |
+
"layer_norm_epsilon": 1e-06,
|
22 |
+
"tie_embedding_weights": false
|
23 |
+
},
|
24 |
+
"registered_name": "keras_nlp>T5Backbone",
|
25 |
+
"assets": [],
|
26 |
+
"weights": "model.weights.h5"
|
27 |
+
}
|
metadata.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"keras_version": "3.0.1",
|
3 |
+
"keras_nlp_version": "0.7.0",
|
4 |
+
"parameter_count": 76961152,
|
5 |
+
"date_saved": "2023-12-27@02:29:58"
|
6 |
+
}
|
model.weights.h5
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b93dc7963d4df574339c947359b312e5590ce3466f663a6acbc610522d188faf
|
3 |
+
size 308492984
|
tokenizer.json
ADDED
@@ -0,0 +1,16 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"module": "keras_nlp.src.models.t5.t5_tokenizer",
|
3 |
+
"class_name": "T5Tokenizer",
|
4 |
+
"config": {
|
5 |
+
"name": "t5_tokenizer",
|
6 |
+
"trainable": true,
|
7 |
+
"dtype": "int32",
|
8 |
+
"proto": null,
|
9 |
+
"sequence_length": null
|
10 |
+
},
|
11 |
+
"registered_name": "keras_nlp>T5Tokenizer",
|
12 |
+
"assets": [
|
13 |
+
"assets/tokenizer/vocabulary.spm"
|
14 |
+
],
|
15 |
+
"weights": null
|
16 |
+
}
|