xianbin commited on
Commit
60b568e
1 Parent(s): 10f60d3

Update model weights to latest version

Browse files
config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
- "_name_or_path": "/home/project/11003280/sft_rlhf/models/7B_model/7B_SafeTensors",
3
  "architectures": [
4
  "MPTForCausalLM"
5
  ],
@@ -53,7 +53,7 @@
53
  "norm_type": "low_precision_layernorm",
54
  "resid_pdrop": 0.0,
55
  "torch_dtype": "bfloat16",
56
- "transformers_version": "4.34.0",
57
  "use_cache": false,
58
  "vocab_size": 256000
59
  }
 
1
  {
2
+ "_name_or_path": "aisingapore/sealion7b",
3
  "architectures": [
4
  "MPTForCausalLM"
5
  ],
 
53
  "norm_type": "low_precision_layernorm",
54
  "resid_pdrop": 0.0,
55
  "torch_dtype": "bfloat16",
56
+ "transformers_version": "4.37.2",
57
  "use_cache": false,
58
  "vocab_size": 256000
59
  }
generation_config.json CHANGED
@@ -1,5 +1,5 @@
1
  {
2
  "_from_model_config": true,
3
- "transformers_version": "4.34.0",
4
  "use_cache": false
5
  }
 
1
  {
2
  "_from_model_config": true,
3
+ "transformers_version": "4.37.2",
4
  "use_cache": false
5
  }
model-00001-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ec05e038ef093118222d561cf5ff110721cebb4033106cbdd9193dc601722cb4
3
+ size 4933505648
model-00002-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9c0acf0dc6025344ced5ab450d75e5bfcf62868e78a0c060c5c9fffdb286deb
3
+ size 4967831752
model-00003-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a887263b29d7b702c46295b48d28c95086ba4b9f666900577b91836e722a420
3
+ size 4967781776
model-00004-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:089b687a1102645e904f39c98094ace78ec3ed638124ec36d2a8ac11d402a047
3
+ size 134242752
model.safetensors.index.json CHANGED
@@ -3,521 +3,521 @@
3
  "total_size": 15003303936
4
  },
5
  "weight_map": {
6
- "transformer.blocks.0.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
7
- "transformer.blocks.0.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
8
- "transformer.blocks.0.attn.k_ln.bias": "model-00001-of-00002.safetensors",
9
- "transformer.blocks.0.attn.k_ln.weight": "model-00001-of-00002.safetensors",
10
- "transformer.blocks.0.attn.out_proj.bias": "model-00001-of-00002.safetensors",
11
- "transformer.blocks.0.attn.out_proj.weight": "model-00001-of-00002.safetensors",
12
- "transformer.blocks.0.attn.q_ln.bias": "model-00001-of-00002.safetensors",
13
- "transformer.blocks.0.attn.q_ln.weight": "model-00001-of-00002.safetensors",
14
- "transformer.blocks.0.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
15
- "transformer.blocks.0.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
16
- "transformer.blocks.0.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
17
- "transformer.blocks.0.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
18
- "transformer.blocks.0.norm_1.bias": "model-00001-of-00002.safetensors",
19
- "transformer.blocks.0.norm_1.weight": "model-00001-of-00002.safetensors",
20
- "transformer.blocks.0.norm_2.bias": "model-00001-of-00002.safetensors",
21
- "transformer.blocks.0.norm_2.weight": "model-00001-of-00002.safetensors",
22
- "transformer.blocks.1.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
23
- "transformer.blocks.1.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
24
- "transformer.blocks.1.attn.k_ln.bias": "model-00001-of-00002.safetensors",
25
- "transformer.blocks.1.attn.k_ln.weight": "model-00001-of-00002.safetensors",
26
- "transformer.blocks.1.attn.out_proj.bias": "model-00001-of-00002.safetensors",
27
- "transformer.blocks.1.attn.out_proj.weight": "model-00001-of-00002.safetensors",
28
- "transformer.blocks.1.attn.q_ln.bias": "model-00001-of-00002.safetensors",
29
- "transformer.blocks.1.attn.q_ln.weight": "model-00001-of-00002.safetensors",
30
- "transformer.blocks.1.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
31
- "transformer.blocks.1.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
32
- "transformer.blocks.1.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
33
- "transformer.blocks.1.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
34
- "transformer.blocks.1.norm_1.bias": "model-00001-of-00002.safetensors",
35
- "transformer.blocks.1.norm_1.weight": "model-00001-of-00002.safetensors",
36
- "transformer.blocks.1.norm_2.bias": "model-00001-of-00002.safetensors",
37
- "transformer.blocks.1.norm_2.weight": "model-00001-of-00002.safetensors",
38
- "transformer.blocks.10.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
39
- "transformer.blocks.10.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
40
- "transformer.blocks.10.attn.k_ln.bias": "model-00001-of-00002.safetensors",
41
- "transformer.blocks.10.attn.k_ln.weight": "model-00001-of-00002.safetensors",
42
- "transformer.blocks.10.attn.out_proj.bias": "model-00001-of-00002.safetensors",
43
- "transformer.blocks.10.attn.out_proj.weight": "model-00001-of-00002.safetensors",
44
- "transformer.blocks.10.attn.q_ln.bias": "model-00001-of-00002.safetensors",
45
- "transformer.blocks.10.attn.q_ln.weight": "model-00001-of-00002.safetensors",
46
- "transformer.blocks.10.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
47
- "transformer.blocks.10.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
48
- "transformer.blocks.10.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
49
- "transformer.blocks.10.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
50
- "transformer.blocks.10.norm_1.bias": "model-00001-of-00002.safetensors",
51
- "transformer.blocks.10.norm_1.weight": "model-00001-of-00002.safetensors",
52
- "transformer.blocks.10.norm_2.bias": "model-00001-of-00002.safetensors",
53
- "transformer.blocks.10.norm_2.weight": "model-00001-of-00002.safetensors",
54
- "transformer.blocks.11.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
55
- "transformer.blocks.11.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
56
- "transformer.blocks.11.attn.k_ln.bias": "model-00001-of-00002.safetensors",
57
- "transformer.blocks.11.attn.k_ln.weight": "model-00001-of-00002.safetensors",
58
- "transformer.blocks.11.attn.out_proj.bias": "model-00001-of-00002.safetensors",
59
- "transformer.blocks.11.attn.out_proj.weight": "model-00001-of-00002.safetensors",
60
- "transformer.blocks.11.attn.q_ln.bias": "model-00001-of-00002.safetensors",
61
- "transformer.blocks.11.attn.q_ln.weight": "model-00001-of-00002.safetensors",
62
- "transformer.blocks.11.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
63
- "transformer.blocks.11.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
64
- "transformer.blocks.11.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
65
- "transformer.blocks.11.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
66
- "transformer.blocks.11.norm_1.bias": "model-00001-of-00002.safetensors",
67
- "transformer.blocks.11.norm_1.weight": "model-00001-of-00002.safetensors",
68
- "transformer.blocks.11.norm_2.bias": "model-00001-of-00002.safetensors",
69
- "transformer.blocks.11.norm_2.weight": "model-00001-of-00002.safetensors",
70
- "transformer.blocks.12.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
71
- "transformer.blocks.12.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
72
- "transformer.blocks.12.attn.k_ln.bias": "model-00001-of-00002.safetensors",
73
- "transformer.blocks.12.attn.k_ln.weight": "model-00001-of-00002.safetensors",
74
- "transformer.blocks.12.attn.out_proj.bias": "model-00001-of-00002.safetensors",
75
- "transformer.blocks.12.attn.out_proj.weight": "model-00001-of-00002.safetensors",
76
- "transformer.blocks.12.attn.q_ln.bias": "model-00001-of-00002.safetensors",
77
- "transformer.blocks.12.attn.q_ln.weight": "model-00001-of-00002.safetensors",
78
- "transformer.blocks.12.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
79
- "transformer.blocks.12.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
80
- "transformer.blocks.12.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
81
- "transformer.blocks.12.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
82
- "transformer.blocks.12.norm_1.bias": "model-00001-of-00002.safetensors",
83
- "transformer.blocks.12.norm_1.weight": "model-00001-of-00002.safetensors",
84
- "transformer.blocks.12.norm_2.bias": "model-00001-of-00002.safetensors",
85
- "transformer.blocks.12.norm_2.weight": "model-00001-of-00002.safetensors",
86
- "transformer.blocks.13.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
87
- "transformer.blocks.13.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
88
- "transformer.blocks.13.attn.k_ln.bias": "model-00001-of-00002.safetensors",
89
- "transformer.blocks.13.attn.k_ln.weight": "model-00001-of-00002.safetensors",
90
- "transformer.blocks.13.attn.out_proj.bias": "model-00001-of-00002.safetensors",
91
- "transformer.blocks.13.attn.out_proj.weight": "model-00001-of-00002.safetensors",
92
- "transformer.blocks.13.attn.q_ln.bias": "model-00001-of-00002.safetensors",
93
- "transformer.blocks.13.attn.q_ln.weight": "model-00001-of-00002.safetensors",
94
- "transformer.blocks.13.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
95
- "transformer.blocks.13.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
96
- "transformer.blocks.13.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
97
- "transformer.blocks.13.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
98
- "transformer.blocks.13.norm_1.bias": "model-00001-of-00002.safetensors",
99
- "transformer.blocks.13.norm_1.weight": "model-00001-of-00002.safetensors",
100
- "transformer.blocks.13.norm_2.bias": "model-00001-of-00002.safetensors",
101
- "transformer.blocks.13.norm_2.weight": "model-00001-of-00002.safetensors",
102
- "transformer.blocks.14.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
103
- "transformer.blocks.14.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
104
- "transformer.blocks.14.attn.k_ln.bias": "model-00001-of-00002.safetensors",
105
- "transformer.blocks.14.attn.k_ln.weight": "model-00001-of-00002.safetensors",
106
- "transformer.blocks.14.attn.out_proj.bias": "model-00001-of-00002.safetensors",
107
- "transformer.blocks.14.attn.out_proj.weight": "model-00001-of-00002.safetensors",
108
- "transformer.blocks.14.attn.q_ln.bias": "model-00001-of-00002.safetensors",
109
- "transformer.blocks.14.attn.q_ln.weight": "model-00001-of-00002.safetensors",
110
- "transformer.blocks.14.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
111
- "transformer.blocks.14.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
112
- "transformer.blocks.14.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
113
- "transformer.blocks.14.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
114
- "transformer.blocks.14.norm_1.bias": "model-00001-of-00002.safetensors",
115
- "transformer.blocks.14.norm_1.weight": "model-00001-of-00002.safetensors",
116
- "transformer.blocks.14.norm_2.bias": "model-00001-of-00002.safetensors",
117
- "transformer.blocks.14.norm_2.weight": "model-00001-of-00002.safetensors",
118
- "transformer.blocks.15.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
119
- "transformer.blocks.15.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
120
- "transformer.blocks.15.attn.k_ln.bias": "model-00001-of-00002.safetensors",
121
- "transformer.blocks.15.attn.k_ln.weight": "model-00001-of-00002.safetensors",
122
- "transformer.blocks.15.attn.out_proj.bias": "model-00001-of-00002.safetensors",
123
- "transformer.blocks.15.attn.out_proj.weight": "model-00001-of-00002.safetensors",
124
- "transformer.blocks.15.attn.q_ln.bias": "model-00001-of-00002.safetensors",
125
- "transformer.blocks.15.attn.q_ln.weight": "model-00001-of-00002.safetensors",
126
- "transformer.blocks.15.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
127
- "transformer.blocks.15.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
128
- "transformer.blocks.15.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
129
- "transformer.blocks.15.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
130
- "transformer.blocks.15.norm_1.bias": "model-00001-of-00002.safetensors",
131
- "transformer.blocks.15.norm_1.weight": "model-00001-of-00002.safetensors",
132
- "transformer.blocks.15.norm_2.bias": "model-00001-of-00002.safetensors",
133
- "transformer.blocks.15.norm_2.weight": "model-00001-of-00002.safetensors",
134
- "transformer.blocks.16.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
135
- "transformer.blocks.16.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
136
- "transformer.blocks.16.attn.k_ln.bias": "model-00001-of-00002.safetensors",
137
- "transformer.blocks.16.attn.k_ln.weight": "model-00001-of-00002.safetensors",
138
- "transformer.blocks.16.attn.out_proj.bias": "model-00001-of-00002.safetensors",
139
- "transformer.blocks.16.attn.out_proj.weight": "model-00001-of-00002.safetensors",
140
- "transformer.blocks.16.attn.q_ln.bias": "model-00001-of-00002.safetensors",
141
- "transformer.blocks.16.attn.q_ln.weight": "model-00001-of-00002.safetensors",
142
- "transformer.blocks.16.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
143
- "transformer.blocks.16.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
144
- "transformer.blocks.16.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
145
- "transformer.blocks.16.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
146
- "transformer.blocks.16.norm_1.bias": "model-00001-of-00002.safetensors",
147
- "transformer.blocks.16.norm_1.weight": "model-00001-of-00002.safetensors",
148
- "transformer.blocks.16.norm_2.bias": "model-00001-of-00002.safetensors",
149
- "transformer.blocks.16.norm_2.weight": "model-00001-of-00002.safetensors",
150
- "transformer.blocks.17.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
151
- "transformer.blocks.17.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
152
- "transformer.blocks.17.attn.k_ln.bias": "model-00001-of-00002.safetensors",
153
- "transformer.blocks.17.attn.k_ln.weight": "model-00001-of-00002.safetensors",
154
- "transformer.blocks.17.attn.out_proj.bias": "model-00001-of-00002.safetensors",
155
- "transformer.blocks.17.attn.out_proj.weight": "model-00001-of-00002.safetensors",
156
- "transformer.blocks.17.attn.q_ln.bias": "model-00001-of-00002.safetensors",
157
- "transformer.blocks.17.attn.q_ln.weight": "model-00001-of-00002.safetensors",
158
- "transformer.blocks.17.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
159
- "transformer.blocks.17.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
160
- "transformer.blocks.17.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
161
- "transformer.blocks.17.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
162
- "transformer.blocks.17.norm_1.bias": "model-00001-of-00002.safetensors",
163
- "transformer.blocks.17.norm_1.weight": "model-00001-of-00002.safetensors",
164
- "transformer.blocks.17.norm_2.bias": "model-00001-of-00002.safetensors",
165
- "transformer.blocks.17.norm_2.weight": "model-00001-of-00002.safetensors",
166
- "transformer.blocks.18.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
167
- "transformer.blocks.18.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
168
- "transformer.blocks.18.attn.k_ln.bias": "model-00001-of-00002.safetensors",
169
- "transformer.blocks.18.attn.k_ln.weight": "model-00001-of-00002.safetensors",
170
- "transformer.blocks.18.attn.out_proj.bias": "model-00001-of-00002.safetensors",
171
- "transformer.blocks.18.attn.out_proj.weight": "model-00001-of-00002.safetensors",
172
- "transformer.blocks.18.attn.q_ln.bias": "model-00001-of-00002.safetensors",
173
- "transformer.blocks.18.attn.q_ln.weight": "model-00001-of-00002.safetensors",
174
- "transformer.blocks.18.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
175
- "transformer.blocks.18.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
176
- "transformer.blocks.18.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
177
- "transformer.blocks.18.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
178
- "transformer.blocks.18.norm_1.bias": "model-00001-of-00002.safetensors",
179
- "transformer.blocks.18.norm_1.weight": "model-00001-of-00002.safetensors",
180
- "transformer.blocks.18.norm_2.bias": "model-00001-of-00002.safetensors",
181
- "transformer.blocks.18.norm_2.weight": "model-00001-of-00002.safetensors",
182
- "transformer.blocks.19.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
183
- "transformer.blocks.19.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
184
- "transformer.blocks.19.attn.k_ln.bias": "model-00001-of-00002.safetensors",
185
- "transformer.blocks.19.attn.k_ln.weight": "model-00001-of-00002.safetensors",
186
- "transformer.blocks.19.attn.out_proj.bias": "model-00001-of-00002.safetensors",
187
- "transformer.blocks.19.attn.out_proj.weight": "model-00001-of-00002.safetensors",
188
- "transformer.blocks.19.attn.q_ln.bias": "model-00001-of-00002.safetensors",
189
- "transformer.blocks.19.attn.q_ln.weight": "model-00001-of-00002.safetensors",
190
- "transformer.blocks.19.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
191
- "transformer.blocks.19.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
192
- "transformer.blocks.19.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
193
- "transformer.blocks.19.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
194
- "transformer.blocks.19.norm_1.bias": "model-00001-of-00002.safetensors",
195
- "transformer.blocks.19.norm_1.weight": "model-00001-of-00002.safetensors",
196
- "transformer.blocks.19.norm_2.bias": "model-00001-of-00002.safetensors",
197
- "transformer.blocks.19.norm_2.weight": "model-00001-of-00002.safetensors",
198
- "transformer.blocks.2.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
199
- "transformer.blocks.2.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
200
- "transformer.blocks.2.attn.k_ln.bias": "model-00001-of-00002.safetensors",
201
- "transformer.blocks.2.attn.k_ln.weight": "model-00001-of-00002.safetensors",
202
- "transformer.blocks.2.attn.out_proj.bias": "model-00001-of-00002.safetensors",
203
- "transformer.blocks.2.attn.out_proj.weight": "model-00001-of-00002.safetensors",
204
- "transformer.blocks.2.attn.q_ln.bias": "model-00001-of-00002.safetensors",
205
- "transformer.blocks.2.attn.q_ln.weight": "model-00001-of-00002.safetensors",
206
- "transformer.blocks.2.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
207
- "transformer.blocks.2.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
208
- "transformer.blocks.2.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
209
- "transformer.blocks.2.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
210
- "transformer.blocks.2.norm_1.bias": "model-00001-of-00002.safetensors",
211
- "transformer.blocks.2.norm_1.weight": "model-00001-of-00002.safetensors",
212
- "transformer.blocks.2.norm_2.bias": "model-00001-of-00002.safetensors",
213
- "transformer.blocks.2.norm_2.weight": "model-00001-of-00002.safetensors",
214
- "transformer.blocks.20.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
215
- "transformer.blocks.20.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
216
- "transformer.blocks.20.attn.k_ln.bias": "model-00002-of-00002.safetensors",
217
- "transformer.blocks.20.attn.k_ln.weight": "model-00002-of-00002.safetensors",
218
- "transformer.blocks.20.attn.out_proj.bias": "model-00002-of-00002.safetensors",
219
- "transformer.blocks.20.attn.out_proj.weight": "model-00002-of-00002.safetensors",
220
- "transformer.blocks.20.attn.q_ln.bias": "model-00002-of-00002.safetensors",
221
- "transformer.blocks.20.attn.q_ln.weight": "model-00002-of-00002.safetensors",
222
- "transformer.blocks.20.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
223
- "transformer.blocks.20.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
224
- "transformer.blocks.20.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
225
- "transformer.blocks.20.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
226
- "transformer.blocks.20.norm_1.bias": "model-00002-of-00002.safetensors",
227
- "transformer.blocks.20.norm_1.weight": "model-00002-of-00002.safetensors",
228
- "transformer.blocks.20.norm_2.bias": "model-00002-of-00002.safetensors",
229
- "transformer.blocks.20.norm_2.weight": "model-00002-of-00002.safetensors",
230
- "transformer.blocks.21.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
231
- "transformer.blocks.21.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
232
- "transformer.blocks.21.attn.k_ln.bias": "model-00002-of-00002.safetensors",
233
- "transformer.blocks.21.attn.k_ln.weight": "model-00002-of-00002.safetensors",
234
- "transformer.blocks.21.attn.out_proj.bias": "model-00002-of-00002.safetensors",
235
- "transformer.blocks.21.attn.out_proj.weight": "model-00002-of-00002.safetensors",
236
- "transformer.blocks.21.attn.q_ln.bias": "model-00002-of-00002.safetensors",
237
- "transformer.blocks.21.attn.q_ln.weight": "model-00002-of-00002.safetensors",
238
- "transformer.blocks.21.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
239
- "transformer.blocks.21.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
240
- "transformer.blocks.21.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
241
- "transformer.blocks.21.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
242
- "transformer.blocks.21.norm_1.bias": "model-00002-of-00002.safetensors",
243
- "transformer.blocks.21.norm_1.weight": "model-00002-of-00002.safetensors",
244
- "transformer.blocks.21.norm_2.bias": "model-00002-of-00002.safetensors",
245
- "transformer.blocks.21.norm_2.weight": "model-00002-of-00002.safetensors",
246
- "transformer.blocks.22.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
247
- "transformer.blocks.22.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
248
- "transformer.blocks.22.attn.k_ln.bias": "model-00002-of-00002.safetensors",
249
- "transformer.blocks.22.attn.k_ln.weight": "model-00002-of-00002.safetensors",
250
- "transformer.blocks.22.attn.out_proj.bias": "model-00002-of-00002.safetensors",
251
- "transformer.blocks.22.attn.out_proj.weight": "model-00002-of-00002.safetensors",
252
- "transformer.blocks.22.attn.q_ln.bias": "model-00002-of-00002.safetensors",
253
- "transformer.blocks.22.attn.q_ln.weight": "model-00002-of-00002.safetensors",
254
- "transformer.blocks.22.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
255
- "transformer.blocks.22.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
256
- "transformer.blocks.22.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
257
- "transformer.blocks.22.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
258
- "transformer.blocks.22.norm_1.bias": "model-00002-of-00002.safetensors",
259
- "transformer.blocks.22.norm_1.weight": "model-00002-of-00002.safetensors",
260
- "transformer.blocks.22.norm_2.bias": "model-00002-of-00002.safetensors",
261
- "transformer.blocks.22.norm_2.weight": "model-00002-of-00002.safetensors",
262
- "transformer.blocks.23.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
263
- "transformer.blocks.23.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
264
- "transformer.blocks.23.attn.k_ln.bias": "model-00002-of-00002.safetensors",
265
- "transformer.blocks.23.attn.k_ln.weight": "model-00002-of-00002.safetensors",
266
- "transformer.blocks.23.attn.out_proj.bias": "model-00002-of-00002.safetensors",
267
- "transformer.blocks.23.attn.out_proj.weight": "model-00002-of-00002.safetensors",
268
- "transformer.blocks.23.attn.q_ln.bias": "model-00002-of-00002.safetensors",
269
- "transformer.blocks.23.attn.q_ln.weight": "model-00002-of-00002.safetensors",
270
- "transformer.blocks.23.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
271
- "transformer.blocks.23.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
272
- "transformer.blocks.23.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
273
- "transformer.blocks.23.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
274
- "transformer.blocks.23.norm_1.bias": "model-00002-of-00002.safetensors",
275
- "transformer.blocks.23.norm_1.weight": "model-00002-of-00002.safetensors",
276
- "transformer.blocks.23.norm_2.bias": "model-00002-of-00002.safetensors",
277
- "transformer.blocks.23.norm_2.weight": "model-00002-of-00002.safetensors",
278
- "transformer.blocks.24.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
279
- "transformer.blocks.24.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
280
- "transformer.blocks.24.attn.k_ln.bias": "model-00002-of-00002.safetensors",
281
- "transformer.blocks.24.attn.k_ln.weight": "model-00002-of-00002.safetensors",
282
- "transformer.blocks.24.attn.out_proj.bias": "model-00002-of-00002.safetensors",
283
- "transformer.blocks.24.attn.out_proj.weight": "model-00002-of-00002.safetensors",
284
- "transformer.blocks.24.attn.q_ln.bias": "model-00002-of-00002.safetensors",
285
- "transformer.blocks.24.attn.q_ln.weight": "model-00002-of-00002.safetensors",
286
- "transformer.blocks.24.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
287
- "transformer.blocks.24.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
288
- "transformer.blocks.24.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
289
- "transformer.blocks.24.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
290
- "transformer.blocks.24.norm_1.bias": "model-00002-of-00002.safetensors",
291
- "transformer.blocks.24.norm_1.weight": "model-00002-of-00002.safetensors",
292
- "transformer.blocks.24.norm_2.bias": "model-00002-of-00002.safetensors",
293
- "transformer.blocks.24.norm_2.weight": "model-00002-of-00002.safetensors",
294
- "transformer.blocks.25.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
295
- "transformer.blocks.25.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
296
- "transformer.blocks.25.attn.k_ln.bias": "model-00002-of-00002.safetensors",
297
- "transformer.blocks.25.attn.k_ln.weight": "model-00002-of-00002.safetensors",
298
- "transformer.blocks.25.attn.out_proj.bias": "model-00002-of-00002.safetensors",
299
- "transformer.blocks.25.attn.out_proj.weight": "model-00002-of-00002.safetensors",
300
- "transformer.blocks.25.attn.q_ln.bias": "model-00002-of-00002.safetensors",
301
- "transformer.blocks.25.attn.q_ln.weight": "model-00002-of-00002.safetensors",
302
- "transformer.blocks.25.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
303
- "transformer.blocks.25.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
304
- "transformer.blocks.25.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
305
- "transformer.blocks.25.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
306
- "transformer.blocks.25.norm_1.bias": "model-00002-of-00002.safetensors",
307
- "transformer.blocks.25.norm_1.weight": "model-00002-of-00002.safetensors",
308
- "transformer.blocks.25.norm_2.bias": "model-00002-of-00002.safetensors",
309
- "transformer.blocks.25.norm_2.weight": "model-00002-of-00002.safetensors",
310
- "transformer.blocks.26.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
311
- "transformer.blocks.26.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
312
- "transformer.blocks.26.attn.k_ln.bias": "model-00002-of-00002.safetensors",
313
- "transformer.blocks.26.attn.k_ln.weight": "model-00002-of-00002.safetensors",
314
- "transformer.blocks.26.attn.out_proj.bias": "model-00002-of-00002.safetensors",
315
- "transformer.blocks.26.attn.out_proj.weight": "model-00002-of-00002.safetensors",
316
- "transformer.blocks.26.attn.q_ln.bias": "model-00002-of-00002.safetensors",
317
- "transformer.blocks.26.attn.q_ln.weight": "model-00002-of-00002.safetensors",
318
- "transformer.blocks.26.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
319
- "transformer.blocks.26.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
320
- "transformer.blocks.26.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
321
- "transformer.blocks.26.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
322
- "transformer.blocks.26.norm_1.bias": "model-00002-of-00002.safetensors",
323
- "transformer.blocks.26.norm_1.weight": "model-00002-of-00002.safetensors",
324
- "transformer.blocks.26.norm_2.bias": "model-00002-of-00002.safetensors",
325
- "transformer.blocks.26.norm_2.weight": "model-00002-of-00002.safetensors",
326
- "transformer.blocks.27.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
327
- "transformer.blocks.27.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
328
- "transformer.blocks.27.attn.k_ln.bias": "model-00002-of-00002.safetensors",
329
- "transformer.blocks.27.attn.k_ln.weight": "model-00002-of-00002.safetensors",
330
- "transformer.blocks.27.attn.out_proj.bias": "model-00002-of-00002.safetensors",
331
- "transformer.blocks.27.attn.out_proj.weight": "model-00002-of-00002.safetensors",
332
- "transformer.blocks.27.attn.q_ln.bias": "model-00002-of-00002.safetensors",
333
- "transformer.blocks.27.attn.q_ln.weight": "model-00002-of-00002.safetensors",
334
- "transformer.blocks.27.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
335
- "transformer.blocks.27.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
336
- "transformer.blocks.27.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
337
- "transformer.blocks.27.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
338
- "transformer.blocks.27.norm_1.bias": "model-00002-of-00002.safetensors",
339
- "transformer.blocks.27.norm_1.weight": "model-00002-of-00002.safetensors",
340
- "transformer.blocks.27.norm_2.bias": "model-00002-of-00002.safetensors",
341
- "transformer.blocks.27.norm_2.weight": "model-00002-of-00002.safetensors",
342
- "transformer.blocks.28.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
343
- "transformer.blocks.28.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
344
- "transformer.blocks.28.attn.k_ln.bias": "model-00002-of-00002.safetensors",
345
- "transformer.blocks.28.attn.k_ln.weight": "model-00002-of-00002.safetensors",
346
- "transformer.blocks.28.attn.out_proj.bias": "model-00002-of-00002.safetensors",
347
- "transformer.blocks.28.attn.out_proj.weight": "model-00002-of-00002.safetensors",
348
- "transformer.blocks.28.attn.q_ln.bias": "model-00002-of-00002.safetensors",
349
- "transformer.blocks.28.attn.q_ln.weight": "model-00002-of-00002.safetensors",
350
- "transformer.blocks.28.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
351
- "transformer.blocks.28.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
352
- "transformer.blocks.28.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
353
- "transformer.blocks.28.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
354
- "transformer.blocks.28.norm_1.bias": "model-00002-of-00002.safetensors",
355
- "transformer.blocks.28.norm_1.weight": "model-00002-of-00002.safetensors",
356
- "transformer.blocks.28.norm_2.bias": "model-00002-of-00002.safetensors",
357
- "transformer.blocks.28.norm_2.weight": "model-00002-of-00002.safetensors",
358
- "transformer.blocks.29.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
359
- "transformer.blocks.29.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
360
- "transformer.blocks.29.attn.k_ln.bias": "model-00002-of-00002.safetensors",
361
- "transformer.blocks.29.attn.k_ln.weight": "model-00002-of-00002.safetensors",
362
- "transformer.blocks.29.attn.out_proj.bias": "model-00002-of-00002.safetensors",
363
- "transformer.blocks.29.attn.out_proj.weight": "model-00002-of-00002.safetensors",
364
- "transformer.blocks.29.attn.q_ln.bias": "model-00002-of-00002.safetensors",
365
- "transformer.blocks.29.attn.q_ln.weight": "model-00002-of-00002.safetensors",
366
- "transformer.blocks.29.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
367
- "transformer.blocks.29.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
368
- "transformer.blocks.29.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
369
- "transformer.blocks.29.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
370
- "transformer.blocks.29.norm_1.bias": "model-00002-of-00002.safetensors",
371
- "transformer.blocks.29.norm_1.weight": "model-00002-of-00002.safetensors",
372
- "transformer.blocks.29.norm_2.bias": "model-00002-of-00002.safetensors",
373
- "transformer.blocks.29.norm_2.weight": "model-00002-of-00002.safetensors",
374
- "transformer.blocks.3.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
375
- "transformer.blocks.3.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
376
- "transformer.blocks.3.attn.k_ln.bias": "model-00001-of-00002.safetensors",
377
- "transformer.blocks.3.attn.k_ln.weight": "model-00001-of-00002.safetensors",
378
- "transformer.blocks.3.attn.out_proj.bias": "model-00001-of-00002.safetensors",
379
- "transformer.blocks.3.attn.out_proj.weight": "model-00001-of-00002.safetensors",
380
- "transformer.blocks.3.attn.q_ln.bias": "model-00001-of-00002.safetensors",
381
- "transformer.blocks.3.attn.q_ln.weight": "model-00001-of-00002.safetensors",
382
- "transformer.blocks.3.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
383
- "transformer.blocks.3.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
384
- "transformer.blocks.3.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
385
- "transformer.blocks.3.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
386
- "transformer.blocks.3.norm_1.bias": "model-00001-of-00002.safetensors",
387
- "transformer.blocks.3.norm_1.weight": "model-00001-of-00002.safetensors",
388
- "transformer.blocks.3.norm_2.bias": "model-00001-of-00002.safetensors",
389
- "transformer.blocks.3.norm_2.weight": "model-00001-of-00002.safetensors",
390
- "transformer.blocks.30.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
391
- "transformer.blocks.30.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
392
- "transformer.blocks.30.attn.k_ln.bias": "model-00002-of-00002.safetensors",
393
- "transformer.blocks.30.attn.k_ln.weight": "model-00002-of-00002.safetensors",
394
- "transformer.blocks.30.attn.out_proj.bias": "model-00002-of-00002.safetensors",
395
- "transformer.blocks.30.attn.out_proj.weight": "model-00002-of-00002.safetensors",
396
- "transformer.blocks.30.attn.q_ln.bias": "model-00002-of-00002.safetensors",
397
- "transformer.blocks.30.attn.q_ln.weight": "model-00002-of-00002.safetensors",
398
- "transformer.blocks.30.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
399
- "transformer.blocks.30.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
400
- "transformer.blocks.30.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
401
- "transformer.blocks.30.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
402
- "transformer.blocks.30.norm_1.bias": "model-00002-of-00002.safetensors",
403
- "transformer.blocks.30.norm_1.weight": "model-00002-of-00002.safetensors",
404
- "transformer.blocks.30.norm_2.bias": "model-00002-of-00002.safetensors",
405
- "transformer.blocks.30.norm_2.weight": "model-00002-of-00002.safetensors",
406
- "transformer.blocks.31.attn.Wqkv.bias": "model-00002-of-00002.safetensors",
407
- "transformer.blocks.31.attn.Wqkv.weight": "model-00002-of-00002.safetensors",
408
- "transformer.blocks.31.attn.k_ln.bias": "model-00002-of-00002.safetensors",
409
- "transformer.blocks.31.attn.k_ln.weight": "model-00002-of-00002.safetensors",
410
- "transformer.blocks.31.attn.out_proj.bias": "model-00002-of-00002.safetensors",
411
- "transformer.blocks.31.attn.out_proj.weight": "model-00002-of-00002.safetensors",
412
- "transformer.blocks.31.attn.q_ln.bias": "model-00002-of-00002.safetensors",
413
- "transformer.blocks.31.attn.q_ln.weight": "model-00002-of-00002.safetensors",
414
- "transformer.blocks.31.ffn.down_proj.bias": "model-00002-of-00002.safetensors",
415
- "transformer.blocks.31.ffn.down_proj.weight": "model-00002-of-00002.safetensors",
416
- "transformer.blocks.31.ffn.up_proj.bias": "model-00002-of-00002.safetensors",
417
- "transformer.blocks.31.ffn.up_proj.weight": "model-00002-of-00002.safetensors",
418
- "transformer.blocks.31.norm_1.bias": "model-00002-of-00002.safetensors",
419
- "transformer.blocks.31.norm_1.weight": "model-00002-of-00002.safetensors",
420
- "transformer.blocks.31.norm_2.bias": "model-00002-of-00002.safetensors",
421
- "transformer.blocks.31.norm_2.weight": "model-00002-of-00002.safetensors",
422
- "transformer.blocks.4.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
423
- "transformer.blocks.4.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
424
- "transformer.blocks.4.attn.k_ln.bias": "model-00001-of-00002.safetensors",
425
- "transformer.blocks.4.attn.k_ln.weight": "model-00001-of-00002.safetensors",
426
- "transformer.blocks.4.attn.out_proj.bias": "model-00001-of-00002.safetensors",
427
- "transformer.blocks.4.attn.out_proj.weight": "model-00001-of-00002.safetensors",
428
- "transformer.blocks.4.attn.q_ln.bias": "model-00001-of-00002.safetensors",
429
- "transformer.blocks.4.attn.q_ln.weight": "model-00001-of-00002.safetensors",
430
- "transformer.blocks.4.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
431
- "transformer.blocks.4.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
432
- "transformer.blocks.4.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
433
- "transformer.blocks.4.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
434
- "transformer.blocks.4.norm_1.bias": "model-00001-of-00002.safetensors",
435
- "transformer.blocks.4.norm_1.weight": "model-00001-of-00002.safetensors",
436
- "transformer.blocks.4.norm_2.bias": "model-00001-of-00002.safetensors",
437
- "transformer.blocks.4.norm_2.weight": "model-00001-of-00002.safetensors",
438
- "transformer.blocks.5.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
439
- "transformer.blocks.5.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
440
- "transformer.blocks.5.attn.k_ln.bias": "model-00001-of-00002.safetensors",
441
- "transformer.blocks.5.attn.k_ln.weight": "model-00001-of-00002.safetensors",
442
- "transformer.blocks.5.attn.out_proj.bias": "model-00001-of-00002.safetensors",
443
- "transformer.blocks.5.attn.out_proj.weight": "model-00001-of-00002.safetensors",
444
- "transformer.blocks.5.attn.q_ln.bias": "model-00001-of-00002.safetensors",
445
- "transformer.blocks.5.attn.q_ln.weight": "model-00001-of-00002.safetensors",
446
- "transformer.blocks.5.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
447
- "transformer.blocks.5.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
448
- "transformer.blocks.5.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
449
- "transformer.blocks.5.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
450
- "transformer.blocks.5.norm_1.bias": "model-00001-of-00002.safetensors",
451
- "transformer.blocks.5.norm_1.weight": "model-00001-of-00002.safetensors",
452
- "transformer.blocks.5.norm_2.bias": "model-00001-of-00002.safetensors",
453
- "transformer.blocks.5.norm_2.weight": "model-00001-of-00002.safetensors",
454
- "transformer.blocks.6.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
455
- "transformer.blocks.6.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
456
- "transformer.blocks.6.attn.k_ln.bias": "model-00001-of-00002.safetensors",
457
- "transformer.blocks.6.attn.k_ln.weight": "model-00001-of-00002.safetensors",
458
- "transformer.blocks.6.attn.out_proj.bias": "model-00001-of-00002.safetensors",
459
- "transformer.blocks.6.attn.out_proj.weight": "model-00001-of-00002.safetensors",
460
- "transformer.blocks.6.attn.q_ln.bias": "model-00001-of-00002.safetensors",
461
- "transformer.blocks.6.attn.q_ln.weight": "model-00001-of-00002.safetensors",
462
- "transformer.blocks.6.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
463
- "transformer.blocks.6.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
464
- "transformer.blocks.6.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
465
- "transformer.blocks.6.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
466
- "transformer.blocks.6.norm_1.bias": "model-00001-of-00002.safetensors",
467
- "transformer.blocks.6.norm_1.weight": "model-00001-of-00002.safetensors",
468
- "transformer.blocks.6.norm_2.bias": "model-00001-of-00002.safetensors",
469
- "transformer.blocks.6.norm_2.weight": "model-00001-of-00002.safetensors",
470
- "transformer.blocks.7.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
471
- "transformer.blocks.7.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
472
- "transformer.blocks.7.attn.k_ln.bias": "model-00001-of-00002.safetensors",
473
- "transformer.blocks.7.attn.k_ln.weight": "model-00001-of-00002.safetensors",
474
- "transformer.blocks.7.attn.out_proj.bias": "model-00001-of-00002.safetensors",
475
- "transformer.blocks.7.attn.out_proj.weight": "model-00001-of-00002.safetensors",
476
- "transformer.blocks.7.attn.q_ln.bias": "model-00001-of-00002.safetensors",
477
- "transformer.blocks.7.attn.q_ln.weight": "model-00001-of-00002.safetensors",
478
- "transformer.blocks.7.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
479
- "transformer.blocks.7.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
480
- "transformer.blocks.7.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
481
- "transformer.blocks.7.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
482
- "transformer.blocks.7.norm_1.bias": "model-00001-of-00002.safetensors",
483
- "transformer.blocks.7.norm_1.weight": "model-00001-of-00002.safetensors",
484
- "transformer.blocks.7.norm_2.bias": "model-00001-of-00002.safetensors",
485
- "transformer.blocks.7.norm_2.weight": "model-00001-of-00002.safetensors",
486
- "transformer.blocks.8.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
487
- "transformer.blocks.8.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
488
- "transformer.blocks.8.attn.k_ln.bias": "model-00001-of-00002.safetensors",
489
- "transformer.blocks.8.attn.k_ln.weight": "model-00001-of-00002.safetensors",
490
- "transformer.blocks.8.attn.out_proj.bias": "model-00001-of-00002.safetensors",
491
- "transformer.blocks.8.attn.out_proj.weight": "model-00001-of-00002.safetensors",
492
- "transformer.blocks.8.attn.q_ln.bias": "model-00001-of-00002.safetensors",
493
- "transformer.blocks.8.attn.q_ln.weight": "model-00001-of-00002.safetensors",
494
- "transformer.blocks.8.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
495
- "transformer.blocks.8.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
496
- "transformer.blocks.8.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
497
- "transformer.blocks.8.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
498
- "transformer.blocks.8.norm_1.bias": "model-00001-of-00002.safetensors",
499
- "transformer.blocks.8.norm_1.weight": "model-00001-of-00002.safetensors",
500
- "transformer.blocks.8.norm_2.bias": "model-00001-of-00002.safetensors",
501
- "transformer.blocks.8.norm_2.weight": "model-00001-of-00002.safetensors",
502
- "transformer.blocks.9.attn.Wqkv.bias": "model-00001-of-00002.safetensors",
503
- "transformer.blocks.9.attn.Wqkv.weight": "model-00001-of-00002.safetensors",
504
- "transformer.blocks.9.attn.k_ln.bias": "model-00001-of-00002.safetensors",
505
- "transformer.blocks.9.attn.k_ln.weight": "model-00001-of-00002.safetensors",
506
- "transformer.blocks.9.attn.out_proj.bias": "model-00001-of-00002.safetensors",
507
- "transformer.blocks.9.attn.out_proj.weight": "model-00001-of-00002.safetensors",
508
- "transformer.blocks.9.attn.q_ln.bias": "model-00001-of-00002.safetensors",
509
- "transformer.blocks.9.attn.q_ln.weight": "model-00001-of-00002.safetensors",
510
- "transformer.blocks.9.ffn.down_proj.bias": "model-00001-of-00002.safetensors",
511
- "transformer.blocks.9.ffn.down_proj.weight": "model-00001-of-00002.safetensors",
512
- "transformer.blocks.9.ffn.up_proj.bias": "model-00001-of-00002.safetensors",
513
- "transformer.blocks.9.ffn.up_proj.weight": "model-00001-of-00002.safetensors",
514
- "transformer.blocks.9.norm_1.bias": "model-00001-of-00002.safetensors",
515
- "transformer.blocks.9.norm_1.weight": "model-00001-of-00002.safetensors",
516
- "transformer.blocks.9.norm_2.bias": "model-00001-of-00002.safetensors",
517
- "transformer.blocks.9.norm_2.weight": "model-00001-of-00002.safetensors",
518
- "transformer.norm_f.bias": "model-00002-of-00002.safetensors",
519
- "transformer.norm_f.weight": "model-00002-of-00002.safetensors",
520
- "transformer.wpe.weight": "model-00001-of-00002.safetensors",
521
- "transformer.wte.weight": "model-00001-of-00002.safetensors"
522
  }
523
  }
 
3
  "total_size": 15003303936
4
  },
5
  "weight_map": {
6
+ "transformer.blocks.0.attn.Wqkv.bias": "model-00001-of-00004.safetensors",
7
+ "transformer.blocks.0.attn.Wqkv.weight": "model-00001-of-00004.safetensors",
8
+ "transformer.blocks.0.attn.k_ln.bias": "model-00001-of-00004.safetensors",
9
+ "transformer.blocks.0.attn.k_ln.weight": "model-00001-of-00004.safetensors",
10
+ "transformer.blocks.0.attn.out_proj.bias": "model-00001-of-00004.safetensors",
11
+ "transformer.blocks.0.attn.out_proj.weight": "model-00001-of-00004.safetensors",
12
+ "transformer.blocks.0.attn.q_ln.bias": "model-00001-of-00004.safetensors",
13
+ "transformer.blocks.0.attn.q_ln.weight": "model-00001-of-00004.safetensors",
14
+ "transformer.blocks.0.ffn.down_proj.bias": "model-00001-of-00004.safetensors",
15
+ "transformer.blocks.0.ffn.down_proj.weight": "model-00001-of-00004.safetensors",
16
+ "transformer.blocks.0.ffn.up_proj.bias": "model-00001-of-00004.safetensors",
17
+ "transformer.blocks.0.ffn.up_proj.weight": "model-00001-of-00004.safetensors",
18
+ "transformer.blocks.0.norm_1.bias": "model-00001-of-00004.safetensors",
19
+ "transformer.blocks.0.norm_1.weight": "model-00001-of-00004.safetensors",
20
+ "transformer.blocks.0.norm_2.bias": "model-00001-of-00004.safetensors",
21
+ "transformer.blocks.0.norm_2.weight": "model-00001-of-00004.safetensors",
22
+ "transformer.blocks.1.attn.Wqkv.bias": "model-00001-of-00004.safetensors",
23
+ "transformer.blocks.1.attn.Wqkv.weight": "model-00001-of-00004.safetensors",
24
+ "transformer.blocks.1.attn.k_ln.bias": "model-00001-of-00004.safetensors",
25
+ "transformer.blocks.1.attn.k_ln.weight": "model-00001-of-00004.safetensors",
26
+ "transformer.blocks.1.attn.out_proj.bias": "model-00001-of-00004.safetensors",
27
+ "transformer.blocks.1.attn.out_proj.weight": "model-00001-of-00004.safetensors",
28
+ "transformer.blocks.1.attn.q_ln.bias": "model-00001-of-00004.safetensors",
29
+ "transformer.blocks.1.attn.q_ln.weight": "model-00001-of-00004.safetensors",
30
+ "transformer.blocks.1.ffn.down_proj.bias": "model-00001-of-00004.safetensors",
31
+ "transformer.blocks.1.ffn.down_proj.weight": "model-00001-of-00004.safetensors",
32
+ "transformer.blocks.1.ffn.up_proj.bias": "model-00001-of-00004.safetensors",
33
+ "transformer.blocks.1.ffn.up_proj.weight": "model-00001-of-00004.safetensors",
34
+ "transformer.blocks.1.norm_1.bias": "model-00001-of-00004.safetensors",
35
+ "transformer.blocks.1.norm_1.weight": "model-00001-of-00004.safetensors",
36
+ "transformer.blocks.1.norm_2.bias": "model-00001-of-00004.safetensors",
37
+ "transformer.blocks.1.norm_2.weight": "model-00001-of-00004.safetensors",
38
+ "transformer.blocks.10.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
39
+ "transformer.blocks.10.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
40
+ "transformer.blocks.10.attn.k_ln.bias": "model-00002-of-00004.safetensors",
41
+ "transformer.blocks.10.attn.k_ln.weight": "model-00002-of-00004.safetensors",
42
+ "transformer.blocks.10.attn.out_proj.bias": "model-00002-of-00004.safetensors",
43
+ "transformer.blocks.10.attn.out_proj.weight": "model-00002-of-00004.safetensors",
44
+ "transformer.blocks.10.attn.q_ln.bias": "model-00002-of-00004.safetensors",
45
+ "transformer.blocks.10.attn.q_ln.weight": "model-00002-of-00004.safetensors",
46
+ "transformer.blocks.10.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
47
+ "transformer.blocks.10.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
48
+ "transformer.blocks.10.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
49
+ "transformer.blocks.10.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
50
+ "transformer.blocks.10.norm_1.bias": "model-00002-of-00004.safetensors",
51
+ "transformer.blocks.10.norm_1.weight": "model-00002-of-00004.safetensors",
52
+ "transformer.blocks.10.norm_2.bias": "model-00002-of-00004.safetensors",
53
+ "transformer.blocks.10.norm_2.weight": "model-00002-of-00004.safetensors",
54
+ "transformer.blocks.11.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
55
+ "transformer.blocks.11.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
56
+ "transformer.blocks.11.attn.k_ln.bias": "model-00002-of-00004.safetensors",
57
+ "transformer.blocks.11.attn.k_ln.weight": "model-00002-of-00004.safetensors",
58
+ "transformer.blocks.11.attn.out_proj.bias": "model-00002-of-00004.safetensors",
59
+ "transformer.blocks.11.attn.out_proj.weight": "model-00002-of-00004.safetensors",
60
+ "transformer.blocks.11.attn.q_ln.bias": "model-00002-of-00004.safetensors",
61
+ "transformer.blocks.11.attn.q_ln.weight": "model-00002-of-00004.safetensors",
62
+ "transformer.blocks.11.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
63
+ "transformer.blocks.11.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
64
+ "transformer.blocks.11.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
65
+ "transformer.blocks.11.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
66
+ "transformer.blocks.11.norm_1.bias": "model-00002-of-00004.safetensors",
67
+ "transformer.blocks.11.norm_1.weight": "model-00002-of-00004.safetensors",
68
+ "transformer.blocks.11.norm_2.bias": "model-00002-of-00004.safetensors",
69
+ "transformer.blocks.11.norm_2.weight": "model-00002-of-00004.safetensors",
70
+ "transformer.blocks.12.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
71
+ "transformer.blocks.12.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
72
+ "transformer.blocks.12.attn.k_ln.bias": "model-00002-of-00004.safetensors",
73
+ "transformer.blocks.12.attn.k_ln.weight": "model-00002-of-00004.safetensors",
74
+ "transformer.blocks.12.attn.out_proj.bias": "model-00002-of-00004.safetensors",
75
+ "transformer.blocks.12.attn.out_proj.weight": "model-00002-of-00004.safetensors",
76
+ "transformer.blocks.12.attn.q_ln.bias": "model-00002-of-00004.safetensors",
77
+ "transformer.blocks.12.attn.q_ln.weight": "model-00002-of-00004.safetensors",
78
+ "transformer.blocks.12.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
79
+ "transformer.blocks.12.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
80
+ "transformer.blocks.12.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
81
+ "transformer.blocks.12.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
82
+ "transformer.blocks.12.norm_1.bias": "model-00002-of-00004.safetensors",
83
+ "transformer.blocks.12.norm_1.weight": "model-00002-of-00004.safetensors",
84
+ "transformer.blocks.12.norm_2.bias": "model-00002-of-00004.safetensors",
85
+ "transformer.blocks.12.norm_2.weight": "model-00002-of-00004.safetensors",
86
+ "transformer.blocks.13.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
87
+ "transformer.blocks.13.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
88
+ "transformer.blocks.13.attn.k_ln.bias": "model-00002-of-00004.safetensors",
89
+ "transformer.blocks.13.attn.k_ln.weight": "model-00002-of-00004.safetensors",
90
+ "transformer.blocks.13.attn.out_proj.bias": "model-00002-of-00004.safetensors",
91
+ "transformer.blocks.13.attn.out_proj.weight": "model-00002-of-00004.safetensors",
92
+ "transformer.blocks.13.attn.q_ln.bias": "model-00002-of-00004.safetensors",
93
+ "transformer.blocks.13.attn.q_ln.weight": "model-00002-of-00004.safetensors",
94
+ "transformer.blocks.13.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
95
+ "transformer.blocks.13.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
96
+ "transformer.blocks.13.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
97
+ "transformer.blocks.13.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
98
+ "transformer.blocks.13.norm_1.bias": "model-00002-of-00004.safetensors",
99
+ "transformer.blocks.13.norm_1.weight": "model-00002-of-00004.safetensors",
100
+ "transformer.blocks.13.norm_2.bias": "model-00002-of-00004.safetensors",
101
+ "transformer.blocks.13.norm_2.weight": "model-00002-of-00004.safetensors",
102
+ "transformer.blocks.14.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
103
+ "transformer.blocks.14.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
104
+ "transformer.blocks.14.attn.k_ln.bias": "model-00002-of-00004.safetensors",
105
+ "transformer.blocks.14.attn.k_ln.weight": "model-00002-of-00004.safetensors",
106
+ "transformer.blocks.14.attn.out_proj.bias": "model-00002-of-00004.safetensors",
107
+ "transformer.blocks.14.attn.out_proj.weight": "model-00002-of-00004.safetensors",
108
+ "transformer.blocks.14.attn.q_ln.bias": "model-00002-of-00004.safetensors",
109
+ "transformer.blocks.14.attn.q_ln.weight": "model-00002-of-00004.safetensors",
110
+ "transformer.blocks.14.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
111
+ "transformer.blocks.14.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
112
+ "transformer.blocks.14.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
113
+ "transformer.blocks.14.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
114
+ "transformer.blocks.14.norm_1.bias": "model-00002-of-00004.safetensors",
115
+ "transformer.blocks.14.norm_1.weight": "model-00002-of-00004.safetensors",
116
+ "transformer.blocks.14.norm_2.bias": "model-00002-of-00004.safetensors",
117
+ "transformer.blocks.14.norm_2.weight": "model-00002-of-00004.safetensors",
118
+ "transformer.blocks.15.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
119
+ "transformer.blocks.15.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
120
+ "transformer.blocks.15.attn.k_ln.bias": "model-00002-of-00004.safetensors",
121
+ "transformer.blocks.15.attn.k_ln.weight": "model-00002-of-00004.safetensors",
122
+ "transformer.blocks.15.attn.out_proj.bias": "model-00002-of-00004.safetensors",
123
+ "transformer.blocks.15.attn.out_proj.weight": "model-00002-of-00004.safetensors",
124
+ "transformer.blocks.15.attn.q_ln.bias": "model-00002-of-00004.safetensors",
125
+ "transformer.blocks.15.attn.q_ln.weight": "model-00002-of-00004.safetensors",
126
+ "transformer.blocks.15.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
127
+ "transformer.blocks.15.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
128
+ "transformer.blocks.15.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
129
+ "transformer.blocks.15.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
130
+ "transformer.blocks.15.norm_1.bias": "model-00002-of-00004.safetensors",
131
+ "transformer.blocks.15.norm_1.weight": "model-00002-of-00004.safetensors",
132
+ "transformer.blocks.15.norm_2.bias": "model-00002-of-00004.safetensors",
133
+ "transformer.blocks.15.norm_2.weight": "model-00002-of-00004.safetensors",
134
+ "transformer.blocks.16.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
135
+ "transformer.blocks.16.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
136
+ "transformer.blocks.16.attn.k_ln.bias": "model-00002-of-00004.safetensors",
137
+ "transformer.blocks.16.attn.k_ln.weight": "model-00002-of-00004.safetensors",
138
+ "transformer.blocks.16.attn.out_proj.bias": "model-00002-of-00004.safetensors",
139
+ "transformer.blocks.16.attn.out_proj.weight": "model-00002-of-00004.safetensors",
140
+ "transformer.blocks.16.attn.q_ln.bias": "model-00002-of-00004.safetensors",
141
+ "transformer.blocks.16.attn.q_ln.weight": "model-00002-of-00004.safetensors",
142
+ "transformer.blocks.16.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
143
+ "transformer.blocks.16.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
144
+ "transformer.blocks.16.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
145
+ "transformer.blocks.16.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
146
+ "transformer.blocks.16.norm_1.bias": "model-00002-of-00004.safetensors",
147
+ "transformer.blocks.16.norm_1.weight": "model-00002-of-00004.safetensors",
148
+ "transformer.blocks.16.norm_2.bias": "model-00002-of-00004.safetensors",
149
+ "transformer.blocks.16.norm_2.weight": "model-00002-of-00004.safetensors",
150
+ "transformer.blocks.17.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
151
+ "transformer.blocks.17.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
152
+ "transformer.blocks.17.attn.k_ln.bias": "model-00002-of-00004.safetensors",
153
+ "transformer.blocks.17.attn.k_ln.weight": "model-00002-of-00004.safetensors",
154
+ "transformer.blocks.17.attn.out_proj.bias": "model-00002-of-00004.safetensors",
155
+ "transformer.blocks.17.attn.out_proj.weight": "model-00002-of-00004.safetensors",
156
+ "transformer.blocks.17.attn.q_ln.bias": "model-00002-of-00004.safetensors",
157
+ "transformer.blocks.17.attn.q_ln.weight": "model-00002-of-00004.safetensors",
158
+ "transformer.blocks.17.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
159
+ "transformer.blocks.17.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
160
+ "transformer.blocks.17.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
161
+ "transformer.blocks.17.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
162
+ "transformer.blocks.17.norm_1.bias": "model-00002-of-00004.safetensors",
163
+ "transformer.blocks.17.norm_1.weight": "model-00002-of-00004.safetensors",
164
+ "transformer.blocks.17.norm_2.bias": "model-00002-of-00004.safetensors",
165
+ "transformer.blocks.17.norm_2.weight": "model-00002-of-00004.safetensors",
166
+ "transformer.blocks.18.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
167
+ "transformer.blocks.18.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
168
+ "transformer.blocks.18.attn.k_ln.bias": "model-00002-of-00004.safetensors",
169
+ "transformer.blocks.18.attn.k_ln.weight": "model-00002-of-00004.safetensors",
170
+ "transformer.blocks.18.attn.out_proj.bias": "model-00002-of-00004.safetensors",
171
+ "transformer.blocks.18.attn.out_proj.weight": "model-00002-of-00004.safetensors",
172
+ "transformer.blocks.18.attn.q_ln.bias": "model-00002-of-00004.safetensors",
173
+ "transformer.blocks.18.attn.q_ln.weight": "model-00002-of-00004.safetensors",
174
+ "transformer.blocks.18.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
175
+ "transformer.blocks.18.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
176
+ "transformer.blocks.18.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
177
+ "transformer.blocks.18.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
178
+ "transformer.blocks.18.norm_1.bias": "model-00002-of-00004.safetensors",
179
+ "transformer.blocks.18.norm_1.weight": "model-00002-of-00004.safetensors",
180
+ "transformer.blocks.18.norm_2.bias": "model-00002-of-00004.safetensors",
181
+ "transformer.blocks.18.norm_2.weight": "model-00002-of-00004.safetensors",
182
+ "transformer.blocks.19.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
183
+ "transformer.blocks.19.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
184
+ "transformer.blocks.19.attn.k_ln.bias": "model-00002-of-00004.safetensors",
185
+ "transformer.blocks.19.attn.k_ln.weight": "model-00002-of-00004.safetensors",
186
+ "transformer.blocks.19.attn.out_proj.bias": "model-00002-of-00004.safetensors",
187
+ "transformer.blocks.19.attn.out_proj.weight": "model-00002-of-00004.safetensors",
188
+ "transformer.blocks.19.attn.q_ln.bias": "model-00002-of-00004.safetensors",
189
+ "transformer.blocks.19.attn.q_ln.weight": "model-00002-of-00004.safetensors",
190
+ "transformer.blocks.19.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
191
+ "transformer.blocks.19.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
192
+ "transformer.blocks.19.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
193
+ "transformer.blocks.19.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
194
+ "transformer.blocks.19.norm_1.bias": "model-00002-of-00004.safetensors",
195
+ "transformer.blocks.19.norm_1.weight": "model-00002-of-00004.safetensors",
196
+ "transformer.blocks.19.norm_2.bias": "model-00002-of-00004.safetensors",
197
+ "transformer.blocks.19.norm_2.weight": "model-00002-of-00004.safetensors",
198
+ "transformer.blocks.2.attn.Wqkv.bias": "model-00001-of-00004.safetensors",
199
+ "transformer.blocks.2.attn.Wqkv.weight": "model-00001-of-00004.safetensors",
200
+ "transformer.blocks.2.attn.k_ln.bias": "model-00001-of-00004.safetensors",
201
+ "transformer.blocks.2.attn.k_ln.weight": "model-00001-of-00004.safetensors",
202
+ "transformer.blocks.2.attn.out_proj.bias": "model-00001-of-00004.safetensors",
203
+ "transformer.blocks.2.attn.out_proj.weight": "model-00001-of-00004.safetensors",
204
+ "transformer.blocks.2.attn.q_ln.bias": "model-00001-of-00004.safetensors",
205
+ "transformer.blocks.2.attn.q_ln.weight": "model-00001-of-00004.safetensors",
206
+ "transformer.blocks.2.ffn.down_proj.bias": "model-00001-of-00004.safetensors",
207
+ "transformer.blocks.2.ffn.down_proj.weight": "model-00001-of-00004.safetensors",
208
+ "transformer.blocks.2.ffn.up_proj.bias": "model-00001-of-00004.safetensors",
209
+ "transformer.blocks.2.ffn.up_proj.weight": "model-00001-of-00004.safetensors",
210
+ "transformer.blocks.2.norm_1.bias": "model-00001-of-00004.safetensors",
211
+ "transformer.blocks.2.norm_1.weight": "model-00001-of-00004.safetensors",
212
+ "transformer.blocks.2.norm_2.bias": "model-00001-of-00004.safetensors",
213
+ "transformer.blocks.2.norm_2.weight": "model-00001-of-00004.safetensors",
214
+ "transformer.blocks.20.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
215
+ "transformer.blocks.20.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
216
+ "transformer.blocks.20.attn.k_ln.bias": "model-00003-of-00004.safetensors",
217
+ "transformer.blocks.20.attn.k_ln.weight": "model-00003-of-00004.safetensors",
218
+ "transformer.blocks.20.attn.out_proj.bias": "model-00003-of-00004.safetensors",
219
+ "transformer.blocks.20.attn.out_proj.weight": "model-00003-of-00004.safetensors",
220
+ "transformer.blocks.20.attn.q_ln.bias": "model-00003-of-00004.safetensors",
221
+ "transformer.blocks.20.attn.q_ln.weight": "model-00003-of-00004.safetensors",
222
+ "transformer.blocks.20.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
223
+ "transformer.blocks.20.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
224
+ "transformer.blocks.20.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
225
+ "transformer.blocks.20.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
226
+ "transformer.blocks.20.norm_1.bias": "model-00003-of-00004.safetensors",
227
+ "transformer.blocks.20.norm_1.weight": "model-00003-of-00004.safetensors",
228
+ "transformer.blocks.20.norm_2.bias": "model-00003-of-00004.safetensors",
229
+ "transformer.blocks.20.norm_2.weight": "model-00003-of-00004.safetensors",
230
+ "transformer.blocks.21.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
231
+ "transformer.blocks.21.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
232
+ "transformer.blocks.21.attn.k_ln.bias": "model-00003-of-00004.safetensors",
233
+ "transformer.blocks.21.attn.k_ln.weight": "model-00003-of-00004.safetensors",
234
+ "transformer.blocks.21.attn.out_proj.bias": "model-00003-of-00004.safetensors",
235
+ "transformer.blocks.21.attn.out_proj.weight": "model-00003-of-00004.safetensors",
236
+ "transformer.blocks.21.attn.q_ln.bias": "model-00003-of-00004.safetensors",
237
+ "transformer.blocks.21.attn.q_ln.weight": "model-00003-of-00004.safetensors",
238
+ "transformer.blocks.21.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
239
+ "transformer.blocks.21.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
240
+ "transformer.blocks.21.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
241
+ "transformer.blocks.21.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
242
+ "transformer.blocks.21.norm_1.bias": "model-00003-of-00004.safetensors",
243
+ "transformer.blocks.21.norm_1.weight": "model-00003-of-00004.safetensors",
244
+ "transformer.blocks.21.norm_2.bias": "model-00003-of-00004.safetensors",
245
+ "transformer.blocks.21.norm_2.weight": "model-00003-of-00004.safetensors",
246
+ "transformer.blocks.22.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
247
+ "transformer.blocks.22.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
248
+ "transformer.blocks.22.attn.k_ln.bias": "model-00003-of-00004.safetensors",
249
+ "transformer.blocks.22.attn.k_ln.weight": "model-00003-of-00004.safetensors",
250
+ "transformer.blocks.22.attn.out_proj.bias": "model-00003-of-00004.safetensors",
251
+ "transformer.blocks.22.attn.out_proj.weight": "model-00003-of-00004.safetensors",
252
+ "transformer.blocks.22.attn.q_ln.bias": "model-00003-of-00004.safetensors",
253
+ "transformer.blocks.22.attn.q_ln.weight": "model-00003-of-00004.safetensors",
254
+ "transformer.blocks.22.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
255
+ "transformer.blocks.22.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
256
+ "transformer.blocks.22.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
257
+ "transformer.blocks.22.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
258
+ "transformer.blocks.22.norm_1.bias": "model-00003-of-00004.safetensors",
259
+ "transformer.blocks.22.norm_1.weight": "model-00003-of-00004.safetensors",
260
+ "transformer.blocks.22.norm_2.bias": "model-00003-of-00004.safetensors",
261
+ "transformer.blocks.22.norm_2.weight": "model-00003-of-00004.safetensors",
262
+ "transformer.blocks.23.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
263
+ "transformer.blocks.23.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
264
+ "transformer.blocks.23.attn.k_ln.bias": "model-00003-of-00004.safetensors",
265
+ "transformer.blocks.23.attn.k_ln.weight": "model-00003-of-00004.safetensors",
266
+ "transformer.blocks.23.attn.out_proj.bias": "model-00003-of-00004.safetensors",
267
+ "transformer.blocks.23.attn.out_proj.weight": "model-00003-of-00004.safetensors",
268
+ "transformer.blocks.23.attn.q_ln.bias": "model-00003-of-00004.safetensors",
269
+ "transformer.blocks.23.attn.q_ln.weight": "model-00003-of-00004.safetensors",
270
+ "transformer.blocks.23.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
271
+ "transformer.blocks.23.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
272
+ "transformer.blocks.23.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
273
+ "transformer.blocks.23.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
274
+ "transformer.blocks.23.norm_1.bias": "model-00003-of-00004.safetensors",
275
+ "transformer.blocks.23.norm_1.weight": "model-00003-of-00004.safetensors",
276
+ "transformer.blocks.23.norm_2.bias": "model-00003-of-00004.safetensors",
277
+ "transformer.blocks.23.norm_2.weight": "model-00003-of-00004.safetensors",
278
+ "transformer.blocks.24.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
279
+ "transformer.blocks.24.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
280
+ "transformer.blocks.24.attn.k_ln.bias": "model-00003-of-00004.safetensors",
281
+ "transformer.blocks.24.attn.k_ln.weight": "model-00003-of-00004.safetensors",
282
+ "transformer.blocks.24.attn.out_proj.bias": "model-00003-of-00004.safetensors",
283
+ "transformer.blocks.24.attn.out_proj.weight": "model-00003-of-00004.safetensors",
284
+ "transformer.blocks.24.attn.q_ln.bias": "model-00003-of-00004.safetensors",
285
+ "transformer.blocks.24.attn.q_ln.weight": "model-00003-of-00004.safetensors",
286
+ "transformer.blocks.24.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
287
+ "transformer.blocks.24.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
288
+ "transformer.blocks.24.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
289
+ "transformer.blocks.24.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
290
+ "transformer.blocks.24.norm_1.bias": "model-00003-of-00004.safetensors",
291
+ "transformer.blocks.24.norm_1.weight": "model-00003-of-00004.safetensors",
292
+ "transformer.blocks.24.norm_2.bias": "model-00003-of-00004.safetensors",
293
+ "transformer.blocks.24.norm_2.weight": "model-00003-of-00004.safetensors",
294
+ "transformer.blocks.25.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
295
+ "transformer.blocks.25.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
296
+ "transformer.blocks.25.attn.k_ln.bias": "model-00003-of-00004.safetensors",
297
+ "transformer.blocks.25.attn.k_ln.weight": "model-00003-of-00004.safetensors",
298
+ "transformer.blocks.25.attn.out_proj.bias": "model-00003-of-00004.safetensors",
299
+ "transformer.blocks.25.attn.out_proj.weight": "model-00003-of-00004.safetensors",
300
+ "transformer.blocks.25.attn.q_ln.bias": "model-00003-of-00004.safetensors",
301
+ "transformer.blocks.25.attn.q_ln.weight": "model-00003-of-00004.safetensors",
302
+ "transformer.blocks.25.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
303
+ "transformer.blocks.25.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
304
+ "transformer.blocks.25.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
305
+ "transformer.blocks.25.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
306
+ "transformer.blocks.25.norm_1.bias": "model-00003-of-00004.safetensors",
307
+ "transformer.blocks.25.norm_1.weight": "model-00003-of-00004.safetensors",
308
+ "transformer.blocks.25.norm_2.bias": "model-00003-of-00004.safetensors",
309
+ "transformer.blocks.25.norm_2.weight": "model-00003-of-00004.safetensors",
310
+ "transformer.blocks.26.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
311
+ "transformer.blocks.26.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
312
+ "transformer.blocks.26.attn.k_ln.bias": "model-00003-of-00004.safetensors",
313
+ "transformer.blocks.26.attn.k_ln.weight": "model-00003-of-00004.safetensors",
314
+ "transformer.blocks.26.attn.out_proj.bias": "model-00003-of-00004.safetensors",
315
+ "transformer.blocks.26.attn.out_proj.weight": "model-00003-of-00004.safetensors",
316
+ "transformer.blocks.26.attn.q_ln.bias": "model-00003-of-00004.safetensors",
317
+ "transformer.blocks.26.attn.q_ln.weight": "model-00003-of-00004.safetensors",
318
+ "transformer.blocks.26.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
319
+ "transformer.blocks.26.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
320
+ "transformer.blocks.26.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
321
+ "transformer.blocks.26.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
322
+ "transformer.blocks.26.norm_1.bias": "model-00003-of-00004.safetensors",
323
+ "transformer.blocks.26.norm_1.weight": "model-00003-of-00004.safetensors",
324
+ "transformer.blocks.26.norm_2.bias": "model-00003-of-00004.safetensors",
325
+ "transformer.blocks.26.norm_2.weight": "model-00003-of-00004.safetensors",
326
+ "transformer.blocks.27.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
327
+ "transformer.blocks.27.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
328
+ "transformer.blocks.27.attn.k_ln.bias": "model-00003-of-00004.safetensors",
329
+ "transformer.blocks.27.attn.k_ln.weight": "model-00003-of-00004.safetensors",
330
+ "transformer.blocks.27.attn.out_proj.bias": "model-00003-of-00004.safetensors",
331
+ "transformer.blocks.27.attn.out_proj.weight": "model-00003-of-00004.safetensors",
332
+ "transformer.blocks.27.attn.q_ln.bias": "model-00003-of-00004.safetensors",
333
+ "transformer.blocks.27.attn.q_ln.weight": "model-00003-of-00004.safetensors",
334
+ "transformer.blocks.27.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
335
+ "transformer.blocks.27.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
336
+ "transformer.blocks.27.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
337
+ "transformer.blocks.27.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
338
+ "transformer.blocks.27.norm_1.bias": "model-00003-of-00004.safetensors",
339
+ "transformer.blocks.27.norm_1.weight": "model-00003-of-00004.safetensors",
340
+ "transformer.blocks.27.norm_2.bias": "model-00003-of-00004.safetensors",
341
+ "transformer.blocks.27.norm_2.weight": "model-00003-of-00004.safetensors",
342
+ "transformer.blocks.28.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
343
+ "transformer.blocks.28.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
344
+ "transformer.blocks.28.attn.k_ln.bias": "model-00003-of-00004.safetensors",
345
+ "transformer.blocks.28.attn.k_ln.weight": "model-00003-of-00004.safetensors",
346
+ "transformer.blocks.28.attn.out_proj.bias": "model-00003-of-00004.safetensors",
347
+ "transformer.blocks.28.attn.out_proj.weight": "model-00003-of-00004.safetensors",
348
+ "transformer.blocks.28.attn.q_ln.bias": "model-00003-of-00004.safetensors",
349
+ "transformer.blocks.28.attn.q_ln.weight": "model-00003-of-00004.safetensors",
350
+ "transformer.blocks.28.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
351
+ "transformer.blocks.28.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
352
+ "transformer.blocks.28.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
353
+ "transformer.blocks.28.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
354
+ "transformer.blocks.28.norm_1.bias": "model-00003-of-00004.safetensors",
355
+ "transformer.blocks.28.norm_1.weight": "model-00003-of-00004.safetensors",
356
+ "transformer.blocks.28.norm_2.bias": "model-00003-of-00004.safetensors",
357
+ "transformer.blocks.28.norm_2.weight": "model-00003-of-00004.safetensors",
358
+ "transformer.blocks.29.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
359
+ "transformer.blocks.29.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
360
+ "transformer.blocks.29.attn.k_ln.bias": "model-00003-of-00004.safetensors",
361
+ "transformer.blocks.29.attn.k_ln.weight": "model-00003-of-00004.safetensors",
362
+ "transformer.blocks.29.attn.out_proj.bias": "model-00003-of-00004.safetensors",
363
+ "transformer.blocks.29.attn.out_proj.weight": "model-00003-of-00004.safetensors",
364
+ "transformer.blocks.29.attn.q_ln.bias": "model-00003-of-00004.safetensors",
365
+ "transformer.blocks.29.attn.q_ln.weight": "model-00003-of-00004.safetensors",
366
+ "transformer.blocks.29.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
367
+ "transformer.blocks.29.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
368
+ "transformer.blocks.29.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
369
+ "transformer.blocks.29.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
370
+ "transformer.blocks.29.norm_1.bias": "model-00003-of-00004.safetensors",
371
+ "transformer.blocks.29.norm_1.weight": "model-00003-of-00004.safetensors",
372
+ "transformer.blocks.29.norm_2.bias": "model-00003-of-00004.safetensors",
373
+ "transformer.blocks.29.norm_2.weight": "model-00003-of-00004.safetensors",
374
+ "transformer.blocks.3.attn.Wqkv.bias": "model-00001-of-00004.safetensors",
375
+ "transformer.blocks.3.attn.Wqkv.weight": "model-00001-of-00004.safetensors",
376
+ "transformer.blocks.3.attn.k_ln.bias": "model-00001-of-00004.safetensors",
377
+ "transformer.blocks.3.attn.k_ln.weight": "model-00001-of-00004.safetensors",
378
+ "transformer.blocks.3.attn.out_proj.bias": "model-00001-of-00004.safetensors",
379
+ "transformer.blocks.3.attn.out_proj.weight": "model-00001-of-00004.safetensors",
380
+ "transformer.blocks.3.attn.q_ln.bias": "model-00001-of-00004.safetensors",
381
+ "transformer.blocks.3.attn.q_ln.weight": "model-00001-of-00004.safetensors",
382
+ "transformer.blocks.3.ffn.down_proj.bias": "model-00001-of-00004.safetensors",
383
+ "transformer.blocks.3.ffn.down_proj.weight": "model-00001-of-00004.safetensors",
384
+ "transformer.blocks.3.ffn.up_proj.bias": "model-00001-of-00004.safetensors",
385
+ "transformer.blocks.3.ffn.up_proj.weight": "model-00001-of-00004.safetensors",
386
+ "transformer.blocks.3.norm_1.bias": "model-00001-of-00004.safetensors",
387
+ "transformer.blocks.3.norm_1.weight": "model-00001-of-00004.safetensors",
388
+ "transformer.blocks.3.norm_2.bias": "model-00001-of-00004.safetensors",
389
+ "transformer.blocks.3.norm_2.weight": "model-00001-of-00004.safetensors",
390
+ "transformer.blocks.30.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
391
+ "transformer.blocks.30.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
392
+ "transformer.blocks.30.attn.k_ln.bias": "model-00003-of-00004.safetensors",
393
+ "transformer.blocks.30.attn.k_ln.weight": "model-00003-of-00004.safetensors",
394
+ "transformer.blocks.30.attn.out_proj.bias": "model-00003-of-00004.safetensors",
395
+ "transformer.blocks.30.attn.out_proj.weight": "model-00003-of-00004.safetensors",
396
+ "transformer.blocks.30.attn.q_ln.bias": "model-00003-of-00004.safetensors",
397
+ "transformer.blocks.30.attn.q_ln.weight": "model-00003-of-00004.safetensors",
398
+ "transformer.blocks.30.ffn.down_proj.bias": "model-00003-of-00004.safetensors",
399
+ "transformer.blocks.30.ffn.down_proj.weight": "model-00003-of-00004.safetensors",
400
+ "transformer.blocks.30.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
401
+ "transformer.blocks.30.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
402
+ "transformer.blocks.30.norm_1.bias": "model-00003-of-00004.safetensors",
403
+ "transformer.blocks.30.norm_1.weight": "model-00003-of-00004.safetensors",
404
+ "transformer.blocks.30.norm_2.bias": "model-00003-of-00004.safetensors",
405
+ "transformer.blocks.30.norm_2.weight": "model-00003-of-00004.safetensors",
406
+ "transformer.blocks.31.attn.Wqkv.bias": "model-00003-of-00004.safetensors",
407
+ "transformer.blocks.31.attn.Wqkv.weight": "model-00003-of-00004.safetensors",
408
+ "transformer.blocks.31.attn.k_ln.bias": "model-00003-of-00004.safetensors",
409
+ "transformer.blocks.31.attn.k_ln.weight": "model-00003-of-00004.safetensors",
410
+ "transformer.blocks.31.attn.out_proj.bias": "model-00003-of-00004.safetensors",
411
+ "transformer.blocks.31.attn.out_proj.weight": "model-00003-of-00004.safetensors",
412
+ "transformer.blocks.31.attn.q_ln.bias": "model-00003-of-00004.safetensors",
413
+ "transformer.blocks.31.attn.q_ln.weight": "model-00003-of-00004.safetensors",
414
+ "transformer.blocks.31.ffn.down_proj.bias": "model-00004-of-00004.safetensors",
415
+ "transformer.blocks.31.ffn.down_proj.weight": "model-00004-of-00004.safetensors",
416
+ "transformer.blocks.31.ffn.up_proj.bias": "model-00003-of-00004.safetensors",
417
+ "transformer.blocks.31.ffn.up_proj.weight": "model-00003-of-00004.safetensors",
418
+ "transformer.blocks.31.norm_1.bias": "model-00003-of-00004.safetensors",
419
+ "transformer.blocks.31.norm_1.weight": "model-00003-of-00004.safetensors",
420
+ "transformer.blocks.31.norm_2.bias": "model-00003-of-00004.safetensors",
421
+ "transformer.blocks.31.norm_2.weight": "model-00003-of-00004.safetensors",
422
+ "transformer.blocks.4.attn.Wqkv.bias": "model-00001-of-00004.safetensors",
423
+ "transformer.blocks.4.attn.Wqkv.weight": "model-00001-of-00004.safetensors",
424
+ "transformer.blocks.4.attn.k_ln.bias": "model-00001-of-00004.safetensors",
425
+ "transformer.blocks.4.attn.k_ln.weight": "model-00001-of-00004.safetensors",
426
+ "transformer.blocks.4.attn.out_proj.bias": "model-00001-of-00004.safetensors",
427
+ "transformer.blocks.4.attn.out_proj.weight": "model-00001-of-00004.safetensors",
428
+ "transformer.blocks.4.attn.q_ln.bias": "model-00001-of-00004.safetensors",
429
+ "transformer.blocks.4.attn.q_ln.weight": "model-00001-of-00004.safetensors",
430
+ "transformer.blocks.4.ffn.down_proj.bias": "model-00001-of-00004.safetensors",
431
+ "transformer.blocks.4.ffn.down_proj.weight": "model-00001-of-00004.safetensors",
432
+ "transformer.blocks.4.ffn.up_proj.bias": "model-00001-of-00004.safetensors",
433
+ "transformer.blocks.4.ffn.up_proj.weight": "model-00001-of-00004.safetensors",
434
+ "transformer.blocks.4.norm_1.bias": "model-00001-of-00004.safetensors",
435
+ "transformer.blocks.4.norm_1.weight": "model-00001-of-00004.safetensors",
436
+ "transformer.blocks.4.norm_2.bias": "model-00001-of-00004.safetensors",
437
+ "transformer.blocks.4.norm_2.weight": "model-00001-of-00004.safetensors",
438
+ "transformer.blocks.5.attn.Wqkv.bias": "model-00001-of-00004.safetensors",
439
+ "transformer.blocks.5.attn.Wqkv.weight": "model-00001-of-00004.safetensors",
440
+ "transformer.blocks.5.attn.k_ln.bias": "model-00001-of-00004.safetensors",
441
+ "transformer.blocks.5.attn.k_ln.weight": "model-00001-of-00004.safetensors",
442
+ "transformer.blocks.5.attn.out_proj.bias": "model-00001-of-00004.safetensors",
443
+ "transformer.blocks.5.attn.out_proj.weight": "model-00001-of-00004.safetensors",
444
+ "transformer.blocks.5.attn.q_ln.bias": "model-00001-of-00004.safetensors",
445
+ "transformer.blocks.5.attn.q_ln.weight": "model-00001-of-00004.safetensors",
446
+ "transformer.blocks.5.ffn.down_proj.bias": "model-00001-of-00004.safetensors",
447
+ "transformer.blocks.5.ffn.down_proj.weight": "model-00001-of-00004.safetensors",
448
+ "transformer.blocks.5.ffn.up_proj.bias": "model-00001-of-00004.safetensors",
449
+ "transformer.blocks.5.ffn.up_proj.weight": "model-00001-of-00004.safetensors",
450
+ "transformer.blocks.5.norm_1.bias": "model-00001-of-00004.safetensors",
451
+ "transformer.blocks.5.norm_1.weight": "model-00001-of-00004.safetensors",
452
+ "transformer.blocks.5.norm_2.bias": "model-00001-of-00004.safetensors",
453
+ "transformer.blocks.5.norm_2.weight": "model-00001-of-00004.safetensors",
454
+ "transformer.blocks.6.attn.Wqkv.bias": "model-00001-of-00004.safetensors",
455
+ "transformer.blocks.6.attn.Wqkv.weight": "model-00001-of-00004.safetensors",
456
+ "transformer.blocks.6.attn.k_ln.bias": "model-00001-of-00004.safetensors",
457
+ "transformer.blocks.6.attn.k_ln.weight": "model-00001-of-00004.safetensors",
458
+ "transformer.blocks.6.attn.out_proj.bias": "model-00001-of-00004.safetensors",
459
+ "transformer.blocks.6.attn.out_proj.weight": "model-00001-of-00004.safetensors",
460
+ "transformer.blocks.6.attn.q_ln.bias": "model-00001-of-00004.safetensors",
461
+ "transformer.blocks.6.attn.q_ln.weight": "model-00001-of-00004.safetensors",
462
+ "transformer.blocks.6.ffn.down_proj.bias": "model-00001-of-00004.safetensors",
463
+ "transformer.blocks.6.ffn.down_proj.weight": "model-00001-of-00004.safetensors",
464
+ "transformer.blocks.6.ffn.up_proj.bias": "model-00001-of-00004.safetensors",
465
+ "transformer.blocks.6.ffn.up_proj.weight": "model-00001-of-00004.safetensors",
466
+ "transformer.blocks.6.norm_1.bias": "model-00001-of-00004.safetensors",
467
+ "transformer.blocks.6.norm_1.weight": "model-00001-of-00004.safetensors",
468
+ "transformer.blocks.6.norm_2.bias": "model-00001-of-00004.safetensors",
469
+ "transformer.blocks.6.norm_2.weight": "model-00001-of-00004.safetensors",
470
+ "transformer.blocks.7.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
471
+ "transformer.blocks.7.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
472
+ "transformer.blocks.7.attn.k_ln.bias": "model-00002-of-00004.safetensors",
473
+ "transformer.blocks.7.attn.k_ln.weight": "model-00002-of-00004.safetensors",
474
+ "transformer.blocks.7.attn.out_proj.bias": "model-00002-of-00004.safetensors",
475
+ "transformer.blocks.7.attn.out_proj.weight": "model-00002-of-00004.safetensors",
476
+ "transformer.blocks.7.attn.q_ln.bias": "model-00002-of-00004.safetensors",
477
+ "transformer.blocks.7.attn.q_ln.weight": "model-00002-of-00004.safetensors",
478
+ "transformer.blocks.7.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
479
+ "transformer.blocks.7.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
480
+ "transformer.blocks.7.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
481
+ "transformer.blocks.7.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
482
+ "transformer.blocks.7.norm_1.bias": "model-00001-of-00004.safetensors",
483
+ "transformer.blocks.7.norm_1.weight": "model-00001-of-00004.safetensors",
484
+ "transformer.blocks.7.norm_2.bias": "model-00002-of-00004.safetensors",
485
+ "transformer.blocks.7.norm_2.weight": "model-00002-of-00004.safetensors",
486
+ "transformer.blocks.8.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
487
+ "transformer.blocks.8.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
488
+ "transformer.blocks.8.attn.k_ln.bias": "model-00002-of-00004.safetensors",
489
+ "transformer.blocks.8.attn.k_ln.weight": "model-00002-of-00004.safetensors",
490
+ "transformer.blocks.8.attn.out_proj.bias": "model-00002-of-00004.safetensors",
491
+ "transformer.blocks.8.attn.out_proj.weight": "model-00002-of-00004.safetensors",
492
+ "transformer.blocks.8.attn.q_ln.bias": "model-00002-of-00004.safetensors",
493
+ "transformer.blocks.8.attn.q_ln.weight": "model-00002-of-00004.safetensors",
494
+ "transformer.blocks.8.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
495
+ "transformer.blocks.8.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
496
+ "transformer.blocks.8.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
497
+ "transformer.blocks.8.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
498
+ "transformer.blocks.8.norm_1.bias": "model-00002-of-00004.safetensors",
499
+ "transformer.blocks.8.norm_1.weight": "model-00002-of-00004.safetensors",
500
+ "transformer.blocks.8.norm_2.bias": "model-00002-of-00004.safetensors",
501
+ "transformer.blocks.8.norm_2.weight": "model-00002-of-00004.safetensors",
502
+ "transformer.blocks.9.attn.Wqkv.bias": "model-00002-of-00004.safetensors",
503
+ "transformer.blocks.9.attn.Wqkv.weight": "model-00002-of-00004.safetensors",
504
+ "transformer.blocks.9.attn.k_ln.bias": "model-00002-of-00004.safetensors",
505
+ "transformer.blocks.9.attn.k_ln.weight": "model-00002-of-00004.safetensors",
506
+ "transformer.blocks.9.attn.out_proj.bias": "model-00002-of-00004.safetensors",
507
+ "transformer.blocks.9.attn.out_proj.weight": "model-00002-of-00004.safetensors",
508
+ "transformer.blocks.9.attn.q_ln.bias": "model-00002-of-00004.safetensors",
509
+ "transformer.blocks.9.attn.q_ln.weight": "model-00002-of-00004.safetensors",
510
+ "transformer.blocks.9.ffn.down_proj.bias": "model-00002-of-00004.safetensors",
511
+ "transformer.blocks.9.ffn.down_proj.weight": "model-00002-of-00004.safetensors",
512
+ "transformer.blocks.9.ffn.up_proj.bias": "model-00002-of-00004.safetensors",
513
+ "transformer.blocks.9.ffn.up_proj.weight": "model-00002-of-00004.safetensors",
514
+ "transformer.blocks.9.norm_1.bias": "model-00002-of-00004.safetensors",
515
+ "transformer.blocks.9.norm_1.weight": "model-00002-of-00004.safetensors",
516
+ "transformer.blocks.9.norm_2.bias": "model-00002-of-00004.safetensors",
517
+ "transformer.blocks.9.norm_2.weight": "model-00002-of-00004.safetensors",
518
+ "transformer.norm_f.bias": "model-00004-of-00004.safetensors",
519
+ "transformer.norm_f.weight": "model-00004-of-00004.safetensors",
520
+ "transformer.wpe.weight": "model-00001-of-00004.safetensors",
521
+ "transformer.wte.weight": "model-00001-of-00004.safetensors"
522
  }
523
  }
special_tokens_map.json CHANGED
@@ -1,11 +1,17 @@
1
  {
2
- "additional_special_tokens": [
3
- "<unk>",
4
- "<|endoftext|>",
5
- "<|endofline|>",
6
- "<|padding|>"
7
- ],
8
- "eos_token": "<|endoftext|>",
9
  "pad_token": "<|padding|>",
10
- "unk_token": "<unk>"
 
 
 
 
 
 
11
  }
 
1
  {
2
+ "eos_token": {
3
+ "content": "<|endoftext|>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
  "pad_token": "<|padding|>",
10
+ "unk_token": {
11
+ "content": "<unk>",
12
+ "lstrip": false,
13
+ "normalized": false,
14
+ "rstrip": false,
15
+ "single_word": false
16
+ }
17
  }
tokenizer_config.json CHANGED
@@ -35,12 +35,6 @@
35
  "special": true
36
  }
37
  },
38
- "additional_special_tokens": [
39
- "<unk>",
40
- "<|endoftext|>",
41
- "<|endofline|>",
42
- "<|padding|>"
43
- ],
44
  "auto_map": {
45
  "AutoTokenizer": [
46
  "tokenization_SEA_BPE.SEABPETokenizer",
@@ -55,6 +49,5 @@
55
  "pad_token": "<|padding|>",
56
  "sp_model_kwargs": {},
57
  "tokenizer_class": "SEABPETokenizer",
58
- "tokenizer_file": null,
59
  "unk_token": "<unk>"
60
  }
 
35
  "special": true
36
  }
37
  },
 
 
 
 
 
 
38
  "auto_map": {
39
  "AutoTokenizer": [
40
  "tokenization_SEA_BPE.SEABPETokenizer",
 
49
  "pad_token": "<|padding|>",
50
  "sp_model_kwargs": {},
51
  "tokenizer_class": "SEABPETokenizer",
 
52
  "unk_token": "<unk>"
53
  }