patrickvonplaten ArthurZ HF staff commited on
Commit
d03bd04
1 Parent(s): e5a8cc1

Add Flax weights (#5)

Browse files

- Add Flax weights (459b4abbcc7d5558196418eb3f770b9b0497cbf4)


Co-authored-by: Arthur Zucker <[email protected]>

flax_model-00001-of-00002.msgpack ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e7d1cdd23c8fc68e26e2ec8f119c7b33e21a9511e8dfd0cd349d222ad24fa79
3
+ size 9976202375
flax_model-00002-of-00002.msgpack ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b408202854dbe2d2957690720ec0f16ae8cbbe5495eda084aab14e110ba83110
3
+ size 4161843861
flax_model.msgpack.index.json ADDED
@@ -0,0 +1,372 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_size": 14138032128
4
+ },
5
+ "weight_map": {
6
+ "transformer/h/0/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
7
+ "transformer/h/0/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
8
+ "transformer/h/0/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
9
+ "transformer/h/0/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
10
+ "transformer/h/0/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
11
+ "transformer/h/0/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
12
+ "transformer/h/0/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
13
+ "transformer/h/0/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
14
+ "transformer/h/0/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
15
+ "transformer/h/0/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
16
+ "transformer/h/0/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
17
+ "transformer/h/0/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
18
+ "transformer/h/1/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
19
+ "transformer/h/1/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
20
+ "transformer/h/1/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
21
+ "transformer/h/1/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
22
+ "transformer/h/1/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
23
+ "transformer/h/1/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
24
+ "transformer/h/1/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
25
+ "transformer/h/1/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
26
+ "transformer/h/1/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
27
+ "transformer/h/1/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
28
+ "transformer/h/1/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
29
+ "transformer/h/1/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
30
+ "transformer/h/10/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
31
+ "transformer/h/10/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
32
+ "transformer/h/10/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
33
+ "transformer/h/10/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
34
+ "transformer/h/10/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
35
+ "transformer/h/10/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
36
+ "transformer/h/10/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
37
+ "transformer/h/10/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
38
+ "transformer/h/10/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
39
+ "transformer/h/10/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
40
+ "transformer/h/10/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
41
+ "transformer/h/10/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
42
+ "transformer/h/11/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
43
+ "transformer/h/11/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
44
+ "transformer/h/11/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
45
+ "transformer/h/11/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
46
+ "transformer/h/11/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
47
+ "transformer/h/11/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
48
+ "transformer/h/11/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
49
+ "transformer/h/11/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
50
+ "transformer/h/11/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
51
+ "transformer/h/11/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
52
+ "transformer/h/11/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
53
+ "transformer/h/11/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
54
+ "transformer/h/12/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
55
+ "transformer/h/12/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
56
+ "transformer/h/12/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
57
+ "transformer/h/12/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
58
+ "transformer/h/12/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
59
+ "transformer/h/12/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
60
+ "transformer/h/12/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
61
+ "transformer/h/12/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
62
+ "transformer/h/12/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
63
+ "transformer/h/12/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
64
+ "transformer/h/12/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
65
+ "transformer/h/12/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
66
+ "transformer/h/13/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
67
+ "transformer/h/13/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
68
+ "transformer/h/13/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
69
+ "transformer/h/13/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
70
+ "transformer/h/13/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
71
+ "transformer/h/13/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
72
+ "transformer/h/13/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
73
+ "transformer/h/13/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
74
+ "transformer/h/13/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
75
+ "transformer/h/13/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
76
+ "transformer/h/13/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
77
+ "transformer/h/13/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
78
+ "transformer/h/14/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
79
+ "transformer/h/14/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
80
+ "transformer/h/14/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
81
+ "transformer/h/14/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
82
+ "transformer/h/14/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
83
+ "transformer/h/14/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
84
+ "transformer/h/14/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
85
+ "transformer/h/14/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
86
+ "transformer/h/14/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
87
+ "transformer/h/14/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
88
+ "transformer/h/14/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
89
+ "transformer/h/14/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
90
+ "transformer/h/15/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
91
+ "transformer/h/15/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
92
+ "transformer/h/15/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
93
+ "transformer/h/15/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
94
+ "transformer/h/15/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
95
+ "transformer/h/15/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
96
+ "transformer/h/15/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
97
+ "transformer/h/15/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
98
+ "transformer/h/15/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
99
+ "transformer/h/15/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
100
+ "transformer/h/15/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
101
+ "transformer/h/15/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
102
+ "transformer/h/16/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
103
+ "transformer/h/16/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
104
+ "transformer/h/16/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
105
+ "transformer/h/16/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
106
+ "transformer/h/16/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
107
+ "transformer/h/16/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
108
+ "transformer/h/16/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
109
+ "transformer/h/16/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
110
+ "transformer/h/16/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
111
+ "transformer/h/16/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
112
+ "transformer/h/16/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
113
+ "transformer/h/16/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
114
+ "transformer/h/17/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
115
+ "transformer/h/17/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
116
+ "transformer/h/17/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
117
+ "transformer/h/17/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
118
+ "transformer/h/17/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
119
+ "transformer/h/17/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
120
+ "transformer/h/17/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
121
+ "transformer/h/17/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
122
+ "transformer/h/17/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
123
+ "transformer/h/17/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
124
+ "transformer/h/17/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
125
+ "transformer/h/17/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
126
+ "transformer/h/18/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
127
+ "transformer/h/18/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
128
+ "transformer/h/18/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
129
+ "transformer/h/18/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
130
+ "transformer/h/18/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
131
+ "transformer/h/18/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
132
+ "transformer/h/18/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
133
+ "transformer/h/18/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
134
+ "transformer/h/18/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
135
+ "transformer/h/18/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
136
+ "transformer/h/18/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
137
+ "transformer/h/18/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
138
+ "transformer/h/19/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
139
+ "transformer/h/19/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
140
+ "transformer/h/19/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00002.msgpack",
141
+ "transformer/h/19/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00002.msgpack",
142
+ "transformer/h/19/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
143
+ "transformer/h/19/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
144
+ "transformer/h/19/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
145
+ "transformer/h/19/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
146
+ "transformer/h/19/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
147
+ "transformer/h/19/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
148
+ "transformer/h/19/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
149
+ "transformer/h/19/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
150
+ "transformer/h/2/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
151
+ "transformer/h/2/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
152
+ "transformer/h/2/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
153
+ "transformer/h/2/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
154
+ "transformer/h/2/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
155
+ "transformer/h/2/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
156
+ "transformer/h/2/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
157
+ "transformer/h/2/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
158
+ "transformer/h/2/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
159
+ "transformer/h/2/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
160
+ "transformer/h/2/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
161
+ "transformer/h/2/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
162
+ "transformer/h/20/input_layernorm/bias": "flax_model-00002-of-00002.msgpack",
163
+ "transformer/h/20/input_layernorm/scale": "flax_model-00002-of-00002.msgpack",
164
+ "transformer/h/20/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00002.msgpack",
165
+ "transformer/h/20/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00002.msgpack",
166
+ "transformer/h/20/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00002.msgpack",
167
+ "transformer/h/20/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00002.msgpack",
168
+ "transformer/h/20/post_attention_layernorm/bias": "flax_model-00002-of-00002.msgpack",
169
+ "transformer/h/20/post_attention_layernorm/scale": "flax_model-00002-of-00002.msgpack",
170
+ "transformer/h/20/self_attention/dense/bias": "flax_model-00002-of-00002.msgpack",
171
+ "transformer/h/20/self_attention/dense/kernel": "flax_model-00002-of-00002.msgpack",
172
+ "transformer/h/20/self_attention/query_key_value/bias": "flax_model-00002-of-00002.msgpack",
173
+ "transformer/h/20/self_attention/query_key_value/kernel": "flax_model-00002-of-00002.msgpack",
174
+ "transformer/h/21/input_layernorm/bias": "flax_model-00002-of-00002.msgpack",
175
+ "transformer/h/21/input_layernorm/scale": "flax_model-00002-of-00002.msgpack",
176
+ "transformer/h/21/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00002.msgpack",
177
+ "transformer/h/21/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00002.msgpack",
178
+ "transformer/h/21/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00002.msgpack",
179
+ "transformer/h/21/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00002.msgpack",
180
+ "transformer/h/21/post_attention_layernorm/bias": "flax_model-00002-of-00002.msgpack",
181
+ "transformer/h/21/post_attention_layernorm/scale": "flax_model-00002-of-00002.msgpack",
182
+ "transformer/h/21/self_attention/dense/bias": "flax_model-00002-of-00002.msgpack",
183
+ "transformer/h/21/self_attention/dense/kernel": "flax_model-00002-of-00002.msgpack",
184
+ "transformer/h/21/self_attention/query_key_value/bias": "flax_model-00002-of-00002.msgpack",
185
+ "transformer/h/21/self_attention/query_key_value/kernel": "flax_model-00002-of-00002.msgpack",
186
+ "transformer/h/22/input_layernorm/bias": "flax_model-00002-of-00002.msgpack",
187
+ "transformer/h/22/input_layernorm/scale": "flax_model-00002-of-00002.msgpack",
188
+ "transformer/h/22/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00002.msgpack",
189
+ "transformer/h/22/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00002.msgpack",
190
+ "transformer/h/22/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00002.msgpack",
191
+ "transformer/h/22/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00002.msgpack",
192
+ "transformer/h/22/post_attention_layernorm/bias": "flax_model-00002-of-00002.msgpack",
193
+ "transformer/h/22/post_attention_layernorm/scale": "flax_model-00002-of-00002.msgpack",
194
+ "transformer/h/22/self_attention/dense/bias": "flax_model-00002-of-00002.msgpack",
195
+ "transformer/h/22/self_attention/dense/kernel": "flax_model-00002-of-00002.msgpack",
196
+ "transformer/h/22/self_attention/query_key_value/bias": "flax_model-00002-of-00002.msgpack",
197
+ "transformer/h/22/self_attention/query_key_value/kernel": "flax_model-00002-of-00002.msgpack",
198
+ "transformer/h/23/input_layernorm/bias": "flax_model-00002-of-00002.msgpack",
199
+ "transformer/h/23/input_layernorm/scale": "flax_model-00002-of-00002.msgpack",
200
+ "transformer/h/23/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00002.msgpack",
201
+ "transformer/h/23/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00002.msgpack",
202
+ "transformer/h/23/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00002.msgpack",
203
+ "transformer/h/23/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00002.msgpack",
204
+ "transformer/h/23/post_attention_layernorm/bias": "flax_model-00002-of-00002.msgpack",
205
+ "transformer/h/23/post_attention_layernorm/scale": "flax_model-00002-of-00002.msgpack",
206
+ "transformer/h/23/self_attention/dense/bias": "flax_model-00002-of-00002.msgpack",
207
+ "transformer/h/23/self_attention/dense/kernel": "flax_model-00002-of-00002.msgpack",
208
+ "transformer/h/23/self_attention/query_key_value/bias": "flax_model-00002-of-00002.msgpack",
209
+ "transformer/h/23/self_attention/query_key_value/kernel": "flax_model-00002-of-00002.msgpack",
210
+ "transformer/h/24/input_layernorm/bias": "flax_model-00002-of-00002.msgpack",
211
+ "transformer/h/24/input_layernorm/scale": "flax_model-00002-of-00002.msgpack",
212
+ "transformer/h/24/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00002.msgpack",
213
+ "transformer/h/24/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00002.msgpack",
214
+ "transformer/h/24/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00002.msgpack",
215
+ "transformer/h/24/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00002.msgpack",
216
+ "transformer/h/24/post_attention_layernorm/bias": "flax_model-00002-of-00002.msgpack",
217
+ "transformer/h/24/post_attention_layernorm/scale": "flax_model-00002-of-00002.msgpack",
218
+ "transformer/h/24/self_attention/dense/bias": "flax_model-00002-of-00002.msgpack",
219
+ "transformer/h/24/self_attention/dense/kernel": "flax_model-00002-of-00002.msgpack",
220
+ "transformer/h/24/self_attention/query_key_value/bias": "flax_model-00002-of-00002.msgpack",
221
+ "transformer/h/24/self_attention/query_key_value/kernel": "flax_model-00002-of-00002.msgpack",
222
+ "transformer/h/25/input_layernorm/bias": "flax_model-00002-of-00002.msgpack",
223
+ "transformer/h/25/input_layernorm/scale": "flax_model-00002-of-00002.msgpack",
224
+ "transformer/h/25/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00002.msgpack",
225
+ "transformer/h/25/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00002.msgpack",
226
+ "transformer/h/25/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00002.msgpack",
227
+ "transformer/h/25/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00002.msgpack",
228
+ "transformer/h/25/post_attention_layernorm/bias": "flax_model-00002-of-00002.msgpack",
229
+ "transformer/h/25/post_attention_layernorm/scale": "flax_model-00002-of-00002.msgpack",
230
+ "transformer/h/25/self_attention/dense/bias": "flax_model-00002-of-00002.msgpack",
231
+ "transformer/h/25/self_attention/dense/kernel": "flax_model-00002-of-00002.msgpack",
232
+ "transformer/h/25/self_attention/query_key_value/bias": "flax_model-00002-of-00002.msgpack",
233
+ "transformer/h/25/self_attention/query_key_value/kernel": "flax_model-00002-of-00002.msgpack",
234
+ "transformer/h/26/input_layernorm/bias": "flax_model-00002-of-00002.msgpack",
235
+ "transformer/h/26/input_layernorm/scale": "flax_model-00002-of-00002.msgpack",
236
+ "transformer/h/26/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00002.msgpack",
237
+ "transformer/h/26/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00002.msgpack",
238
+ "transformer/h/26/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00002.msgpack",
239
+ "transformer/h/26/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00002.msgpack",
240
+ "transformer/h/26/post_attention_layernorm/bias": "flax_model-00002-of-00002.msgpack",
241
+ "transformer/h/26/post_attention_layernorm/scale": "flax_model-00002-of-00002.msgpack",
242
+ "transformer/h/26/self_attention/dense/bias": "flax_model-00002-of-00002.msgpack",
243
+ "transformer/h/26/self_attention/dense/kernel": "flax_model-00002-of-00002.msgpack",
244
+ "transformer/h/26/self_attention/query_key_value/bias": "flax_model-00002-of-00002.msgpack",
245
+ "transformer/h/26/self_attention/query_key_value/kernel": "flax_model-00002-of-00002.msgpack",
246
+ "transformer/h/27/input_layernorm/bias": "flax_model-00002-of-00002.msgpack",
247
+ "transformer/h/27/input_layernorm/scale": "flax_model-00002-of-00002.msgpack",
248
+ "transformer/h/27/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00002.msgpack",
249
+ "transformer/h/27/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00002.msgpack",
250
+ "transformer/h/27/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00002.msgpack",
251
+ "transformer/h/27/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00002.msgpack",
252
+ "transformer/h/27/post_attention_layernorm/bias": "flax_model-00002-of-00002.msgpack",
253
+ "transformer/h/27/post_attention_layernorm/scale": "flax_model-00002-of-00002.msgpack",
254
+ "transformer/h/27/self_attention/dense/bias": "flax_model-00002-of-00002.msgpack",
255
+ "transformer/h/27/self_attention/dense/kernel": "flax_model-00002-of-00002.msgpack",
256
+ "transformer/h/27/self_attention/query_key_value/bias": "flax_model-00002-of-00002.msgpack",
257
+ "transformer/h/27/self_attention/query_key_value/kernel": "flax_model-00002-of-00002.msgpack",
258
+ "transformer/h/28/input_layernorm/bias": "flax_model-00002-of-00002.msgpack",
259
+ "transformer/h/28/input_layernorm/scale": "flax_model-00002-of-00002.msgpack",
260
+ "transformer/h/28/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00002.msgpack",
261
+ "transformer/h/28/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00002.msgpack",
262
+ "transformer/h/28/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00002.msgpack",
263
+ "transformer/h/28/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00002.msgpack",
264
+ "transformer/h/28/post_attention_layernorm/bias": "flax_model-00002-of-00002.msgpack",
265
+ "transformer/h/28/post_attention_layernorm/scale": "flax_model-00002-of-00002.msgpack",
266
+ "transformer/h/28/self_attention/dense/bias": "flax_model-00002-of-00002.msgpack",
267
+ "transformer/h/28/self_attention/dense/kernel": "flax_model-00002-of-00002.msgpack",
268
+ "transformer/h/28/self_attention/query_key_value/bias": "flax_model-00002-of-00002.msgpack",
269
+ "transformer/h/28/self_attention/query_key_value/kernel": "flax_model-00002-of-00002.msgpack",
270
+ "transformer/h/29/input_layernorm/bias": "flax_model-00002-of-00002.msgpack",
271
+ "transformer/h/29/input_layernorm/scale": "flax_model-00002-of-00002.msgpack",
272
+ "transformer/h/29/mlp/dense_4h_to_h/bias": "flax_model-00002-of-00002.msgpack",
273
+ "transformer/h/29/mlp/dense_4h_to_h/kernel": "flax_model-00002-of-00002.msgpack",
274
+ "transformer/h/29/mlp/dense_h_to_4h/bias": "flax_model-00002-of-00002.msgpack",
275
+ "transformer/h/29/mlp/dense_h_to_4h/kernel": "flax_model-00002-of-00002.msgpack",
276
+ "transformer/h/29/post_attention_layernorm/bias": "flax_model-00002-of-00002.msgpack",
277
+ "transformer/h/29/post_attention_layernorm/scale": "flax_model-00002-of-00002.msgpack",
278
+ "transformer/h/29/self_attention/dense/bias": "flax_model-00002-of-00002.msgpack",
279
+ "transformer/h/29/self_attention/dense/kernel": "flax_model-00002-of-00002.msgpack",
280
+ "transformer/h/29/self_attention/query_key_value/bias": "flax_model-00002-of-00002.msgpack",
281
+ "transformer/h/29/self_attention/query_key_value/kernel": "flax_model-00002-of-00002.msgpack",
282
+ "transformer/h/3/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
283
+ "transformer/h/3/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
284
+ "transformer/h/3/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
285
+ "transformer/h/3/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
286
+ "transformer/h/3/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
287
+ "transformer/h/3/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
288
+ "transformer/h/3/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
289
+ "transformer/h/3/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
290
+ "transformer/h/3/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
291
+ "transformer/h/3/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
292
+ "transformer/h/3/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
293
+ "transformer/h/3/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
294
+ "transformer/h/4/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
295
+ "transformer/h/4/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
296
+ "transformer/h/4/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
297
+ "transformer/h/4/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
298
+ "transformer/h/4/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
299
+ "transformer/h/4/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
300
+ "transformer/h/4/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
301
+ "transformer/h/4/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
302
+ "transformer/h/4/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
303
+ "transformer/h/4/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
304
+ "transformer/h/4/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
305
+ "transformer/h/4/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
306
+ "transformer/h/5/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
307
+ "transformer/h/5/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
308
+ "transformer/h/5/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
309
+ "transformer/h/5/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
310
+ "transformer/h/5/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
311
+ "transformer/h/5/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
312
+ "transformer/h/5/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
313
+ "transformer/h/5/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
314
+ "transformer/h/5/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
315
+ "transformer/h/5/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
316
+ "transformer/h/5/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
317
+ "transformer/h/5/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
318
+ "transformer/h/6/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
319
+ "transformer/h/6/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
320
+ "transformer/h/6/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
321
+ "transformer/h/6/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
322
+ "transformer/h/6/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
323
+ "transformer/h/6/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
324
+ "transformer/h/6/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
325
+ "transformer/h/6/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
326
+ "transformer/h/6/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
327
+ "transformer/h/6/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
328
+ "transformer/h/6/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
329
+ "transformer/h/6/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
330
+ "transformer/h/7/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
331
+ "transformer/h/7/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
332
+ "transformer/h/7/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
333
+ "transformer/h/7/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
334
+ "transformer/h/7/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
335
+ "transformer/h/7/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
336
+ "transformer/h/7/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
337
+ "transformer/h/7/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
338
+ "transformer/h/7/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
339
+ "transformer/h/7/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
340
+ "transformer/h/7/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
341
+ "transformer/h/7/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
342
+ "transformer/h/8/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
343
+ "transformer/h/8/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
344
+ "transformer/h/8/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
345
+ "transformer/h/8/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
346
+ "transformer/h/8/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
347
+ "transformer/h/8/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
348
+ "transformer/h/8/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
349
+ "transformer/h/8/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
350
+ "transformer/h/8/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
351
+ "transformer/h/8/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
352
+ "transformer/h/8/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
353
+ "transformer/h/8/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
354
+ "transformer/h/9/input_layernorm/bias": "flax_model-00001-of-00002.msgpack",
355
+ "transformer/h/9/input_layernorm/scale": "flax_model-00001-of-00002.msgpack",
356
+ "transformer/h/9/mlp/dense_4h_to_h/bias": "flax_model-00001-of-00002.msgpack",
357
+ "transformer/h/9/mlp/dense_4h_to_h/kernel": "flax_model-00001-of-00002.msgpack",
358
+ "transformer/h/9/mlp/dense_h_to_4h/bias": "flax_model-00001-of-00002.msgpack",
359
+ "transformer/h/9/mlp/dense_h_to_4h/kernel": "flax_model-00001-of-00002.msgpack",
360
+ "transformer/h/9/post_attention_layernorm/bias": "flax_model-00001-of-00002.msgpack",
361
+ "transformer/h/9/post_attention_layernorm/scale": "flax_model-00001-of-00002.msgpack",
362
+ "transformer/h/9/self_attention/dense/bias": "flax_model-00001-of-00002.msgpack",
363
+ "transformer/h/9/self_attention/dense/kernel": "flax_model-00001-of-00002.msgpack",
364
+ "transformer/h/9/self_attention/query_key_value/bias": "flax_model-00001-of-00002.msgpack",
365
+ "transformer/h/9/self_attention/query_key_value/kernel": "flax_model-00001-of-00002.msgpack",
366
+ "transformer/ln_f/bias": "flax_model-00002-of-00002.msgpack",
367
+ "transformer/ln_f/scale": "flax_model-00002-of-00002.msgpack",
368
+ "transformer/word_embeddings/embedding": "flax_model-00001-of-00002.msgpack",
369
+ "transformer/word_embeddings_layernorm/bias": "flax_model-00001-of-00002.msgpack",
370
+ "transformer/word_embeddings_layernorm/scale": "flax_model-00001-of-00002.msgpack"
371
+ }
372
+ }