shadeMe-explosion commited on
Commit
9aa9cc6
1 Parent(s): 806e598

Initial commit

Browse files

Hyperparameters are similar to those of trl-internal-testing/tiny-random-GPTNeoXForCausalLM

config.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "BertForPreTraining"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "classifier_dropout": null,
7
+ "hidden_act": "gelu",
8
+ "hidden_dropout_prob": 0.1,
9
+ "hidden_size": 32,
10
+ "initializer_range": 0.02,
11
+ "intermediate_size": 37,
12
+ "layer_norm_eps": 1e-12,
13
+ "max_position_embeddings": 512,
14
+ "model_type": "bert",
15
+ "num_attention_heads": 4,
16
+ "num_hidden_layers": 5,
17
+ "pad_token_id": 0,
18
+ "position_embedding_type": "absolute",
19
+ "torch_dtype": "float32",
20
+ "transformers_version": "4.28.1",
21
+ "type_vocab_size": 16,
22
+ "use_cache": true,
23
+ "vocab_size": 1024
24
+ }
pytorch_model-00001-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:29c3f9216f7cae1f06dd27f0037423a75cba16c43135f3d9e0cd28044c3789f9
3
+ size 5034
pytorch_model-00002-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2bfcfe65c163ae2ee1e6ba7fb9964ed155baf11f23ce17fee1f7b2f0d7e68e3a
3
+ size 132010
pytorch_model-00003-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:37668b1aae41c6c7fcffbd25864006accfe935b88b33cab5d9bd017b514e186d
3
+ size 106715
pytorch_model-00004-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2348bd5eb0de6551a5b1359c638bb2f03eb269cccc7b58d2137da304ae379dd4
3
+ size 118755
pytorch_model-00005-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:654a4f3abd9495b0e0659febcbb85d71821be77dcd563c843d00fd28e383694a
3
+ size 21845
pytorch_model-00006-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d10da8220149bb879f73e21e232c9ba6d168dfd8f9210b76ffde4874cda0e8ea
3
+ size 132010
pytorch_model-00007-of-00007.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:990f7ce6c752bb603a05f22ca365c6dbf535ab870bedda29822ecace7fe0d183
3
+ size 5836
pytorch_model.bin.index.json ADDED
@@ -0,0 +1,104 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 487020
4
+ },
5
+ "weight_map": {
6
+ "bert.embeddings.LayerNorm.bias": "pytorch_model-00003-of-00007.bin",
7
+ "bert.embeddings.LayerNorm.weight": "pytorch_model-00003-of-00007.bin",
8
+ "bert.embeddings.position_embeddings.weight": "pytorch_model-00003-of-00007.bin",
9
+ "bert.embeddings.position_ids": "pytorch_model-00001-of-00007.bin",
10
+ "bert.embeddings.token_type_embeddings.weight": "pytorch_model-00003-of-00007.bin",
11
+ "bert.embeddings.word_embeddings.weight": "pytorch_model-00002-of-00007.bin",
12
+ "bert.encoder.layer.0.attention.output.LayerNorm.bias": "pytorch_model-00003-of-00007.bin",
13
+ "bert.encoder.layer.0.attention.output.LayerNorm.weight": "pytorch_model-00003-of-00007.bin",
14
+ "bert.encoder.layer.0.attention.output.dense.bias": "pytorch_model-00003-of-00007.bin",
15
+ "bert.encoder.layer.0.attention.output.dense.weight": "pytorch_model-00003-of-00007.bin",
16
+ "bert.encoder.layer.0.attention.self.key.bias": "pytorch_model-00003-of-00007.bin",
17
+ "bert.encoder.layer.0.attention.self.key.weight": "pytorch_model-00003-of-00007.bin",
18
+ "bert.encoder.layer.0.attention.self.query.bias": "pytorch_model-00003-of-00007.bin",
19
+ "bert.encoder.layer.0.attention.self.query.weight": "pytorch_model-00003-of-00007.bin",
20
+ "bert.encoder.layer.0.attention.self.value.bias": "pytorch_model-00003-of-00007.bin",
21
+ "bert.encoder.layer.0.attention.self.value.weight": "pytorch_model-00003-of-00007.bin",
22
+ "bert.encoder.layer.0.intermediate.dense.bias": "pytorch_model-00003-of-00007.bin",
23
+ "bert.encoder.layer.0.intermediate.dense.weight": "pytorch_model-00003-of-00007.bin",
24
+ "bert.encoder.layer.0.output.LayerNorm.bias": "pytorch_model-00003-of-00007.bin",
25
+ "bert.encoder.layer.0.output.LayerNorm.weight": "pytorch_model-00003-of-00007.bin",
26
+ "bert.encoder.layer.0.output.dense.bias": "pytorch_model-00003-of-00007.bin",
27
+ "bert.encoder.layer.0.output.dense.weight": "pytorch_model-00003-of-00007.bin",
28
+ "bert.encoder.layer.1.attention.output.LayerNorm.bias": "pytorch_model-00004-of-00007.bin",
29
+ "bert.encoder.layer.1.attention.output.LayerNorm.weight": "pytorch_model-00004-of-00007.bin",
30
+ "bert.encoder.layer.1.attention.output.dense.bias": "pytorch_model-00004-of-00007.bin",
31
+ "bert.encoder.layer.1.attention.output.dense.weight": "pytorch_model-00004-of-00007.bin",
32
+ "bert.encoder.layer.1.attention.self.key.bias": "pytorch_model-00004-of-00007.bin",
33
+ "bert.encoder.layer.1.attention.self.key.weight": "pytorch_model-00004-of-00007.bin",
34
+ "bert.encoder.layer.1.attention.self.query.bias": "pytorch_model-00003-of-00007.bin",
35
+ "bert.encoder.layer.1.attention.self.query.weight": "pytorch_model-00003-of-00007.bin",
36
+ "bert.encoder.layer.1.attention.self.value.bias": "pytorch_model-00004-of-00007.bin",
37
+ "bert.encoder.layer.1.attention.self.value.weight": "pytorch_model-00004-of-00007.bin",
38
+ "bert.encoder.layer.1.intermediate.dense.bias": "pytorch_model-00004-of-00007.bin",
39
+ "bert.encoder.layer.1.intermediate.dense.weight": "pytorch_model-00004-of-00007.bin",
40
+ "bert.encoder.layer.1.output.LayerNorm.bias": "pytorch_model-00004-of-00007.bin",
41
+ "bert.encoder.layer.1.output.LayerNorm.weight": "pytorch_model-00004-of-00007.bin",
42
+ "bert.encoder.layer.1.output.dense.bias": "pytorch_model-00004-of-00007.bin",
43
+ "bert.encoder.layer.1.output.dense.weight": "pytorch_model-00004-of-00007.bin",
44
+ "bert.encoder.layer.2.attention.output.LayerNorm.bias": "pytorch_model-00004-of-00007.bin",
45
+ "bert.encoder.layer.2.attention.output.LayerNorm.weight": "pytorch_model-00004-of-00007.bin",
46
+ "bert.encoder.layer.2.attention.output.dense.bias": "pytorch_model-00004-of-00007.bin",
47
+ "bert.encoder.layer.2.attention.output.dense.weight": "pytorch_model-00004-of-00007.bin",
48
+ "bert.encoder.layer.2.attention.self.key.bias": "pytorch_model-00004-of-00007.bin",
49
+ "bert.encoder.layer.2.attention.self.key.weight": "pytorch_model-00004-of-00007.bin",
50
+ "bert.encoder.layer.2.attention.self.query.bias": "pytorch_model-00004-of-00007.bin",
51
+ "bert.encoder.layer.2.attention.self.query.weight": "pytorch_model-00004-of-00007.bin",
52
+ "bert.encoder.layer.2.attention.self.value.bias": "pytorch_model-00004-of-00007.bin",
53
+ "bert.encoder.layer.2.attention.self.value.weight": "pytorch_model-00004-of-00007.bin",
54
+ "bert.encoder.layer.2.intermediate.dense.bias": "pytorch_model-00004-of-00007.bin",
55
+ "bert.encoder.layer.2.intermediate.dense.weight": "pytorch_model-00004-of-00007.bin",
56
+ "bert.encoder.layer.2.output.LayerNorm.bias": "pytorch_model-00004-of-00007.bin",
57
+ "bert.encoder.layer.2.output.LayerNorm.weight": "pytorch_model-00004-of-00007.bin",
58
+ "bert.encoder.layer.2.output.dense.bias": "pytorch_model-00004-of-00007.bin",
59
+ "bert.encoder.layer.2.output.dense.weight": "pytorch_model-00004-of-00007.bin",
60
+ "bert.encoder.layer.3.attention.output.LayerNorm.bias": "pytorch_model-00004-of-00007.bin",
61
+ "bert.encoder.layer.3.attention.output.LayerNorm.weight": "pytorch_model-00004-of-00007.bin",
62
+ "bert.encoder.layer.3.attention.output.dense.bias": "pytorch_model-00004-of-00007.bin",
63
+ "bert.encoder.layer.3.attention.output.dense.weight": "pytorch_model-00004-of-00007.bin",
64
+ "bert.encoder.layer.3.attention.self.key.bias": "pytorch_model-00004-of-00007.bin",
65
+ "bert.encoder.layer.3.attention.self.key.weight": "pytorch_model-00004-of-00007.bin",
66
+ "bert.encoder.layer.3.attention.self.query.bias": "pytorch_model-00004-of-00007.bin",
67
+ "bert.encoder.layer.3.attention.self.query.weight": "pytorch_model-00004-of-00007.bin",
68
+ "bert.encoder.layer.3.attention.self.value.bias": "pytorch_model-00004-of-00007.bin",
69
+ "bert.encoder.layer.3.attention.self.value.weight": "pytorch_model-00004-of-00007.bin",
70
+ "bert.encoder.layer.3.intermediate.dense.bias": "pytorch_model-00004-of-00007.bin",
71
+ "bert.encoder.layer.3.intermediate.dense.weight": "pytorch_model-00004-of-00007.bin",
72
+ "bert.encoder.layer.3.output.LayerNorm.bias": "pytorch_model-00004-of-00007.bin",
73
+ "bert.encoder.layer.3.output.LayerNorm.weight": "pytorch_model-00004-of-00007.bin",
74
+ "bert.encoder.layer.3.output.dense.bias": "pytorch_model-00004-of-00007.bin",
75
+ "bert.encoder.layer.3.output.dense.weight": "pytorch_model-00004-of-00007.bin",
76
+ "bert.encoder.layer.4.attention.output.LayerNorm.bias": "pytorch_model-00004-of-00007.bin",
77
+ "bert.encoder.layer.4.attention.output.LayerNorm.weight": "pytorch_model-00004-of-00007.bin",
78
+ "bert.encoder.layer.4.attention.output.dense.bias": "pytorch_model-00004-of-00007.bin",
79
+ "bert.encoder.layer.4.attention.output.dense.weight": "pytorch_model-00004-of-00007.bin",
80
+ "bert.encoder.layer.4.attention.self.key.bias": "pytorch_model-00004-of-00007.bin",
81
+ "bert.encoder.layer.4.attention.self.key.weight": "pytorch_model-00004-of-00007.bin",
82
+ "bert.encoder.layer.4.attention.self.query.bias": "pytorch_model-00004-of-00007.bin",
83
+ "bert.encoder.layer.4.attention.self.query.weight": "pytorch_model-00004-of-00007.bin",
84
+ "bert.encoder.layer.4.attention.self.value.bias": "pytorch_model-00004-of-00007.bin",
85
+ "bert.encoder.layer.4.attention.self.value.weight": "pytorch_model-00004-of-00007.bin",
86
+ "bert.encoder.layer.4.intermediate.dense.bias": "pytorch_model-00004-of-00007.bin",
87
+ "bert.encoder.layer.4.intermediate.dense.weight": "pytorch_model-00004-of-00007.bin",
88
+ "bert.encoder.layer.4.output.LayerNorm.bias": "pytorch_model-00005-of-00007.bin",
89
+ "bert.encoder.layer.4.output.LayerNorm.weight": "pytorch_model-00005-of-00007.bin",
90
+ "bert.encoder.layer.4.output.dense.bias": "pytorch_model-00005-of-00007.bin",
91
+ "bert.encoder.layer.4.output.dense.weight": "pytorch_model-00005-of-00007.bin",
92
+ "bert.pooler.dense.bias": "pytorch_model-00005-of-00007.bin",
93
+ "bert.pooler.dense.weight": "pytorch_model-00005-of-00007.bin",
94
+ "cls.predictions.bias": "pytorch_model-00005-of-00007.bin",
95
+ "cls.predictions.decoder.bias": "pytorch_model-00007-of-00007.bin",
96
+ "cls.predictions.decoder.weight": "pytorch_model-00006-of-00007.bin",
97
+ "cls.predictions.transform.LayerNorm.bias": "pytorch_model-00005-of-00007.bin",
98
+ "cls.predictions.transform.LayerNorm.weight": "pytorch_model-00005-of-00007.bin",
99
+ "cls.predictions.transform.dense.bias": "pytorch_model-00005-of-00007.bin",
100
+ "cls.predictions.transform.dense.weight": "pytorch_model-00005-of-00007.bin",
101
+ "cls.seq_relationship.bias": "pytorch_model-00007-of-00007.bin",
102
+ "cls.seq_relationship.weight": "pytorch_model-00007-of-00007.bin"
103
+ }
104
+ }