model.safetensors.index.json
31.9 KB · 367 lines · json Raw
1 {
2 "metadata": {
3 "total_parameters": 4561852416,
4 "total_size": 9123704832
5 },
6 "weight_map": {
7 "audio_decoder.codebook_embeddings.weight": "model-00002-of-00002.safetensors",
8 "audio_decoder.embeddings.weight": "model-00002-of-00002.safetensors",
9 "audio_decoder.layers.0.attention.wo.weight": "model-00002-of-00002.safetensors",
10 "audio_decoder.layers.0.attention.wqkv.weight": "model-00002-of-00002.safetensors",
11 "audio_decoder.layers.0.attention_norm.weight": "model-00002-of-00002.safetensors",
12 "audio_decoder.layers.0.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
13 "audio_decoder.layers.0.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
14 "audio_decoder.layers.0.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
15 "audio_decoder.layers.0.ffn_norm.weight": "model-00002-of-00002.safetensors",
16 "audio_decoder.layers.1.attention.wo.weight": "model-00002-of-00002.safetensors",
17 "audio_decoder.layers.1.attention.wqkv.weight": "model-00002-of-00002.safetensors",
18 "audio_decoder.layers.1.attention_norm.weight": "model-00002-of-00002.safetensors",
19 "audio_decoder.layers.1.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
20 "audio_decoder.layers.1.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
21 "audio_decoder.layers.1.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
22 "audio_decoder.layers.1.ffn_norm.weight": "model-00002-of-00002.safetensors",
23 "audio_decoder.layers.2.attention.wo.weight": "model-00002-of-00002.safetensors",
24 "audio_decoder.layers.2.attention.wqkv.weight": "model-00002-of-00002.safetensors",
25 "audio_decoder.layers.2.attention_norm.weight": "model-00002-of-00002.safetensors",
26 "audio_decoder.layers.2.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
27 "audio_decoder.layers.2.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
28 "audio_decoder.layers.2.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
29 "audio_decoder.layers.2.ffn_norm.weight": "model-00002-of-00002.safetensors",
30 "audio_decoder.layers.3.attention.wo.weight": "model-00002-of-00002.safetensors",
31 "audio_decoder.layers.3.attention.wqkv.weight": "model-00002-of-00002.safetensors",
32 "audio_decoder.layers.3.attention_norm.weight": "model-00002-of-00002.safetensors",
33 "audio_decoder.layers.3.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
34 "audio_decoder.layers.3.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
35 "audio_decoder.layers.3.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
36 "audio_decoder.layers.3.ffn_norm.weight": "model-00002-of-00002.safetensors",
37 "audio_decoder.norm.weight": "model-00002-of-00002.safetensors",
38 "audio_decoder.output.weight": "model-00002-of-00002.safetensors",
39 "text_model.model.embeddings.weight": "model-00001-of-00002.safetensors",
40 "text_model.model.layers.0.attention.k_norm.weight": "model-00001-of-00002.safetensors",
41 "text_model.model.layers.0.attention.q_norm.weight": "model-00001-of-00002.safetensors",
42 "text_model.model.layers.0.attention.wo.weight": "model-00001-of-00002.safetensors",
43 "text_model.model.layers.0.attention.wqkv.weight": "model-00001-of-00002.safetensors",
44 "text_model.model.layers.0.attention_norm.weight": "model-00001-of-00002.safetensors",
45 "text_model.model.layers.0.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
46 "text_model.model.layers.0.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
47 "text_model.model.layers.0.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
48 "text_model.model.layers.0.ffn_norm.weight": "model-00001-of-00002.safetensors",
49 "text_model.model.layers.1.attention.k_norm.weight": "model-00001-of-00002.safetensors",
50 "text_model.model.layers.1.attention.q_norm.weight": "model-00001-of-00002.safetensors",
51 "text_model.model.layers.1.attention.wo.weight": "model-00001-of-00002.safetensors",
52 "text_model.model.layers.1.attention.wqkv.weight": "model-00001-of-00002.safetensors",
53 "text_model.model.layers.1.attention_norm.weight": "model-00001-of-00002.safetensors",
54 "text_model.model.layers.1.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
55 "text_model.model.layers.1.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
56 "text_model.model.layers.1.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
57 "text_model.model.layers.1.ffn_norm.weight": "model-00001-of-00002.safetensors",
58 "text_model.model.layers.10.attention.k_norm.weight": "model-00001-of-00002.safetensors",
59 "text_model.model.layers.10.attention.q_norm.weight": "model-00001-of-00002.safetensors",
60 "text_model.model.layers.10.attention.wo.weight": "model-00001-of-00002.safetensors",
61 "text_model.model.layers.10.attention.wqkv.weight": "model-00001-of-00002.safetensors",
62 "text_model.model.layers.10.attention_norm.weight": "model-00001-of-00002.safetensors",
63 "text_model.model.layers.10.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
64 "text_model.model.layers.10.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
65 "text_model.model.layers.10.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
66 "text_model.model.layers.10.ffn_norm.weight": "model-00001-of-00002.safetensors",
67 "text_model.model.layers.11.attention.k_norm.weight": "model-00001-of-00002.safetensors",
68 "text_model.model.layers.11.attention.q_norm.weight": "model-00001-of-00002.safetensors",
69 "text_model.model.layers.11.attention.wo.weight": "model-00001-of-00002.safetensors",
70 "text_model.model.layers.11.attention.wqkv.weight": "model-00001-of-00002.safetensors",
71 "text_model.model.layers.11.attention_norm.weight": "model-00001-of-00002.safetensors",
72 "text_model.model.layers.11.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
73 "text_model.model.layers.11.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
74 "text_model.model.layers.11.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
75 "text_model.model.layers.11.ffn_norm.weight": "model-00001-of-00002.safetensors",
76 "text_model.model.layers.12.attention.k_norm.weight": "model-00001-of-00002.safetensors",
77 "text_model.model.layers.12.attention.q_norm.weight": "model-00001-of-00002.safetensors",
78 "text_model.model.layers.12.attention.wo.weight": "model-00001-of-00002.safetensors",
79 "text_model.model.layers.12.attention.wqkv.weight": "model-00001-of-00002.safetensors",
80 "text_model.model.layers.12.attention_norm.weight": "model-00001-of-00002.safetensors",
81 "text_model.model.layers.12.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
82 "text_model.model.layers.12.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
83 "text_model.model.layers.12.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
84 "text_model.model.layers.12.ffn_norm.weight": "model-00001-of-00002.safetensors",
85 "text_model.model.layers.13.attention.k_norm.weight": "model-00001-of-00002.safetensors",
86 "text_model.model.layers.13.attention.q_norm.weight": "model-00001-of-00002.safetensors",
87 "text_model.model.layers.13.attention.wo.weight": "model-00001-of-00002.safetensors",
88 "text_model.model.layers.13.attention.wqkv.weight": "model-00001-of-00002.safetensors",
89 "text_model.model.layers.13.attention_norm.weight": "model-00001-of-00002.safetensors",
90 "text_model.model.layers.13.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
91 "text_model.model.layers.13.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
92 "text_model.model.layers.13.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
93 "text_model.model.layers.13.ffn_norm.weight": "model-00001-of-00002.safetensors",
94 "text_model.model.layers.14.attention.k_norm.weight": "model-00001-of-00002.safetensors",
95 "text_model.model.layers.14.attention.q_norm.weight": "model-00001-of-00002.safetensors",
96 "text_model.model.layers.14.attention.wo.weight": "model-00001-of-00002.safetensors",
97 "text_model.model.layers.14.attention.wqkv.weight": "model-00001-of-00002.safetensors",
98 "text_model.model.layers.14.attention_norm.weight": "model-00001-of-00002.safetensors",
99 "text_model.model.layers.14.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
100 "text_model.model.layers.14.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
101 "text_model.model.layers.14.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
102 "text_model.model.layers.14.ffn_norm.weight": "model-00001-of-00002.safetensors",
103 "text_model.model.layers.15.attention.k_norm.weight": "model-00001-of-00002.safetensors",
104 "text_model.model.layers.15.attention.q_norm.weight": "model-00001-of-00002.safetensors",
105 "text_model.model.layers.15.attention.wo.weight": "model-00001-of-00002.safetensors",
106 "text_model.model.layers.15.attention.wqkv.weight": "model-00001-of-00002.safetensors",
107 "text_model.model.layers.15.attention_norm.weight": "model-00001-of-00002.safetensors",
108 "text_model.model.layers.15.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
109 "text_model.model.layers.15.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
110 "text_model.model.layers.15.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
111 "text_model.model.layers.15.ffn_norm.weight": "model-00001-of-00002.safetensors",
112 "text_model.model.layers.16.attention.k_norm.weight": "model-00001-of-00002.safetensors",
113 "text_model.model.layers.16.attention.q_norm.weight": "model-00001-of-00002.safetensors",
114 "text_model.model.layers.16.attention.wo.weight": "model-00001-of-00002.safetensors",
115 "text_model.model.layers.16.attention.wqkv.weight": "model-00001-of-00002.safetensors",
116 "text_model.model.layers.16.attention_norm.weight": "model-00001-of-00002.safetensors",
117 "text_model.model.layers.16.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
118 "text_model.model.layers.16.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
119 "text_model.model.layers.16.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
120 "text_model.model.layers.16.ffn_norm.weight": "model-00001-of-00002.safetensors",
121 "text_model.model.layers.17.attention.k_norm.weight": "model-00001-of-00002.safetensors",
122 "text_model.model.layers.17.attention.q_norm.weight": "model-00001-of-00002.safetensors",
123 "text_model.model.layers.17.attention.wo.weight": "model-00001-of-00002.safetensors",
124 "text_model.model.layers.17.attention.wqkv.weight": "model-00001-of-00002.safetensors",
125 "text_model.model.layers.17.attention_norm.weight": "model-00001-of-00002.safetensors",
126 "text_model.model.layers.17.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
127 "text_model.model.layers.17.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
128 "text_model.model.layers.17.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
129 "text_model.model.layers.17.ffn_norm.weight": "model-00001-of-00002.safetensors",
130 "text_model.model.layers.18.attention.k_norm.weight": "model-00001-of-00002.safetensors",
131 "text_model.model.layers.18.attention.q_norm.weight": "model-00001-of-00002.safetensors",
132 "text_model.model.layers.18.attention.wo.weight": "model-00001-of-00002.safetensors",
133 "text_model.model.layers.18.attention.wqkv.weight": "model-00001-of-00002.safetensors",
134 "text_model.model.layers.18.attention_norm.weight": "model-00001-of-00002.safetensors",
135 "text_model.model.layers.18.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
136 "text_model.model.layers.18.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
137 "text_model.model.layers.18.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
138 "text_model.model.layers.18.ffn_norm.weight": "model-00001-of-00002.safetensors",
139 "text_model.model.layers.19.attention.k_norm.weight": "model-00001-of-00002.safetensors",
140 "text_model.model.layers.19.attention.q_norm.weight": "model-00001-of-00002.safetensors",
141 "text_model.model.layers.19.attention.wo.weight": "model-00001-of-00002.safetensors",
142 "text_model.model.layers.19.attention.wqkv.weight": "model-00001-of-00002.safetensors",
143 "text_model.model.layers.19.attention_norm.weight": "model-00001-of-00002.safetensors",
144 "text_model.model.layers.19.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
145 "text_model.model.layers.19.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
146 "text_model.model.layers.19.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
147 "text_model.model.layers.19.ffn_norm.weight": "model-00001-of-00002.safetensors",
148 "text_model.model.layers.2.attention.k_norm.weight": "model-00001-of-00002.safetensors",
149 "text_model.model.layers.2.attention.q_norm.weight": "model-00001-of-00002.safetensors",
150 "text_model.model.layers.2.attention.wo.weight": "model-00001-of-00002.safetensors",
151 "text_model.model.layers.2.attention.wqkv.weight": "model-00001-of-00002.safetensors",
152 "text_model.model.layers.2.attention_norm.weight": "model-00001-of-00002.safetensors",
153 "text_model.model.layers.2.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
154 "text_model.model.layers.2.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
155 "text_model.model.layers.2.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
156 "text_model.model.layers.2.ffn_norm.weight": "model-00001-of-00002.safetensors",
157 "text_model.model.layers.20.attention.k_norm.weight": "model-00001-of-00002.safetensors",
158 "text_model.model.layers.20.attention.q_norm.weight": "model-00001-of-00002.safetensors",
159 "text_model.model.layers.20.attention.wo.weight": "model-00001-of-00002.safetensors",
160 "text_model.model.layers.20.attention.wqkv.weight": "model-00001-of-00002.safetensors",
161 "text_model.model.layers.20.attention_norm.weight": "model-00002-of-00002.safetensors",
162 "text_model.model.layers.20.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
163 "text_model.model.layers.20.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
164 "text_model.model.layers.20.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
165 "text_model.model.layers.20.ffn_norm.weight": "model-00002-of-00002.safetensors",
166 "text_model.model.layers.21.attention.k_norm.weight": "model-00002-of-00002.safetensors",
167 "text_model.model.layers.21.attention.q_norm.weight": "model-00002-of-00002.safetensors",
168 "text_model.model.layers.21.attention.wo.weight": "model-00002-of-00002.safetensors",
169 "text_model.model.layers.21.attention.wqkv.weight": "model-00002-of-00002.safetensors",
170 "text_model.model.layers.21.attention_norm.weight": "model-00002-of-00002.safetensors",
171 "text_model.model.layers.21.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
172 "text_model.model.layers.21.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
173 "text_model.model.layers.21.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
174 "text_model.model.layers.21.ffn_norm.weight": "model-00002-of-00002.safetensors",
175 "text_model.model.layers.22.attention.k_norm.weight": "model-00002-of-00002.safetensors",
176 "text_model.model.layers.22.attention.q_norm.weight": "model-00002-of-00002.safetensors",
177 "text_model.model.layers.22.attention.wo.weight": "model-00002-of-00002.safetensors",
178 "text_model.model.layers.22.attention.wqkv.weight": "model-00002-of-00002.safetensors",
179 "text_model.model.layers.22.attention_norm.weight": "model-00002-of-00002.safetensors",
180 "text_model.model.layers.22.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
181 "text_model.model.layers.22.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
182 "text_model.model.layers.22.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
183 "text_model.model.layers.22.ffn_norm.weight": "model-00002-of-00002.safetensors",
184 "text_model.model.layers.23.attention.k_norm.weight": "model-00002-of-00002.safetensors",
185 "text_model.model.layers.23.attention.q_norm.weight": "model-00002-of-00002.safetensors",
186 "text_model.model.layers.23.attention.wo.weight": "model-00002-of-00002.safetensors",
187 "text_model.model.layers.23.attention.wqkv.weight": "model-00002-of-00002.safetensors",
188 "text_model.model.layers.23.attention_norm.weight": "model-00002-of-00002.safetensors",
189 "text_model.model.layers.23.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
190 "text_model.model.layers.23.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
191 "text_model.model.layers.23.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
192 "text_model.model.layers.23.ffn_norm.weight": "model-00002-of-00002.safetensors",
193 "text_model.model.layers.24.attention.k_norm.weight": "model-00002-of-00002.safetensors",
194 "text_model.model.layers.24.attention.q_norm.weight": "model-00002-of-00002.safetensors",
195 "text_model.model.layers.24.attention.wo.weight": "model-00002-of-00002.safetensors",
196 "text_model.model.layers.24.attention.wqkv.weight": "model-00002-of-00002.safetensors",
197 "text_model.model.layers.24.attention_norm.weight": "model-00002-of-00002.safetensors",
198 "text_model.model.layers.24.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
199 "text_model.model.layers.24.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
200 "text_model.model.layers.24.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
201 "text_model.model.layers.24.ffn_norm.weight": "model-00002-of-00002.safetensors",
202 "text_model.model.layers.25.attention.k_norm.weight": "model-00002-of-00002.safetensors",
203 "text_model.model.layers.25.attention.q_norm.weight": "model-00002-of-00002.safetensors",
204 "text_model.model.layers.25.attention.wo.weight": "model-00002-of-00002.safetensors",
205 "text_model.model.layers.25.attention.wqkv.weight": "model-00002-of-00002.safetensors",
206 "text_model.model.layers.25.attention_norm.weight": "model-00002-of-00002.safetensors",
207 "text_model.model.layers.25.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
208 "text_model.model.layers.25.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
209 "text_model.model.layers.25.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
210 "text_model.model.layers.25.ffn_norm.weight": "model-00002-of-00002.safetensors",
211 "text_model.model.layers.26.attention.k_norm.weight": "model-00002-of-00002.safetensors",
212 "text_model.model.layers.26.attention.q_norm.weight": "model-00002-of-00002.safetensors",
213 "text_model.model.layers.26.attention.wo.weight": "model-00002-of-00002.safetensors",
214 "text_model.model.layers.26.attention.wqkv.weight": "model-00002-of-00002.safetensors",
215 "text_model.model.layers.26.attention_norm.weight": "model-00002-of-00002.safetensors",
216 "text_model.model.layers.26.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
217 "text_model.model.layers.26.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
218 "text_model.model.layers.26.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
219 "text_model.model.layers.26.ffn_norm.weight": "model-00002-of-00002.safetensors",
220 "text_model.model.layers.27.attention.k_norm.weight": "model-00002-of-00002.safetensors",
221 "text_model.model.layers.27.attention.q_norm.weight": "model-00002-of-00002.safetensors",
222 "text_model.model.layers.27.attention.wo.weight": "model-00002-of-00002.safetensors",
223 "text_model.model.layers.27.attention.wqkv.weight": "model-00002-of-00002.safetensors",
224 "text_model.model.layers.27.attention_norm.weight": "model-00002-of-00002.safetensors",
225 "text_model.model.layers.27.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
226 "text_model.model.layers.27.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
227 "text_model.model.layers.27.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
228 "text_model.model.layers.27.ffn_norm.weight": "model-00002-of-00002.safetensors",
229 "text_model.model.layers.28.attention.k_norm.weight": "model-00002-of-00002.safetensors",
230 "text_model.model.layers.28.attention.q_norm.weight": "model-00002-of-00002.safetensors",
231 "text_model.model.layers.28.attention.wo.weight": "model-00002-of-00002.safetensors",
232 "text_model.model.layers.28.attention.wqkv.weight": "model-00002-of-00002.safetensors",
233 "text_model.model.layers.28.attention_norm.weight": "model-00002-of-00002.safetensors",
234 "text_model.model.layers.28.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
235 "text_model.model.layers.28.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
236 "text_model.model.layers.28.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
237 "text_model.model.layers.28.ffn_norm.weight": "model-00002-of-00002.safetensors",
238 "text_model.model.layers.29.attention.k_norm.weight": "model-00002-of-00002.safetensors",
239 "text_model.model.layers.29.attention.q_norm.weight": "model-00002-of-00002.safetensors",
240 "text_model.model.layers.29.attention.wo.weight": "model-00002-of-00002.safetensors",
241 "text_model.model.layers.29.attention.wqkv.weight": "model-00002-of-00002.safetensors",
242 "text_model.model.layers.29.attention_norm.weight": "model-00002-of-00002.safetensors",
243 "text_model.model.layers.29.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
244 "text_model.model.layers.29.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
245 "text_model.model.layers.29.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
246 "text_model.model.layers.29.ffn_norm.weight": "model-00002-of-00002.safetensors",
247 "text_model.model.layers.3.attention.k_norm.weight": "model-00001-of-00002.safetensors",
248 "text_model.model.layers.3.attention.q_norm.weight": "model-00001-of-00002.safetensors",
249 "text_model.model.layers.3.attention.wo.weight": "model-00001-of-00002.safetensors",
250 "text_model.model.layers.3.attention.wqkv.weight": "model-00001-of-00002.safetensors",
251 "text_model.model.layers.3.attention_norm.weight": "model-00001-of-00002.safetensors",
252 "text_model.model.layers.3.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
253 "text_model.model.layers.3.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
254 "text_model.model.layers.3.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
255 "text_model.model.layers.3.ffn_norm.weight": "model-00001-of-00002.safetensors",
256 "text_model.model.layers.30.attention.k_norm.weight": "model-00002-of-00002.safetensors",
257 "text_model.model.layers.30.attention.q_norm.weight": "model-00002-of-00002.safetensors",
258 "text_model.model.layers.30.attention.wo.weight": "model-00002-of-00002.safetensors",
259 "text_model.model.layers.30.attention.wqkv.weight": "model-00002-of-00002.safetensors",
260 "text_model.model.layers.30.attention_norm.weight": "model-00002-of-00002.safetensors",
261 "text_model.model.layers.30.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
262 "text_model.model.layers.30.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
263 "text_model.model.layers.30.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
264 "text_model.model.layers.30.ffn_norm.weight": "model-00002-of-00002.safetensors",
265 "text_model.model.layers.31.attention.k_norm.weight": "model-00002-of-00002.safetensors",
266 "text_model.model.layers.31.attention.q_norm.weight": "model-00002-of-00002.safetensors",
267 "text_model.model.layers.31.attention.wo.weight": "model-00002-of-00002.safetensors",
268 "text_model.model.layers.31.attention.wqkv.weight": "model-00002-of-00002.safetensors",
269 "text_model.model.layers.31.attention_norm.weight": "model-00002-of-00002.safetensors",
270 "text_model.model.layers.31.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
271 "text_model.model.layers.31.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
272 "text_model.model.layers.31.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
273 "text_model.model.layers.31.ffn_norm.weight": "model-00002-of-00002.safetensors",
274 "text_model.model.layers.32.attention.k_norm.weight": "model-00002-of-00002.safetensors",
275 "text_model.model.layers.32.attention.q_norm.weight": "model-00002-of-00002.safetensors",
276 "text_model.model.layers.32.attention.wo.weight": "model-00002-of-00002.safetensors",
277 "text_model.model.layers.32.attention.wqkv.weight": "model-00002-of-00002.safetensors",
278 "text_model.model.layers.32.attention_norm.weight": "model-00002-of-00002.safetensors",
279 "text_model.model.layers.32.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
280 "text_model.model.layers.32.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
281 "text_model.model.layers.32.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
282 "text_model.model.layers.32.ffn_norm.weight": "model-00002-of-00002.safetensors",
283 "text_model.model.layers.33.attention.k_norm.weight": "model-00002-of-00002.safetensors",
284 "text_model.model.layers.33.attention.q_norm.weight": "model-00002-of-00002.safetensors",
285 "text_model.model.layers.33.attention.wo.weight": "model-00002-of-00002.safetensors",
286 "text_model.model.layers.33.attention.wqkv.weight": "model-00002-of-00002.safetensors",
287 "text_model.model.layers.33.attention_norm.weight": "model-00002-of-00002.safetensors",
288 "text_model.model.layers.33.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
289 "text_model.model.layers.33.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
290 "text_model.model.layers.33.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
291 "text_model.model.layers.33.ffn_norm.weight": "model-00002-of-00002.safetensors",
292 "text_model.model.layers.34.attention.k_norm.weight": "model-00002-of-00002.safetensors",
293 "text_model.model.layers.34.attention.q_norm.weight": "model-00002-of-00002.safetensors",
294 "text_model.model.layers.34.attention.wo.weight": "model-00002-of-00002.safetensors",
295 "text_model.model.layers.34.attention.wqkv.weight": "model-00002-of-00002.safetensors",
296 "text_model.model.layers.34.attention_norm.weight": "model-00002-of-00002.safetensors",
297 "text_model.model.layers.34.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
298 "text_model.model.layers.34.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
299 "text_model.model.layers.34.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
300 "text_model.model.layers.34.ffn_norm.weight": "model-00002-of-00002.safetensors",
301 "text_model.model.layers.35.attention.k_norm.weight": "model-00002-of-00002.safetensors",
302 "text_model.model.layers.35.attention.q_norm.weight": "model-00002-of-00002.safetensors",
303 "text_model.model.layers.35.attention.wo.weight": "model-00002-of-00002.safetensors",
304 "text_model.model.layers.35.attention.wqkv.weight": "model-00002-of-00002.safetensors",
305 "text_model.model.layers.35.attention_norm.weight": "model-00002-of-00002.safetensors",
306 "text_model.model.layers.35.feed_forward.w1.weight": "model-00002-of-00002.safetensors",
307 "text_model.model.layers.35.feed_forward.w2.weight": "model-00002-of-00002.safetensors",
308 "text_model.model.layers.35.feed_forward.w3.weight": "model-00002-of-00002.safetensors",
309 "text_model.model.layers.35.ffn_norm.weight": "model-00002-of-00002.safetensors",
310 "text_model.model.layers.4.attention.k_norm.weight": "model-00001-of-00002.safetensors",
311 "text_model.model.layers.4.attention.q_norm.weight": "model-00001-of-00002.safetensors",
312 "text_model.model.layers.4.attention.wo.weight": "model-00001-of-00002.safetensors",
313 "text_model.model.layers.4.attention.wqkv.weight": "model-00001-of-00002.safetensors",
314 "text_model.model.layers.4.attention_norm.weight": "model-00001-of-00002.safetensors",
315 "text_model.model.layers.4.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
316 "text_model.model.layers.4.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
317 "text_model.model.layers.4.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
318 "text_model.model.layers.4.ffn_norm.weight": "model-00001-of-00002.safetensors",
319 "text_model.model.layers.5.attention.k_norm.weight": "model-00001-of-00002.safetensors",
320 "text_model.model.layers.5.attention.q_norm.weight": "model-00001-of-00002.safetensors",
321 "text_model.model.layers.5.attention.wo.weight": "model-00001-of-00002.safetensors",
322 "text_model.model.layers.5.attention.wqkv.weight": "model-00001-of-00002.safetensors",
323 "text_model.model.layers.5.attention_norm.weight": "model-00001-of-00002.safetensors",
324 "text_model.model.layers.5.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
325 "text_model.model.layers.5.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
326 "text_model.model.layers.5.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
327 "text_model.model.layers.5.ffn_norm.weight": "model-00001-of-00002.safetensors",
328 "text_model.model.layers.6.attention.k_norm.weight": "model-00001-of-00002.safetensors",
329 "text_model.model.layers.6.attention.q_norm.weight": "model-00001-of-00002.safetensors",
330 "text_model.model.layers.6.attention.wo.weight": "model-00001-of-00002.safetensors",
331 "text_model.model.layers.6.attention.wqkv.weight": "model-00001-of-00002.safetensors",
332 "text_model.model.layers.6.attention_norm.weight": "model-00001-of-00002.safetensors",
333 "text_model.model.layers.6.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
334 "text_model.model.layers.6.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
335 "text_model.model.layers.6.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
336 "text_model.model.layers.6.ffn_norm.weight": "model-00001-of-00002.safetensors",
337 "text_model.model.layers.7.attention.k_norm.weight": "model-00001-of-00002.safetensors",
338 "text_model.model.layers.7.attention.q_norm.weight": "model-00001-of-00002.safetensors",
339 "text_model.model.layers.7.attention.wo.weight": "model-00001-of-00002.safetensors",
340 "text_model.model.layers.7.attention.wqkv.weight": "model-00001-of-00002.safetensors",
341 "text_model.model.layers.7.attention_norm.weight": "model-00001-of-00002.safetensors",
342 "text_model.model.layers.7.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
343 "text_model.model.layers.7.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
344 "text_model.model.layers.7.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
345 "text_model.model.layers.7.ffn_norm.weight": "model-00001-of-00002.safetensors",
346 "text_model.model.layers.8.attention.k_norm.weight": "model-00001-of-00002.safetensors",
347 "text_model.model.layers.8.attention.q_norm.weight": "model-00001-of-00002.safetensors",
348 "text_model.model.layers.8.attention.wo.weight": "model-00001-of-00002.safetensors",
349 "text_model.model.layers.8.attention.wqkv.weight": "model-00001-of-00002.safetensors",
350 "text_model.model.layers.8.attention_norm.weight": "model-00001-of-00002.safetensors",
351 "text_model.model.layers.8.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
352 "text_model.model.layers.8.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
353 "text_model.model.layers.8.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
354 "text_model.model.layers.8.ffn_norm.weight": "model-00001-of-00002.safetensors",
355 "text_model.model.layers.9.attention.k_norm.weight": "model-00001-of-00002.safetensors",
356 "text_model.model.layers.9.attention.q_norm.weight": "model-00001-of-00002.safetensors",
357 "text_model.model.layers.9.attention.wo.weight": "model-00001-of-00002.safetensors",
358 "text_model.model.layers.9.attention.wqkv.weight": "model-00001-of-00002.safetensors",
359 "text_model.model.layers.9.attention_norm.weight": "model-00001-of-00002.safetensors",
360 "text_model.model.layers.9.feed_forward.w1.weight": "model-00001-of-00002.safetensors",
361 "text_model.model.layers.9.feed_forward.w2.weight": "model-00001-of-00002.safetensors",
362 "text_model.model.layers.9.feed_forward.w3.weight": "model-00001-of-00002.safetensors",
363 "text_model.model.layers.9.ffn_norm.weight": "model-00001-of-00002.safetensors",
364 "text_model.model.norm.weight": "model-00002-of-00002.safetensors"
365 }
366 }
367