howard-hou
commited on
Commit
•
87f587a
1
Parent(s):
18a2239
Upload RankingPrompterForPreTraining
Browse files- config.json +3 -2
- configuration_rankingprompter.py +2 -0
- pytorch_model.bin +1 -1
config.json
CHANGED
@@ -1,5 +1,5 @@
|
|
1 |
{
|
2 |
-
"_name_or_path": "
|
3 |
"architectures": [
|
4 |
"RankingPrompterForPreTraining"
|
5 |
],
|
@@ -22,6 +22,7 @@
|
|
22 |
"layer_norm_epsilon": 1e-06,
|
23 |
"max_new_tokens": 64,
|
24 |
"model_type": "umt5",
|
|
|
25 |
"num_decoder_layers": 8,
|
26 |
"num_heads": 6,
|
27 |
"num_layers": 8,
|
@@ -32,7 +33,7 @@
|
|
32 |
"tie_word_embeddings": false,
|
33 |
"tokenizer_class": "T5Tokenizer",
|
34 |
"torch_dtype": "float32",
|
35 |
-
"transformers_version": "4.32.
|
36 |
"use_cache": true,
|
37 |
"vocab_size": 256384
|
38 |
}
|
|
|
1 |
{
|
2 |
+
"_name_or_path": "..\\..\\..\\huggingface_model\\RankingPrompterForPreTraining-small\\",
|
3 |
"architectures": [
|
4 |
"RankingPrompterForPreTraining"
|
5 |
],
|
|
|
22 |
"layer_norm_epsilon": 1e-06,
|
23 |
"max_new_tokens": 64,
|
24 |
"model_type": "umt5",
|
25 |
+
"num_answer_query": 128,
|
26 |
"num_decoder_layers": 8,
|
27 |
"num_heads": 6,
|
28 |
"num_layers": 8,
|
|
|
33 |
"tie_word_embeddings": false,
|
34 |
"tokenizer_class": "T5Tokenizer",
|
35 |
"torch_dtype": "float32",
|
36 |
+
"transformers_version": "4.32.1",
|
37 |
"use_cache": true,
|
38 |
"vocab_size": 256384
|
39 |
}
|
configuration_rankingprompter.py
CHANGED
@@ -14,6 +14,7 @@ class RankingPrompterConfig(PretrainedConfig):
|
|
14 |
num_heads=6,
|
15 |
relative_attention_num_buckets=32,
|
16 |
relative_attention_max_distance=128,
|
|
|
17 |
dropout_rate=0.1,
|
18 |
layer_norm_epsilon=1e-6,
|
19 |
initializer_factor=1.0,
|
@@ -48,6 +49,7 @@ class RankingPrompterConfig(PretrainedConfig):
|
|
48 |
self.num_heads = num_heads
|
49 |
self.relative_attention_num_buckets = relative_attention_num_buckets
|
50 |
self.relative_attention_max_distance = relative_attention_max_distance
|
|
|
51 |
self.dropout_rate = dropout_rate
|
52 |
self.classifier_dropout = classifier_dropout
|
53 |
self.layer_norm_epsilon = layer_norm_epsilon
|
|
|
14 |
num_heads=6,
|
15 |
relative_attention_num_buckets=32,
|
16 |
relative_attention_max_distance=128,
|
17 |
+
num_answer_query=128,
|
18 |
dropout_rate=0.1,
|
19 |
layer_norm_epsilon=1e-6,
|
20 |
initializer_factor=1.0,
|
|
|
49 |
self.num_heads = num_heads
|
50 |
self.relative_attention_num_buckets = relative_attention_num_buckets
|
51 |
self.relative_attention_max_distance = relative_attention_max_distance
|
52 |
+
self.num_answer_query = num_answer_query
|
53 |
self.dropout_rate = dropout_rate
|
54 |
self.classifier_dropout = classifier_dropout
|
55 |
self.layer_norm_epsilon = layer_norm_epsilon
|
pytorch_model.bin
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 701403585
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:aab6831f798bedd1b458b5cd3b77c941d58564bb70941bcc2aed16ad8cdee75d
|
3 |
size 701403585
|