Hugging Face's logo

Dahoas
/
gptneox-rm-static

Model card Files Community
Dahoas commited on
Commit
5e05c58
•
1 Parent(s): a4aebab

Upload index.json

Browse files
Files changed (1) hide show
  1. index.json +1330 -0
index.json ADDED
@@ -0,0 +1,1330 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "model.gpt_neox.embed_in.weight": "hf_ckpt_0.pt",
3
+ "model.gpt_neox.layers.0.input_layernorm.weight": "hf_ckpt_0.pt",
4
+ "model.gpt_neox.layers.0.input_layernorm.bias": "hf_ckpt_0.pt",
5
+ "model.gpt_neox.layers.0.post_attention_layernorm.weight": "hf_ckpt_0.pt",
6
+ "model.gpt_neox.layers.0.post_attention_layernorm.bias": "hf_ckpt_0.pt",
7
+ "model.gpt_neox.layers.0.attention.bias": "hf_ckpt_0.pt",
8
+ "model.gpt_neox.layers.0.attention.masked_bias": "hf_ckpt_0.pt",
9
+ "model.gpt_neox.layers.0.attention.rotary_emb.inv_freq": "hf_ckpt_0.pt",
10
+ "model.gpt_neox.layers.0.attention.query_key_value.weight": "hf_ckpt_0.pt",
11
+ "model.gpt_neox.layers.0.attention.query_key_value.bias": "hf_ckpt_0.pt",
12
+ "model.gpt_neox.layers.0.attention.dense.weight": "hf_ckpt_0.pt",
13
+ "model.gpt_neox.layers.0.attention.dense.bias": "hf_ckpt_0.pt",
14
+ "model.gpt_neox.layers.0.mlp.dense_h_to_4h.weight": "hf_ckpt_0.pt",
15
+ "model.gpt_neox.layers.0.mlp.dense_h_to_4h.bias": "hf_ckpt_0.pt",
16
+ "model.gpt_neox.layers.0.mlp.dense_4h_to_h.weight": "hf_ckpt_0.pt",
17
+ "model.gpt_neox.layers.0.mlp.dense_4h_to_h.bias": "hf_ckpt_0.pt",
18
+ "model.gpt_neox.layers.1.input_layernorm.weight": "hf_ckpt_0.pt",
19
+ "model.gpt_neox.layers.1.input_layernorm.bias": "hf_ckpt_0.pt",
20
+ "model.gpt_neox.layers.1.post_attention_layernorm.weight": "hf_ckpt_0.pt",
21
+ "model.gpt_neox.layers.1.post_attention_layernorm.bias": "hf_ckpt_0.pt",
22
+ "model.gpt_neox.layers.1.attention.bias": "hf_ckpt_0.pt",
23
+ "model.gpt_neox.layers.1.attention.masked_bias": "hf_ckpt_0.pt",
24
+ "model.gpt_neox.layers.1.attention.rotary_emb.inv_freq": "hf_ckpt_0.pt",
25
+ "model.gpt_neox.layers.1.attention.query_key_value.weight": "hf_ckpt_0.pt",
26
+ "model.gpt_neox.layers.1.attention.query_key_value.bias": "hf_ckpt_0.pt",
27
+ "model.gpt_neox.layers.1.attention.dense.weight": "hf_ckpt_0.pt",
28
+ "model.gpt_neox.layers.1.attention.dense.bias": "hf_ckpt_0.pt",
29
+ "model.gpt_neox.layers.1.mlp.dense_h_to_4h.weight": "hf_ckpt_0.pt",
30
+ "model.gpt_neox.layers.1.mlp.dense_h_to_4h.bias": "hf_ckpt_1.pt",
31
+ "model.gpt_neox.layers.1.mlp.dense_4h_to_h.weight": "hf_ckpt_1.pt",
32
+ "model.gpt_neox.layers.1.mlp.dense_4h_to_h.bias": "hf_ckpt_1.pt",
33
+ "model.gpt_neox.layers.2.input_layernorm.weight": "hf_ckpt_1.pt",
34
+ "model.gpt_neox.layers.2.input_layernorm.bias": "hf_ckpt_1.pt",
35
+ "model.gpt_neox.layers.2.post_attention_layernorm.weight": "hf_ckpt_1.pt",
36
+ "model.gpt_neox.layers.2.post_attention_layernorm.bias": "hf_ckpt_1.pt",
37
+ "model.gpt_neox.layers.2.attention.bias": "hf_ckpt_1.pt",
38
+ "model.gpt_neox.layers.2.attention.masked_bias": "hf_ckpt_1.pt",
39
+ "model.gpt_neox.layers.2.attention.rotary_emb.inv_freq": "hf_ckpt_1.pt",
40
+ "model.gpt_neox.layers.2.attention.query_key_value.weight": "hf_ckpt_1.pt",
41
+ "model.gpt_neox.layers.2.attention.query_key_value.bias": "hf_ckpt_1.pt",
42
+ "model.gpt_neox.layers.2.attention.dense.weight": "hf_ckpt_1.pt",
43
+ "model.gpt_neox.layers.2.attention.dense.bias": "hf_ckpt_1.pt",
44
+ "model.gpt_neox.layers.2.mlp.dense_h_to_4h.weight": "hf_ckpt_1.pt",
45
+ "model.gpt_neox.layers.2.mlp.dense_h_to_4h.bias": "hf_ckpt_1.pt",
46
+ "model.gpt_neox.layers.2.mlp.dense_4h_to_h.weight": "hf_ckpt_1.pt",
47
+ "model.gpt_neox.layers.2.mlp.dense_4h_to_h.bias": "hf_ckpt_1.pt",
48
+ "model.gpt_neox.layers.3.input_layernorm.weight": "hf_ckpt_1.pt",
49
+ "model.gpt_neox.layers.3.input_layernorm.bias": "hf_ckpt_1.pt",
50
+ "model.gpt_neox.layers.3.post_attention_layernorm.weight": "hf_ckpt_1.pt",
51
+ "model.gpt_neox.layers.3.post_attention_layernorm.bias": "hf_ckpt_1.pt",
52
+ "model.gpt_neox.layers.3.attention.bias": "hf_ckpt_1.pt",
53
+ "model.gpt_neox.layers.3.attention.masked_bias": "hf_ckpt_1.pt",
54
+ "model.gpt_neox.layers.3.attention.rotary_emb.inv_freq": "hf_ckpt_1.pt",
55
+ "model.gpt_neox.layers.3.attention.query_key_value.weight": "hf_ckpt_1.pt",
56
+ "model.gpt_neox.layers.3.attention.query_key_value.bias": "hf_ckpt_1.pt",
57
+ "model.gpt_neox.layers.3.attention.dense.weight": "hf_ckpt_1.pt",
58
+ "model.gpt_neox.layers.3.attention.dense.bias": "hf_ckpt_2.pt",
59
+ "model.gpt_neox.layers.3.mlp.dense_h_to_4h.weight": "hf_ckpt_2.pt",
60
+ "model.gpt_neox.layers.3.mlp.dense_h_to_4h.bias": "hf_ckpt_2.pt",
61
+ "model.gpt_neox.layers.3.mlp.dense_4h_to_h.weight": "hf_ckpt_2.pt",
62
+ "model.gpt_neox.layers.3.mlp.dense_4h_to_h.bias": "hf_ckpt_2.pt",
63
+ "model.gpt_neox.layers.4.input_layernorm.weight": "hf_ckpt_2.pt",
64
+ "model.gpt_neox.layers.4.input_layernorm.bias": "hf_ckpt_2.pt",
65
+ "model.gpt_neox.layers.4.post_attention_layernorm.weight": "hf_ckpt_2.pt",
66
+ "model.gpt_neox.layers.4.post_attention_layernorm.bias": "hf_ckpt_2.pt",
67
+ "model.gpt_neox.layers.4.attention.bias": "hf_ckpt_2.pt",
68
+ "model.gpt_neox.layers.4.attention.masked_bias": "hf_ckpt_2.pt",
69
+ "model.gpt_neox.layers.4.attention.rotary_emb.inv_freq": "hf_ckpt_2.pt",
70
+ "model.gpt_neox.layers.4.attention.query_key_value.weight": "hf_ckpt_2.pt",
71
+ "model.gpt_neox.layers.4.attention.query_key_value.bias": "hf_ckpt_2.pt",
72
+ "model.gpt_neox.layers.4.attention.dense.weight": "hf_ckpt_2.pt",
73
+ "model.gpt_neox.layers.4.attention.dense.bias": "hf_ckpt_2.pt",
74
+ "model.gpt_neox.layers.4.mlp.dense_h_to_4h.weight": "hf_ckpt_2.pt",
75
+ "model.gpt_neox.layers.4.mlp.dense_h_to_4h.bias": "hf_ckpt_2.pt",
76
+ "model.gpt_neox.layers.4.mlp.dense_4h_to_h.weight": "hf_ckpt_2.pt",
77
+ "model.gpt_neox.layers.4.mlp.dense_4h_to_h.bias": "hf_ckpt_2.pt",
78
+ "model.gpt_neox.layers.5.input_layernorm.weight": "hf_ckpt_2.pt",
79
+ "model.gpt_neox.layers.5.input_layernorm.bias": "hf_ckpt_2.pt",
80
+ "model.gpt_neox.layers.5.post_attention_layernorm.weight": "hf_ckpt_2.pt",
81
+ "model.gpt_neox.layers.5.post_attention_layernorm.bias": "hf_ckpt_2.pt",
82
+ "model.gpt_neox.layers.5.attention.bias": "hf_ckpt_2.pt",
83
+ "model.gpt_neox.layers.5.attention.masked_bias": "hf_ckpt_2.pt",
84
+ "model.gpt_neox.layers.5.attention.rotary_emb.inv_freq": "hf_ckpt_2.pt",
85
+ "model.gpt_neox.layers.5.attention.query_key_value.weight": "hf_ckpt_2.pt",
86
+ "model.gpt_neox.layers.5.attention.query_key_value.bias": "hf_ckpt_3.pt",
87
+ "model.gpt_neox.layers.5.attention.dense.weight": "hf_ckpt_3.pt",
88
+ "model.gpt_neox.layers.5.attention.dense.bias": "hf_ckpt_3.pt",
89
+ "model.gpt_neox.layers.5.mlp.dense_h_to_4h.weight": "hf_ckpt_3.pt",
90
+ "model.gpt_neox.layers.5.mlp.dense_h_to_4h.bias": "hf_ckpt_3.pt",
91
+ "model.gpt_neox.layers.5.mlp.dense_4h_to_h.weight": "hf_ckpt_3.pt",
92
+ "model.gpt_neox.layers.5.mlp.dense_4h_to_h.bias": "hf_ckpt_3.pt",
93
+ "model.gpt_neox.layers.6.input_layernorm.weight": "hf_ckpt_3.pt",
94
+ "model.gpt_neox.layers.6.input_layernorm.bias": "hf_ckpt_3.pt",
95
+ "model.gpt_neox.layers.6.post_attention_layernorm.weight": "hf_ckpt_3.pt",
96
+ "model.gpt_neox.layers.6.post_attention_layernorm.bias": "hf_ckpt_3.pt",
97
+ "model.gpt_neox.layers.6.attention.bias": "hf_ckpt_3.pt",
98
+ "model.gpt_neox.layers.6.attention.masked_bias": "hf_ckpt_3.pt",
99
+ "model.gpt_neox.layers.6.attention.rotary_emb.inv_freq": "hf_ckpt_3.pt",
100
+ "model.gpt_neox.layers.6.attention.query_key_value.weight": "hf_ckpt_3.pt",
101
+ "model.gpt_neox.layers.6.attention.query_key_value.bias": "hf_ckpt_3.pt",
102
+ "model.gpt_neox.layers.6.attention.dense.weight": "hf_ckpt_3.pt",
103
+ "model.gpt_neox.layers.6.attention.dense.bias": "hf_ckpt_3.pt",
104
+ "model.gpt_neox.layers.6.mlp.dense_h_to_4h.weight": "hf_ckpt_3.pt",
105
+ "model.gpt_neox.layers.6.mlp.dense_h_to_4h.bias": "hf_ckpt_3.pt",
106
+ "model.gpt_neox.layers.6.mlp.dense_4h_to_h.weight": "hf_ckpt_3.pt",
107
+ "model.gpt_neox.layers.6.mlp.dense_4h_to_h.bias": "hf_ckpt_3.pt",
108
+ "model.gpt_neox.layers.7.input_layernorm.weight": "hf_ckpt_3.pt",
109
+ "model.gpt_neox.layers.7.input_layernorm.bias": "hf_ckpt_3.pt",
110
+ "model.gpt_neox.layers.7.post_attention_layernorm.weight": "hf_ckpt_3.pt",
111
+ "model.gpt_neox.layers.7.post_attention_layernorm.bias": "hf_ckpt_3.pt",
112
+ "model.gpt_neox.layers.7.attention.bias": "hf_ckpt_3.pt",
113
+ "model.gpt_neox.layers.7.attention.masked_bias": "hf_ckpt_3.pt",
114
+ "model.gpt_neox.layers.7.attention.rotary_emb.inv_freq": "hf_ckpt_4.pt",
115
+ "model.gpt_neox.layers.7.attention.query_key_value.weight": "hf_ckpt_4.pt",
116
+ "model.gpt_neox.layers.7.attention.query_key_value.bias": "hf_ckpt_4.pt",
117
+ "model.gpt_neox.layers.7.attention.dense.weight": "hf_ckpt_4.pt",
118
+ "model.gpt_neox.layers.7.attention.dense.bias": "hf_ckpt_4.pt",
119
+ "model.gpt_neox.layers.7.mlp.dense_h_to_4h.weight": "hf_ckpt_4.pt",
120
+ "model.gpt_neox.layers.7.mlp.dense_h_to_4h.bias": "hf_ckpt_4.pt",
121
+ "model.gpt_neox.layers.7.mlp.dense_4h_to_h.weight": "hf_ckpt_4.pt",
122
+ "model.gpt_neox.layers.7.mlp.dense_4h_to_h.bias": "hf_ckpt_4.pt",
123
+ "model.gpt_neox.layers.8.input_layernorm.weight": "hf_ckpt_4.pt",
124
+ "model.gpt_neox.layers.8.input_layernorm.bias": "hf_ckpt_4.pt",
125
+ "model.gpt_neox.layers.8.post_attention_layernorm.weight": "hf_ckpt_4.pt",
126
+ "model.gpt_neox.layers.8.post_attention_layernorm.bias": "hf_ckpt_4.pt",
127
+ "model.gpt_neox.layers.8.attention.bias": "hf_ckpt_4.pt",
128
+ "model.gpt_neox.layers.8.attention.masked_bias": "hf_ckpt_4.pt",
129
+ "model.gpt_neox.layers.8.attention.rotary_emb.inv_freq": "hf_ckpt_4.pt",
130
+ "model.gpt_neox.layers.8.attention.query_key_value.weight": "hf_ckpt_4.pt",
131
+ "model.gpt_neox.layers.8.attention.query_key_value.bias": "hf_ckpt_4.pt",
132
+ "model.gpt_neox.layers.8.attention.dense.weight": "hf_ckpt_4.pt",
133
+ "model.gpt_neox.layers.8.attention.dense.bias": "hf_ckpt_4.pt",
134
+ "model.gpt_neox.layers.8.mlp.dense_h_to_4h.weight": "hf_ckpt_4.pt",
135
+ "model.gpt_neox.layers.8.mlp.dense_h_to_4h.bias": "hf_ckpt_4.pt",
136
+ "model.gpt_neox.layers.8.mlp.dense_4h_to_h.weight": "hf_ckpt_4.pt",
137
+ "model.gpt_neox.layers.8.mlp.dense_4h_to_h.bias": "hf_ckpt_4.pt",
138
+ "model.gpt_neox.layers.9.input_layernorm.weight": "hf_ckpt_4.pt",
139
+ "model.gpt_neox.layers.9.input_layernorm.bias": "hf_ckpt_4.pt",
140
+ "model.gpt_neox.layers.9.post_attention_layernorm.weight": "hf_ckpt_4.pt",
141
+ "model.gpt_neox.layers.9.post_attention_layernorm.bias": "hf_ckpt_4.pt",
142
+ "model.gpt_neox.layers.9.attention.bias": "hf_ckpt_5.pt",
143
+ "model.gpt_neox.layers.9.attention.masked_bias": "hf_ckpt_5.pt",
144
+ "model.gpt_neox.layers.9.attention.rotary_emb.inv_freq": "hf_ckpt_5.pt",
145
+ "model.gpt_neox.layers.9.attention.query_key_value.weight": "hf_ckpt_5.pt",
146
+ "model.gpt_neox.layers.9.attention.query_key_value.bias": "hf_ckpt_5.pt",
147
+ "model.gpt_neox.layers.9.attention.dense.weight": "hf_ckpt_5.pt",
148
+ "model.gpt_neox.layers.9.attention.dense.bias": "hf_ckpt_5.pt",
149
+ "model.gpt_neox.layers.9.mlp.dense_h_to_4h.weight": "hf_ckpt_5.pt",
150
+ "model.gpt_neox.layers.9.mlp.dense_h_to_4h.bias": "hf_ckpt_5.pt",
151
+ "model.gpt_neox.layers.9.mlp.dense_4h_to_h.weight": "hf_ckpt_5.pt",
152
+ "model.gpt_neox.layers.9.mlp.dense_4h_to_h.bias": "hf_ckpt_5.pt",
153
+ "model.gpt_neox.layers.10.input_layernorm.weight": "hf_ckpt_5.pt",
154
+ "model.gpt_neox.layers.10.input_layernorm.bias": "hf_ckpt_5.pt",
155
+ "model.gpt_neox.layers.10.post_attention_layernorm.weight": "hf_ckpt_5.pt",
156
+ "model.gpt_neox.layers.10.post_attention_layernorm.bias": "hf_ckpt_5.pt",
157
+ "model.gpt_neox.layers.10.attention.bias": "hf_ckpt_5.pt",
158
+ "model.gpt_neox.layers.10.attention.masked_bias": "hf_ckpt_5.pt",
159
+ "model.gpt_neox.layers.10.attention.rotary_emb.inv_freq": "hf_ckpt_5.pt",
160
+ "model.gpt_neox.layers.10.attention.query_key_value.weight": "hf_ckpt_5.pt",
161
+ "model.gpt_neox.layers.10.attention.query_key_value.bias": "hf_ckpt_5.pt",
162
+ "model.gpt_neox.layers.10.attention.dense.weight": "hf_ckpt_5.pt",
163
+ "model.gpt_neox.layers.10.attention.dense.bias": "hf_ckpt_5.pt",
164
+ "model.gpt_neox.layers.10.mlp.dense_h_to_4h.weight": "hf_ckpt_5.pt",
165
+ "model.gpt_neox.layers.10.mlp.dense_h_to_4h.bias": "hf_ckpt_5.pt",
166
+ "model.gpt_neox.layers.10.mlp.dense_4h_to_h.weight": "hf_ckpt_5.pt",
167
+ "model.gpt_neox.layers.10.mlp.dense_4h_to_h.bias": "hf_ckpt_5.pt",
168
+ "model.gpt_neox.layers.11.input_layernorm.weight": "hf_ckpt_5.pt",
169
+ "model.gpt_neox.layers.11.input_layernorm.bias": "hf_ckpt_5.pt",
170
+ "model.gpt_neox.layers.11.post_attention_layernorm.weight": "hf_ckpt_6.pt",
171
+ "model.gpt_neox.layers.11.post_attention_layernorm.bias": "hf_ckpt_6.pt",
172
+ "model.gpt_neox.layers.11.attention.bias": "hf_ckpt_6.pt",
173
+ "model.gpt_neox.layers.11.attention.masked_bias": "hf_ckpt_6.pt",
174
+ "model.gpt_neox.layers.11.attention.rotary_emb.inv_freq": "hf_ckpt_6.pt",
175
+ "model.gpt_neox.layers.11.attention.query_key_value.weight": "hf_ckpt_6.pt",
176
+ "model.gpt_neox.layers.11.attention.query_key_value.bias": "hf_ckpt_6.pt",
177
+ "model.gpt_neox.layers.11.attention.dense.weight": "hf_ckpt_6.pt",
178
+ "model.gpt_neox.layers.11.attention.dense.bias": "hf_ckpt_6.pt",
179
+ "model.gpt_neox.layers.11.mlp.dense_h_to_4h.weight": "hf_ckpt_6.pt",
180
+ "model.gpt_neox.layers.11.mlp.dense_h_to_4h.bias": "hf_ckpt_6.pt",
181
+ "model.gpt_neox.layers.11.mlp.dense_4h_to_h.weight": "hf_ckpt_6.pt",
182
+ "model.gpt_neox.layers.11.mlp.dense_4h_to_h.bias": "hf_ckpt_6.pt",
183
+ "model.gpt_neox.layers.12.input_layernorm.weight": "hf_ckpt_6.pt",
184
+ "model.gpt_neox.layers.12.input_layernorm.bias": "hf_ckpt_6.pt",
185
+ "model.gpt_neox.layers.12.post_attention_layernorm.weight": "hf_ckpt_6.pt",
186
+ "model.gpt_neox.layers.12.post_attention_layernorm.bias": "hf_ckpt_6.pt",
187
+ "model.gpt_neox.layers.12.attention.bias": "hf_ckpt_6.pt",
188
+ "model.gpt_neox.layers.12.attention.masked_bias": "hf_ckpt_6.pt",
189
+ "model.gpt_neox.layers.12.attention.rotary_emb.inv_freq": "hf_ckpt_6.pt",
190
+ "model.gpt_neox.layers.12.attention.query_key_value.weight": "hf_ckpt_6.pt",
191
+ "model.gpt_neox.layers.12.attention.query_key_value.bias": "hf_ckpt_6.pt",
192
+ "model.gpt_neox.layers.12.attention.dense.weight": "hf_ckpt_6.pt",
193
+ "model.gpt_neox.layers.12.attention.dense.bias": "hf_ckpt_6.pt",
194
+ "model.gpt_neox.layers.12.mlp.dense_h_to_4h.weight": "hf_ckpt_6.pt",
195
+ "model.gpt_neox.layers.12.mlp.dense_h_to_4h.bias": "hf_ckpt_6.pt",
196
+ "model.gpt_neox.layers.12.mlp.dense_4h_to_h.weight": "hf_ckpt_6.pt",
197
+ "model.gpt_neox.layers.12.mlp.dense_4h_to_h.bias": "hf_ckpt_6.pt",
198
+ "model.gpt_neox.layers.13.input_layernorm.weight": "hf_ckpt_7.pt",
199
+ "model.gpt_neox.layers.13.input_layernorm.bias": "hf_ckpt_7.pt",
200
+ "model.gpt_neox.layers.13.post_attention_layernorm.weight": "hf_ckpt_7.pt",
201
+ "model.gpt_neox.layers.13.post_attention_layernorm.bias": "hf_ckpt_7.pt",
202
+ "model.gpt_neox.layers.13.attention.bias": "hf_ckpt_7.pt",
203
+ "model.gpt_neox.layers.13.attention.masked_bias": "hf_ckpt_7.pt",
204
+ "model.gpt_neox.layers.13.attention.rotary_emb.inv_freq": "hf_ckpt_7.pt",
205
+ "model.gpt_neox.layers.13.attention.query_key_value.weight": "hf_ckpt_7.pt",
206
+ "model.gpt_neox.layers.13.attention.query_key_value.bias": "hf_ckpt_7.pt",
207
+ "model.gpt_neox.layers.13.attention.dense.weight": "hf_ckpt_7.pt",
208
+ "model.gpt_neox.layers.13.attention.dense.bias": "hf_ckpt_7.pt",
209
+ "model.gpt_neox.layers.13.mlp.dense_h_to_4h.weight": "hf_ckpt_7.pt",
210
+ "model.gpt_neox.layers.13.mlp.dense_h_to_4h.bias": "hf_ckpt_7.pt",
211
+ "model.gpt_neox.layers.13.mlp.dense_4h_to_h.weight": "hf_ckpt_7.pt",
212
+ "model.gpt_neox.layers.13.mlp.dense_4h_to_h.bias": "hf_ckpt_7.pt",
213
+ "model.gpt_neox.layers.14.input_layernorm.weight": "hf_ckpt_7.pt",
214
+ "model.gpt_neox.layers.14.input_layernorm.bias": "hf_ckpt_7.pt",
215
+ "model.gpt_neox.layers.14.post_attention_layernorm.weight": "hf_ckpt_7.pt",
216
+ "model.gpt_neox.layers.14.post_attention_layernorm.bias": "hf_ckpt_7.pt",
217
+ "model.gpt_neox.layers.14.attention.bias": "hf_ckpt_7.pt",
218
+ "model.gpt_neox.layers.14.attention.masked_bias": "hf_ckpt_7.pt",
219
+ "model.gpt_neox.layers.14.attention.rotary_emb.inv_freq": "hf_ckpt_7.pt",
220
+ "model.gpt_neox.layers.14.attention.query_key_value.weight": "hf_ckpt_7.pt",
221
+ "model.gpt_neox.layers.14.attention.query_key_value.bias": "hf_ckpt_7.pt",
222
+ "model.gpt_neox.layers.14.attention.dense.weight": "hf_ckpt_7.pt",
223
+ "model.gpt_neox.layers.14.attention.dense.bias": "hf_ckpt_7.pt",
224
+ "model.gpt_neox.layers.14.mlp.dense_h_to_4h.weight": "hf_ckpt_7.pt",
225
+ "model.gpt_neox.layers.14.mlp.dense_h_to_4h.bias": "hf_ckpt_7.pt",
226
+ "model.gpt_neox.layers.14.mlp.dense_4h_to_h.weight": "hf_ckpt_8.pt",
227
+ "model.gpt_neox.layers.14.mlp.dense_4h_to_h.bias": "hf_ckpt_8.pt",
228
+ "model.gpt_neox.layers.15.input_layernorm.weight": "hf_ckpt_8.pt",
229
+ "model.gpt_neox.layers.15.input_layernorm.bias": "hf_ckpt_8.pt",
230
+ "model.gpt_neox.layers.15.post_attention_layernorm.weight": "hf_ckpt_8.pt",
231
+ "model.gpt_neox.layers.15.post_attention_layernorm.bias": "hf_ckpt_8.pt",
232
+ "model.gpt_neox.layers.15.attention.bias": "hf_ckpt_8.pt",
233
+ "model.gpt_neox.layers.15.attention.masked_bias": "hf_ckpt_8.pt",
234
+ "model.gpt_neox.layers.15.attention.rotary_emb.inv_freq": "hf_ckpt_8.pt",
235
+ "model.gpt_neox.layers.15.attention.query_key_value.weight": "hf_ckpt_8.pt",
236
+ "model.gpt_neox.layers.15.attention.query_key_value.bias": "hf_ckpt_8.pt",
237
+ "model.gpt_neox.layers.15.attention.dense.weight": "hf_ckpt_8.pt",
238
+ "model.gpt_neox.layers.15.attention.dense.bias": "hf_ckpt_8.pt",
239
+ "model.gpt_neox.layers.15.mlp.dense_h_to_4h.weight": "hf_ckpt_8.pt",
240
+ "model.gpt_neox.layers.15.mlp.dense_h_to_4h.bias": "hf_ckpt_8.pt",
241
+ "model.gpt_neox.layers.15.mlp.dense_4h_to_h.weight": "hf_ckpt_8.pt",
242
+ "model.gpt_neox.layers.15.mlp.dense_4h_to_h.bias": "hf_ckpt_8.pt",
243
+ "model.gpt_neox.layers.16.input_layernorm.weight": "hf_ckpt_8.pt",
244
+ "model.gpt_neox.layers.16.input_layernorm.bias": "hf_ckpt_8.pt",
245
+ "model.gpt_neox.layers.16.post_attention_layernorm.weight": "hf_ckpt_8.pt",
246
+ "model.gpt_neox.layers.16.post_attention_layernorm.bias": "hf_ckpt_8.pt",
247
+ "model.gpt_neox.layers.16.attention.bias": "hf_ckpt_8.pt",
248
+ "model.gpt_neox.layers.16.attention.masked_bias": "hf_ckpt_8.pt",
249
+ "model.gpt_neox.layers.16.attention.rotary_emb.inv_freq": "hf_ckpt_8.pt",
250
+ "model.gpt_neox.layers.16.attention.query_key_value.weight": "hf_ckpt_8.pt",
251
+ "model.gpt_neox.layers.16.attention.query_key_value.bias": "hf_ckpt_8.pt",
252
+ "model.gpt_neox.layers.16.attention.dense.weight": "hf_ckpt_8.pt",
253
+ "model.gpt_neox.layers.16.attention.dense.bias": "hf_ckpt_8.pt",
254
+ "model.gpt_neox.layers.16.mlp.dense_h_to_4h.weight": "hf_ckpt_9.pt",
255
+ "model.gpt_neox.layers.16.mlp.dense_h_to_4h.bias": "hf_ckpt_9.pt",
256
+ "model.gpt_neox.layers.16.mlp.dense_4h_to_h.weight": "hf_ckpt_9.pt",
257
+ "model.gpt_neox.layers.16.mlp.dense_4h_to_h.bias": "hf_ckpt_9.pt",
258
+ "model.gpt_neox.layers.17.input_layernorm.weight": "hf_ckpt_9.pt",
259
+ "model.gpt_neox.layers.17.input_layernorm.bias": "hf_ckpt_9.pt",
260
+ "model.gpt_neox.layers.17.post_attention_layernorm.weight": "hf_ckpt_9.pt",
261
+ "model.gpt_neox.layers.17.post_attention_layernorm.bias": "hf_ckpt_9.pt",
262
+ "model.gpt_neox.layers.17.attention.bias": "hf_ckpt_9.pt",
263
+ "model.gpt_neox.layers.17.attention.masked_bias": "hf_ckpt_9.pt",
264
+ "model.gpt_neox.layers.17.attention.rotary_emb.inv_freq": "hf_ckpt_9.pt",
265
+ "model.gpt_neox.layers.17.attention.query_key_value.weight": "hf_ckpt_9.pt",
266
+ "model.gpt_neox.layers.17.attention.query_key_value.bias": "hf_ckpt_9.pt",
267
+ "model.gpt_neox.layers.17.attention.dense.weight": "hf_ckpt_9.pt",
268
+ "model.gpt_neox.layers.17.attention.dense.bias": "hf_ckpt_9.pt",
269
+ "model.gpt_neox.layers.17.mlp.dense_h_to_4h.weight": "hf_ckpt_9.pt",
270
+ "model.gpt_neox.layers.17.mlp.dense_h_to_4h.bias": "hf_ckpt_9.pt",
271
+ "model.gpt_neox.layers.17.mlp.dense_4h_to_h.weight": "hf_ckpt_9.pt",
272
+ "model.gpt_neox.layers.17.mlp.dense_4h_to_h.bias": "hf_ckpt_9.pt",
273
+ "model.gpt_neox.layers.18.input_layernorm.weight": "hf_ckpt_9.pt",
274
+ "model.gpt_neox.layers.18.input_layernorm.bias": "hf_ckpt_9.pt",
275
+ "model.gpt_neox.layers.18.post_attention_layernorm.weight": "hf_ckpt_9.pt",
276
+ "model.gpt_neox.layers.18.post_attention_layernorm.bias": "hf_ckpt_9.pt",
277
+ "model.gpt_neox.layers.18.attention.bias": "hf_ckpt_9.pt",
278
+ "model.gpt_neox.layers.18.attention.masked_bias": "hf_ckpt_9.pt",
279
+ "model.gpt_neox.layers.18.attention.rotary_emb.inv_freq": "hf_ckpt_9.pt",
280
+ "model.gpt_neox.layers.18.attention.query_key_value.weight": "hf_ckpt_9.pt",
281
+ "model.gpt_neox.layers.18.attention.query_key_value.bias": "hf_ckpt_9.pt",
282
+ "model.gpt_neox.layers.18.attention.dense.weight": "hf_ckpt_10.pt",
283
+ "model.gpt_neox.layers.18.attention.dense.bias": "hf_ckpt_10.pt",
284
+ "model.gpt_neox.layers.18.mlp.dense_h_to_4h.weight": "hf_ckpt_10.pt",
285
+ "model.gpt_neox.layers.18.mlp.dense_h_to_4h.bias": "hf_ckpt_10.pt",
286
+ "model.gpt_neox.layers.18.mlp.dense_4h_to_h.weight": "hf_ckpt_10.pt",
287
+ "model.gpt_neox.layers.18.mlp.dense_4h_to_h.bias": "hf_ckpt_10.pt",
288
+ "model.gpt_neox.layers.19.input_layernorm.weight": "hf_ckpt_10.pt",
289
+ "model.gpt_neox.layers.19.input_layernorm.bias": "hf_ckpt_10.pt",
290
+ "model.gpt_neox.layers.19.post_attention_layernorm.weight": "hf_ckpt_10.pt",
291
+ "model.gpt_neox.layers.19.post_attention_layernorm.bias": "hf_ckpt_10.pt",
292
+ "model.gpt_neox.layers.19.attention.bias": "hf_ckpt_10.pt",
293
+ "model.gpt_neox.layers.19.attention.masked_bias": "hf_ckpt_10.pt",
294
+ "model.gpt_neox.layers.19.attention.rotary_emb.inv_freq": "hf_ckpt_10.pt",
295
+ "model.gpt_neox.layers.19.attention.query_key_value.weight": "hf_ckpt_10.pt",
296
+ "model.gpt_neox.layers.19.attention.query_key_value.bias": "hf_ckpt_10.pt",
297
+ "model.gpt_neox.layers.19.attention.dense.weight": "hf_ckpt_10.pt",
298
+ "model.gpt_neox.layers.19.attention.dense.bias": "hf_ckpt_10.pt",
299
+ "model.gpt_neox.layers.19.mlp.dense_h_to_4h.weight": "hf_ckpt_10.pt",
300
+ "model.gpt_neox.layers.19.mlp.dense_h_to_4h.bias": "hf_ckpt_10.pt",
301
+ "model.gpt_neox.layers.19.mlp.dense_4h_to_h.weight": "hf_ckpt_10.pt",
302
+ "model.gpt_neox.layers.19.mlp.dense_4h_to_h.bias": "hf_ckpt_10.pt",
303
+ "model.gpt_neox.layers.20.input_layernorm.weight": "hf_ckpt_10.pt",
304
+ "model.gpt_neox.layers.20.input_layernorm.bias": "hf_ckpt_10.pt",
305
+ "model.gpt_neox.layers.20.post_attention_layernorm.weight": "hf_ckpt_10.pt",
306
+ "model.gpt_neox.layers.20.post_attention_layernorm.bias": "hf_ckpt_10.pt",
307
+ "model.gpt_neox.layers.20.attention.bias": "hf_ckpt_10.pt",
308
+ "model.gpt_neox.layers.20.attention.masked_bias": "hf_ckpt_10.pt",
309
+ "model.gpt_neox.layers.20.attention.rotary_emb.inv_freq": "hf_ckpt_10.pt",
310
+ "model.gpt_neox.layers.20.attention.query_key_value.weight": "hf_ckpt_11.pt",
311
+ "model.gpt_neox.layers.20.attention.query_key_value.bias": "hf_ckpt_11.pt",
312
+ "model.gpt_neox.layers.20.attention.dense.weight": "hf_ckpt_11.pt",
313
+ "model.gpt_neox.layers.20.attention.dense.bias": "hf_ckpt_11.pt",
314
+ "model.gpt_neox.layers.20.mlp.dense_h_to_4h.weight": "hf_ckpt_11.pt",
315
+ "model.gpt_neox.layers.20.mlp.dense_h_to_4h.bias": "hf_ckpt_11.pt",
316
+ "model.gpt_neox.layers.20.mlp.dense_4h_to_h.weight": "hf_ckpt_11.pt",
317
+ "model.gpt_neox.layers.20.mlp.dense_4h_to_h.bias": "hf_ckpt_11.pt",
318
+ "model.gpt_neox.layers.21.input_layernorm.weight": "hf_ckpt_11.pt",
319
+ "model.gpt_neox.layers.21.input_layernorm.bias": "hf_ckpt_11.pt",
320
+ "model.gpt_neox.layers.21.post_attention_layernorm.weight": "hf_ckpt_11.pt",
321
+ "model.gpt_neox.layers.21.post_attention_layernorm.bias": "hf_ckpt_11.pt",
322
+ "model.gpt_neox.layers.21.attention.bias": "hf_ckpt_11.pt",
323
+ "model.gpt_neox.layers.21.attention.masked_bias": "hf_ckpt_11.pt",
324
+ "model.gpt_neox.layers.21.attention.rotary_emb.inv_freq": "hf_ckpt_11.pt",
325
+ "model.gpt_neox.layers.21.attention.query_key_value.weight": "hf_ckpt_11.pt",
326
+ "model.gpt_neox.layers.21.attention.query_key_value.bias": "hf_ckpt_11.pt",
327
+ "model.gpt_neox.layers.21.attention.dense.weight": "hf_ckpt_11.pt",
328
+ "model.gpt_neox.layers.21.attention.dense.bias": "hf_ckpt_11.pt",
329
+ "model.gpt_neox.layers.21.mlp.dense_h_to_4h.weight": "hf_ckpt_11.pt",
330
+ "model.gpt_neox.layers.21.mlp.dense_h_to_4h.bias": "hf_ckpt_11.pt",
331
+ "model.gpt_neox.layers.21.mlp.dense_4h_to_h.weight": "hf_ckpt_11.pt",
332
+ "model.gpt_neox.layers.21.mlp.dense_4h_to_h.bias": "hf_ckpt_11.pt",
333
+ "model.gpt_neox.layers.22.input_layernorm.weight": "hf_ckpt_11.pt",
334
+ "model.gpt_neox.layers.22.input_layernorm.bias": "hf_ckpt_11.pt",
335
+ "model.gpt_neox.layers.22.post_attention_layernorm.weight": "hf_ckpt_11.pt",
336
+ "model.gpt_neox.layers.22.post_attention_layernorm.bias": "hf_ckpt_11.pt",
337
+ "model.gpt_neox.layers.22.attention.bias": "hf_ckpt_11.pt",
338
+ "model.gpt_neox.layers.22.attention.masked_bias": "hf_ckpt_12.pt",
339
+ "model.gpt_neox.layers.22.attention.rotary_emb.inv_freq": "hf_ckpt_12.pt",
340
+ "model.gpt_neox.layers.22.attention.query_key_value.weight": "hf_ckpt_12.pt",
341
+ "model.gpt_neox.layers.22.attention.query_key_value.bias": "hf_ckpt_12.pt",
342
+ "model.gpt_neox.layers.22.attention.dense.weight": "hf_ckpt_12.pt",
343
+ "model.gpt_neox.layers.22.attention.dense.bias": "hf_ckpt_12.pt",
344
+ "model.gpt_neox.layers.22.mlp.dense_h_to_4h.weight": "hf_ckpt_12.pt",
345
+ "model.gpt_neox.layers.22.mlp.dense_h_to_4h.bias": "hf_ckpt_12.pt",
346
+ "model.gpt_neox.layers.22.mlp.dense_4h_to_h.weight": "hf_ckpt_12.pt",
347
+ "model.gpt_neox.layers.22.mlp.dense_4h_to_h.bias": "hf_ckpt_12.pt",
348
+ "model.gpt_neox.layers.23.input_layernorm.weight": "hf_ckpt_12.pt",
349
+ "model.gpt_neox.layers.23.input_layernorm.bias": "hf_ckpt_12.pt",
350
+ "model.gpt_neox.layers.23.post_attention_layernorm.weight": "hf_ckpt_12.pt",
351
+ "model.gpt_neox.layers.23.post_attention_layernorm.bias": "hf_ckpt_12.pt",
352
+ "model.gpt_neox.layers.23.attention.bias": "hf_ckpt_12.pt",
353
+ "model.gpt_neox.layers.23.attention.masked_bias": "hf_ckpt_12.pt",
354
+ "model.gpt_neox.layers.23.attention.rotary_emb.inv_freq": "hf_ckpt_12.pt",
355
+ "model.gpt_neox.layers.23.attention.query_key_value.weight": "hf_ckpt_12.pt",
356
+ "model.gpt_neox.layers.23.attention.query_key_value.bias": "hf_ckpt_12.pt",
357
+ "model.gpt_neox.layers.23.attention.dense.weight": "hf_ckpt_12.pt",
358
+ "model.gpt_neox.layers.23.attention.dense.bias": "hf_ckpt_12.pt",
359
+ "model.gpt_neox.layers.23.mlp.dense_h_to_4h.weight": "hf_ckpt_12.pt",
360
+ "model.gpt_neox.layers.23.mlp.dense_h_to_4h.bias": "hf_ckpt_12.pt",
361
+ "model.gpt_neox.layers.23.mlp.dense_4h_to_h.weight": "hf_ckpt_12.pt",
362
+ "model.gpt_neox.layers.23.mlp.dense_4h_to_h.bias": "hf_ckpt_12.pt",
363
+ "model.gpt_neox.layers.24.input_layernorm.weight": "hf_ckpt_12.pt",
364
+ "model.gpt_neox.layers.24.input_layernorm.bias": "hf_ckpt_12.pt",
365
+ "model.gpt_neox.layers.24.post_attention_layernorm.weight": "hf_ckpt_12.pt",
366
+ "model.gpt_neox.layers.24.post_attention_layernorm.bias": "hf_ckpt_13.pt",
367
+ "model.gpt_neox.layers.24.attention.bias": "hf_ckpt_13.pt",
368
+ "model.gpt_neox.layers.24.attention.masked_bias": "hf_ckpt_13.pt",
369
+ "model.gpt_neox.layers.24.attention.rotary_emb.inv_freq": "hf_ckpt_13.pt",
370
+ "model.gpt_neox.layers.24.attention.query_key_value.weight": "hf_ckpt_13.pt",
371
+ "model.gpt_neox.layers.24.attention.query_key_value.bias": "hf_ckpt_13.pt",
372
+ "model.gpt_neox.layers.24.attention.dense.weight": "hf_ckpt_13.pt",
373
+ "model.gpt_neox.layers.24.attention.dense.bias": "hf_ckpt_13.pt",
374
+ "model.gpt_neox.layers.24.mlp.dense_h_to_4h.weight": "hf_ckpt_13.pt",
375
+ "model.gpt_neox.layers.24.mlp.dense_h_to_4h.bias": "hf_ckpt_13.pt",
376
+ "model.gpt_neox.layers.24.mlp.dense_4h_to_h.weight": "hf_ckpt_13.pt",
377
+ "model.gpt_neox.layers.24.mlp.dense_4h_to_h.bias": "hf_ckpt_13.pt",
378
+ "model.gpt_neox.layers.25.input_layernorm.weight": "hf_ckpt_13.pt",
379
+ "model.gpt_neox.layers.25.input_layernorm.bias": "hf_ckpt_13.pt",
380
+ "model.gpt_neox.layers.25.post_attention_layernorm.weight": "hf_ckpt_13.pt",
381
+ "model.gpt_neox.layers.25.post_attention_layernorm.bias": "hf_ckpt_13.pt",
382
+ "model.gpt_neox.layers.25.attention.bias": "hf_ckpt_13.pt",
383
+ "model.gpt_neox.layers.25.attention.masked_bias": "hf_ckpt_13.pt",
384
+ "model.gpt_neox.layers.25.attention.rotary_emb.inv_freq": "hf_ckpt_13.pt",
385
+ "model.gpt_neox.layers.25.attention.query_key_value.weight": "hf_ckpt_13.pt",
386
+ "model.gpt_neox.layers.25.attention.query_key_value.bias": "hf_ckpt_13.pt",
387
+ "model.gpt_neox.layers.25.attention.dense.weight": "hf_ckpt_13.pt",
388
+ "model.gpt_neox.layers.25.attention.dense.bias": "hf_ckpt_13.pt",
389
+ "model.gpt_neox.layers.25.mlp.dense_h_to_4h.weight": "hf_ckpt_13.pt",
390
+ "model.gpt_neox.layers.25.mlp.dense_h_to_4h.bias": "hf_ckpt_13.pt",
391
+ "model.gpt_neox.layers.25.mlp.dense_4h_to_h.weight": "hf_ckpt_13.pt",
392
+ "model.gpt_neox.layers.25.mlp.dense_4h_to_h.bias": "hf_ckpt_13.pt",
393
+ "model.gpt_neox.layers.26.input_layernorm.weight": "hf_ckpt_13.pt",
394
+ "model.gpt_neox.layers.26.input_layernorm.bias": "hf_ckpt_14.pt",
395
+ "model.gpt_neox.layers.26.post_attention_layernorm.weight": "hf_ckpt_14.pt",
396
+ "model.gpt_neox.layers.26.post_attention_layernorm.bias": "hf_ckpt_14.pt",
397
+ "model.gpt_neox.layers.26.attention.bias": "hf_ckpt_14.pt",
398
+ "model.gpt_neox.layers.26.attention.masked_bias": "hf_ckpt_14.pt",
399
+ "model.gpt_neox.layers.26.attention.rotary_emb.inv_freq": "hf_ckpt_14.pt",
400
+ "model.gpt_neox.layers.26.attention.query_key_value.weight": "hf_ckpt_14.pt",
401
+ "model.gpt_neox.layers.26.attention.query_key_value.bias": "hf_ckpt_14.pt",
402
+ "model.gpt_neox.layers.26.attention.dense.weight": "hf_ckpt_14.pt",
403
+ "model.gpt_neox.layers.26.attention.dense.bias": "hf_ckpt_14.pt",
404
+ "model.gpt_neox.layers.26.mlp.dense_h_to_4h.weight": "hf_ckpt_14.pt",
405
+ "model.gpt_neox.layers.26.mlp.dense_h_to_4h.bias": "hf_ckpt_14.pt",
406
+ "model.gpt_neox.layers.26.mlp.dense_4h_to_h.weight": "hf_ckpt_14.pt",
407
+ "model.gpt_neox.layers.26.mlp.dense_4h_to_h.bias": "hf_ckpt_14.pt",
408
+ "model.gpt_neox.layers.27.input_layernorm.weight": "hf_ckpt_14.pt",
409
+ "model.gpt_neox.layers.27.input_layernorm.bias": "hf_ckpt_14.pt",
410
+ "model.gpt_neox.layers.27.post_attention_layernorm.weight": "hf_ckpt_14.pt",
411
+ "model.gpt_neox.layers.27.post_attention_layernorm.bias": "hf_ckpt_14.pt",
412
+ "model.gpt_neox.layers.27.attention.bias": "hf_ckpt_14.pt",
413
+ "model.gpt_neox.layers.27.attention.masked_bias": "hf_ckpt_14.pt",
414
+ "model.gpt_neox.layers.27.attention.rotary_emb.inv_freq": "hf_ckpt_14.pt",
415
+ "model.gpt_neox.layers.27.attention.query_key_value.weight": "hf_ckpt_14.pt",
416
+ "model.gpt_neox.layers.27.attention.query_key_value.bias": "hf_ckpt_14.pt",
417
+ "model.gpt_neox.layers.27.attention.dense.weight": "hf_ckpt_14.pt",
418
+ "model.gpt_neox.layers.27.attention.dense.bias": "hf_ckpt_14.pt",
419
+ "model.gpt_neox.layers.27.mlp.dense_h_to_4h.weight": "hf_ckpt_14.pt",
420
+ "model.gpt_neox.layers.27.mlp.dense_h_to_4h.bias": "hf_ckpt_14.pt",
421
+ "model.gpt_neox.layers.27.mlp.dense_4h_to_h.weight": "hf_ckpt_14.pt",
422
+ "model.gpt_neox.layers.27.mlp.dense_4h_to_h.bias": "hf_ckpt_15.pt",
423
+ "model.gpt_neox.layers.28.input_layernorm.weight": "hf_ckpt_15.pt",
424
+ "model.gpt_neox.layers.28.input_layernorm.bias": "hf_ckpt_15.pt",
425
+ "model.gpt_neox.layers.28.post_attention_layernorm.weight": "hf_ckpt_15.pt",
426
+ "model.gpt_neox.layers.28.post_attention_layernorm.bias": "hf_ckpt_15.pt",
427
+ "model.gpt_neox.layers.28.attention.bias": "hf_ckpt_15.pt",
428
+ "model.gpt_neox.layers.28.attention.masked_bias": "hf_ckpt_15.pt",
429
+ "model.gpt_neox.layers.28.attention.rotary_emb.inv_freq": "hf_ckpt_15.pt",
430
+ "model.gpt_neox.layers.28.attention.query_key_value.weight": "hf_ckpt_15.pt",
431
+ "model.gpt_neox.layers.28.attention.query_key_value.bias": "hf_ckpt_15.pt",
432
+ "model.gpt_neox.layers.28.attention.dense.weight": "hf_ckpt_15.pt",
433
+ "model.gpt_neox.layers.28.attention.dense.bias": "hf_ckpt_15.pt",
434
+ "model.gpt_neox.layers.28.mlp.dense_h_to_4h.weight": "hf_ckpt_15.pt",
435
+ "model.gpt_neox.layers.28.mlp.dense_h_to_4h.bias": "hf_ckpt_15.pt",
436
+ "model.gpt_neox.layers.28.mlp.dense_4h_to_h.weight": "hf_ckpt_15.pt",
437
+ "model.gpt_neox.layers.28.mlp.dense_4h_to_h.bias": "hf_ckpt_15.pt",
438
+ "model.gpt_neox.layers.29.input_layernorm.weight": "hf_ckpt_15.pt",
439
+ "model.gpt_neox.layers.29.input_layernorm.bias": "hf_ckpt_15.pt",
440
+ "model.gpt_neox.layers.29.post_attention_layernorm.weight": "hf_ckpt_15.pt",
441
+ "model.gpt_neox.layers.29.post_attention_layernorm.bias": "hf_ckpt_15.pt",
442
+ "model.gpt_neox.layers.29.attention.bias": "hf_ckpt_15.pt",
443
+ "model.gpt_neox.layers.29.attention.masked_bias": "hf_ckpt_15.pt",
444
+ "model.gpt_neox.layers.29.attention.rotary_emb.inv_freq": "hf_ckpt_15.pt",
445
+ "model.gpt_neox.layers.29.attention.query_key_value.weight": "hf_ckpt_15.pt",
446
+ "model.gpt_neox.layers.29.attention.query_key_value.bias": "hf_ckpt_15.pt",
447
+ "model.gpt_neox.layers.29.attention.dense.weight": "hf_ckpt_15.pt",
448
+ "model.gpt_neox.layers.29.attention.dense.bias": "hf_ckpt_15.pt",
449
+ "model.gpt_neox.layers.29.mlp.dense_h_to_4h.weight": "hf_ckpt_15.pt",
450
+ "model.gpt_neox.layers.29.mlp.dense_h_to_4h.bias": "hf_ckpt_16.pt",
451
+ "model.gpt_neox.layers.29.mlp.dense_4h_to_h.weight": "hf_ckpt_16.pt",
452
+ "model.gpt_neox.layers.29.mlp.dense_4h_to_h.bias": "hf_ckpt_16.pt",
453
+ "model.gpt_neox.layers.30.input_layernorm.weight": "hf_ckpt_16.pt",
454
+ "model.gpt_neox.layers.30.input_layernorm.bias": "hf_ckpt_16.pt",
455
+ "model.gpt_neox.layers.30.post_attention_layernorm.weight": "hf_ckpt_16.pt",
456
+ "model.gpt_neox.layers.30.post_attention_layernorm.bias": "hf_ckpt_16.pt",
457
+ "model.gpt_neox.layers.30.attention.bias": "hf_ckpt_16.pt",
458
+ "model.gpt_neox.layers.30.attention.masked_bias": "hf_ckpt_16.pt",
459
+ "model.gpt_neox.layers.30.attention.rotary_emb.inv_freq": "hf_ckpt_16.pt",
460
+ "model.gpt_neox.layers.30.attention.query_key_value.weight": "hf_ckpt_16.pt",
461
+ "model.gpt_neox.layers.30.attention.query_key_value.bias": "hf_ckpt_16.pt",
462
+ "model.gpt_neox.layers.30.attention.dense.weight": "hf_ckpt_16.pt",
463
+ "model.gpt_neox.layers.30.attention.dense.bias": "hf_ckpt_16.pt",
464
+ "model.gpt_neox.layers.30.mlp.dense_h_to_4h.weight": "hf_ckpt_16.pt",
465
+ "model.gpt_neox.layers.30.mlp.dense_h_to_4h.bias": "hf_ckpt_16.pt",
466
+ "model.gpt_neox.layers.30.mlp.dense_4h_to_h.weight": "hf_ckpt_16.pt",
467
+ "model.gpt_neox.layers.30.mlp.dense_4h_to_h.bias": "hf_ckpt_16.pt",
468
+ "model.gpt_neox.layers.31.input_layernorm.weight": "hf_ckpt_16.pt",
469
+ "model.gpt_neox.layers.31.input_layernorm.bias": "hf_ckpt_16.pt",
470
+ "model.gpt_neox.layers.31.post_attention_layernorm.weight": "hf_ckpt_16.pt",
471
+ "model.gpt_neox.layers.31.post_attention_layernorm.bias": "hf_ckpt_16.pt",
472
+ "model.gpt_neox.layers.31.attention.bias": "hf_ckpt_16.pt",
473
+ "model.gpt_neox.layers.31.attention.masked_bias": "hf_ckpt_16.pt",
474
+ "model.gpt_neox.layers.31.attention.rotary_emb.inv_freq": "hf_ckpt_16.pt",
475
+ "model.gpt_neox.layers.31.attention.query_key_value.weight": "hf_ckpt_16.pt",
476
+ "model.gpt_neox.layers.31.attention.query_key_value.bias": "hf_ckpt_16.pt",
477
+ "model.gpt_neox.layers.31.attention.dense.weight": "hf_ckpt_16.pt",
478
+ "model.gpt_neox.layers.31.attention.dense.bias": "hf_ckpt_17.pt",
479
+ "model.gpt_neox.layers.31.mlp.dense_h_to_4h.weight": "hf_ckpt_17.pt",
480
+ "model.gpt_neox.layers.31.mlp.dense_h_to_4h.bias": "hf_ckpt_17.pt",
481
+ "model.gpt_neox.layers.31.mlp.dense_4h_to_h.weight": "hf_ckpt_17.pt",
482
+ "model.gpt_neox.layers.31.mlp.dense_4h_to_h.bias": "hf_ckpt_17.pt",
483
+ "model.gpt_neox.layers.32.input_layernorm.weight": "hf_ckpt_17.pt",
484
+ "model.gpt_neox.layers.32.input_layernorm.bias": "hf_ckpt_17.pt",
485
+ "model.gpt_neox.layers.32.post_attention_layernorm.weight": "hf_ckpt_17.pt",
486
+ "model.gpt_neox.layers.32.post_attention_layernorm.bias": "hf_ckpt_17.pt",
487
+ "model.gpt_neox.layers.32.attention.bias": "hf_ckpt_17.pt",
488
+ "model.gpt_neox.layers.32.attention.masked_bias": "hf_ckpt_17.pt",
489
+ "model.gpt_neox.layers.32.attention.rotary_emb.inv_freq": "hf_ckpt_17.pt",
490
+ "model.gpt_neox.layers.32.attention.query_key_value.weight": "hf_ckpt_17.pt",
491
+ "model.gpt_neox.layers.32.attention.query_key_value.bias": "hf_ckpt_17.pt",
492
+ "model.gpt_neox.layers.32.attention.dense.weight": "hf_ckpt_17.pt",
493
+ "model.gpt_neox.layers.32.attention.dense.bias": "hf_ckpt_17.pt",
494
+ "model.gpt_neox.layers.32.mlp.dense_h_to_4h.weight": "hf_ckpt_17.pt",
495
+ "model.gpt_neox.layers.32.mlp.dense_h_to_4h.bias": "hf_ckpt_17.pt",
496
+ "model.gpt_neox.layers.32.mlp.dense_4h_to_h.weight": "hf_ckpt_17.pt",
497
+ "model.gpt_neox.layers.32.mlp.dense_4h_to_h.bias": "hf_ckpt_17.pt",
498
+ "model.gpt_neox.layers.33.input_layernorm.weight": "hf_ckpt_17.pt",
499
+ "model.gpt_neox.layers.33.input_layernorm.bias": "hf_ckpt_17.pt",
500
+ "model.gpt_neox.layers.33.post_attention_layernorm.weight": "hf_ckpt_17.pt",
501
+ "model.gpt_neox.layers.33.post_attention_layernorm.bias": "hf_ckpt_17.pt",
502
+ "model.gpt_neox.layers.33.attention.bias": "hf_ckpt_17.pt",
503
+ "model.gpt_neox.layers.33.attention.masked_bias": "hf_ckpt_17.pt",
504
+ "model.gpt_neox.layers.33.attention.rotary_emb.inv_freq": "hf_ckpt_17.pt",
505
+ "model.gpt_neox.layers.33.attention.query_key_value.weight": "hf_ckpt_17.pt",
506
+ "model.gpt_neox.layers.33.attention.query_key_value.bias": "hf_ckpt_18.pt",
507
+ "model.gpt_neox.layers.33.attention.dense.weight": "hf_ckpt_18.pt",
508
+ "model.gpt_neox.layers.33.attention.dense.bias": "hf_ckpt_18.pt",
509
+ "model.gpt_neox.layers.33.mlp.dense_h_to_4h.weight": "hf_ckpt_18.pt",
510
+ "model.gpt_neox.layers.33.mlp.dense_h_to_4h.bias": "hf_ckpt_18.pt",
511
+ "model.gpt_neox.layers.33.mlp.dense_4h_to_h.weight": "hf_ckpt_18.pt",
512
+ "model.gpt_neox.layers.33.mlp.dense_4h_to_h.bias": "hf_ckpt_18.pt",
513
+ "model.gpt_neox.layers.34.input_layernorm.weight": "hf_ckpt_18.pt",
514
+ "model.gpt_neox.layers.34.input_layernorm.bias": "hf_ckpt_18.pt",
515
+ "model.gpt_neox.layers.34.post_attention_layernorm.weight": "hf_ckpt_18.pt",
516
+ "model.gpt_neox.layers.34.post_attention_layernorm.bias": "hf_ckpt_18.pt",
517
+ "model.gpt_neox.layers.34.attention.bias": "hf_ckpt_18.pt",
518
+ "model.gpt_neox.layers.34.attention.masked_bias": "hf_ckpt_18.pt",
519
+ "model.gpt_neox.layers.34.attention.rotary_emb.inv_freq": "hf_ckpt_18.pt",
520
+ "model.gpt_neox.layers.34.attention.query_key_value.weight": "hf_ckpt_18.pt",
521
+ "model.gpt_neox.layers.34.attention.query_key_value.bias": "hf_ckpt_18.pt",
522
+ "model.gpt_neox.layers.34.attention.dense.weight": "hf_ckpt_18.pt",
523
+ "model.gpt_neox.layers.34.attention.dense.bias": "hf_ckpt_18.pt",
524
+ "model.gpt_neox.layers.34.mlp.dense_h_to_4h.weight": "hf_ckpt_18.pt",
525
+ "model.gpt_neox.layers.34.mlp.dense_h_to_4h.bias": "hf_ckpt_18.pt",
526
+ "model.gpt_neox.layers.34.mlp.dense_4h_to_h.weight": "hf_ckpt_18.pt",
527
+ "model.gpt_neox.layers.34.mlp.dense_4h_to_h.bias": "hf_ckpt_18.pt",
528
+ "model.gpt_neox.layers.35.input_layernorm.weight": "hf_ckpt_18.pt",
529
+ "model.gpt_neox.layers.35.input_layernorm.bias": "hf_ckpt_18.pt",
530
+ "model.gpt_neox.layers.35.post_attention_layernorm.weight": "hf_ckpt_18.pt",
531
+ "model.gpt_neox.layers.35.post_attention_layernorm.bias": "hf_ckpt_18.pt",
532
+ "model.gpt_neox.layers.35.attention.bias": "hf_ckpt_18.pt",
533
+ "model.gpt_neox.layers.35.attention.masked_bias": "hf_ckpt_18.pt",
534
+ "model.gpt_neox.layers.35.attention.rotary_emb.inv_freq": "hf_ckpt_19.pt",
535
+ "model.gpt_neox.layers.35.attention.query_key_value.weight": "hf_ckpt_19.pt",
536
+ "model.gpt_neox.layers.35.attention.query_key_value.bias": "hf_ckpt_19.pt",
537
+ "model.gpt_neox.layers.35.attention.dense.weight": "hf_ckpt_19.pt",
538
+ "model.gpt_neox.layers.35.attention.dense.bias": "hf_ckpt_19.pt",
539
+ "model.gpt_neox.layers.35.mlp.dense_h_to_4h.weight": "hf_ckpt_19.pt",
540
+ "model.gpt_neox.layers.35.mlp.dense_h_to_4h.bias": "hf_ckpt_19.pt",
541
+ "model.gpt_neox.layers.35.mlp.dense_4h_to_h.weight": "hf_ckpt_19.pt",
542
+ "model.gpt_neox.layers.35.mlp.dense_4h_to_h.bias": "hf_ckpt_19.pt",
543
+ "model.gpt_neox.layers.36.input_layernorm.weight": "hf_ckpt_19.pt",
544
+ "model.gpt_neox.layers.36.input_layernorm.bias": "hf_ckpt_19.pt",
545
+ "model.gpt_neox.layers.36.post_attention_layernorm.weight": "hf_ckpt_19.pt",
546
+ "model.gpt_neox.layers.36.post_attention_layernorm.bias": "hf_ckpt_19.pt",
547
+ "model.gpt_neox.layers.36.attention.bias": "hf_ckpt_19.pt",
548
+ "model.gpt_neox.layers.36.attention.masked_bias": "hf_ckpt_19.pt",
549
+ "model.gpt_neox.layers.36.attention.rotary_emb.inv_freq": "hf_ckpt_19.pt",
550
+ "model.gpt_neox.layers.36.attention.query_key_value.weight": "hf_ckpt_19.pt",
551
+ "model.gpt_neox.layers.36.attention.query_key_value.bias": "hf_ckpt_19.pt",
552
+ "model.gpt_neox.layers.36.attention.dense.weight": "hf_ckpt_19.pt",
553
+ "model.gpt_neox.layers.36.attention.dense.bias": "hf_ckpt_19.pt",
554
+ "model.gpt_neox.layers.36.mlp.dense_h_to_4h.weight": "hf_ckpt_19.pt",
555
+ "model.gpt_neox.layers.36.mlp.dense_h_to_4h.bias": "hf_ckpt_19.pt",
556
+ "model.gpt_neox.layers.36.mlp.dense_4h_to_h.weight": "hf_ckpt_19.pt",
557
+ "model.gpt_neox.layers.36.mlp.dense_4h_to_h.bias": "hf_ckpt_19.pt",
558
+ "model.gpt_neox.layers.37.input_layernorm.weight": "hf_ckpt_19.pt",
559
+ "model.gpt_neox.layers.37.input_layernorm.bias": "hf_ckpt_19.pt",
560
+ "model.gpt_neox.layers.37.post_attention_layernorm.weight": "hf_ckpt_19.pt",
561
+ "model.gpt_neox.layers.37.post_attention_layernorm.bias": "hf_ckpt_19.pt",
562
+ "model.gpt_neox.layers.37.attention.bias": "hf_ckpt_20.pt",
563
+ "model.gpt_neox.layers.37.attention.masked_bias": "hf_ckpt_20.pt",
564
+ "model.gpt_neox.layers.37.attention.rotary_emb.inv_freq": "hf_ckpt_20.pt",
565
+ "model.gpt_neox.layers.37.attention.query_key_value.weight": "hf_ckpt_20.pt",
566
+ "model.gpt_neox.layers.37.attention.query_key_value.bias": "hf_ckpt_20.pt",
567
+ "model.gpt_neox.layers.37.attention.dense.weight": "hf_ckpt_20.pt",
568
+ "model.gpt_neox.layers.37.attention.dense.bias": "hf_ckpt_20.pt",
569
+ "model.gpt_neox.layers.37.mlp.dense_h_to_4h.weight": "hf_ckpt_20.pt",
570
+ "model.gpt_neox.layers.37.mlp.dense_h_to_4h.bias": "hf_ckpt_20.pt",
571
+ "model.gpt_neox.layers.37.mlp.dense_4h_to_h.weight": "hf_ckpt_20.pt",
572
+ "model.gpt_neox.layers.37.mlp.dense_4h_to_h.bias": "hf_ckpt_20.pt",
573
+ "model.gpt_neox.layers.38.input_layernorm.weight": "hf_ckpt_20.pt",
574
+ "model.gpt_neox.layers.38.input_layernorm.bias": "hf_ckpt_20.pt",
575
+ "model.gpt_neox.layers.38.post_attention_layernorm.weight": "hf_ckpt_20.pt",
576
+ "model.gpt_neox.layers.38.post_attention_layernorm.bias": "hf_ckpt_20.pt",
577
+ "model.gpt_neox.layers.38.attention.bias": "hf_ckpt_20.pt",
578
+ "model.gpt_neox.layers.38.attention.masked_bias": "hf_ckpt_20.pt",
579
+ "model.gpt_neox.layers.38.attention.rotary_emb.inv_freq": "hf_ckpt_20.pt",
580
+ "model.gpt_neox.layers.38.attention.query_key_value.weight": "hf_ckpt_20.pt",
581
+ "model.gpt_neox.layers.38.attention.query_key_value.bias": "hf_ckpt_20.pt",
582
+ "model.gpt_neox.layers.38.attention.dense.weight": "hf_ckpt_20.pt",
583
+ "model.gpt_neox.layers.38.attention.dense.bias": "hf_ckpt_20.pt",
584
+ "model.gpt_neox.layers.38.mlp.dense_h_to_4h.weight": "hf_ckpt_20.pt",
585
+ "model.gpt_neox.layers.38.mlp.dense_h_to_4h.bias": "hf_ckpt_20.pt",
586
+ "model.gpt_neox.layers.38.mlp.dense_4h_to_h.weight": "hf_ckpt_20.pt",
587
+ "model.gpt_neox.layers.38.mlp.dense_4h_to_h.bias": "hf_ckpt_20.pt",
588
+ "model.gpt_neox.layers.39.input_layernorm.weight": "hf_ckpt_20.pt",
589
+ "model.gpt_neox.layers.39.input_layernorm.bias": "hf_ckpt_20.pt",
590
+ "model.gpt_neox.layers.39.post_attention_layernorm.weight": "hf_ckpt_21.pt",
591
+ "model.gpt_neox.layers.39.post_attention_layernorm.bias": "hf_ckpt_21.pt",
592
+ "model.gpt_neox.layers.39.attention.bias": "hf_ckpt_21.pt",
593
+ "model.gpt_neox.layers.39.attention.masked_bias": "hf_ckpt_21.pt",
594
+ "model.gpt_neox.layers.39.attention.rotary_emb.inv_freq": "hf_ckpt_21.pt",
595
+ "model.gpt_neox.layers.39.attention.query_key_value.weight": "hf_ckpt_21.pt",
596
+ "model.gpt_neox.layers.39.attention.query_key_value.bias": "hf_ckpt_21.pt",
597
+ "model.gpt_neox.layers.39.attention.dense.weight": "hf_ckpt_21.pt",
598
+ "model.gpt_neox.layers.39.attention.dense.bias": "hf_ckpt_21.pt",
599
+ "model.gpt_neox.layers.39.mlp.dense_h_to_4h.weight": "hf_ckpt_21.pt",
600
+ "model.gpt_neox.layers.39.mlp.dense_h_to_4h.bias": "hf_ckpt_21.pt",
601
+ "model.gpt_neox.layers.39.mlp.dense_4h_to_h.weight": "hf_ckpt_21.pt",
602
+ "model.gpt_neox.layers.39.mlp.dense_4h_to_h.bias": "hf_ckpt_21.pt",
603
+ "model.gpt_neox.layers.40.input_layernorm.weight": "hf_ckpt_21.pt",
604
+ "model.gpt_neox.layers.40.input_layernorm.bias": "hf_ckpt_21.pt",
605
+ "model.gpt_neox.layers.40.post_attention_layernorm.weight": "hf_ckpt_21.pt",
606
+ "model.gpt_neox.layers.40.post_attention_layernorm.bias": "hf_ckpt_21.pt",
607
+ "model.gpt_neox.layers.40.attention.bias": "hf_ckpt_21.pt",
608
+ "model.gpt_neox.layers.40.attention.masked_bias": "hf_ckpt_21.pt",
609
+ "model.gpt_neox.layers.40.attention.rotary_emb.inv_freq": "hf_ckpt_21.pt",
610
+ "model.gpt_neox.layers.40.attention.query_key_value.weight": "hf_ckpt_21.pt",
611
+ "model.gpt_neox.layers.40.attention.query_key_value.bias": "hf_ckpt_21.pt",
612
+ "model.gpt_neox.layers.40.attention.dense.weight": "hf_ckpt_21.pt",
613
+ "model.gpt_neox.layers.40.attention.dense.bias": "hf_ckpt_21.pt",
614
+ "model.gpt_neox.layers.40.mlp.dense_h_to_4h.weight": "hf_ckpt_21.pt",
615
+ "model.gpt_neox.layers.40.mlp.dense_h_to_4h.bias": "hf_ckpt_21.pt",
616
+ "model.gpt_neox.layers.40.mlp.dense_4h_to_h.weight": "hf_ckpt_21.pt",
617
+ "model.gpt_neox.layers.40.mlp.dense_4h_to_h.bias": "hf_ckpt_21.pt",
618
+ "model.gpt_neox.layers.41.input_layernorm.weight": "hf_ckpt_22.pt",
619
+ "model.gpt_neox.layers.41.input_layernorm.bias": "hf_ckpt_22.pt",
620
+ "model.gpt_neox.layers.41.post_attention_layernorm.weight": "hf_ckpt_22.pt",
621
+ "model.gpt_neox.layers.41.post_attention_layernorm.bias": "hf_ckpt_22.pt",
622
+ "model.gpt_neox.layers.41.attention.bias": "hf_ckpt_22.pt",
623
+ "model.gpt_neox.layers.41.attention.masked_bias": "hf_ckpt_22.pt",
624
+ "model.gpt_neox.layers.41.attention.rotary_emb.inv_freq": "hf_ckpt_22.pt",
625
+ "model.gpt_neox.layers.41.attention.query_key_value.weight": "hf_ckpt_22.pt",
626
+ "model.gpt_neox.layers.41.attention.query_key_value.bias": "hf_ckpt_22.pt",
627
+ "model.gpt_neox.layers.41.attention.dense.weight": "hf_ckpt_22.pt",
628
+ "model.gpt_neox.layers.41.attention.dense.bias": "hf_ckpt_22.pt",
629
+ "model.gpt_neox.layers.41.mlp.dense_h_to_4h.weight": "hf_ckpt_22.pt",
630
+ "model.gpt_neox.layers.41.mlp.dense_h_to_4h.bias": "hf_ckpt_22.pt",
631
+ "model.gpt_neox.layers.41.mlp.dense_4h_to_h.weight": "hf_ckpt_22.pt",
632
+ "model.gpt_neox.layers.41.mlp.dense_4h_to_h.bias": "hf_ckpt_22.pt",
633
+ "model.gpt_neox.layers.42.input_layernorm.weight": "hf_ckpt_22.pt",
634
+ "model.gpt_neox.layers.42.input_layernorm.bias": "hf_ckpt_22.pt",
635
+ "model.gpt_neox.layers.42.post_attention_layernorm.weight": "hf_ckpt_22.pt",
636
+ "model.gpt_neox.layers.42.post_attention_layernorm.bias": "hf_ckpt_22.pt",
637
+ "model.gpt_neox.layers.42.attention.bias": "hf_ckpt_22.pt",
638
+ "model.gpt_neox.layers.42.attention.masked_bias": "hf_ckpt_22.pt",
639
+ "model.gpt_neox.layers.42.attention.rotary_emb.inv_freq": "hf_ckpt_22.pt",
640
+ "model.gpt_neox.layers.42.attention.query_key_value.weight": "hf_ckpt_22.pt",
641
+ "model.gpt_neox.layers.42.attention.query_key_value.bias": "hf_ckpt_22.pt",
642
+ "model.gpt_neox.layers.42.attention.dense.weight": "hf_ckpt_22.pt",
643
+ "model.gpt_neox.layers.42.attention.dense.bias": "hf_ckpt_22.pt",
644
+ "model.gpt_neox.layers.42.mlp.dense_h_to_4h.weight": "hf_ckpt_22.pt",
645
+ "model.gpt_neox.layers.42.mlp.dense_h_to_4h.bias": "hf_ckpt_22.pt",
646
+ "model.gpt_neox.layers.42.mlp.dense_4h_to_h.weight": "hf_ckpt_23.pt",
647
+ "model.gpt_neox.layers.42.mlp.dense_4h_to_h.bias": "hf_ckpt_23.pt",
648
+ "model.gpt_neox.layers.43.input_layernorm.weight": "hf_ckpt_23.pt",
649
+ "model.gpt_neox.layers.43.input_layernorm.bias": "hf_ckpt_23.pt",
650
+ "model.gpt_neox.layers.43.post_attention_layernorm.weight": "hf_ckpt_23.pt",
651
+ "model.gpt_neox.layers.43.post_attention_layernorm.bias": "hf_ckpt_23.pt",
652
+ "model.gpt_neox.layers.43.attention.bias": "hf_ckpt_23.pt",
653
+ "model.gpt_neox.layers.43.attention.masked_bias": "hf_ckpt_23.pt",
654
+ "model.gpt_neox.layers.43.attention.rotary_emb.inv_freq": "hf_ckpt_23.pt",
655
+ "model.gpt_neox.layers.43.attention.query_key_value.weight": "hf_ckpt_23.pt",
656
+ "model.gpt_neox.layers.43.attention.query_key_value.bias": "hf_ckpt_23.pt",
657
+ "model.gpt_neox.layers.43.attention.dense.weight": "hf_ckpt_23.pt",
658
+ "model.gpt_neox.layers.43.attention.dense.bias": "hf_ckpt_23.pt",
659
+ "model.gpt_neox.layers.43.mlp.dense_h_to_4h.weight": "hf_ckpt_23.pt",
660
+ "model.gpt_neox.layers.43.mlp.dense_h_to_4h.bias": "hf_ckpt_23.pt",
661
+ "model.gpt_neox.layers.43.mlp.dense_4h_to_h.weight": "hf_ckpt_23.pt",
662
+ "model.gpt_neox.layers.43.mlp.dense_4h_to_h.bias": "hf_ckpt_23.pt",
663
+ "model.gpt_neox.final_layer_norm.weight": "hf_ckpt_23.pt",
664
+ "model.gpt_neox.final_layer_norm.bias": "hf_ckpt_23.pt",
665
+ "model.embed_out.weight": "hf_ckpt_23.pt",
666
+ "transformer.embed_in.weight": "hf_ckpt_23.pt",
667
+ "transformer.layers.0.input_layernorm.weight": "hf_ckpt_23.pt",
668
+ "transformer.layers.0.input_layernorm.bias": "hf_ckpt_23.pt",
669
+ "transformer.layers.0.post_attention_layernorm.weight": "hf_ckpt_23.pt",
670
+ "transformer.layers.0.post_attention_layernorm.bias": "hf_ckpt_23.pt",
671
+ "transformer.layers.0.attention.bias": "hf_ckpt_23.pt",
672
+ "transformer.layers.0.attention.masked_bias": "hf_ckpt_23.pt",
673
+ "transformer.layers.0.attention.rotary_emb.inv_freq": "hf_ckpt_23.pt",
674
+ "transformer.layers.0.attention.query_key_value.weight": "hf_ckpt_24.pt",
675
+ "transformer.layers.0.attention.query_key_value.bias": "hf_ckpt_24.pt",
676
+ "transformer.layers.0.attention.dense.weight": "hf_ckpt_24.pt",
677
+ "transformer.layers.0.attention.dense.bias": "hf_ckpt_24.pt",
678
+ "transformer.layers.0.mlp.dense_h_to_4h.weight": "hf_ckpt_24.pt",
679
+ "transformer.layers.0.mlp.dense_h_to_4h.bias": "hf_ckpt_24.pt",
680
+ "transformer.layers.0.mlp.dense_4h_to_h.weight": "hf_ckpt_24.pt",
681
+ "transformer.layers.0.mlp.dense_4h_to_h.bias": "hf_ckpt_24.pt",
682
+ "transformer.layers.1.input_layernorm.weight": "hf_ckpt_24.pt",
683
+ "transformer.layers.1.input_layernorm.bias": "hf_ckpt_24.pt",
684
+ "transformer.layers.1.post_attention_layernorm.weight": "hf_ckpt_24.pt",
685
+ "transformer.layers.1.post_attention_layernorm.bias": "hf_ckpt_24.pt",
686
+ "transformer.layers.1.attention.bias": "hf_ckpt_24.pt",
687
+ "transformer.layers.1.attention.masked_bias": "hf_ckpt_24.pt",
688
+ "transformer.layers.1.attention.rotary_emb.inv_freq": "hf_ckpt_24.pt",
689
+ "transformer.layers.1.attention.query_key_value.weight": "hf_ckpt_24.pt",
690
+ "transformer.layers.1.attention.query_key_value.bias": "hf_ckpt_24.pt",
691
+ "transformer.layers.1.attention.dense.weight": "hf_ckpt_24.pt",
692
+ "transformer.layers.1.attention.dense.bias": "hf_ckpt_24.pt",
693
+ "transformer.layers.1.mlp.dense_h_to_4h.weight": "hf_ckpt_24.pt",
694
+ "transformer.layers.1.mlp.dense_h_to_4h.bias": "hf_ckpt_24.pt",
695
+ "transformer.layers.1.mlp.dense_4h_to_h.weight": "hf_ckpt_24.pt",
696
+ "transformer.layers.1.mlp.dense_4h_to_h.bias": "hf_ckpt_24.pt",
697
+ "transformer.layers.2.input_layernorm.weight": "hf_ckpt_24.pt",
698
+ "transformer.layers.2.input_layernorm.bias": "hf_ckpt_24.pt",
699
+ "transformer.layers.2.post_attention_layernorm.weight": "hf_ckpt_24.pt",
700
+ "transformer.layers.2.post_attention_layernorm.bias": "hf_ckpt_24.pt",
701
+ "transformer.layers.2.attention.bias": "hf_ckpt_24.pt",
702
+ "transformer.layers.2.attention.masked_bias": "hf_ckpt_25.pt",
703
+ "transformer.layers.2.attention.rotary_emb.inv_freq": "hf_ckpt_25.pt",
704
+ "transformer.layers.2.attention.query_key_value.weight": "hf_ckpt_25.pt",
705
+ "transformer.layers.2.attention.query_key_value.bias": "hf_ckpt_25.pt",
706
+ "transformer.layers.2.attention.dense.weight": "hf_ckpt_25.pt",
707
+ "transformer.layers.2.attention.dense.bias": "hf_ckpt_25.pt",
708
+ "transformer.layers.2.mlp.dense_h_to_4h.weight": "hf_ckpt_25.pt",
709
+ "transformer.layers.2.mlp.dense_h_to_4h.bias": "hf_ckpt_25.pt",
710
+ "transformer.layers.2.mlp.dense_4h_to_h.weight": "hf_ckpt_25.pt",
711
+ "transformer.layers.2.mlp.dense_4h_to_h.bias": "hf_ckpt_25.pt",
712
+ "transformer.layers.3.input_layernorm.weight": "hf_ckpt_25.pt",
713
+ "transformer.layers.3.input_layernorm.bias": "hf_ckpt_25.pt",
714
+ "transformer.layers.3.post_attention_layernorm.weight": "hf_ckpt_25.pt",
715
+ "transformer.layers.3.post_attention_layernorm.bias": "hf_ckpt_25.pt",
716
+ "transformer.layers.3.attention.bias": "hf_ckpt_25.pt",
717
+ "transformer.layers.3.attention.masked_bias": "hf_ckpt_25.pt",
718
+ "transformer.layers.3.attention.rotary_emb.inv_freq": "hf_ckpt_25.pt",
719
+ "transformer.layers.3.attention.query_key_value.weight": "hf_ckpt_25.pt",
720
+ "transformer.layers.3.attention.query_key_value.bias": "hf_ckpt_25.pt",
721
+ "transformer.layers.3.attention.dense.weight": "hf_ckpt_25.pt",
722
+ "transformer.layers.3.attention.dense.bias": "hf_ckpt_25.pt",
723
+ "transformer.layers.3.mlp.dense_h_to_4h.weight": "hf_ckpt_25.pt",
724
+ "transformer.layers.3.mlp.dense_h_to_4h.bias": "hf_ckpt_25.pt",
725
+ "transformer.layers.3.mlp.dense_4h_to_h.weight": "hf_ckpt_25.pt",
726
+ "transformer.layers.3.mlp.dense_4h_to_h.bias": "hf_ckpt_25.pt",
727
+ "transformer.layers.4.input_layernorm.weight": "hf_ckpt_25.pt",
728
+ "transformer.layers.4.input_layernorm.bias": "hf_ckpt_25.pt",
729
+ "transformer.layers.4.post_attention_layernorm.weight": "hf_ckpt_25.pt",
730
+ "transformer.layers.4.post_attention_layernorm.bias": "hf_ckpt_26.pt",
731
+ "transformer.layers.4.attention.bias": "hf_ckpt_26.pt",
732
+ "transformer.layers.4.attention.masked_bias": "hf_ckpt_26.pt",
733
+ "transformer.layers.4.attention.rotary_emb.inv_freq": "hf_ckpt_26.pt",
734
+ "transformer.layers.4.attention.query_key_value.weight": "hf_ckpt_26.pt",
735
+ "transformer.layers.4.attention.query_key_value.bias": "hf_ckpt_26.pt",
736
+ "transformer.layers.4.attention.dense.weight": "hf_ckpt_26.pt",
737
+ "transformer.layers.4.attention.dense.bias": "hf_ckpt_26.pt",
738
+ "transformer.layers.4.mlp.dense_h_to_4h.weight": "hf_ckpt_26.pt",
739
+ "transformer.layers.4.mlp.dense_h_to_4h.bias": "hf_ckpt_26.pt",
740
+ "transformer.layers.4.mlp.dense_4h_to_h.weight": "hf_ckpt_26.pt",
741
+ "transformer.layers.4.mlp.dense_4h_to_h.bias": "hf_ckpt_26.pt",
742
+ "transformer.layers.5.input_layernorm.weight": "hf_ckpt_26.pt",
743
+ "transformer.layers.5.input_layernorm.bias": "hf_ckpt_26.pt",
744
+ "transformer.layers.5.post_attention_layernorm.weight": "hf_ckpt_26.pt",
745
+ "transformer.layers.5.post_attention_layernorm.bias": "hf_ckpt_26.pt",
746
+ "transformer.layers.5.attention.bias": "hf_ckpt_26.pt",
747
+ "transformer.layers.5.attention.masked_bias": "hf_ckpt_26.pt",
748
+ "transformer.layers.5.attention.rotary_emb.inv_freq": "hf_ckpt_26.pt",
749
+ "transformer.layers.5.attention.query_key_value.weight": "hf_ckpt_26.pt",
750
+ "transformer.layers.5.attention.query_key_value.bias": "hf_ckpt_26.pt",
751
+ "transformer.layers.5.attention.dense.weight": "hf_ckpt_26.pt",
752
+ "transformer.layers.5.attention.dense.bias": "hf_ckpt_26.pt",
753
+ "transformer.layers.5.mlp.dense_h_to_4h.weight": "hf_ckpt_26.pt",
754
+ "transformer.layers.5.mlp.dense_h_to_4h.bias": "hf_ckpt_26.pt",
755
+ "transformer.layers.5.mlp.dense_4h_to_h.weight": "hf_ckpt_26.pt",
756
+ "transformer.layers.5.mlp.dense_4h_to_h.bias": "hf_ckpt_26.pt",
757
+ "transformer.layers.6.input_layernorm.weight": "hf_ckpt_26.pt",
758
+ "transformer.layers.6.input_layernorm.bias": "hf_ckpt_27.pt",
759
+ "transformer.layers.6.post_attention_layernorm.weight": "hf_ckpt_27.pt",
760
+ "transformer.layers.6.post_attention_layernorm.bias": "hf_ckpt_27.pt",
761
+ "transformer.layers.6.attention.bias": "hf_ckpt_27.pt",
762
+ "transformer.layers.6.attention.masked_bias": "hf_ckpt_27.pt",
763
+ "transformer.layers.6.attention.rotary_emb.inv_freq": "hf_ckpt_27.pt",
764
+ "transformer.layers.6.attention.query_key_value.weight": "hf_ckpt_27.pt",
765
+ "transformer.layers.6.attention.query_key_value.bias": "hf_ckpt_27.pt",
766
+ "transformer.layers.6.attention.dense.weight": "hf_ckpt_27.pt",
767
+ "transformer.layers.6.attention.dense.bias": "hf_ckpt_27.pt",
768
+ "transformer.layers.6.mlp.dense_h_to_4h.weight": "hf_ckpt_27.pt",
769
+ "transformer.layers.6.mlp.dense_h_to_4h.bias": "hf_ckpt_27.pt",
770
+ "transformer.layers.6.mlp.dense_4h_to_h.weight": "hf_ckpt_27.pt",
771
+ "transformer.layers.6.mlp.dense_4h_to_h.bias": "hf_ckpt_27.pt",
772
+ "transformer.layers.7.input_layernorm.weight": "hf_ckpt_27.pt",
773
+ "transformer.layers.7.input_layernorm.bias": "hf_ckpt_27.pt",
774
+ "transformer.layers.7.post_attention_layernorm.weight": "hf_ckpt_27.pt",
775
+ "transformer.layers.7.post_attention_layernorm.bias": "hf_ckpt_27.pt",
776
+ "transformer.layers.7.attention.bias": "hf_ckpt_27.pt",
777
+ "transformer.layers.7.attention.masked_bias": "hf_ckpt_27.pt",
778
+ "transformer.layers.7.attention.rotary_emb.inv_freq": "hf_ckpt_27.pt",
779
+ "transformer.layers.7.attention.query_key_value.weight": "hf_ckpt_27.pt",
780
+ "transformer.layers.7.attention.query_key_value.bias": "hf_ckpt_27.pt",
781
+ "transformer.layers.7.attention.dense.weight": "hf_ckpt_27.pt",
782
+ "transformer.layers.7.attention.dense.bias": "hf_ckpt_27.pt",
783
+ "transformer.layers.7.mlp.dense_h_to_4h.weight": "hf_ckpt_27.pt",
784
+ "transformer.layers.7.mlp.dense_h_to_4h.bias": "hf_ckpt_27.pt",
785
+ "transformer.layers.7.mlp.dense_4h_to_h.weight": "hf_ckpt_27.pt",
786
+ "transformer.layers.7.mlp.dense_4h_to_h.bias": "hf_ckpt_28.pt",
787
+ "transformer.layers.8.input_layernorm.weight": "hf_ckpt_28.pt",
788
+ "transformer.layers.8.input_layernorm.bias": "hf_ckpt_28.pt",
789
+ "transformer.layers.8.post_attention_layernorm.weight": "hf_ckpt_28.pt",
790
+ "transformer.layers.8.post_attention_layernorm.bias": "hf_ckpt_28.pt",
791
+ "transformer.layers.8.attention.bias": "hf_ckpt_28.pt",
792
+ "transformer.layers.8.attention.masked_bias": "hf_ckpt_28.pt",
793
+ "transformer.layers.8.attention.rotary_emb.inv_freq": "hf_ckpt_28.pt",
794
+ "transformer.layers.8.attention.query_key_value.weight": "hf_ckpt_28.pt",
795
+ "transformer.layers.8.attention.query_key_value.bias": "hf_ckpt_28.pt",
796
+ "transformer.layers.8.attention.dense.weight": "hf_ckpt_28.pt",
797
+ "transformer.layers.8.attention.dense.bias": "hf_ckpt_28.pt",
798
+ "transformer.layers.8.mlp.dense_h_to_4h.weight": "hf_ckpt_28.pt",
799
+ "transformer.layers.8.mlp.dense_h_to_4h.bias": "hf_ckpt_28.pt",
800
+ "transformer.layers.8.mlp.dense_4h_to_h.weight": "hf_ckpt_28.pt",
801
+ "transformer.layers.8.mlp.dense_4h_to_h.bias": "hf_ckpt_28.pt",
802
+ "transformer.layers.9.input_layernorm.weight": "hf_ckpt_28.pt",
803
+ "transformer.layers.9.input_layernorm.bias": "hf_ckpt_28.pt",
804
+ "transformer.layers.9.post_attention_layernorm.weight": "hf_ckpt_28.pt",
805
+ "transformer.layers.9.post_attention_layernorm.bias": "hf_ckpt_28.pt",
806
+ "transformer.layers.9.attention.bias": "hf_ckpt_28.pt",
807
+ "transformer.layers.9.attention.masked_bias": "hf_ckpt_28.pt",
808
+ "transformer.layers.9.attention.rotary_emb.inv_freq": "hf_ckpt_28.pt",
809
+ "transformer.layers.9.attention.query_key_value.weight": "hf_ckpt_28.pt",
810
+ "transformer.layers.9.attention.query_key_value.bias": "hf_ckpt_28.pt",
811
+ "transformer.layers.9.attention.dense.weight": "hf_ckpt_28.pt",
812
+ "transformer.layers.9.attention.dense.bias": "hf_ckpt_28.pt",
813
+ "transformer.layers.9.mlp.dense_h_to_4h.weight": "hf_ckpt_28.pt",
814
+ "transformer.layers.9.mlp.dense_h_to_4h.bias": "hf_ckpt_29.pt",
815
+ "transformer.layers.9.mlp.dense_4h_to_h.weight": "hf_ckpt_29.pt",
816
+ "transformer.layers.9.mlp.dense_4h_to_h.bias": "hf_ckpt_29.pt",
817
+ "transformer.layers.10.input_layernorm.weight": "hf_ckpt_29.pt",
818
+ "transformer.layers.10.input_layernorm.bias": "hf_ckpt_29.pt",
819
+ "transformer.layers.10.post_attention_layernorm.weight": "hf_ckpt_29.pt",
820
+ "transformer.layers.10.post_attention_layernorm.bias": "hf_ckpt_29.pt",
821
+ "transformer.layers.10.attention.bias": "hf_ckpt_29.pt",
822
+ "transformer.layers.10.attention.masked_bias": "hf_ckpt_29.pt",
823
+ "transformer.layers.10.attention.rotary_emb.inv_freq": "hf_ckpt_29.pt",
824
+ "transformer.layers.10.attention.query_key_value.weight": "hf_ckpt_29.pt",
825
+ "transformer.layers.10.attention.query_key_value.bias": "hf_ckpt_29.pt",
826
+ "transformer.layers.10.attention.dense.weight": "hf_ckpt_29.pt",
827
+ "transformer.layers.10.attention.dense.bias": "hf_ckpt_29.pt",
828
+ "transformer.layers.10.mlp.dense_h_to_4h.weight": "hf_ckpt_29.pt",
829
+ "transformer.layers.10.mlp.dense_h_to_4h.bias": "hf_ckpt_29.pt",
830
+ "transformer.layers.10.mlp.dense_4h_to_h.weight": "hf_ckpt_29.pt",
831
+ "transformer.layers.10.mlp.dense_4h_to_h.bias": "hf_ckpt_29.pt",
832
+ "transformer.layers.11.input_layernorm.weight": "hf_ckpt_29.pt",
833
+ "transformer.layers.11.input_layernorm.bias": "hf_ckpt_29.pt",
834
+ "transformer.layers.11.post_attention_layernorm.weight": "hf_ckpt_29.pt",
835
+ "transformer.layers.11.post_attention_layernorm.bias": "hf_ckpt_29.pt",
836
+ "transformer.layers.11.attention.bias": "hf_ckpt_29.pt",
837
+ "transformer.layers.11.attention.masked_bias": "hf_ckpt_29.pt",
838
+ "transformer.layers.11.attention.rotary_emb.inv_freq": "hf_ckpt_29.pt",
839
+ "transformer.layers.11.attention.query_key_value.weight": "hf_ckpt_29.pt",
840
+ "transformer.layers.11.attention.query_key_value.bias": "hf_ckpt_29.pt",
841
+ "transformer.layers.11.attention.dense.weight": "hf_ckpt_29.pt",
842
+ "transformer.layers.11.attention.dense.bias": "hf_ckpt_30.pt",
843
+ "transformer.layers.11.mlp.dense_h_to_4h.weight": "hf_ckpt_30.pt",
844
+ "transformer.layers.11.mlp.dense_h_to_4h.bias": "hf_ckpt_30.pt",
845
+ "transformer.layers.11.mlp.dense_4h_to_h.weight": "hf_ckpt_30.pt",
846
+ "transformer.layers.11.mlp.dense_4h_to_h.bias": "hf_ckpt_30.pt",
847
+ "transformer.layers.12.input_layernorm.weight": "hf_ckpt_30.pt",
848
+ "transformer.layers.12.input_layernorm.bias": "hf_ckpt_30.pt",
849
+ "transformer.layers.12.post_attention_layernorm.weight": "hf_ckpt_30.pt",
850
+ "transformer.layers.12.post_attention_layernorm.bias": "hf_ckpt_30.pt",
851
+ "transformer.layers.12.attention.bias": "hf_ckpt_30.pt",
852
+ "transformer.layers.12.attention.masked_bias": "hf_ckpt_30.pt",
853
+ "transformer.layers.12.attention.rotary_emb.inv_freq": "hf_ckpt_30.pt",
854
+ "transformer.layers.12.attention.query_key_value.weight": "hf_ckpt_30.pt",
855
+ "transformer.layers.12.attention.query_key_value.bias": "hf_ckpt_30.pt",
856
+ "transformer.layers.12.attention.dense.weight": "hf_ckpt_30.pt",
857
+ "transformer.layers.12.attention.dense.bias": "hf_ckpt_30.pt",
858
+ "transformer.layers.12.mlp.dense_h_to_4h.weight": "hf_ckpt_30.pt",
859
+ "transformer.layers.12.mlp.dense_h_to_4h.bias": "hf_ckpt_30.pt",
860
+ "transformer.layers.12.mlp.dense_4h_to_h.weight": "hf_ckpt_30.pt",
861
+ "transformer.layers.12.mlp.dense_4h_to_h.bias": "hf_ckpt_30.pt",
862
+ "transformer.layers.13.input_layernorm.weight": "hf_ckpt_30.pt",
863
+ "transformer.layers.13.input_layernorm.bias": "hf_ckpt_30.pt",
864
+ "transformer.layers.13.post_attention_layernorm.weight": "hf_ckpt_30.pt",
865
+ "transformer.layers.13.post_attention_layernorm.bias": "hf_ckpt_30.pt",
866
+ "transformer.layers.13.attention.bias": "hf_ckpt_30.pt",
867
+ "transformer.layers.13.attention.masked_bias": "hf_ckpt_30.pt",
868
+ "transformer.layers.13.attention.rotary_emb.inv_freq": "hf_ckpt_30.pt",
869
+ "transformer.layers.13.attention.query_key_value.weight": "hf_ckpt_30.pt",
870
+ "transformer.layers.13.attention.query_key_value.bias": "hf_ckpt_31.pt",
871
+ "transformer.layers.13.attention.dense.weight": "hf_ckpt_31.pt",
872
+ "transformer.layers.13.attention.dense.bias": "hf_ckpt_31.pt",
873
+ "transformer.layers.13.mlp.dense_h_to_4h.weight": "hf_ckpt_31.pt",
874
+ "transformer.layers.13.mlp.dense_h_to_4h.bias": "hf_ckpt_31.pt",
875
+ "transformer.layers.13.mlp.dense_4h_to_h.weight": "hf_ckpt_31.pt",
876
+ "transformer.layers.13.mlp.dense_4h_to_h.bias": "hf_ckpt_31.pt",
877
+ "transformer.layers.14.input_layernorm.weight": "hf_ckpt_31.pt",
878
+ "transformer.layers.14.input_layernorm.bias": "hf_ckpt_31.pt",
879
+ "transformer.layers.14.post_attention_layernorm.weight": "hf_ckpt_31.pt",
880
+ "transformer.layers.14.post_attention_layernorm.bias": "hf_ckpt_31.pt",
881
+ "transformer.layers.14.attention.bias": "hf_ckpt_31.pt",
882
+ "transformer.layers.14.attention.masked_bias": "hf_ckpt_31.pt",
883
+ "transformer.layers.14.attention.rotary_emb.inv_freq": "hf_ckpt_31.pt",
884
+ "transformer.layers.14.attention.query_key_value.weight": "hf_ckpt_31.pt",
885
+ "transformer.layers.14.attention.query_key_value.bias": "hf_ckpt_31.pt",
886
+ "transformer.layers.14.attention.dense.weight": "hf_ckpt_31.pt",
887
+ "transformer.layers.14.attention.dense.bias": "hf_ckpt_31.pt",
888
+ "transformer.layers.14.mlp.dense_h_to_4h.weight": "hf_ckpt_31.pt",
889
+ "transformer.layers.14.mlp.dense_h_to_4h.bias": "hf_ckpt_31.pt",
890
+ "transformer.layers.14.mlp.dense_4h_to_h.weight": "hf_ckpt_31.pt",
891
+ "transformer.layers.14.mlp.dense_4h_to_h.bias": "hf_ckpt_31.pt",
892
+ "transformer.layers.15.input_layernorm.weight": "hf_ckpt_31.pt",
893
+ "transformer.layers.15.input_layernorm.bias": "hf_ckpt_31.pt",
894
+ "transformer.layers.15.post_attention_layernorm.weight": "hf_ckpt_31.pt",
895
+ "transformer.layers.15.post_attention_layernorm.bias": "hf_ckpt_31.pt",
896
+ "transformer.layers.15.attention.bias": "hf_ckpt_31.pt",
897
+ "transformer.layers.15.attention.masked_bias": "hf_ckpt_31.pt",
898
+ "transformer.layers.15.attention.rotary_emb.inv_freq": "hf_ckpt_32.pt",
899
+ "transformer.layers.15.attention.query_key_value.weight": "hf_ckpt_32.pt",
900
+ "transformer.layers.15.attention.query_key_value.bias": "hf_ckpt_32.pt",
901
+ "transformer.layers.15.attention.dense.weight": "hf_ckpt_32.pt",
902
+ "transformer.layers.15.attention.dense.bias": "hf_ckpt_32.pt",
903
+ "transformer.layers.15.mlp.dense_h_to_4h.weight": "hf_ckpt_32.pt",
904
+ "transformer.layers.15.mlp.dense_h_to_4h.bias": "hf_ckpt_32.pt",
905
+ "transformer.layers.15.mlp.dense_4h_to_h.weight": "hf_ckpt_32.pt",
906
+ "transformer.layers.15.mlp.dense_4h_to_h.bias": "hf_ckpt_32.pt",
907
+ "transformer.layers.16.input_layernorm.weight": "hf_ckpt_32.pt",
908
+ "transformer.layers.16.input_layernorm.bias": "hf_ckpt_32.pt",
909
+ "transformer.layers.16.post_attention_layernorm.weight": "hf_ckpt_32.pt",
910
+ "transformer.layers.16.post_attention_layernorm.bias": "hf_ckpt_32.pt",
911
+ "transformer.layers.16.attention.bias": "hf_ckpt_32.pt",
912
+ "transformer.layers.16.attention.masked_bias": "hf_ckpt_32.pt",
913
+ "transformer.layers.16.attention.rotary_emb.inv_freq": "hf_ckpt_32.pt",
914
+ "transformer.layers.16.attention.query_key_value.weight": "hf_ckpt_32.pt",
915
+ "transformer.layers.16.attention.query_key_value.bias": "hf_ckpt_32.pt",
916
+ "transformer.layers.16.attention.dense.weight": "hf_ckpt_32.pt",
917
+ "transformer.layers.16.attention.dense.bias": "hf_ckpt_32.pt",
918
+ "transformer.layers.16.mlp.dense_h_to_4h.weight": "hf_ckpt_32.pt",
919
+ "transformer.layers.16.mlp.dense_h_to_4h.bias": "hf_ckpt_32.pt",
920
+ "transformer.layers.16.mlp.dense_4h_to_h.weight": "hf_ckpt_32.pt",
921
+ "transformer.layers.16.mlp.dense_4h_to_h.bias": "hf_ckpt_32.pt",
922
+ "transformer.layers.17.input_layernorm.weight": "hf_ckpt_32.pt",
923
+ "transformer.layers.17.input_layernorm.bias": "hf_ckpt_32.pt",
924
+ "transformer.layers.17.post_attention_layernorm.weight": "hf_ckpt_32.pt",
925
+ "transformer.layers.17.post_attention_layernorm.bias": "hf_ckpt_32.pt",
926
+ "transformer.layers.17.attention.bias": "hf_ckpt_33.pt",
927
+ "transformer.layers.17.attention.masked_bias": "hf_ckpt_33.pt",
928
+ "transformer.layers.17.attention.rotary_emb.inv_freq": "hf_ckpt_33.pt",
929
+ "transformer.layers.17.attention.query_key_value.weight": "hf_ckpt_33.pt",
930
+ "transformer.layers.17.attention.query_key_value.bias": "hf_ckpt_33.pt",
931
+ "transformer.layers.17.attention.dense.weight": "hf_ckpt_33.pt",
932
+ "transformer.layers.17.attention.dense.bias": "hf_ckpt_33.pt",
933
+ "transformer.layers.17.mlp.dense_h_to_4h.weight": "hf_ckpt_33.pt",
934
+ "transformer.layers.17.mlp.dense_h_to_4h.bias": "hf_ckpt_33.pt",
935
+ "transformer.layers.17.mlp.dense_4h_to_h.weight": "hf_ckpt_33.pt",
936
+ "transformer.layers.17.mlp.dense_4h_to_h.bias": "hf_ckpt_33.pt",
937
+ "transformer.layers.18.input_layernorm.weight": "hf_ckpt_33.pt",
938
+ "transformer.layers.18.input_layernorm.bias": "hf_ckpt_33.pt",
939
+ "transformer.layers.18.post_attention_layernorm.weight": "hf_ckpt_33.pt",
940
+ "transformer.layers.18.post_attention_layernorm.bias": "hf_ckpt_33.pt",
941
+ "transformer.layers.18.attention.bias": "hf_ckpt_33.pt",
942
+ "transformer.layers.18.attention.masked_bias": "hf_ckpt_33.pt",
943
+ "transformer.layers.18.attention.rotary_emb.inv_freq": "hf_ckpt_33.pt",
944
+ "transformer.layers.18.attention.query_key_value.weight": "hf_ckpt_33.pt",
945
+ "transformer.layers.18.attention.query_key_value.bias": "hf_ckpt_33.pt",
946
+ "transformer.layers.18.attention.dense.weight": "hf_ckpt_33.pt",
947
+ "transformer.layers.18.attention.dense.bias": "hf_ckpt_33.pt",
948
+ "transformer.layers.18.mlp.dense_h_to_4h.weight": "hf_ckpt_33.pt",
949
+ "transformer.layers.18.mlp.dense_h_to_4h.bias": "hf_ckpt_33.pt",
950
+ "transformer.layers.18.mlp.dense_4h_to_h.weight": "hf_ckpt_33.pt",
951
+ "transformer.layers.18.mlp.dense_4h_to_h.bias": "hf_ckpt_33.pt",
952
+ "transformer.layers.19.input_layernorm.weight": "hf_ckpt_33.pt",
953
+ "transformer.layers.19.input_layernorm.bias": "hf_ckpt_33.pt",
954
+ "transformer.layers.19.post_attention_layernorm.weight": "hf_ckpt_34.pt",
955
+ "transformer.layers.19.post_attention_layernorm.bias": "hf_ckpt_34.pt",
956
+ "transformer.layers.19.attention.bias": "hf_ckpt_34.pt",
957
+ "transformer.layers.19.attention.masked_bias": "hf_ckpt_34.pt",
958
+ "transformer.layers.19.attention.rotary_emb.inv_freq": "hf_ckpt_34.pt",
959
+ "transformer.layers.19.attention.query_key_value.weight": "hf_ckpt_34.pt",
960
+ "transformer.layers.19.attention.query_key_value.bias": "hf_ckpt_34.pt",
961
+ "transformer.layers.19.attention.dense.weight": "hf_ckpt_34.pt",
962
+ "transformer.layers.19.attention.dense.bias": "hf_ckpt_34.pt",
963
+ "transformer.layers.19.mlp.dense_h_to_4h.weight": "hf_ckpt_34.pt",
964
+ "transformer.layers.19.mlp.dense_h_to_4h.bias": "hf_ckpt_34.pt",
965
+ "transformer.layers.19.mlp.dense_4h_to_h.weight": "hf_ckpt_34.pt",
966
+ "transformer.layers.19.mlp.dense_4h_to_h.bias": "hf_ckpt_34.pt",
967
+ "transformer.layers.20.input_layernorm.weight": "hf_ckpt_34.pt",
968
+ "transformer.layers.20.input_layernorm.bias": "hf_ckpt_34.pt",
969
+ "transformer.layers.20.post_attention_layernorm.weight": "hf_ckpt_34.pt",
970
+ "transformer.layers.20.post_attention_layernorm.bias": "hf_ckpt_34.pt",
971
+ "transformer.layers.20.attention.bias": "hf_ckpt_34.pt",
972
+ "transformer.layers.20.attention.masked_bias": "hf_ckpt_34.pt",
973
+ "transformer.layers.20.attention.rotary_emb.inv_freq": "hf_ckpt_34.pt",
974
+ "transformer.layers.20.attention.query_key_value.weight": "hf_ckpt_34.pt",
975
+ "transformer.layers.20.attention.query_key_value.bias": "hf_ckpt_34.pt",
976
+ "transformer.layers.20.attention.dense.weight": "hf_ckpt_34.pt",
977
+ "transformer.layers.20.attention.dense.bias": "hf_ckpt_34.pt",
978
+ "transformer.layers.20.mlp.dense_h_to_4h.weight": "hf_ckpt_34.pt",
979
+ "transformer.layers.20.mlp.dense_h_to_4h.bias": "hf_ckpt_34.pt",
980
+ "transformer.layers.20.mlp.dense_4h_to_h.weight": "hf_ckpt_34.pt",
981
+ "transformer.layers.20.mlp.dense_4h_to_h.bias": "hf_ckpt_34.pt",
982
+ "transformer.layers.21.input_layernorm.weight": "hf_ckpt_35.pt",
983
+ "transformer.layers.21.input_layernorm.bias": "hf_ckpt_35.pt",
984
+ "transformer.layers.21.post_attention_layernorm.weight": "hf_ckpt_35.pt",
985
+ "transformer.layers.21.post_attention_layernorm.bias": "hf_ckpt_35.pt",
986
+ "transformer.layers.21.attention.bias": "hf_ckpt_35.pt",
987
+ "transformer.layers.21.attention.masked_bias": "hf_ckpt_35.pt",
988
+ "transformer.layers.21.attention.rotary_emb.inv_freq": "hf_ckpt_35.pt",
989
+ "transformer.layers.21.attention.query_key_value.weight": "hf_ckpt_35.pt",
990
+ "transformer.layers.21.attention.query_key_value.bias": "hf_ckpt_35.pt",
991
+ "transformer.layers.21.attention.dense.weight": "hf_ckpt_35.pt",
992
+ "transformer.layers.21.attention.dense.bias": "hf_ckpt_35.pt",
993
+ "transformer.layers.21.mlp.dense_h_to_4h.weight": "hf_ckpt_35.pt",
994
+ "transformer.layers.21.mlp.dense_h_to_4h.bias": "hf_ckpt_35.pt",
995
+ "transformer.layers.21.mlp.dense_4h_to_h.weight": "hf_ckpt_35.pt",
996
+ "transformer.layers.21.mlp.dense_4h_to_h.bias": "hf_ckpt_35.pt",
997
+ "transformer.layers.22.input_layernorm.weight": "hf_ckpt_35.pt",
998
+ "transformer.layers.22.input_layernorm.bias": "hf_ckpt_35.pt",
999
+ "transformer.layers.22.post_attention_layernorm.weight": "hf_ckpt_35.pt",
1000
+ "transformer.layers.22.post_attention_layernorm.bias": "hf_ckpt_35.pt",
1001
+ "transformer.layers.22.attention.bias": "hf_ckpt_35.pt",
1002
+ "transformer.layers.22.attention.masked_bias": "hf_ckpt_35.pt",
1003
+ "transformer.layers.22.attention.rotary_emb.inv_freq": "hf_ckpt_35.pt",
1004
+ "transformer.layers.22.attention.query_key_value.weight": "hf_ckpt_35.pt",
1005
+ "transformer.layers.22.attention.query_key_value.bias": "hf_ckpt_35.pt",
1006
+ "transformer.layers.22.attention.dense.weight": "hf_ckpt_35.pt",
1007
+ "transformer.layers.22.attention.dense.bias": "hf_ckpt_35.pt",
1008
+ "transformer.layers.22.mlp.dense_h_to_4h.weight": "hf_ckpt_35.pt",
1009
+ "transformer.layers.22.mlp.dense_h_to_4h.bias": "hf_ckpt_35.pt",
1010
+ "transformer.layers.22.mlp.dense_4h_to_h.weight": "hf_ckpt_36.pt",
1011
+ "transformer.layers.22.mlp.dense_4h_to_h.bias": "hf_ckpt_36.pt",
1012
+ "transformer.layers.23.input_layernorm.weight": "hf_ckpt_36.pt",
1013
+ "transformer.layers.23.input_layernorm.bias": "hf_ckpt_36.pt",
1014
+ "transformer.layers.23.post_attention_layernorm.weight": "hf_ckpt_36.pt",
1015
+ "transformer.layers.23.post_attention_layernorm.bias": "hf_ckpt_36.pt",
1016
+ "transformer.layers.23.attention.bias": "hf_ckpt_36.pt",
1017
+ "transformer.layers.23.attention.masked_bias": "hf_ckpt_36.pt",
1018
+ "transformer.layers.23.attention.rotary_emb.inv_freq": "hf_ckpt_36.pt",
1019
+ "transformer.layers.23.attention.query_key_value.weight": "hf_ckpt_36.pt",
1020
+ "transformer.layers.23.attention.query_key_value.bias": "hf_ckpt_36.pt",
1021
+ "transformer.layers.23.attention.dense.weight": "hf_ckpt_36.pt",
1022
+ "transformer.layers.23.attention.dense.bias": "hf_ckpt_36.pt",
1023
+ "transformer.layers.23.mlp.dense_h_to_4h.weight": "hf_ckpt_36.pt",
1024
+ "transformer.layers.23.mlp.dense_h_to_4h.bias": "hf_ckpt_36.pt",
1025
+ "transformer.layers.23.mlp.dense_4h_to_h.weight": "hf_ckpt_36.pt",
1026
+ "transformer.layers.23.mlp.dense_4h_to_h.bias": "hf_ckpt_36.pt",
1027
+ "transformer.layers.24.input_layernorm.weight": "hf_ckpt_36.pt",
1028
+ "transformer.layers.24.input_layernorm.bias": "hf_ckpt_36.pt",
1029
+ "transformer.layers.24.post_attention_layernorm.weight": "hf_ckpt_36.pt",
1030
+ "transformer.layers.24.post_attention_layernorm.bias": "hf_ckpt_36.pt",
1031
+ "transformer.layers.24.attention.bias": "hf_ckpt_36.pt",
1032
+ "transformer.layers.24.attention.masked_bias": "hf_ckpt_36.pt",
1033
+ "transformer.layers.24.attention.rotary_emb.inv_freq": "hf_ckpt_36.pt",
1034
+ "transformer.layers.24.attention.query_key_value.weight": "hf_ckpt_36.pt",
1035
+ "transformer.layers.24.attention.query_key_value.bias": "hf_ckpt_36.pt",
1036
+ "transformer.layers.24.attention.dense.weight": "hf_ckpt_36.pt",
1037
+ "transformer.layers.24.attention.dense.bias": "hf_ckpt_36.pt",
1038
+ "transformer.layers.24.mlp.dense_h_to_4h.weight": "hf_ckpt_37.pt",
1039
+ "transformer.layers.24.mlp.dense_h_to_4h.bias": "hf_ckpt_37.pt",
1040
+ "transformer.layers.24.mlp.dense_4h_to_h.weight": "hf_ckpt_37.pt",
1041
+ "transformer.layers.24.mlp.dense_4h_to_h.bias": "hf_ckpt_37.pt",
1042
+ "transformer.layers.25.input_layernorm.weight": "hf_ckpt_37.pt",
1043
+ "transformer.layers.25.input_layernorm.bias": "hf_ckpt_37.pt",
1044
+ "transformer.layers.25.post_attention_layernorm.weight": "hf_ckpt_37.pt",
1045
+ "transformer.layers.25.post_attention_layernorm.bias": "hf_ckpt_37.pt",
1046
+ "transformer.layers.25.attention.bias": "hf_ckpt_37.pt",
1047
+ "transformer.layers.25.attention.masked_bias": "hf_ckpt_37.pt",
1048
+ "transformer.layers.25.attention.rotary_emb.inv_freq": "hf_ckpt_37.pt",
1049
+ "transformer.layers.25.attention.query_key_value.weight": "hf_ckpt_37.pt",
1050
+ "transformer.layers.25.attention.query_key_value.bias": "hf_ckpt_37.pt",
1051
+ "transformer.layers.25.attention.dense.weight": "hf_ckpt_37.pt",
1052
+ "transformer.layers.25.attention.dense.bias": "hf_ckpt_37.pt",
1053
+ "transformer.layers.25.mlp.dense_h_to_4h.weight": "hf_ckpt_37.pt",
1054
+ "transformer.layers.25.mlp.dense_h_to_4h.bias": "hf_ckpt_37.pt",
1055
+ "transformer.layers.25.mlp.dense_4h_to_h.weight": "hf_ckpt_37.pt",
1056
+ "transformer.layers.25.mlp.dense_4h_to_h.bias": "hf_ckpt_37.pt",
1057
+ "transformer.layers.26.input_layernorm.weight": "hf_ckpt_37.pt",
1058
+ "transformer.layers.26.input_layernorm.bias": "hf_ckpt_37.pt",
1059
+ "transformer.layers.26.post_attention_layernorm.weight": "hf_ckpt_37.pt",
1060
+ "transformer.layers.26.post_attention_layernorm.bias": "hf_ckpt_37.pt",
1061
+ "transformer.layers.26.attention.bias": "hf_ckpt_37.pt",
1062
+ "transformer.layers.26.attention.masked_bias": "hf_ckpt_37.pt",
1063
+ "transformer.layers.26.attention.rotary_emb.inv_freq": "hf_ckpt_37.pt",
1064
+ "transformer.layers.26.attention.query_key_value.weight": "hf_ckpt_37.pt",
1065
+ "transformer.layers.26.attention.query_key_value.bias": "hf_ckpt_37.pt",
1066
+ "transformer.layers.26.attention.dense.weight": "hf_ckpt_38.pt",
1067
+ "transformer.layers.26.attention.dense.bias": "hf_ckpt_38.pt",
1068
+ "transformer.layers.26.mlp.dense_h_to_4h.weight": "hf_ckpt_38.pt",
1069
+ "transformer.layers.26.mlp.dense_h_to_4h.bias": "hf_ckpt_38.pt",
1070
+ "transformer.layers.26.mlp.dense_4h_to_h.weight": "hf_ckpt_38.pt",
1071
+ "transformer.layers.26.mlp.dense_4h_to_h.bias": "hf_ckpt_38.pt",
1072
+ "transformer.layers.27.input_layernorm.weight": "hf_ckpt_38.pt",
1073
+ "transformer.layers.27.input_layernorm.bias": "hf_ckpt_38.pt",
1074
+ "transformer.layers.27.post_attention_layernorm.weight": "hf_ckpt_38.pt",
1075
+ "transformer.layers.27.post_attention_layernorm.bias": "hf_ckpt_38.pt",
1076
+ "transformer.layers.27.attention.bias": "hf_ckpt_38.pt",
1077
+ "transformer.layers.27.attention.masked_bias": "hf_ckpt_38.pt",
1078
+ "transformer.layers.27.attention.rotary_emb.inv_freq": "hf_ckpt_38.pt",
1079
+ "transformer.layers.27.attention.query_key_value.weight": "hf_ckpt_38.pt",
1080
+ "transformer.layers.27.attention.query_key_value.bias": "hf_ckpt_38.pt",
1081
+ "transformer.layers.27.attention.dense.weight": "hf_ckpt_38.pt",
1082
+ "transformer.layers.27.attention.dense.bias": "hf_ckpt_38.pt",
1083
+ "transformer.layers.27.mlp.dense_h_to_4h.weight": "hf_ckpt_38.pt",
1084
+ "transformer.layers.27.mlp.dense_h_to_4h.bias": "hf_ckpt_38.pt",
1085
+ "transformer.layers.27.mlp.dense_4h_to_h.weight": "hf_ckpt_38.pt",
1086
+ "transformer.layers.27.mlp.dense_4h_to_h.bias": "hf_ckpt_38.pt",
1087
+ "transformer.layers.28.input_layernorm.weight": "hf_ckpt_38.pt",
1088
+ "transformer.layers.28.input_layernorm.bias": "hf_ckpt_38.pt",
1089
+ "transformer.layers.28.post_attention_layernorm.weight": "hf_ckpt_38.pt",
1090
+ "transformer.layers.28.post_attention_layernorm.bias": "hf_ckpt_38.pt",
1091
+ "transformer.layers.28.attention.bias": "hf_ckpt_38.pt",
1092
+ "transformer.layers.28.attention.masked_bias": "hf_ckpt_38.pt",
1093
+ "transformer.layers.28.attention.rotary_emb.inv_freq": "hf_ckpt_38.pt",
1094
+ "transformer.layers.28.attention.query_key_value.weight": "hf_ckpt_39.pt",
1095
+ "transformer.layers.28.attention.query_key_value.bias": "hf_ckpt_39.pt",
1096
+ "transformer.layers.28.attention.dense.weight": "hf_ckpt_39.pt",
1097
+ "transformer.layers.28.attention.dense.bias": "hf_ckpt_39.pt",
1098
+ "transformer.layers.28.mlp.dense_h_to_4h.weight": "hf_ckpt_39.pt",
1099
+ "transformer.layers.28.mlp.dense_h_to_4h.bias": "hf_ckpt_39.pt",
1100
+ "transformer.layers.28.mlp.dense_4h_to_h.weight": "hf_ckpt_39.pt",
1101
+ "transformer.layers.28.mlp.dense_4h_to_h.bias": "hf_ckpt_39.pt",
1102
+ "transformer.layers.29.input_layernorm.weight": "hf_ckpt_39.pt",
1103
+ "transformer.layers.29.input_layernorm.bias": "hf_ckpt_39.pt",
1104
+ "transformer.layers.29.post_attention_layernorm.weight": "hf_ckpt_39.pt",
1105
+ "transformer.layers.29.post_attention_layernorm.bias": "hf_ckpt_39.pt",
1106
+ "transformer.layers.29.attention.bias": "hf_ckpt_39.pt",
1107
+ "transformer.layers.29.attention.masked_bias": "hf_ckpt_39.pt",
1108
+ "transformer.layers.29.attention.rotary_emb.inv_freq": "hf_ckpt_39.pt",
1109
+ "transformer.layers.29.attention.query_key_value.weight": "hf_ckpt_39.pt",
1110
+ "transformer.layers.29.attention.query_key_value.bias": "hf_ckpt_39.pt",
1111
+ "transformer.layers.29.attention.dense.weight": "hf_ckpt_39.pt",
1112
+ "transformer.layers.29.attention.dense.bias": "hf_ckpt_39.pt",
1113
+ "transformer.layers.29.mlp.dense_h_to_4h.weight": "hf_ckpt_39.pt",
1114
+ "transformer.layers.29.mlp.dense_h_to_4h.bias": "hf_ckpt_39.pt",
1115
+ "transformer.layers.29.mlp.dense_4h_to_h.weight": "hf_ckpt_39.pt",
1116
+ "transformer.layers.29.mlp.dense_4h_to_h.bias": "hf_ckpt_39.pt",
1117
+ "transformer.layers.30.input_layernorm.weight": "hf_ckpt_39.pt",
1118
+ "transformer.layers.30.input_layernorm.bias": "hf_ckpt_39.pt",
1119
+ "transformer.layers.30.post_attention_layernorm.weight": "hf_ckpt_39.pt",
1120
+ "transformer.layers.30.post_attention_layernorm.bias": "hf_ckpt_39.pt",
1121
+ "transformer.layers.30.attention.bias": "hf_ckpt_39.pt",
1122
+ "transformer.layers.30.attention.masked_bias": "hf_ckpt_40.pt",
1123
+ "transformer.layers.30.attention.rotary_emb.inv_freq": "hf_ckpt_40.pt",
1124
+ "transformer.layers.30.attention.query_key_value.weight": "hf_ckpt_40.pt",
1125
+ "transformer.layers.30.attention.query_key_value.bias": "hf_ckpt_40.pt",
1126
+ "transformer.layers.30.attention.dense.weight": "hf_ckpt_40.pt",
1127
+ "transformer.layers.30.attention.dense.bias": "hf_ckpt_40.pt",
1128
+ "transformer.layers.30.mlp.dense_h_to_4h.weight": "hf_ckpt_40.pt",
1129
+ "transformer.layers.30.mlp.dense_h_to_4h.bias": "hf_ckpt_40.pt",
1130
+ "transformer.layers.30.mlp.dense_4h_to_h.weight": "hf_ckpt_40.pt",
1131
+ "transformer.layers.30.mlp.dense_4h_to_h.bias": "hf_ckpt_40.pt",
1132
+ "transformer.layers.31.input_layernorm.weight": "hf_ckpt_40.pt",
1133
+ "transformer.layers.31.input_layernorm.bias": "hf_ckpt_40.pt",
1134
+ "transformer.layers.31.post_attention_layernorm.weight": "hf_ckpt_40.pt",
1135
+ "transformer.layers.31.post_attention_layernorm.bias": "hf_ckpt_40.pt",
1136
+ "transformer.layers.31.attention.bias": "hf_ckpt_40.pt",
1137
+ "transformer.layers.31.attention.masked_bias": "hf_ckpt_40.pt",
1138
+ "transformer.layers.31.attention.rotary_emb.inv_freq": "hf_ckpt_40.pt",
1139
+ "transformer.layers.31.attention.query_key_value.weight": "hf_ckpt_40.pt",
1140
+ "transformer.layers.31.attention.query_key_value.bias": "hf_ckpt_40.pt",
1141
+ "transformer.layers.31.attention.dense.weight": "hf_ckpt_40.pt",
1142
+ "transformer.layers.31.attention.dense.bias": "hf_ckpt_40.pt",
1143
+ "transformer.layers.31.mlp.dense_h_to_4h.weight": "hf_ckpt_40.pt",
1144
+ "transformer.layers.31.mlp.dense_h_to_4h.bias": "hf_ckpt_40.pt",
1145
+ "transformer.layers.31.mlp.dense_4h_to_h.weight": "hf_ckpt_40.pt",
1146
+ "transformer.layers.31.mlp.dense_4h_to_h.bias": "hf_ckpt_40.pt",
1147
+ "transformer.layers.32.input_layernorm.weight": "hf_ckpt_40.pt",
1148
+ "transformer.layers.32.input_layernorm.bias": "hf_ckpt_40.pt",
1149
+ "transformer.layers.32.post_attention_layernorm.weight": "hf_ckpt_40.pt",
1150
+ "transformer.layers.32.post_attention_layernorm.bias": "hf_ckpt_41.pt",
1151
+ "transformer.layers.32.attention.bias": "hf_ckpt_41.pt",
1152
+ "transformer.layers.32.attention.masked_bias": "hf_ckpt_41.pt",
1153
+ "transformer.layers.32.attention.rotary_emb.inv_freq": "hf_ckpt_41.pt",
1154
+ "transformer.layers.32.attention.query_key_value.weight": "hf_ckpt_41.pt",
1155
+ "transformer.layers.32.attention.query_key_value.bias": "hf_ckpt_41.pt",
1156
+ "transformer.layers.32.attention.dense.weight": "hf_ckpt_41.pt",
1157
+ "transformer.layers.32.attention.dense.bias": "hf_ckpt_41.pt",
1158
+ "transformer.layers.32.mlp.dense_h_to_4h.weight": "hf_ckpt_41.pt",
1159
+ "transformer.layers.32.mlp.dense_h_to_4h.bias": "hf_ckpt_41.pt",
1160
+ "transformer.layers.32.mlp.dense_4h_to_h.weight": "hf_ckpt_41.pt",
1161
+ "transformer.layers.32.mlp.dense_4h_to_h.bias": "hf_ckpt_41.pt",
1162
+ "transformer.layers.33.input_layernorm.weight": "hf_ckpt_41.pt",
1163
+ "transformer.layers.33.input_layernorm.bias": "hf_ckpt_41.pt",
1164
+ "transformer.layers.33.post_attention_layernorm.weight": "hf_ckpt_41.pt",
1165
+ "transformer.layers.33.post_attention_layernorm.bias": "hf_ckpt_41.pt",
1166
+ "transformer.layers.33.attention.bias": "hf_ckpt_41.pt",
1167
+ "transformer.layers.33.attention.masked_bias": "hf_ckpt_41.pt",
1168
+ "transformer.layers.33.attention.rotary_emb.inv_freq": "hf_ckpt_41.pt",
1169
+ "transformer.layers.33.attention.query_key_value.weight": "hf_ckpt_41.pt",
1170
+ "transformer.layers.33.attention.query_key_value.bias": "hf_ckpt_41.pt",
1171
+ "transformer.layers.33.attention.dense.weight": "hf_ckpt_41.pt",
1172
+ "transformer.layers.33.attention.dense.bias": "hf_ckpt_41.pt",
1173
+ "transformer.layers.33.mlp.dense_h_to_4h.weight": "hf_ckpt_41.pt",
1174
+ "transformer.layers.33.mlp.dense_h_to_4h.bias": "hf_ckpt_41.pt",
1175
+ "transformer.layers.33.mlp.dense_4h_to_h.weight": "hf_ckpt_41.pt",
1176
+ "transformer.layers.33.mlp.dense_4h_to_h.bias": "hf_ckpt_41.pt",
1177
+ "transformer.layers.34.input_layernorm.weight": "hf_ckpt_41.pt",
1178
+ "transformer.layers.34.input_layernorm.bias": "hf_ckpt_42.pt",
1179
+ "transformer.layers.34.post_attention_layernorm.weight": "hf_ckpt_42.pt",
1180
+ "transformer.layers.34.post_attention_layernorm.bias": "hf_ckpt_42.pt",
1181
+ "transformer.layers.34.attention.bias": "hf_ckpt_42.pt",
1182
+ "transformer.layers.34.attention.masked_bias": "hf_ckpt_42.pt",
1183
+ "transformer.layers.34.attention.rotary_emb.inv_freq": "hf_ckpt_42.pt",
1184
+ "transformer.layers.34.attention.query_key_value.weight": "hf_ckpt_42.pt",
1185
+ "transformer.layers.34.attention.query_key_value.bias": "hf_ckpt_42.pt",
1186
+ "transformer.layers.34.attention.dense.weight": "hf_ckpt_42.pt",
1187
+ "transformer.layers.34.attention.dense.bias": "hf_ckpt_42.pt",
1188
+ "transformer.layers.34.mlp.dense_h_to_4h.weight": "hf_ckpt_42.pt",
1189
+ "transformer.layers.34.mlp.dense_h_to_4h.bias": "hf_ckpt_42.pt",
1190
+ "transformer.layers.34.mlp.dense_4h_to_h.weight": "hf_ckpt_42.pt",
1191
+ "transformer.layers.34.mlp.dense_4h_to_h.bias": "hf_ckpt_42.pt",
1192
+ "transformer.layers.35.input_layernorm.weight": "hf_ckpt_42.pt",
1193
+ "transformer.layers.35.input_layernorm.bias": "hf_ckpt_42.pt",
1194
+ "transformer.layers.35.post_attention_layernorm.weight": "hf_ckpt_42.pt",
1195
+ "transformer.layers.35.post_attention_layernorm.bias": "hf_ckpt_42.pt",
1196
+ "transformer.layers.35.attention.bias": "hf_ckpt_42.pt",
1197
+ "transformer.layers.35.attention.masked_bias": "hf_ckpt_42.pt",
1198
+ "transformer.layers.35.attention.rotary_emb.inv_freq": "hf_ckpt_42.pt",
1199
+ "transformer.layers.35.attention.query_key_value.weight": "hf_ckpt_42.pt",
1200
+ "transformer.layers.35.attention.query_key_value.bias": "hf_ckpt_42.pt",
1201
+ "transformer.layers.35.attention.dense.weight": "hf_ckpt_42.pt",
1202
+ "transformer.layers.35.attention.dense.bias": "hf_ckpt_42.pt",
1203
+ "transformer.layers.35.mlp.dense_h_to_4h.weight": "hf_ckpt_42.pt",
1204
+ "transformer.layers.35.mlp.dense_h_to_4h.bias": "hf_ckpt_42.pt",
1205
+ "transformer.layers.35.mlp.dense_4h_to_h.weight": "hf_ckpt_42.pt",
1206
+ "transformer.layers.35.mlp.dense_4h_to_h.bias": "hf_ckpt_43.pt",
1207
+ "transformer.layers.36.input_layernorm.weight": "hf_ckpt_43.pt",
1208
+ "transformer.layers.36.input_layernorm.bias": "hf_ckpt_43.pt",
1209
+ "transformer.layers.36.post_attention_layernorm.weight": "hf_ckpt_43.pt",
1210
+ "transformer.layers.36.post_attention_layernorm.bias": "hf_ckpt_43.pt",
1211
+ "transformer.layers.36.attention.bias": "hf_ckpt_43.pt",
1212
+ "transformer.layers.36.attention.masked_bias": "hf_ckpt_43.pt",
1213
+ "transformer.layers.36.attention.rotary_emb.inv_freq": "hf_ckpt_43.pt",
1214
+ "transformer.layers.36.attention.query_key_value.weight": "hf_ckpt_43.pt",
1215
+ "transformer.layers.36.attention.query_key_value.bias": "hf_ckpt_43.pt",
1216
+ "transformer.layers.36.attention.dense.weight": "hf_ckpt_43.pt",
1217
+ "transformer.layers.36.attention.dense.bias": "hf_ckpt_43.pt",
1218
+ "transformer.layers.36.mlp.dense_h_to_4h.weight": "hf_ckpt_43.pt",
1219
+ "transformer.layers.36.mlp.dense_h_to_4h.bias": "hf_ckpt_43.pt",
1220
+ "transformer.layers.36.mlp.dense_4h_to_h.weight": "hf_ckpt_43.pt",
1221
+ "transformer.layers.36.mlp.dense_4h_to_h.bias": "hf_ckpt_43.pt",
1222
+ "transformer.layers.37.input_layernorm.weight": "hf_ckpt_43.pt",
1223
+ "transformer.layers.37.input_layernorm.bias": "hf_ckpt_43.pt",
1224
+ "transformer.layers.37.post_attention_layernorm.weight": "hf_ckpt_43.pt",
1225
+ "transformer.layers.37.post_attention_layernorm.bias": "hf_ckpt_43.pt",
1226
+ "transformer.layers.37.attention.bias": "hf_ckpt_43.pt",
1227
+ "transformer.layers.37.attention.masked_bias": "hf_ckpt_43.pt",
1228
+ "transformer.layers.37.attention.rotary_emb.inv_freq": "hf_ckpt_43.pt",
1229
+ "transformer.layers.37.attention.query_key_value.weight": "hf_ckpt_43.pt",
1230
+ "transformer.layers.37.attention.query_key_value.bias": "hf_ckpt_43.pt",
1231
+ "transformer.layers.37.attention.dense.weight": "hf_ckpt_43.pt",
1232
+ "transformer.layers.37.attention.dense.bias": "hf_ckpt_43.pt",
1233
+ "transformer.layers.37.mlp.dense_h_to_4h.weight": "hf_ckpt_43.pt",
1234
+ "transformer.layers.37.mlp.dense_h_to_4h.bias": "hf_ckpt_44.pt",
1235
+ "transformer.layers.37.mlp.dense_4h_to_h.weight": "hf_ckpt_44.pt",
1236
+ "transformer.layers.37.mlp.dense_4h_to_h.bias": "hf_ckpt_44.pt",
1237
+ "transformer.layers.38.input_layernorm.weight": "hf_ckpt_44.pt",
1238
+ "transformer.layers.38.input_layernorm.bias": "hf_ckpt_44.pt",
1239
+ "transformer.layers.38.post_attention_layernorm.weight": "hf_ckpt_44.pt",
1240
+ "transformer.layers.38.post_attention_layernorm.bias": "hf_ckpt_44.pt",
1241
+ "transformer.layers.38.attention.bias": "hf_ckpt_44.pt",
1242
+ "transformer.layers.38.attention.masked_bias": "hf_ckpt_44.pt",
1243
+ "transformer.layers.38.attention.rotary_emb.inv_freq": "hf_ckpt_44.pt",
1244
+ "transformer.layers.38.attention.query_key_value.weight": "hf_ckpt_44.pt",
1245
+ "transformer.layers.38.attention.query_key_value.bias": "hf_ckpt_44.pt",
1246
+ "transformer.layers.38.attention.dense.weight": "hf_ckpt_44.pt",
1247
+ "transformer.layers.38.attention.dense.bias": "hf_ckpt_44.pt",
1248
+ "transformer.layers.38.mlp.dense_h_to_4h.weight": "hf_ckpt_44.pt",
1249
+ "transformer.layers.38.mlp.dense_h_to_4h.bias": "hf_ckpt_44.pt",
1250
+ "transformer.layers.38.mlp.dense_4h_to_h.weight": "hf_ckpt_44.pt",
1251
+ "transformer.layers.38.mlp.dense_4h_to_h.bias": "hf_ckpt_44.pt",
1252
+ "transformer.layers.39.input_layernorm.weight": "hf_ckpt_44.pt",
1253
+ "transformer.layers.39.input_layernorm.bias": "hf_ckpt_44.pt",
1254
+ "transformer.layers.39.post_attention_layernorm.weight": "hf_ckpt_44.pt",
1255
+ "transformer.layers.39.post_attention_layernorm.bias": "hf_ckpt_44.pt",
1256
+ "transformer.layers.39.attention.bias": "hf_ckpt_44.pt",
1257
+ "transformer.layers.39.attention.masked_bias": "hf_ckpt_44.pt",
1258
+ "transformer.layers.39.attention.rotary_emb.inv_freq": "hf_ckpt_44.pt",
1259
+ "transformer.layers.39.attention.query_key_value.weight": "hf_ckpt_44.pt",
1260
+ "transformer.layers.39.attention.query_key_value.bias": "hf_ckpt_44.pt",
1261
+ "transformer.layers.39.attention.dense.weight": "hf_ckpt_44.pt",
1262
+ "transformer.layers.39.attention.dense.bias": "hf_ckpt_45.pt",
1263
+ "transformer.layers.39.mlp.dense_h_to_4h.weight": "hf_ckpt_45.pt",
1264
+ "transformer.layers.39.mlp.dense_h_to_4h.bias": "hf_ckpt_45.pt",
1265
+ "transformer.layers.39.mlp.dense_4h_to_h.weight": "hf_ckpt_45.pt",
1266
+ "transformer.layers.39.mlp.dense_4h_to_h.bias": "hf_ckpt_45.pt",
1267
+ "transformer.layers.40.input_layernorm.weight": "hf_ckpt_45.pt",
1268
+ "transformer.layers.40.input_layernorm.bias": "hf_ckpt_45.pt",
1269
+ "transformer.layers.40.post_attention_layernorm.weight": "hf_ckpt_45.pt",
1270
+ "transformer.layers.40.post_attention_layernorm.bias": "hf_ckpt_45.pt",
1271
+ "transformer.layers.40.attention.bias": "hf_ckpt_45.pt",
1272
+ "transformer.layers.40.attention.masked_bias": "hf_ckpt_45.pt",
1273
+ "transformer.layers.40.attention.rotary_emb.inv_freq": "hf_ckpt_45.pt",
1274
+ "transformer.layers.40.attention.query_key_value.weight": "hf_ckpt_45.pt",
1275
+ "transformer.layers.40.attention.query_key_value.bias": "hf_ckpt_45.pt",
1276
+ "transformer.layers.40.attention.dense.weight": "hf_ckpt_45.pt",
1277
+ "transformer.layers.40.attention.dense.bias": "hf_ckpt_45.pt",
1278
+ "transformer.layers.40.mlp.dense_h_to_4h.weight": "hf_ckpt_45.pt",
1279
+ "transformer.layers.40.mlp.dense_h_to_4h.bias": "hf_ckpt_45.pt",
1280
+ "transformer.layers.40.mlp.dense_4h_to_h.weight": "hf_ckpt_45.pt",
1281
+ "transformer.layers.40.mlp.dense_4h_to_h.bias": "hf_ckpt_45.pt",
1282
+ "transformer.layers.41.input_layernorm.weight": "hf_ckpt_45.pt",
1283
+ "transformer.layers.41.input_layernorm.bias": "hf_ckpt_45.pt",
1284
+ "transformer.layers.41.post_attention_layernorm.weight": "hf_ckpt_45.pt",
1285
+ "transformer.layers.41.post_attention_layernorm.bias": "hf_ckpt_45.pt",
1286
+ "transformer.layers.41.attention.bias": "hf_ckpt_45.pt",
1287
+ "transformer.layers.41.attention.masked_bias": "hf_ckpt_45.pt",
1288
+ "transformer.layers.41.attention.rotary_emb.inv_freq": "hf_ckpt_45.pt",
1289
+ "transformer.layers.41.attention.query_key_value.weight": "hf_ckpt_45.pt",
1290
+ "transformer.layers.41.attention.query_key_value.bias": "hf_ckpt_46.pt",
1291
+ "transformer.layers.41.attention.dense.weight": "hf_ckpt_46.pt",
1292
+ "transformer.layers.41.attention.dense.bias": "hf_ckpt_46.pt",
1293
+ "transformer.layers.41.mlp.dense_h_to_4h.weight": "hf_ckpt_46.pt",
1294
+ "transformer.layers.41.mlp.dense_h_to_4h.bias": "hf_ckpt_46.pt",
1295
+ "transformer.layers.41.mlp.dense_4h_to_h.weight": "hf_ckpt_46.pt",
1296
+ "transformer.layers.41.mlp.dense_4h_to_h.bias": "hf_ckpt_46.pt",
1297
+ "transformer.layers.42.input_layernorm.weight": "hf_ckpt_46.pt",
1298
+ "transformer.layers.42.input_layernorm.bias": "hf_ckpt_46.pt",
1299
+ "transformer.layers.42.post_attention_layernorm.weight": "hf_ckpt_46.pt",
1300
+ "transformer.layers.42.post_attention_layernorm.bias": "hf_ckpt_46.pt",
1301
+ "transformer.layers.42.attention.bias": "hf_ckpt_46.pt",
1302
+ "transformer.layers.42.attention.masked_bias": "hf_ckpt_46.pt",
1303
+ "transformer.layers.42.attention.rotary_emb.inv_freq": "hf_ckpt_46.pt",
1304
+ "transformer.layers.42.attention.query_key_value.weight": "hf_ckpt_46.pt",
1305
+ "transformer.layers.42.attention.query_key_value.bias": "hf_ckpt_46.pt",
1306
+ "transformer.layers.42.attention.dense.weight": "hf_ckpt_46.pt",
1307
+ "transformer.layers.42.attention.dense.bias": "hf_ckpt_46.pt",
1308
+ "transformer.layers.42.mlp.dense_h_to_4h.weight": "hf_ckpt_46.pt",
1309
+ "transformer.layers.42.mlp.dense_h_to_4h.bias": "hf_ckpt_46.pt",
1310
+ "transformer.layers.42.mlp.dense_4h_to_h.weight": "hf_ckpt_46.pt",
1311
+ "transformer.layers.42.mlp.dense_4h_to_h.bias": "hf_ckpt_46.pt",
1312
+ "transformer.layers.43.input_layernorm.weight": "hf_ckpt_46.pt",
1313
+ "transformer.layers.43.input_layernorm.bias": "hf_ckpt_46.pt",
1314
+ "transformer.layers.43.post_attention_layernorm.weight": "hf_ckpt_46.pt",
1315
+ "transformer.layers.43.post_attention_layernorm.bias": "hf_ckpt_46.pt",
1316
+ "transformer.layers.43.attention.bias": "hf_ckpt_46.pt",
1317
+ "transformer.layers.43.attention.masked_bias": "hf_ckpt_46.pt",
1318
+ "transformer.layers.43.attention.rotary_emb.inv_freq": "hf_ckpt_47.pt",
1319
+ "transformer.layers.43.attention.query_key_value.weight": "hf_ckpt_47.pt",
1320
+ "transformer.layers.43.attention.query_key_value.bias": "hf_ckpt_47.pt",
1321
+ "transformer.layers.43.attention.dense.weight": "hf_ckpt_47.pt",
1322
+ "transformer.layers.43.attention.dense.bias": "hf_ckpt_47.pt",
1323
+ "transformer.layers.43.mlp.dense_h_to_4h.weight": "hf_ckpt_47.pt",
1324
+ "transformer.layers.43.mlp.dense_h_to_4h.bias": "hf_ckpt_47.pt",
1325
+ "transformer.layers.43.mlp.dense_4h_to_h.weight": "hf_ckpt_47.pt",
1326
+ "transformer.layers.43.mlp.dense_4h_to_h.bias": "hf_ckpt_47.pt",
1327
+ "transformer.final_layer_norm.weight": "hf_ckpt_47.pt",
1328
+ "transformer.final_layer_norm.bias": "hf_ckpt_47.pt",
1329
+ "v_head.weight": "hf_ckpt_47.pt"
1330
+ }