amgadhasan commited on
Commit
9a2c2e0
1 Parent(s): 6d9ed28

Upload MixFormerSequentialForCausalLM

Browse files
config.json CHANGED
@@ -31,6 +31,6 @@
31
  "rotary_dim": 32,
32
  "tie_word_embeddings": false,
33
  "torch_dtype": "float32",
34
- "transformers_version": "4.36.0",
35
  "vocab_size": 51200
36
  }
 
31
  "rotary_dim": 32,
32
  "tie_word_embeddings": false,
33
  "torch_dtype": "float32",
34
+ "transformers_version": "4.35.2",
35
  "vocab_size": 51200
36
  }
generation_config.json CHANGED
@@ -1,4 +1,4 @@
1
  {
2
  "_from_model_config": true,
3
- "transformers_version": "4.36.0"
4
  }
 
1
  {
2
  "_from_model_config": true,
3
+ "transformers_version": "4.35.2"
4
  }
model-00001-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:62db27fd4de2e20c6bd339fd5cee1d881e03a1afc5cc007dde2a097fd9aeefe5
3
+ size 2936911448
model-00002-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f4f87b0e123c112c8c2cac5996d3c941238b93f23f11748d91360b0685b32302
3
+ size 2937066528
model-00003-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:07bb65cdf594a8c1f8d94c36328041c795b1df5fc28d226c0ff7d4d93df434da
3
+ size 2937076832
model-00004-of-00004.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0def7f178583dc2aae4207f7a4e13d38f511e3850ccce35ae18cbe918843fc95
3
+ size 2307712584
model.safetensors.index.json CHANGED
@@ -3,330 +3,330 @@
3
  "total_size": 11118735360
4
  },
5
  "weight_map": {
6
- "layers.0.wte.weight": "model-00001-of-00003.safetensors",
7
- "layers.1.ln.bias": "model-00001-of-00003.safetensors",
8
- "layers.1.ln.weight": "model-00001-of-00003.safetensors",
9
- "layers.1.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
10
- "layers.1.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
11
- "layers.1.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
12
- "layers.1.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
13
- "layers.1.mlp.fc1.bias": "model-00001-of-00003.safetensors",
14
- "layers.1.mlp.fc1.weight": "model-00001-of-00003.safetensors",
15
- "layers.1.mlp.fc2.bias": "model-00001-of-00003.safetensors",
16
- "layers.1.mlp.fc2.weight": "model-00001-of-00003.safetensors",
17
- "layers.10.ln.bias": "model-00001-of-00003.safetensors",
18
- "layers.10.ln.weight": "model-00001-of-00003.safetensors",
19
- "layers.10.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
20
- "layers.10.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
21
- "layers.10.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
22
- "layers.10.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
23
- "layers.10.mlp.fc1.bias": "model-00001-of-00003.safetensors",
24
- "layers.10.mlp.fc1.weight": "model-00001-of-00003.safetensors",
25
- "layers.10.mlp.fc2.bias": "model-00001-of-00003.safetensors",
26
- "layers.10.mlp.fc2.weight": "model-00001-of-00003.safetensors",
27
- "layers.11.ln.bias": "model-00001-of-00003.safetensors",
28
- "layers.11.ln.weight": "model-00001-of-00003.safetensors",
29
- "layers.11.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
30
- "layers.11.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
31
- "layers.11.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
32
- "layers.11.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
33
- "layers.11.mlp.fc1.bias": "model-00001-of-00003.safetensors",
34
- "layers.11.mlp.fc1.weight": "model-00001-of-00003.safetensors",
35
- "layers.11.mlp.fc2.bias": "model-00001-of-00003.safetensors",
36
- "layers.11.mlp.fc2.weight": "model-00001-of-00003.safetensors",
37
- "layers.12.ln.bias": "model-00001-of-00003.safetensors",
38
- "layers.12.ln.weight": "model-00001-of-00003.safetensors",
39
- "layers.12.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
40
- "layers.12.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
41
- "layers.12.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
42
- "layers.12.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
43
- "layers.12.mlp.fc1.bias": "model-00001-of-00003.safetensors",
44
- "layers.12.mlp.fc1.weight": "model-00001-of-00003.safetensors",
45
- "layers.12.mlp.fc2.bias": "model-00001-of-00003.safetensors",
46
- "layers.12.mlp.fc2.weight": "model-00001-of-00003.safetensors",
47
- "layers.13.ln.bias": "model-00001-of-00003.safetensors",
48
- "layers.13.ln.weight": "model-00001-of-00003.safetensors",
49
- "layers.13.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
50
- "layers.13.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
51
- "layers.13.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
52
- "layers.13.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
53
- "layers.13.mlp.fc1.bias": "model-00001-of-00003.safetensors",
54
- "layers.13.mlp.fc1.weight": "model-00001-of-00003.safetensors",
55
- "layers.13.mlp.fc2.bias": "model-00001-of-00003.safetensors",
56
- "layers.13.mlp.fc2.weight": "model-00001-of-00003.safetensors",
57
- "layers.14.ln.bias": "model-00001-of-00003.safetensors",
58
- "layers.14.ln.weight": "model-00001-of-00003.safetensors",
59
- "layers.14.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
60
- "layers.14.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
61
- "layers.14.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
62
- "layers.14.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
63
- "layers.14.mlp.fc1.bias": "model-00001-of-00003.safetensors",
64
- "layers.14.mlp.fc1.weight": "model-00001-of-00003.safetensors",
65
- "layers.14.mlp.fc2.bias": "model-00001-of-00003.safetensors",
66
- "layers.14.mlp.fc2.weight": "model-00001-of-00003.safetensors",
67
- "layers.15.ln.bias": "model-00001-of-00003.safetensors",
68
- "layers.15.ln.weight": "model-00001-of-00003.safetensors",
69
- "layers.15.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
70
- "layers.15.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
71
- "layers.15.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
72
- "layers.15.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
73
- "layers.15.mlp.fc1.bias": "model-00002-of-00003.safetensors",
74
- "layers.15.mlp.fc1.weight": "model-00002-of-00003.safetensors",
75
- "layers.15.mlp.fc2.bias": "model-00002-of-00003.safetensors",
76
- "layers.15.mlp.fc2.weight": "model-00002-of-00003.safetensors",
77
- "layers.16.ln.bias": "model-00002-of-00003.safetensors",
78
- "layers.16.ln.weight": "model-00002-of-00003.safetensors",
79
- "layers.16.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
80
- "layers.16.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
81
- "layers.16.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
82
- "layers.16.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
83
- "layers.16.mlp.fc1.bias": "model-00002-of-00003.safetensors",
84
- "layers.16.mlp.fc1.weight": "model-00002-of-00003.safetensors",
85
- "layers.16.mlp.fc2.bias": "model-00002-of-00003.safetensors",
86
- "layers.16.mlp.fc2.weight": "model-00002-of-00003.safetensors",
87
- "layers.17.ln.bias": "model-00002-of-00003.safetensors",
88
- "layers.17.ln.weight": "model-00002-of-00003.safetensors",
89
- "layers.17.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
90
- "layers.17.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
91
- "layers.17.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
92
- "layers.17.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
93
- "layers.17.mlp.fc1.bias": "model-00002-of-00003.safetensors",
94
- "layers.17.mlp.fc1.weight": "model-00002-of-00003.safetensors",
95
- "layers.17.mlp.fc2.bias": "model-00002-of-00003.safetensors",
96
- "layers.17.mlp.fc2.weight": "model-00002-of-00003.safetensors",
97
- "layers.18.ln.bias": "model-00002-of-00003.safetensors",
98
- "layers.18.ln.weight": "model-00002-of-00003.safetensors",
99
- "layers.18.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
100
- "layers.18.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
101
- "layers.18.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
102
- "layers.18.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
103
- "layers.18.mlp.fc1.bias": "model-00002-of-00003.safetensors",
104
- "layers.18.mlp.fc1.weight": "model-00002-of-00003.safetensors",
105
- "layers.18.mlp.fc2.bias": "model-00002-of-00003.safetensors",
106
- "layers.18.mlp.fc2.weight": "model-00002-of-00003.safetensors",
107
- "layers.19.ln.bias": "model-00002-of-00003.safetensors",
108
- "layers.19.ln.weight": "model-00002-of-00003.safetensors",
109
- "layers.19.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
110
- "layers.19.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
111
- "layers.19.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
112
- "layers.19.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
113
- "layers.19.mlp.fc1.bias": "model-00002-of-00003.safetensors",
114
- "layers.19.mlp.fc1.weight": "model-00002-of-00003.safetensors",
115
- "layers.19.mlp.fc2.bias": "model-00002-of-00003.safetensors",
116
- "layers.19.mlp.fc2.weight": "model-00002-of-00003.safetensors",
117
- "layers.2.ln.bias": "model-00001-of-00003.safetensors",
118
- "layers.2.ln.weight": "model-00001-of-00003.safetensors",
119
- "layers.2.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
120
- "layers.2.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
121
- "layers.2.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
122
- "layers.2.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
123
- "layers.2.mlp.fc1.bias": "model-00001-of-00003.safetensors",
124
- "layers.2.mlp.fc1.weight": "model-00001-of-00003.safetensors",
125
- "layers.2.mlp.fc2.bias": "model-00001-of-00003.safetensors",
126
- "layers.2.mlp.fc2.weight": "model-00001-of-00003.safetensors",
127
- "layers.20.ln.bias": "model-00002-of-00003.safetensors",
128
- "layers.20.ln.weight": "model-00002-of-00003.safetensors",
129
- "layers.20.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
130
- "layers.20.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
131
- "layers.20.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
132
- "layers.20.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
133
- "layers.20.mlp.fc1.bias": "model-00002-of-00003.safetensors",
134
- "layers.20.mlp.fc1.weight": "model-00002-of-00003.safetensors",
135
- "layers.20.mlp.fc2.bias": "model-00002-of-00003.safetensors",
136
- "layers.20.mlp.fc2.weight": "model-00002-of-00003.safetensors",
137
- "layers.21.ln.bias": "model-00002-of-00003.safetensors",
138
- "layers.21.ln.weight": "model-00002-of-00003.safetensors",
139
- "layers.21.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
140
- "layers.21.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
141
- "layers.21.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
142
- "layers.21.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
143
- "layers.21.mlp.fc1.bias": "model-00002-of-00003.safetensors",
144
- "layers.21.mlp.fc1.weight": "model-00002-of-00003.safetensors",
145
- "layers.21.mlp.fc2.bias": "model-00002-of-00003.safetensors",
146
- "layers.21.mlp.fc2.weight": "model-00002-of-00003.safetensors",
147
- "layers.22.ln.bias": "model-00002-of-00003.safetensors",
148
- "layers.22.ln.weight": "model-00002-of-00003.safetensors",
149
- "layers.22.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
150
- "layers.22.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
151
- "layers.22.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
152
- "layers.22.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
153
- "layers.22.mlp.fc1.bias": "model-00002-of-00003.safetensors",
154
- "layers.22.mlp.fc1.weight": "model-00002-of-00003.safetensors",
155
- "layers.22.mlp.fc2.bias": "model-00002-of-00003.safetensors",
156
- "layers.22.mlp.fc2.weight": "model-00002-of-00003.safetensors",
157
- "layers.23.ln.bias": "model-00002-of-00003.safetensors",
158
- "layers.23.ln.weight": "model-00002-of-00003.safetensors",
159
- "layers.23.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
160
- "layers.23.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
161
- "layers.23.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
162
- "layers.23.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
163
- "layers.23.mlp.fc1.bias": "model-00002-of-00003.safetensors",
164
- "layers.23.mlp.fc1.weight": "model-00002-of-00003.safetensors",
165
- "layers.23.mlp.fc2.bias": "model-00002-of-00003.safetensors",
166
- "layers.23.mlp.fc2.weight": "model-00002-of-00003.safetensors",
167
- "layers.24.ln.bias": "model-00002-of-00003.safetensors",
168
- "layers.24.ln.weight": "model-00002-of-00003.safetensors",
169
- "layers.24.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
170
- "layers.24.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
171
- "layers.24.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
172
- "layers.24.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
173
- "layers.24.mlp.fc1.bias": "model-00002-of-00003.safetensors",
174
- "layers.24.mlp.fc1.weight": "model-00002-of-00003.safetensors",
175
- "layers.24.mlp.fc2.bias": "model-00002-of-00003.safetensors",
176
- "layers.24.mlp.fc2.weight": "model-00002-of-00003.safetensors",
177
- "layers.25.ln.bias": "model-00002-of-00003.safetensors",
178
- "layers.25.ln.weight": "model-00002-of-00003.safetensors",
179
- "layers.25.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
180
- "layers.25.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
181
- "layers.25.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
182
- "layers.25.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
183
- "layers.25.mlp.fc1.bias": "model-00002-of-00003.safetensors",
184
- "layers.25.mlp.fc1.weight": "model-00002-of-00003.safetensors",
185
- "layers.25.mlp.fc2.bias": "model-00002-of-00003.safetensors",
186
- "layers.25.mlp.fc2.weight": "model-00002-of-00003.safetensors",
187
- "layers.26.ln.bias": "model-00002-of-00003.safetensors",
188
- "layers.26.ln.weight": "model-00002-of-00003.safetensors",
189
- "layers.26.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
190
- "layers.26.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
191
- "layers.26.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
192
- "layers.26.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
193
- "layers.26.mlp.fc1.bias": "model-00002-of-00003.safetensors",
194
- "layers.26.mlp.fc1.weight": "model-00002-of-00003.safetensors",
195
- "layers.26.mlp.fc2.bias": "model-00002-of-00003.safetensors",
196
- "layers.26.mlp.fc2.weight": "model-00002-of-00003.safetensors",
197
- "layers.27.ln.bias": "model-00002-of-00003.safetensors",
198
- "layers.27.ln.weight": "model-00002-of-00003.safetensors",
199
- "layers.27.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
200
- "layers.27.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
201
- "layers.27.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
202
- "layers.27.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
203
- "layers.27.mlp.fc1.bias": "model-00002-of-00003.safetensors",
204
- "layers.27.mlp.fc1.weight": "model-00002-of-00003.safetensors",
205
- "layers.27.mlp.fc2.bias": "model-00002-of-00003.safetensors",
206
- "layers.27.mlp.fc2.weight": "model-00002-of-00003.safetensors",
207
- "layers.28.ln.bias": "model-00002-of-00003.safetensors",
208
- "layers.28.ln.weight": "model-00002-of-00003.safetensors",
209
- "layers.28.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
210
- "layers.28.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
211
- "layers.28.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
212
- "layers.28.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
213
- "layers.28.mlp.fc1.bias": "model-00002-of-00003.safetensors",
214
- "layers.28.mlp.fc1.weight": "model-00002-of-00003.safetensors",
215
- "layers.28.mlp.fc2.bias": "model-00002-of-00003.safetensors",
216
- "layers.28.mlp.fc2.weight": "model-00002-of-00003.safetensors",
217
- "layers.29.ln.bias": "model-00002-of-00003.safetensors",
218
- "layers.29.ln.weight": "model-00002-of-00003.safetensors",
219
- "layers.29.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
220
- "layers.29.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
221
- "layers.29.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
222
- "layers.29.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
223
- "layers.29.mlp.fc1.bias": "model-00002-of-00003.safetensors",
224
- "layers.29.mlp.fc1.weight": "model-00002-of-00003.safetensors",
225
- "layers.29.mlp.fc2.bias": "model-00002-of-00003.safetensors",
226
- "layers.29.mlp.fc2.weight": "model-00002-of-00003.safetensors",
227
- "layers.3.ln.bias": "model-00001-of-00003.safetensors",
228
- "layers.3.ln.weight": "model-00001-of-00003.safetensors",
229
- "layers.3.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
230
- "layers.3.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
231
- "layers.3.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
232
- "layers.3.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
233
- "layers.3.mlp.fc1.bias": "model-00001-of-00003.safetensors",
234
- "layers.3.mlp.fc1.weight": "model-00001-of-00003.safetensors",
235
- "layers.3.mlp.fc2.bias": "model-00001-of-00003.safetensors",
236
- "layers.3.mlp.fc2.weight": "model-00001-of-00003.safetensors",
237
- "layers.30.ln.bias": "model-00002-of-00003.safetensors",
238
- "layers.30.ln.weight": "model-00002-of-00003.safetensors",
239
- "layers.30.mixer.Wqkv.bias": "model-00002-of-00003.safetensors",
240
- "layers.30.mixer.Wqkv.weight": "model-00002-of-00003.safetensors",
241
- "layers.30.mixer.out_proj.bias": "model-00002-of-00003.safetensors",
242
- "layers.30.mixer.out_proj.weight": "model-00002-of-00003.safetensors",
243
- "layers.30.mlp.fc1.bias": "model-00002-of-00003.safetensors",
244
- "layers.30.mlp.fc1.weight": "model-00002-of-00003.safetensors",
245
- "layers.30.mlp.fc2.bias": "model-00003-of-00003.safetensors",
246
- "layers.30.mlp.fc2.weight": "model-00003-of-00003.safetensors",
247
- "layers.31.ln.bias": "model-00003-of-00003.safetensors",
248
- "layers.31.ln.weight": "model-00003-of-00003.safetensors",
249
- "layers.31.mixer.Wqkv.bias": "model-00003-of-00003.safetensors",
250
- "layers.31.mixer.Wqkv.weight": "model-00003-of-00003.safetensors",
251
- "layers.31.mixer.out_proj.bias": "model-00003-of-00003.safetensors",
252
- "layers.31.mixer.out_proj.weight": "model-00003-of-00003.safetensors",
253
- "layers.31.mlp.fc1.bias": "model-00003-of-00003.safetensors",
254
- "layers.31.mlp.fc1.weight": "model-00003-of-00003.safetensors",
255
- "layers.31.mlp.fc2.bias": "model-00003-of-00003.safetensors",
256
- "layers.31.mlp.fc2.weight": "model-00003-of-00003.safetensors",
257
- "layers.32.ln.bias": "model-00003-of-00003.safetensors",
258
- "layers.32.ln.weight": "model-00003-of-00003.safetensors",
259
- "layers.32.mixer.Wqkv.bias": "model-00003-of-00003.safetensors",
260
- "layers.32.mixer.Wqkv.weight": "model-00003-of-00003.safetensors",
261
- "layers.32.mixer.out_proj.bias": "model-00003-of-00003.safetensors",
262
- "layers.32.mixer.out_proj.weight": "model-00003-of-00003.safetensors",
263
- "layers.32.mlp.fc1.bias": "model-00003-of-00003.safetensors",
264
- "layers.32.mlp.fc1.weight": "model-00003-of-00003.safetensors",
265
- "layers.32.mlp.fc2.bias": "model-00003-of-00003.safetensors",
266
- "layers.32.mlp.fc2.weight": "model-00003-of-00003.safetensors",
267
- "layers.33.linear.bias": "model-00003-of-00003.safetensors",
268
- "layers.33.linear.weight": "model-00003-of-00003.safetensors",
269
- "layers.33.ln.bias": "model-00003-of-00003.safetensors",
270
- "layers.33.ln.weight": "model-00003-of-00003.safetensors",
271
- "layers.4.ln.bias": "model-00001-of-00003.safetensors",
272
- "layers.4.ln.weight": "model-00001-of-00003.safetensors",
273
- "layers.4.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
274
- "layers.4.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
275
- "layers.4.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
276
- "layers.4.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
277
- "layers.4.mlp.fc1.bias": "model-00001-of-00003.safetensors",
278
- "layers.4.mlp.fc1.weight": "model-00001-of-00003.safetensors",
279
- "layers.4.mlp.fc2.bias": "model-00001-of-00003.safetensors",
280
- "layers.4.mlp.fc2.weight": "model-00001-of-00003.safetensors",
281
- "layers.5.ln.bias": "model-00001-of-00003.safetensors",
282
- "layers.5.ln.weight": "model-00001-of-00003.safetensors",
283
- "layers.5.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
284
- "layers.5.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
285
- "layers.5.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
286
- "layers.5.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
287
- "layers.5.mlp.fc1.bias": "model-00001-of-00003.safetensors",
288
- "layers.5.mlp.fc1.weight": "model-00001-of-00003.safetensors",
289
- "layers.5.mlp.fc2.bias": "model-00001-of-00003.safetensors",
290
- "layers.5.mlp.fc2.weight": "model-00001-of-00003.safetensors",
291
- "layers.6.ln.bias": "model-00001-of-00003.safetensors",
292
- "layers.6.ln.weight": "model-00001-of-00003.safetensors",
293
- "layers.6.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
294
- "layers.6.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
295
- "layers.6.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
296
- "layers.6.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
297
- "layers.6.mlp.fc1.bias": "model-00001-of-00003.safetensors",
298
- "layers.6.mlp.fc1.weight": "model-00001-of-00003.safetensors",
299
- "layers.6.mlp.fc2.bias": "model-00001-of-00003.safetensors",
300
- "layers.6.mlp.fc2.weight": "model-00001-of-00003.safetensors",
301
- "layers.7.ln.bias": "model-00001-of-00003.safetensors",
302
- "layers.7.ln.weight": "model-00001-of-00003.safetensors",
303
- "layers.7.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
304
- "layers.7.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
305
- "layers.7.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
306
- "layers.7.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
307
- "layers.7.mlp.fc1.bias": "model-00001-of-00003.safetensors",
308
- "layers.7.mlp.fc1.weight": "model-00001-of-00003.safetensors",
309
- "layers.7.mlp.fc2.bias": "model-00001-of-00003.safetensors",
310
- "layers.7.mlp.fc2.weight": "model-00001-of-00003.safetensors",
311
- "layers.8.ln.bias": "model-00001-of-00003.safetensors",
312
- "layers.8.ln.weight": "model-00001-of-00003.safetensors",
313
- "layers.8.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
314
- "layers.8.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
315
- "layers.8.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
316
- "layers.8.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
317
- "layers.8.mlp.fc1.bias": "model-00001-of-00003.safetensors",
318
- "layers.8.mlp.fc1.weight": "model-00001-of-00003.safetensors",
319
- "layers.8.mlp.fc2.bias": "model-00001-of-00003.safetensors",
320
- "layers.8.mlp.fc2.weight": "model-00001-of-00003.safetensors",
321
- "layers.9.ln.bias": "model-00001-of-00003.safetensors",
322
- "layers.9.ln.weight": "model-00001-of-00003.safetensors",
323
- "layers.9.mixer.Wqkv.bias": "model-00001-of-00003.safetensors",
324
- "layers.9.mixer.Wqkv.weight": "model-00001-of-00003.safetensors",
325
- "layers.9.mixer.out_proj.bias": "model-00001-of-00003.safetensors",
326
- "layers.9.mixer.out_proj.weight": "model-00001-of-00003.safetensors",
327
- "layers.9.mlp.fc1.bias": "model-00001-of-00003.safetensors",
328
- "layers.9.mlp.fc1.weight": "model-00001-of-00003.safetensors",
329
- "layers.9.mlp.fc2.bias": "model-00001-of-00003.safetensors",
330
- "layers.9.mlp.fc2.weight": "model-00001-of-00003.safetensors"
331
  }
332
  }
 
3
  "total_size": 11118735360
4
  },
5
  "weight_map": {
6
+ "layers.0.wte.weight": "model-00001-of-00004.safetensors",
7
+ "layers.1.ln.bias": "model-00001-of-00004.safetensors",
8
+ "layers.1.ln.weight": "model-00001-of-00004.safetensors",
9
+ "layers.1.mixer.Wqkv.bias": "model-00001-of-00004.safetensors",
10
+ "layers.1.mixer.Wqkv.weight": "model-00001-of-00004.safetensors",
11
+ "layers.1.mixer.out_proj.bias": "model-00001-of-00004.safetensors",
12
+ "layers.1.mixer.out_proj.weight": "model-00001-of-00004.safetensors",
13
+ "layers.1.mlp.fc1.bias": "model-00001-of-00004.safetensors",
14
+ "layers.1.mlp.fc1.weight": "model-00001-of-00004.safetensors",
15
+ "layers.1.mlp.fc2.bias": "model-00001-of-00004.safetensors",
16
+ "layers.1.mlp.fc2.weight": "model-00001-of-00004.safetensors",
17
+ "layers.10.ln.bias": "model-00002-of-00004.safetensors",
18
+ "layers.10.ln.weight": "model-00002-of-00004.safetensors",
19
+ "layers.10.mixer.Wqkv.bias": "model-00002-of-00004.safetensors",
20
+ "layers.10.mixer.Wqkv.weight": "model-00002-of-00004.safetensors",
21
+ "layers.10.mixer.out_proj.bias": "model-00002-of-00004.safetensors",
22
+ "layers.10.mixer.out_proj.weight": "model-00002-of-00004.safetensors",
23
+ "layers.10.mlp.fc1.bias": "model-00002-of-00004.safetensors",
24
+ "layers.10.mlp.fc1.weight": "model-00002-of-00004.safetensors",
25
+ "layers.10.mlp.fc2.bias": "model-00002-of-00004.safetensors",
26
+ "layers.10.mlp.fc2.weight": "model-00002-of-00004.safetensors",
27
+ "layers.11.ln.bias": "model-00002-of-00004.safetensors",
28
+ "layers.11.ln.weight": "model-00002-of-00004.safetensors",
29
+ "layers.11.mixer.Wqkv.bias": "model-00002-of-00004.safetensors",
30
+ "layers.11.mixer.Wqkv.weight": "model-00002-of-00004.safetensors",
31
+ "layers.11.mixer.out_proj.bias": "model-00002-of-00004.safetensors",
32
+ "layers.11.mixer.out_proj.weight": "model-00002-of-00004.safetensors",
33
+ "layers.11.mlp.fc1.bias": "model-00002-of-00004.safetensors",
34
+ "layers.11.mlp.fc1.weight": "model-00002-of-00004.safetensors",
35
+ "layers.11.mlp.fc2.bias": "model-00002-of-00004.safetensors",
36
+ "layers.11.mlp.fc2.weight": "model-00002-of-00004.safetensors",
37
+ "layers.12.ln.bias": "model-00002-of-00004.safetensors",
38
+ "layers.12.ln.weight": "model-00002-of-00004.safetensors",
39
+ "layers.12.mixer.Wqkv.bias": "model-00002-of-00004.safetensors",
40
+ "layers.12.mixer.Wqkv.weight": "model-00002-of-00004.safetensors",
41
+ "layers.12.mixer.out_proj.bias": "model-00002-of-00004.safetensors",
42
+ "layers.12.mixer.out_proj.weight": "model-00002-of-00004.safetensors",
43
+ "layers.12.mlp.fc1.bias": "model-00002-of-00004.safetensors",
44
+ "layers.12.mlp.fc1.weight": "model-00002-of-00004.safetensors",
45
+ "layers.12.mlp.fc2.bias": "model-00002-of-00004.safetensors",
46
+ "layers.12.mlp.fc2.weight": "model-00002-of-00004.safetensors",
47
+ "layers.13.ln.bias": "model-00002-of-00004.safetensors",
48
+ "layers.13.ln.weight": "model-00002-of-00004.safetensors",
49
+ "layers.13.mixer.Wqkv.bias": "model-00002-of-00004.safetensors",
50
+ "layers.13.mixer.Wqkv.weight": "model-00002-of-00004.safetensors",
51
+ "layers.13.mixer.out_proj.bias": "model-00002-of-00004.safetensors",
52
+ "layers.13.mixer.out_proj.weight": "model-00002-of-00004.safetensors",
53
+ "layers.13.mlp.fc1.bias": "model-00002-of-00004.safetensors",
54
+ "layers.13.mlp.fc1.weight": "model-00002-of-00004.safetensors",
55
+ "layers.13.mlp.fc2.bias": "model-00002-of-00004.safetensors",
56
+ "layers.13.mlp.fc2.weight": "model-00002-of-00004.safetensors",
57
+ "layers.14.ln.bias": "model-00002-of-00004.safetensors",
58
+ "layers.14.ln.weight": "model-00002-of-00004.safetensors",
59
+ "layers.14.mixer.Wqkv.bias": "model-00002-of-00004.safetensors",
60
+ "layers.14.mixer.Wqkv.weight": "model-00002-of-00004.safetensors",
61
+ "layers.14.mixer.out_proj.bias": "model-00002-of-00004.safetensors",
62
+ "layers.14.mixer.out_proj.weight": "model-00002-of-00004.safetensors",
63
+ "layers.14.mlp.fc1.bias": "model-00002-of-00004.safetensors",
64
+ "layers.14.mlp.fc1.weight": "model-00002-of-00004.safetensors",
65
+ "layers.14.mlp.fc2.bias": "model-00002-of-00004.safetensors",
66
+ "layers.14.mlp.fc2.weight": "model-00002-of-00004.safetensors",
67
+ "layers.15.ln.bias": "model-00002-of-00004.safetensors",
68
+ "layers.15.ln.weight": "model-00002-of-00004.safetensors",
69
+ "layers.15.mixer.Wqkv.bias": "model-00002-of-00004.safetensors",
70
+ "layers.15.mixer.Wqkv.weight": "model-00002-of-00004.safetensors",
71
+ "layers.15.mixer.out_proj.bias": "model-00002-of-00004.safetensors",
72
+ "layers.15.mixer.out_proj.weight": "model-00002-of-00004.safetensors",
73
+ "layers.15.mlp.fc1.bias": "model-00002-of-00004.safetensors",
74
+ "layers.15.mlp.fc1.weight": "model-00002-of-00004.safetensors",
75
+ "layers.15.mlp.fc2.bias": "model-00002-of-00004.safetensors",
76
+ "layers.15.mlp.fc2.weight": "model-00002-of-00004.safetensors",
77
+ "layers.16.ln.bias": "model-00002-of-00004.safetensors",
78
+ "layers.16.ln.weight": "model-00002-of-00004.safetensors",
79
+ "layers.16.mixer.Wqkv.bias": "model-00002-of-00004.safetensors",
80
+ "layers.16.mixer.Wqkv.weight": "model-00002-of-00004.safetensors",
81
+ "layers.16.mixer.out_proj.bias": "model-00002-of-00004.safetensors",
82
+ "layers.16.mixer.out_proj.weight": "model-00002-of-00004.safetensors",
83
+ "layers.16.mlp.fc1.bias": "model-00002-of-00004.safetensors",
84
+ "layers.16.mlp.fc1.weight": "model-00002-of-00004.safetensors",
85
+ "layers.16.mlp.fc2.bias": "model-00002-of-00004.safetensors",
86
+ "layers.16.mlp.fc2.weight": "model-00002-of-00004.safetensors",
87
+ "layers.17.ln.bias": "model-00002-of-00004.safetensors",
88
+ "layers.17.ln.weight": "model-00002-of-00004.safetensors",
89
+ "layers.17.mixer.Wqkv.bias": "model-00002-of-00004.safetensors",
90
+ "layers.17.mixer.Wqkv.weight": "model-00002-of-00004.safetensors",
91
+ "layers.17.mixer.out_proj.bias": "model-00002-of-00004.safetensors",
92
+ "layers.17.mixer.out_proj.weight": "model-00002-of-00004.safetensors",
93
+ "layers.17.mlp.fc1.bias": "model-00002-of-00004.safetensors",
94
+ "layers.17.mlp.fc1.weight": "model-00002-of-00004.safetensors",
95
+ "layers.17.mlp.fc2.bias": "model-00002-of-00004.safetensors",
96
+ "layers.17.mlp.fc2.weight": "model-00002-of-00004.safetensors",
97
+ "layers.18.ln.bias": "model-00002-of-00004.safetensors",
98
+ "layers.18.ln.weight": "model-00002-of-00004.safetensors",
99
+ "layers.18.mixer.Wqkv.bias": "model-00003-of-00004.safetensors",
100
+ "layers.18.mixer.Wqkv.weight": "model-00003-of-00004.safetensors",
101
+ "layers.18.mixer.out_proj.bias": "model-00003-of-00004.safetensors",
102
+ "layers.18.mixer.out_proj.weight": "model-00003-of-00004.safetensors",
103
+ "layers.18.mlp.fc1.bias": "model-00003-of-00004.safetensors",
104
+ "layers.18.mlp.fc1.weight": "model-00003-of-00004.safetensors",
105
+ "layers.18.mlp.fc2.bias": "model-00003-of-00004.safetensors",
106
+ "layers.18.mlp.fc2.weight": "model-00003-of-00004.safetensors",
107
+ "layers.19.ln.bias": "model-00003-of-00004.safetensors",
108
+ "layers.19.ln.weight": "model-00003-of-00004.safetensors",
109
+ "layers.19.mixer.Wqkv.bias": "model-00003-of-00004.safetensors",
110
+ "layers.19.mixer.Wqkv.weight": "model-00003-of-00004.safetensors",
111
+ "layers.19.mixer.out_proj.bias": "model-00003-of-00004.safetensors",
112
+ "layers.19.mixer.out_proj.weight": "model-00003-of-00004.safetensors",
113
+ "layers.19.mlp.fc1.bias": "model-00003-of-00004.safetensors",
114
+ "layers.19.mlp.fc1.weight": "model-00003-of-00004.safetensors",
115
+ "layers.19.mlp.fc2.bias": "model-00003-of-00004.safetensors",
116
+ "layers.19.mlp.fc2.weight": "model-00003-of-00004.safetensors",
117
+ "layers.2.ln.bias": "model-00001-of-00004.safetensors",
118
+ "layers.2.ln.weight": "model-00001-of-00004.safetensors",
119
+ "layers.2.mixer.Wqkv.bias": "model-00001-of-00004.safetensors",
120
+ "layers.2.mixer.Wqkv.weight": "model-00001-of-00004.safetensors",
121
+ "layers.2.mixer.out_proj.bias": "model-00001-of-00004.safetensors",
122
+ "layers.2.mixer.out_proj.weight": "model-00001-of-00004.safetensors",
123
+ "layers.2.mlp.fc1.bias": "model-00001-of-00004.safetensors",
124
+ "layers.2.mlp.fc1.weight": "model-00001-of-00004.safetensors",
125
+ "layers.2.mlp.fc2.bias": "model-00001-of-00004.safetensors",
126
+ "layers.2.mlp.fc2.weight": "model-00001-of-00004.safetensors",
127
+ "layers.20.ln.bias": "model-00003-of-00004.safetensors",
128
+ "layers.20.ln.weight": "model-00003-of-00004.safetensors",
129
+ "layers.20.mixer.Wqkv.bias": "model-00003-of-00004.safetensors",
130
+ "layers.20.mixer.Wqkv.weight": "model-00003-of-00004.safetensors",
131
+ "layers.20.mixer.out_proj.bias": "model-00003-of-00004.safetensors",
132
+ "layers.20.mixer.out_proj.weight": "model-00003-of-00004.safetensors",
133
+ "layers.20.mlp.fc1.bias": "model-00003-of-00004.safetensors",
134
+ "layers.20.mlp.fc1.weight": "model-00003-of-00004.safetensors",
135
+ "layers.20.mlp.fc2.bias": "model-00003-of-00004.safetensors",
136
+ "layers.20.mlp.fc2.weight": "model-00003-of-00004.safetensors",
137
+ "layers.21.ln.bias": "model-00003-of-00004.safetensors",
138
+ "layers.21.ln.weight": "model-00003-of-00004.safetensors",
139
+ "layers.21.mixer.Wqkv.bias": "model-00003-of-00004.safetensors",
140
+ "layers.21.mixer.Wqkv.weight": "model-00003-of-00004.safetensors",
141
+ "layers.21.mixer.out_proj.bias": "model-00003-of-00004.safetensors",
142
+ "layers.21.mixer.out_proj.weight": "model-00003-of-00004.safetensors",
143
+ "layers.21.mlp.fc1.bias": "model-00003-of-00004.safetensors",
144
+ "layers.21.mlp.fc1.weight": "model-00003-of-00004.safetensors",
145
+ "layers.21.mlp.fc2.bias": "model-00003-of-00004.safetensors",
146
+ "layers.21.mlp.fc2.weight": "model-00003-of-00004.safetensors",
147
+ "layers.22.ln.bias": "model-00003-of-00004.safetensors",
148
+ "layers.22.ln.weight": "model-00003-of-00004.safetensors",
149
+ "layers.22.mixer.Wqkv.bias": "model-00003-of-00004.safetensors",
150
+ "layers.22.mixer.Wqkv.weight": "model-00003-of-00004.safetensors",
151
+ "layers.22.mixer.out_proj.bias": "model-00003-of-00004.safetensors",
152
+ "layers.22.mixer.out_proj.weight": "model-00003-of-00004.safetensors",
153
+ "layers.22.mlp.fc1.bias": "model-00003-of-00004.safetensors",
154
+ "layers.22.mlp.fc1.weight": "model-00003-of-00004.safetensors",
155
+ "layers.22.mlp.fc2.bias": "model-00003-of-00004.safetensors",
156
+ "layers.22.mlp.fc2.weight": "model-00003-of-00004.safetensors",
157
+ "layers.23.ln.bias": "model-00003-of-00004.safetensors",
158
+ "layers.23.ln.weight": "model-00003-of-00004.safetensors",
159
+ "layers.23.mixer.Wqkv.bias": "model-00003-of-00004.safetensors",
160
+ "layers.23.mixer.Wqkv.weight": "model-00003-of-00004.safetensors",
161
+ "layers.23.mixer.out_proj.bias": "model-00003-of-00004.safetensors",
162
+ "layers.23.mixer.out_proj.weight": "model-00003-of-00004.safetensors",
163
+ "layers.23.mlp.fc1.bias": "model-00003-of-00004.safetensors",
164
+ "layers.23.mlp.fc1.weight": "model-00003-of-00004.safetensors",
165
+ "layers.23.mlp.fc2.bias": "model-00003-of-00004.safetensors",
166
+ "layers.23.mlp.fc2.weight": "model-00003-of-00004.safetensors",
167
+ "layers.24.ln.bias": "model-00003-of-00004.safetensors",
168
+ "layers.24.ln.weight": "model-00003-of-00004.safetensors",
169
+ "layers.24.mixer.Wqkv.bias": "model-00003-of-00004.safetensors",
170
+ "layers.24.mixer.Wqkv.weight": "model-00003-of-00004.safetensors",
171
+ "layers.24.mixer.out_proj.bias": "model-00003-of-00004.safetensors",
172
+ "layers.24.mixer.out_proj.weight": "model-00003-of-00004.safetensors",
173
+ "layers.24.mlp.fc1.bias": "model-00003-of-00004.safetensors",
174
+ "layers.24.mlp.fc1.weight": "model-00003-of-00004.safetensors",
175
+ "layers.24.mlp.fc2.bias": "model-00003-of-00004.safetensors",
176
+ "layers.24.mlp.fc2.weight": "model-00003-of-00004.safetensors",
177
+ "layers.25.ln.bias": "model-00003-of-00004.safetensors",
178
+ "layers.25.ln.weight": "model-00003-of-00004.safetensors",
179
+ "layers.25.mixer.Wqkv.bias": "model-00003-of-00004.safetensors",
180
+ "layers.25.mixer.Wqkv.weight": "model-00003-of-00004.safetensors",
181
+ "layers.25.mixer.out_proj.bias": "model-00003-of-00004.safetensors",
182
+ "layers.25.mixer.out_proj.weight": "model-00003-of-00004.safetensors",
183
+ "layers.25.mlp.fc1.bias": "model-00003-of-00004.safetensors",
184
+ "layers.25.mlp.fc1.weight": "model-00003-of-00004.safetensors",
185
+ "layers.25.mlp.fc2.bias": "model-00003-of-00004.safetensors",
186
+ "layers.25.mlp.fc2.weight": "model-00003-of-00004.safetensors",
187
+ "layers.26.ln.bias": "model-00003-of-00004.safetensors",
188
+ "layers.26.ln.weight": "model-00003-of-00004.safetensors",
189
+ "layers.26.mixer.Wqkv.bias": "model-00003-of-00004.safetensors",
190
+ "layers.26.mixer.Wqkv.weight": "model-00003-of-00004.safetensors",
191
+ "layers.26.mixer.out_proj.bias": "model-00003-of-00004.safetensors",
192
+ "layers.26.mixer.out_proj.weight": "model-00003-of-00004.safetensors",
193
+ "layers.26.mlp.fc1.bias": "model-00003-of-00004.safetensors",
194
+ "layers.26.mlp.fc1.weight": "model-00003-of-00004.safetensors",
195
+ "layers.26.mlp.fc2.bias": "model-00003-of-00004.safetensors",
196
+ "layers.26.mlp.fc2.weight": "model-00003-of-00004.safetensors",
197
+ "layers.27.ln.bias": "model-00003-of-00004.safetensors",
198
+ "layers.27.ln.weight": "model-00003-of-00004.safetensors",
199
+ "layers.27.mixer.Wqkv.bias": "model-00003-of-00004.safetensors",
200
+ "layers.27.mixer.Wqkv.weight": "model-00003-of-00004.safetensors",
201
+ "layers.27.mixer.out_proj.bias": "model-00003-of-00004.safetensors",
202
+ "layers.27.mixer.out_proj.weight": "model-00003-of-00004.safetensors",
203
+ "layers.27.mlp.fc1.bias": "model-00004-of-00004.safetensors",
204
+ "layers.27.mlp.fc1.weight": "model-00004-of-00004.safetensors",
205
+ "layers.27.mlp.fc2.bias": "model-00004-of-00004.safetensors",
206
+ "layers.27.mlp.fc2.weight": "model-00004-of-00004.safetensors",
207
+ "layers.28.ln.bias": "model-00004-of-00004.safetensors",
208
+ "layers.28.ln.weight": "model-00004-of-00004.safetensors",
209
+ "layers.28.mixer.Wqkv.bias": "model-00004-of-00004.safetensors",
210
+ "layers.28.mixer.Wqkv.weight": "model-00004-of-00004.safetensors",
211
+ "layers.28.mixer.out_proj.bias": "model-00004-of-00004.safetensors",
212
+ "layers.28.mixer.out_proj.weight": "model-00004-of-00004.safetensors",
213
+ "layers.28.mlp.fc1.bias": "model-00004-of-00004.safetensors",
214
+ "layers.28.mlp.fc1.weight": "model-00004-of-00004.safetensors",
215
+ "layers.28.mlp.fc2.bias": "model-00004-of-00004.safetensors",
216
+ "layers.28.mlp.fc2.weight": "model-00004-of-00004.safetensors",
217
+ "layers.29.ln.bias": "model-00004-of-00004.safetensors",
218
+ "layers.29.ln.weight": "model-00004-of-00004.safetensors",
219
+ "layers.29.mixer.Wqkv.bias": "model-00004-of-00004.safetensors",
220
+ "layers.29.mixer.Wqkv.weight": "model-00004-of-00004.safetensors",
221
+ "layers.29.mixer.out_proj.bias": "model-00004-of-00004.safetensors",
222
+ "layers.29.mixer.out_proj.weight": "model-00004-of-00004.safetensors",
223
+ "layers.29.mlp.fc1.bias": "model-00004-of-00004.safetensors",
224
+ "layers.29.mlp.fc1.weight": "model-00004-of-00004.safetensors",
225
+ "layers.29.mlp.fc2.bias": "model-00004-of-00004.safetensors",
226
+ "layers.29.mlp.fc2.weight": "model-00004-of-00004.safetensors",
227
+ "layers.3.ln.bias": "model-00001-of-00004.safetensors",
228
+ "layers.3.ln.weight": "model-00001-of-00004.safetensors",
229
+ "layers.3.mixer.Wqkv.bias": "model-00001-of-00004.safetensors",
230
+ "layers.3.mixer.Wqkv.weight": "model-00001-of-00004.safetensors",
231
+ "layers.3.mixer.out_proj.bias": "model-00001-of-00004.safetensors",
232
+ "layers.3.mixer.out_proj.weight": "model-00001-of-00004.safetensors",
233
+ "layers.3.mlp.fc1.bias": "model-00001-of-00004.safetensors",
234
+ "layers.3.mlp.fc1.weight": "model-00001-of-00004.safetensors",
235
+ "layers.3.mlp.fc2.bias": "model-00001-of-00004.safetensors",
236
+ "layers.3.mlp.fc2.weight": "model-00001-of-00004.safetensors",
237
+ "layers.30.ln.bias": "model-00004-of-00004.safetensors",
238
+ "layers.30.ln.weight": "model-00004-of-00004.safetensors",
239
+ "layers.30.mixer.Wqkv.bias": "model-00004-of-00004.safetensors",
240
+ "layers.30.mixer.Wqkv.weight": "model-00004-of-00004.safetensors",
241
+ "layers.30.mixer.out_proj.bias": "model-00004-of-00004.safetensors",
242
+ "layers.30.mixer.out_proj.weight": "model-00004-of-00004.safetensors",
243
+ "layers.30.mlp.fc1.bias": "model-00004-of-00004.safetensors",
244
+ "layers.30.mlp.fc1.weight": "model-00004-of-00004.safetensors",
245
+ "layers.30.mlp.fc2.bias": "model-00004-of-00004.safetensors",
246
+ "layers.30.mlp.fc2.weight": "model-00004-of-00004.safetensors",
247
+ "layers.31.ln.bias": "model-00004-of-00004.safetensors",
248
+ "layers.31.ln.weight": "model-00004-of-00004.safetensors",
249
+ "layers.31.mixer.Wqkv.bias": "model-00004-of-00004.safetensors",
250
+ "layers.31.mixer.Wqkv.weight": "model-00004-of-00004.safetensors",
251
+ "layers.31.mixer.out_proj.bias": "model-00004-of-00004.safetensors",
252
+ "layers.31.mixer.out_proj.weight": "model-00004-of-00004.safetensors",
253
+ "layers.31.mlp.fc1.bias": "model-00004-of-00004.safetensors",
254
+ "layers.31.mlp.fc1.weight": "model-00004-of-00004.safetensors",
255
+ "layers.31.mlp.fc2.bias": "model-00004-of-00004.safetensors",
256
+ "layers.31.mlp.fc2.weight": "model-00004-of-00004.safetensors",
257
+ "layers.32.ln.bias": "model-00004-of-00004.safetensors",
258
+ "layers.32.ln.weight": "model-00004-of-00004.safetensors",
259
+ "layers.32.mixer.Wqkv.bias": "model-00004-of-00004.safetensors",
260
+ "layers.32.mixer.Wqkv.weight": "model-00004-of-00004.safetensors",
261
+ "layers.32.mixer.out_proj.bias": "model-00004-of-00004.safetensors",
262
+ "layers.32.mixer.out_proj.weight": "model-00004-of-00004.safetensors",
263
+ "layers.32.mlp.fc1.bias": "model-00004-of-00004.safetensors",
264
+ "layers.32.mlp.fc1.weight": "model-00004-of-00004.safetensors",
265
+ "layers.32.mlp.fc2.bias": "model-00004-of-00004.safetensors",
266
+ "layers.32.mlp.fc2.weight": "model-00004-of-00004.safetensors",
267
+ "layers.33.linear.bias": "model-00004-of-00004.safetensors",
268
+ "layers.33.linear.weight": "model-00004-of-00004.safetensors",
269
+ "layers.33.ln.bias": "model-00004-of-00004.safetensors",
270
+ "layers.33.ln.weight": "model-00004-of-00004.safetensors",
271
+ "layers.4.ln.bias": "model-00001-of-00004.safetensors",
272
+ "layers.4.ln.weight": "model-00001-of-00004.safetensors",
273
+ "layers.4.mixer.Wqkv.bias": "model-00001-of-00004.safetensors",
274
+ "layers.4.mixer.Wqkv.weight": "model-00001-of-00004.safetensors",
275
+ "layers.4.mixer.out_proj.bias": "model-00001-of-00004.safetensors",
276
+ "layers.4.mixer.out_proj.weight": "model-00001-of-00004.safetensors",
277
+ "layers.4.mlp.fc1.bias": "model-00001-of-00004.safetensors",
278
+ "layers.4.mlp.fc1.weight": "model-00001-of-00004.safetensors",
279
+ "layers.4.mlp.fc2.bias": "model-00001-of-00004.safetensors",
280
+ "layers.4.mlp.fc2.weight": "model-00001-of-00004.safetensors",
281
+ "layers.5.ln.bias": "model-00001-of-00004.safetensors",
282
+ "layers.5.ln.weight": "model-00001-of-00004.safetensors",
283
+ "layers.5.mixer.Wqkv.bias": "model-00001-of-00004.safetensors",
284
+ "layers.5.mixer.Wqkv.weight": "model-00001-of-00004.safetensors",
285
+ "layers.5.mixer.out_proj.bias": "model-00001-of-00004.safetensors",
286
+ "layers.5.mixer.out_proj.weight": "model-00001-of-00004.safetensors",
287
+ "layers.5.mlp.fc1.bias": "model-00001-of-00004.safetensors",
288
+ "layers.5.mlp.fc1.weight": "model-00001-of-00004.safetensors",
289
+ "layers.5.mlp.fc2.bias": "model-00001-of-00004.safetensors",
290
+ "layers.5.mlp.fc2.weight": "model-00001-of-00004.safetensors",
291
+ "layers.6.ln.bias": "model-00001-of-00004.safetensors",
292
+ "layers.6.ln.weight": "model-00001-of-00004.safetensors",
293
+ "layers.6.mixer.Wqkv.bias": "model-00001-of-00004.safetensors",
294
+ "layers.6.mixer.Wqkv.weight": "model-00001-of-00004.safetensors",
295
+ "layers.6.mixer.out_proj.bias": "model-00001-of-00004.safetensors",
296
+ "layers.6.mixer.out_proj.weight": "model-00001-of-00004.safetensors",
297
+ "layers.6.mlp.fc1.bias": "model-00001-of-00004.safetensors",
298
+ "layers.6.mlp.fc1.weight": "model-00001-of-00004.safetensors",
299
+ "layers.6.mlp.fc2.bias": "model-00001-of-00004.safetensors",
300
+ "layers.6.mlp.fc2.weight": "model-00001-of-00004.safetensors",
301
+ "layers.7.ln.bias": "model-00001-of-00004.safetensors",
302
+ "layers.7.ln.weight": "model-00001-of-00004.safetensors",
303
+ "layers.7.mixer.Wqkv.bias": "model-00001-of-00004.safetensors",
304
+ "layers.7.mixer.Wqkv.weight": "model-00001-of-00004.safetensors",
305
+ "layers.7.mixer.out_proj.bias": "model-00001-of-00004.safetensors",
306
+ "layers.7.mixer.out_proj.weight": "model-00001-of-00004.safetensors",
307
+ "layers.7.mlp.fc1.bias": "model-00001-of-00004.safetensors",
308
+ "layers.7.mlp.fc1.weight": "model-00001-of-00004.safetensors",
309
+ "layers.7.mlp.fc2.bias": "model-00001-of-00004.safetensors",
310
+ "layers.7.mlp.fc2.weight": "model-00001-of-00004.safetensors",
311
+ "layers.8.ln.bias": "model-00001-of-00004.safetensors",
312
+ "layers.8.ln.weight": "model-00001-of-00004.safetensors",
313
+ "layers.8.mixer.Wqkv.bias": "model-00001-of-00004.safetensors",
314
+ "layers.8.mixer.Wqkv.weight": "model-00001-of-00004.safetensors",
315
+ "layers.8.mixer.out_proj.bias": "model-00001-of-00004.safetensors",
316
+ "layers.8.mixer.out_proj.weight": "model-00001-of-00004.safetensors",
317
+ "layers.8.mlp.fc1.bias": "model-00001-of-00004.safetensors",
318
+ "layers.8.mlp.fc1.weight": "model-00001-of-00004.safetensors",
319
+ "layers.8.mlp.fc2.bias": "model-00002-of-00004.safetensors",
320
+ "layers.8.mlp.fc2.weight": "model-00002-of-00004.safetensors",
321
+ "layers.9.ln.bias": "model-00002-of-00004.safetensors",
322
+ "layers.9.ln.weight": "model-00002-of-00004.safetensors",
323
+ "layers.9.mixer.Wqkv.bias": "model-00002-of-00004.safetensors",
324
+ "layers.9.mixer.Wqkv.weight": "model-00002-of-00004.safetensors",
325
+ "layers.9.mixer.out_proj.bias": "model-00002-of-00004.safetensors",
326
+ "layers.9.mixer.out_proj.weight": "model-00002-of-00004.safetensors",
327
+ "layers.9.mlp.fc1.bias": "model-00002-of-00004.safetensors",
328
+ "layers.9.mlp.fc1.weight": "model-00002-of-00004.safetensors",
329
+ "layers.9.mlp.fc2.bias": "model-00002-of-00004.safetensors",
330
+ "layers.9.mlp.fc2.weight": "model-00002-of-00004.safetensors"
331
  }
332
  }