Upload folder using huggingface_hub
Browse files
model-00001-of-00002.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7a6d7ef044775c8fb2ab485f33a87de2dde4a3c44968ebfe03e53617691ffd84
|
3 |
+
size 4998337960
|
model-00002-of-00002.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
-
size
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:bd49c9b55dd05b309de0173327cf383bae563b8e9733476bf75257703eb583ab
|
3 |
+
size 543078144
|
model.safetensors.index.json
CHANGED
@@ -72,7 +72,7 @@
|
|
72 |
"model.mamba_layers.0.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
73 |
"model.mamba_layers.0.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
74 |
"model.mamba_layers.0.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
75 |
-
"model.mamba_layers.0.mamba.in_proj.
|
76 |
"model.mamba_layers.0.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
77 |
"model.mamba_layers.0.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
78 |
"model.mamba_layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -81,7 +81,7 @@
|
|
81 |
"model.mamba_layers.1.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
82 |
"model.mamba_layers.1.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
83 |
"model.mamba_layers.1.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
84 |
-
"model.mamba_layers.1.mamba.in_proj.
|
85 |
"model.mamba_layers.1.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
86 |
"model.mamba_layers.1.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
87 |
"model.mamba_layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -90,7 +90,7 @@
|
|
90 |
"model.mamba_layers.10.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
91 |
"model.mamba_layers.10.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
92 |
"model.mamba_layers.10.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
93 |
-
"model.mamba_layers.10.mamba.in_proj.
|
94 |
"model.mamba_layers.10.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
95 |
"model.mamba_layers.10.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
96 |
"model.mamba_layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -99,7 +99,7 @@
|
|
99 |
"model.mamba_layers.11.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
100 |
"model.mamba_layers.11.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
101 |
"model.mamba_layers.11.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
102 |
-
"model.mamba_layers.11.mamba.in_proj.
|
103 |
"model.mamba_layers.11.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
104 |
"model.mamba_layers.11.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
105 |
"model.mamba_layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -108,7 +108,7 @@
|
|
108 |
"model.mamba_layers.12.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
109 |
"model.mamba_layers.12.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
110 |
"model.mamba_layers.12.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
111 |
-
"model.mamba_layers.12.mamba.in_proj.
|
112 |
"model.mamba_layers.12.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
113 |
"model.mamba_layers.12.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
114 |
"model.mamba_layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -117,7 +117,7 @@
|
|
117 |
"model.mamba_layers.13.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
118 |
"model.mamba_layers.13.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
119 |
"model.mamba_layers.13.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
120 |
-
"model.mamba_layers.13.mamba.in_proj.
|
121 |
"model.mamba_layers.13.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
122 |
"model.mamba_layers.13.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
123 |
"model.mamba_layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -126,7 +126,7 @@
|
|
126 |
"model.mamba_layers.14.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
127 |
"model.mamba_layers.14.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
128 |
"model.mamba_layers.14.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
129 |
-
"model.mamba_layers.14.mamba.in_proj.
|
130 |
"model.mamba_layers.14.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
131 |
"model.mamba_layers.14.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
132 |
"model.mamba_layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -135,7 +135,7 @@
|
|
135 |
"model.mamba_layers.15.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
136 |
"model.mamba_layers.15.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
137 |
"model.mamba_layers.15.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
138 |
-
"model.mamba_layers.15.mamba.in_proj.
|
139 |
"model.mamba_layers.15.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
140 |
"model.mamba_layers.15.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
141 |
"model.mamba_layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -144,7 +144,7 @@
|
|
144 |
"model.mamba_layers.16.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
145 |
"model.mamba_layers.16.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
146 |
"model.mamba_layers.16.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
147 |
-
"model.mamba_layers.16.mamba.in_proj.
|
148 |
"model.mamba_layers.16.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
149 |
"model.mamba_layers.16.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
150 |
"model.mamba_layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -153,7 +153,7 @@
|
|
153 |
"model.mamba_layers.17.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
154 |
"model.mamba_layers.17.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
155 |
"model.mamba_layers.17.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
156 |
-
"model.mamba_layers.17.mamba.in_proj.
|
157 |
"model.mamba_layers.17.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
158 |
"model.mamba_layers.17.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
159 |
"model.mamba_layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -162,7 +162,7 @@
|
|
162 |
"model.mamba_layers.18.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
163 |
"model.mamba_layers.18.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
164 |
"model.mamba_layers.18.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
165 |
-
"model.mamba_layers.18.mamba.in_proj.
|
166 |
"model.mamba_layers.18.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
167 |
"model.mamba_layers.18.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
168 |
"model.mamba_layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -171,7 +171,7 @@
|
|
171 |
"model.mamba_layers.19.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
172 |
"model.mamba_layers.19.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
173 |
"model.mamba_layers.19.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
174 |
-
"model.mamba_layers.19.mamba.in_proj.
|
175 |
"model.mamba_layers.19.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
176 |
"model.mamba_layers.19.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
177 |
"model.mamba_layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -180,7 +180,7 @@
|
|
180 |
"model.mamba_layers.2.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
181 |
"model.mamba_layers.2.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
182 |
"model.mamba_layers.2.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
183 |
-
"model.mamba_layers.2.mamba.in_proj.
|
184 |
"model.mamba_layers.2.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
185 |
"model.mamba_layers.2.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
186 |
"model.mamba_layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -189,7 +189,7 @@
|
|
189 |
"model.mamba_layers.20.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
190 |
"model.mamba_layers.20.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
191 |
"model.mamba_layers.20.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
192 |
-
"model.mamba_layers.20.mamba.in_proj.
|
193 |
"model.mamba_layers.20.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
194 |
"model.mamba_layers.20.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
195 |
"model.mamba_layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -198,7 +198,7 @@
|
|
198 |
"model.mamba_layers.21.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
199 |
"model.mamba_layers.21.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
200 |
"model.mamba_layers.21.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
201 |
-
"model.mamba_layers.21.mamba.in_proj.
|
202 |
"model.mamba_layers.21.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
203 |
"model.mamba_layers.21.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
204 |
"model.mamba_layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -207,7 +207,7 @@
|
|
207 |
"model.mamba_layers.22.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
208 |
"model.mamba_layers.22.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
209 |
"model.mamba_layers.22.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
210 |
-
"model.mamba_layers.22.mamba.in_proj.
|
211 |
"model.mamba_layers.22.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
212 |
"model.mamba_layers.22.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
213 |
"model.mamba_layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -216,7 +216,7 @@
|
|
216 |
"model.mamba_layers.23.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
217 |
"model.mamba_layers.23.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
218 |
"model.mamba_layers.23.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
219 |
-
"model.mamba_layers.23.mamba.in_proj.
|
220 |
"model.mamba_layers.23.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
221 |
"model.mamba_layers.23.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
222 |
"model.mamba_layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -225,7 +225,7 @@
|
|
225 |
"model.mamba_layers.24.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
226 |
"model.mamba_layers.24.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
227 |
"model.mamba_layers.24.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
228 |
-
"model.mamba_layers.24.mamba.in_proj.
|
229 |
"model.mamba_layers.24.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
230 |
"model.mamba_layers.24.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
231 |
"model.mamba_layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -234,7 +234,7 @@
|
|
234 |
"model.mamba_layers.25.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
235 |
"model.mamba_layers.25.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
236 |
"model.mamba_layers.25.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
237 |
-
"model.mamba_layers.25.mamba.in_proj.
|
238 |
"model.mamba_layers.25.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
239 |
"model.mamba_layers.25.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
240 |
"model.mamba_layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -243,7 +243,7 @@
|
|
243 |
"model.mamba_layers.26.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
244 |
"model.mamba_layers.26.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
245 |
"model.mamba_layers.26.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
246 |
-
"model.mamba_layers.26.mamba.in_proj.
|
247 |
"model.mamba_layers.26.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
248 |
"model.mamba_layers.26.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
249 |
"model.mamba_layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -252,7 +252,7 @@
|
|
252 |
"model.mamba_layers.27.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
253 |
"model.mamba_layers.27.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
254 |
"model.mamba_layers.27.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
255 |
-
"model.mamba_layers.27.mamba.in_proj.
|
256 |
"model.mamba_layers.27.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
257 |
"model.mamba_layers.27.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
258 |
"model.mamba_layers.28.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -261,7 +261,7 @@
|
|
261 |
"model.mamba_layers.28.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
262 |
"model.mamba_layers.28.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
263 |
"model.mamba_layers.28.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
264 |
-
"model.mamba_layers.28.mamba.in_proj.
|
265 |
"model.mamba_layers.28.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
266 |
"model.mamba_layers.28.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
267 |
"model.mamba_layers.29.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -270,7 +270,7 @@
|
|
270 |
"model.mamba_layers.29.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
271 |
"model.mamba_layers.29.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
272 |
"model.mamba_layers.29.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
273 |
-
"model.mamba_layers.29.mamba.in_proj.
|
274 |
"model.mamba_layers.29.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
275 |
"model.mamba_layers.29.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
276 |
"model.mamba_layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -279,7 +279,7 @@
|
|
279 |
"model.mamba_layers.3.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
280 |
"model.mamba_layers.3.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
281 |
"model.mamba_layers.3.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
282 |
-
"model.mamba_layers.3.mamba.in_proj.
|
283 |
"model.mamba_layers.3.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
284 |
"model.mamba_layers.3.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
285 |
"model.mamba_layers.30.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -288,7 +288,7 @@
|
|
288 |
"model.mamba_layers.30.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
289 |
"model.mamba_layers.30.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
290 |
"model.mamba_layers.30.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
291 |
-
"model.mamba_layers.30.mamba.in_proj.
|
292 |
"model.mamba_layers.30.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
293 |
"model.mamba_layers.30.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
294 |
"model.mamba_layers.31.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -297,7 +297,7 @@
|
|
297 |
"model.mamba_layers.31.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
298 |
"model.mamba_layers.31.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
299 |
"model.mamba_layers.31.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
300 |
-
"model.mamba_layers.31.mamba.in_proj.
|
301 |
"model.mamba_layers.31.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
302 |
"model.mamba_layers.31.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
303 |
"model.mamba_layers.32.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -306,7 +306,7 @@
|
|
306 |
"model.mamba_layers.32.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
307 |
"model.mamba_layers.32.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
308 |
"model.mamba_layers.32.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
309 |
-
"model.mamba_layers.32.mamba.in_proj.
|
310 |
"model.mamba_layers.32.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
311 |
"model.mamba_layers.32.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
312 |
"model.mamba_layers.33.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -315,7 +315,7 @@
|
|
315 |
"model.mamba_layers.33.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
316 |
"model.mamba_layers.33.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
317 |
"model.mamba_layers.33.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
318 |
-
"model.mamba_layers.33.mamba.in_proj.
|
319 |
"model.mamba_layers.33.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
320 |
"model.mamba_layers.33.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
321 |
"model.mamba_layers.34.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -324,7 +324,7 @@
|
|
324 |
"model.mamba_layers.34.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
325 |
"model.mamba_layers.34.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
326 |
"model.mamba_layers.34.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
327 |
-
"model.mamba_layers.34.mamba.in_proj.
|
328 |
"model.mamba_layers.34.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
329 |
"model.mamba_layers.34.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
330 |
"model.mamba_layers.35.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -333,7 +333,7 @@
|
|
333 |
"model.mamba_layers.35.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
334 |
"model.mamba_layers.35.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
335 |
"model.mamba_layers.35.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
336 |
-
"model.mamba_layers.35.mamba.in_proj.
|
337 |
"model.mamba_layers.35.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
338 |
"model.mamba_layers.35.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
339 |
"model.mamba_layers.36.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -342,7 +342,7 @@
|
|
342 |
"model.mamba_layers.36.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
343 |
"model.mamba_layers.36.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
344 |
"model.mamba_layers.36.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
345 |
-
"model.mamba_layers.36.mamba.in_proj.
|
346 |
"model.mamba_layers.36.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
347 |
"model.mamba_layers.36.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
348 |
"model.mamba_layers.37.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -351,7 +351,7 @@
|
|
351 |
"model.mamba_layers.37.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
352 |
"model.mamba_layers.37.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
353 |
"model.mamba_layers.37.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
354 |
-
"model.mamba_layers.37.mamba.in_proj.
|
355 |
"model.mamba_layers.37.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
356 |
"model.mamba_layers.37.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
357 |
"model.mamba_layers.38.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -360,7 +360,7 @@
|
|
360 |
"model.mamba_layers.38.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
361 |
"model.mamba_layers.38.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
362 |
"model.mamba_layers.38.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
363 |
-
"model.mamba_layers.38.mamba.in_proj.
|
364 |
"model.mamba_layers.38.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
365 |
"model.mamba_layers.38.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
366 |
"model.mamba_layers.39.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -369,7 +369,7 @@
|
|
369 |
"model.mamba_layers.39.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
370 |
"model.mamba_layers.39.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
371 |
"model.mamba_layers.39.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
372 |
-
"model.mamba_layers.39.mamba.in_proj.
|
373 |
"model.mamba_layers.39.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
374 |
"model.mamba_layers.39.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
375 |
"model.mamba_layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -378,7 +378,7 @@
|
|
378 |
"model.mamba_layers.4.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
379 |
"model.mamba_layers.4.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
380 |
"model.mamba_layers.4.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
381 |
-
"model.mamba_layers.4.mamba.in_proj.
|
382 |
"model.mamba_layers.4.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
383 |
"model.mamba_layers.4.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
384 |
"model.mamba_layers.40.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -387,7 +387,7 @@
|
|
387 |
"model.mamba_layers.40.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
388 |
"model.mamba_layers.40.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
389 |
"model.mamba_layers.40.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
390 |
-
"model.mamba_layers.40.mamba.in_proj.
|
391 |
"model.mamba_layers.40.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
392 |
"model.mamba_layers.40.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
393 |
"model.mamba_layers.41.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -396,7 +396,7 @@
|
|
396 |
"model.mamba_layers.41.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
397 |
"model.mamba_layers.41.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
398 |
"model.mamba_layers.41.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
399 |
-
"model.mamba_layers.41.mamba.in_proj.
|
400 |
"model.mamba_layers.41.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
401 |
"model.mamba_layers.41.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
402 |
"model.mamba_layers.42.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -405,7 +405,7 @@
|
|
405 |
"model.mamba_layers.42.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
406 |
"model.mamba_layers.42.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
407 |
"model.mamba_layers.42.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
408 |
-
"model.mamba_layers.42.mamba.in_proj.
|
409 |
"model.mamba_layers.42.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
410 |
"model.mamba_layers.42.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
411 |
"model.mamba_layers.43.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -414,7 +414,7 @@
|
|
414 |
"model.mamba_layers.43.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
415 |
"model.mamba_layers.43.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
416 |
"model.mamba_layers.43.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
417 |
-
"model.mamba_layers.43.mamba.in_proj.
|
418 |
"model.mamba_layers.43.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
419 |
"model.mamba_layers.43.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
420 |
"model.mamba_layers.44.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -423,7 +423,7 @@
|
|
423 |
"model.mamba_layers.44.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
424 |
"model.mamba_layers.44.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
425 |
"model.mamba_layers.44.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
426 |
-
"model.mamba_layers.44.mamba.in_proj.
|
427 |
"model.mamba_layers.44.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
428 |
"model.mamba_layers.44.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
429 |
"model.mamba_layers.45.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -432,7 +432,7 @@
|
|
432 |
"model.mamba_layers.45.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
433 |
"model.mamba_layers.45.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
434 |
"model.mamba_layers.45.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
435 |
-
"model.mamba_layers.45.mamba.in_proj.
|
436 |
"model.mamba_layers.45.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
437 |
"model.mamba_layers.45.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
438 |
"model.mamba_layers.46.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -441,7 +441,7 @@
|
|
441 |
"model.mamba_layers.46.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
442 |
"model.mamba_layers.46.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
443 |
"model.mamba_layers.46.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
444 |
-
"model.mamba_layers.46.mamba.in_proj.
|
445 |
"model.mamba_layers.46.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
446 |
"model.mamba_layers.46.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
447 |
"model.mamba_layers.47.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -450,7 +450,7 @@
|
|
450 |
"model.mamba_layers.47.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
451 |
"model.mamba_layers.47.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
452 |
"model.mamba_layers.47.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
453 |
-
"model.mamba_layers.47.mamba.in_proj.
|
454 |
"model.mamba_layers.47.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
455 |
"model.mamba_layers.47.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
456 |
"model.mamba_layers.48.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
@@ -459,7 +459,7 @@
|
|
459 |
"model.mamba_layers.48.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
460 |
"model.mamba_layers.48.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
461 |
"model.mamba_layers.48.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
462 |
-
"model.mamba_layers.48.mamba.in_proj.
|
463 |
"model.mamba_layers.48.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
464 |
"model.mamba_layers.48.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
465 |
"model.mamba_layers.49.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
@@ -468,7 +468,7 @@
|
|
468 |
"model.mamba_layers.49.mamba.conv1d.bias": "model-00002-of-00002.safetensors",
|
469 |
"model.mamba_layers.49.mamba.conv1d.weight": "model-00002-of-00002.safetensors",
|
470 |
"model.mamba_layers.49.mamba.dt_bias": "model-00002-of-00002.safetensors",
|
471 |
-
"model.mamba_layers.49.mamba.in_proj.
|
472 |
"model.mamba_layers.49.mamba.norm.weight": "model-00002-of-00002.safetensors",
|
473 |
"model.mamba_layers.49.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
474 |
"model.mamba_layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -477,7 +477,7 @@
|
|
477 |
"model.mamba_layers.5.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
478 |
"model.mamba_layers.5.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
479 |
"model.mamba_layers.5.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
480 |
-
"model.mamba_layers.5.mamba.in_proj.
|
481 |
"model.mamba_layers.5.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
482 |
"model.mamba_layers.5.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
483 |
"model.mamba_layers.50.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
@@ -486,7 +486,7 @@
|
|
486 |
"model.mamba_layers.50.mamba.conv1d.bias": "model-00002-of-00002.safetensors",
|
487 |
"model.mamba_layers.50.mamba.conv1d.weight": "model-00002-of-00002.safetensors",
|
488 |
"model.mamba_layers.50.mamba.dt_bias": "model-00002-of-00002.safetensors",
|
489 |
-
"model.mamba_layers.50.mamba.in_proj.
|
490 |
"model.mamba_layers.50.mamba.norm.weight": "model-00002-of-00002.safetensors",
|
491 |
"model.mamba_layers.50.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
492 |
"model.mamba_layers.51.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
@@ -495,7 +495,7 @@
|
|
495 |
"model.mamba_layers.51.mamba.conv1d.bias": "model-00002-of-00002.safetensors",
|
496 |
"model.mamba_layers.51.mamba.conv1d.weight": "model-00002-of-00002.safetensors",
|
497 |
"model.mamba_layers.51.mamba.dt_bias": "model-00002-of-00002.safetensors",
|
498 |
-
"model.mamba_layers.51.mamba.in_proj.
|
499 |
"model.mamba_layers.51.mamba.norm.weight": "model-00002-of-00002.safetensors",
|
500 |
"model.mamba_layers.51.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
501 |
"model.mamba_layers.52.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
@@ -504,7 +504,7 @@
|
|
504 |
"model.mamba_layers.52.mamba.conv1d.bias": "model-00002-of-00002.safetensors",
|
505 |
"model.mamba_layers.52.mamba.conv1d.weight": "model-00002-of-00002.safetensors",
|
506 |
"model.mamba_layers.52.mamba.dt_bias": "model-00002-of-00002.safetensors",
|
507 |
-
"model.mamba_layers.52.mamba.in_proj.
|
508 |
"model.mamba_layers.52.mamba.norm.weight": "model-00002-of-00002.safetensors",
|
509 |
"model.mamba_layers.52.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
510 |
"model.mamba_layers.53.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
@@ -513,7 +513,7 @@
|
|
513 |
"model.mamba_layers.53.mamba.conv1d.bias": "model-00002-of-00002.safetensors",
|
514 |
"model.mamba_layers.53.mamba.conv1d.weight": "model-00002-of-00002.safetensors",
|
515 |
"model.mamba_layers.53.mamba.dt_bias": "model-00002-of-00002.safetensors",
|
516 |
-
"model.mamba_layers.53.mamba.in_proj.
|
517 |
"model.mamba_layers.53.mamba.norm.weight": "model-00002-of-00002.safetensors",
|
518 |
"model.mamba_layers.53.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
519 |
"model.mamba_layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -522,7 +522,7 @@
|
|
522 |
"model.mamba_layers.6.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
523 |
"model.mamba_layers.6.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
524 |
"model.mamba_layers.6.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
525 |
-
"model.mamba_layers.6.mamba.in_proj.
|
526 |
"model.mamba_layers.6.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
527 |
"model.mamba_layers.6.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
528 |
"model.mamba_layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -531,7 +531,7 @@
|
|
531 |
"model.mamba_layers.7.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
532 |
"model.mamba_layers.7.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
533 |
"model.mamba_layers.7.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
534 |
-
"model.mamba_layers.7.mamba.in_proj.
|
535 |
"model.mamba_layers.7.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
536 |
"model.mamba_layers.7.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
537 |
"model.mamba_layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -540,7 +540,7 @@
|
|
540 |
"model.mamba_layers.8.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
541 |
"model.mamba_layers.8.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
542 |
"model.mamba_layers.8.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
543 |
-
"model.mamba_layers.8.mamba.in_proj.
|
544 |
"model.mamba_layers.8.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
545 |
"model.mamba_layers.8.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
546 |
"model.mamba_layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
@@ -549,8 +549,8 @@
|
|
549 |
"model.mamba_layers.9.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
550 |
"model.mamba_layers.9.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
551 |
"model.mamba_layers.9.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
552 |
-
"model.mamba_layers.9.mamba.in_proj.
|
553 |
"model.mamba_layers.9.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
554 |
"model.mamba_layers.9.mamba.out_proj.weight": "model-00001-of-00002.safetensors"
|
555 |
}
|
556 |
-
}
|
|
|
72 |
"model.mamba_layers.0.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
73 |
"model.mamba_layers.0.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
74 |
"model.mamba_layers.0.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
75 |
+
"model.mamba_layers.0.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
76 |
"model.mamba_layers.0.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
77 |
"model.mamba_layers.0.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
78 |
"model.mamba_layers.1.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
81 |
"model.mamba_layers.1.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
82 |
"model.mamba_layers.1.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
83 |
"model.mamba_layers.1.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
84 |
+
"model.mamba_layers.1.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
85 |
"model.mamba_layers.1.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
86 |
"model.mamba_layers.1.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
87 |
"model.mamba_layers.10.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
90 |
"model.mamba_layers.10.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
91 |
"model.mamba_layers.10.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
92 |
"model.mamba_layers.10.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
93 |
+
"model.mamba_layers.10.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
94 |
"model.mamba_layers.10.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
95 |
"model.mamba_layers.10.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
96 |
"model.mamba_layers.11.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
99 |
"model.mamba_layers.11.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
100 |
"model.mamba_layers.11.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
101 |
"model.mamba_layers.11.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
102 |
+
"model.mamba_layers.11.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
103 |
"model.mamba_layers.11.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
104 |
"model.mamba_layers.11.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
105 |
"model.mamba_layers.12.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
108 |
"model.mamba_layers.12.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
109 |
"model.mamba_layers.12.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
110 |
"model.mamba_layers.12.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
111 |
+
"model.mamba_layers.12.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
112 |
"model.mamba_layers.12.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
113 |
"model.mamba_layers.12.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
114 |
"model.mamba_layers.13.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
117 |
"model.mamba_layers.13.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
118 |
"model.mamba_layers.13.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
119 |
"model.mamba_layers.13.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
120 |
+
"model.mamba_layers.13.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
121 |
"model.mamba_layers.13.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
122 |
"model.mamba_layers.13.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
123 |
"model.mamba_layers.14.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
126 |
"model.mamba_layers.14.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
127 |
"model.mamba_layers.14.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
128 |
"model.mamba_layers.14.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
129 |
+
"model.mamba_layers.14.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
130 |
"model.mamba_layers.14.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
131 |
"model.mamba_layers.14.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
132 |
"model.mamba_layers.15.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
135 |
"model.mamba_layers.15.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
136 |
"model.mamba_layers.15.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
137 |
"model.mamba_layers.15.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
138 |
+
"model.mamba_layers.15.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
139 |
"model.mamba_layers.15.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
140 |
"model.mamba_layers.15.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
141 |
"model.mamba_layers.16.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
144 |
"model.mamba_layers.16.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
145 |
"model.mamba_layers.16.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
146 |
"model.mamba_layers.16.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
147 |
+
"model.mamba_layers.16.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
148 |
"model.mamba_layers.16.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
149 |
"model.mamba_layers.16.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
150 |
"model.mamba_layers.17.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
153 |
"model.mamba_layers.17.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
154 |
"model.mamba_layers.17.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
155 |
"model.mamba_layers.17.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
156 |
+
"model.mamba_layers.17.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
157 |
"model.mamba_layers.17.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
158 |
"model.mamba_layers.17.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
159 |
"model.mamba_layers.18.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
162 |
"model.mamba_layers.18.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
163 |
"model.mamba_layers.18.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
164 |
"model.mamba_layers.18.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
165 |
+
"model.mamba_layers.18.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
166 |
"model.mamba_layers.18.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
167 |
"model.mamba_layers.18.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
168 |
"model.mamba_layers.19.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
171 |
"model.mamba_layers.19.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
172 |
"model.mamba_layers.19.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
173 |
"model.mamba_layers.19.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
174 |
+
"model.mamba_layers.19.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
175 |
"model.mamba_layers.19.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
176 |
"model.mamba_layers.19.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
177 |
"model.mamba_layers.2.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
180 |
"model.mamba_layers.2.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
181 |
"model.mamba_layers.2.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
182 |
"model.mamba_layers.2.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
183 |
+
"model.mamba_layers.2.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
184 |
"model.mamba_layers.2.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
185 |
"model.mamba_layers.2.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
186 |
"model.mamba_layers.20.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
189 |
"model.mamba_layers.20.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
190 |
"model.mamba_layers.20.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
191 |
"model.mamba_layers.20.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
192 |
+
"model.mamba_layers.20.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
193 |
"model.mamba_layers.20.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
194 |
"model.mamba_layers.20.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
195 |
"model.mamba_layers.21.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
198 |
"model.mamba_layers.21.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
199 |
"model.mamba_layers.21.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
200 |
"model.mamba_layers.21.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
201 |
+
"model.mamba_layers.21.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
202 |
"model.mamba_layers.21.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
203 |
"model.mamba_layers.21.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
204 |
"model.mamba_layers.22.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
207 |
"model.mamba_layers.22.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
208 |
"model.mamba_layers.22.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
209 |
"model.mamba_layers.22.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
210 |
+
"model.mamba_layers.22.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
211 |
"model.mamba_layers.22.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
212 |
"model.mamba_layers.22.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
213 |
"model.mamba_layers.23.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
216 |
"model.mamba_layers.23.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
217 |
"model.mamba_layers.23.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
218 |
"model.mamba_layers.23.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
219 |
+
"model.mamba_layers.23.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
220 |
"model.mamba_layers.23.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
221 |
"model.mamba_layers.23.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
222 |
"model.mamba_layers.24.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
225 |
"model.mamba_layers.24.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
226 |
"model.mamba_layers.24.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
227 |
"model.mamba_layers.24.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
228 |
+
"model.mamba_layers.24.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
229 |
"model.mamba_layers.24.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
230 |
"model.mamba_layers.24.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
231 |
"model.mamba_layers.25.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
234 |
"model.mamba_layers.25.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
235 |
"model.mamba_layers.25.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
236 |
"model.mamba_layers.25.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
237 |
+
"model.mamba_layers.25.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
238 |
"model.mamba_layers.25.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
239 |
"model.mamba_layers.25.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
240 |
"model.mamba_layers.26.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
243 |
"model.mamba_layers.26.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
244 |
"model.mamba_layers.26.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
245 |
"model.mamba_layers.26.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
246 |
+
"model.mamba_layers.26.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
247 |
"model.mamba_layers.26.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
248 |
"model.mamba_layers.26.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
249 |
"model.mamba_layers.27.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
252 |
"model.mamba_layers.27.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
253 |
"model.mamba_layers.27.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
254 |
"model.mamba_layers.27.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
255 |
+
"model.mamba_layers.27.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
256 |
"model.mamba_layers.27.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
257 |
"model.mamba_layers.27.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
258 |
"model.mamba_layers.28.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
261 |
"model.mamba_layers.28.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
262 |
"model.mamba_layers.28.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
263 |
"model.mamba_layers.28.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
264 |
+
"model.mamba_layers.28.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
265 |
"model.mamba_layers.28.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
266 |
"model.mamba_layers.28.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
267 |
"model.mamba_layers.29.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
270 |
"model.mamba_layers.29.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
271 |
"model.mamba_layers.29.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
272 |
"model.mamba_layers.29.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
273 |
+
"model.mamba_layers.29.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
274 |
"model.mamba_layers.29.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
275 |
"model.mamba_layers.29.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
276 |
"model.mamba_layers.3.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
279 |
"model.mamba_layers.3.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
280 |
"model.mamba_layers.3.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
281 |
"model.mamba_layers.3.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
282 |
+
"model.mamba_layers.3.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
283 |
"model.mamba_layers.3.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
284 |
"model.mamba_layers.3.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
285 |
"model.mamba_layers.30.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
288 |
"model.mamba_layers.30.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
289 |
"model.mamba_layers.30.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
290 |
"model.mamba_layers.30.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
291 |
+
"model.mamba_layers.30.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
292 |
"model.mamba_layers.30.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
293 |
"model.mamba_layers.30.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
294 |
"model.mamba_layers.31.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
297 |
"model.mamba_layers.31.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
298 |
"model.mamba_layers.31.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
299 |
"model.mamba_layers.31.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
300 |
+
"model.mamba_layers.31.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
301 |
"model.mamba_layers.31.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
302 |
"model.mamba_layers.31.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
303 |
"model.mamba_layers.32.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
306 |
"model.mamba_layers.32.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
307 |
"model.mamba_layers.32.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
308 |
"model.mamba_layers.32.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
309 |
+
"model.mamba_layers.32.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
310 |
"model.mamba_layers.32.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
311 |
"model.mamba_layers.32.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
312 |
"model.mamba_layers.33.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
315 |
"model.mamba_layers.33.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
316 |
"model.mamba_layers.33.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
317 |
"model.mamba_layers.33.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
318 |
+
"model.mamba_layers.33.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
319 |
"model.mamba_layers.33.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
320 |
"model.mamba_layers.33.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
321 |
"model.mamba_layers.34.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
324 |
"model.mamba_layers.34.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
325 |
"model.mamba_layers.34.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
326 |
"model.mamba_layers.34.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
327 |
+
"model.mamba_layers.34.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
328 |
"model.mamba_layers.34.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
329 |
"model.mamba_layers.34.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
330 |
"model.mamba_layers.35.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
333 |
"model.mamba_layers.35.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
334 |
"model.mamba_layers.35.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
335 |
"model.mamba_layers.35.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
336 |
+
"model.mamba_layers.35.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
337 |
"model.mamba_layers.35.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
338 |
"model.mamba_layers.35.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
339 |
"model.mamba_layers.36.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
342 |
"model.mamba_layers.36.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
343 |
"model.mamba_layers.36.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
344 |
"model.mamba_layers.36.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
345 |
+
"model.mamba_layers.36.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
346 |
"model.mamba_layers.36.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
347 |
"model.mamba_layers.36.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
348 |
"model.mamba_layers.37.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
351 |
"model.mamba_layers.37.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
352 |
"model.mamba_layers.37.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
353 |
"model.mamba_layers.37.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
354 |
+
"model.mamba_layers.37.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
355 |
"model.mamba_layers.37.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
356 |
"model.mamba_layers.37.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
357 |
"model.mamba_layers.38.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
360 |
"model.mamba_layers.38.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
361 |
"model.mamba_layers.38.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
362 |
"model.mamba_layers.38.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
363 |
+
"model.mamba_layers.38.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
364 |
"model.mamba_layers.38.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
365 |
"model.mamba_layers.38.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
366 |
"model.mamba_layers.39.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
369 |
"model.mamba_layers.39.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
370 |
"model.mamba_layers.39.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
371 |
"model.mamba_layers.39.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
372 |
+
"model.mamba_layers.39.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
373 |
"model.mamba_layers.39.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
374 |
"model.mamba_layers.39.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
375 |
"model.mamba_layers.4.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
378 |
"model.mamba_layers.4.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
379 |
"model.mamba_layers.4.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
380 |
"model.mamba_layers.4.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
381 |
+
"model.mamba_layers.4.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
382 |
"model.mamba_layers.4.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
383 |
"model.mamba_layers.4.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
384 |
"model.mamba_layers.40.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
387 |
"model.mamba_layers.40.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
388 |
"model.mamba_layers.40.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
389 |
"model.mamba_layers.40.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
390 |
+
"model.mamba_layers.40.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
391 |
"model.mamba_layers.40.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
392 |
"model.mamba_layers.40.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
393 |
"model.mamba_layers.41.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
396 |
"model.mamba_layers.41.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
397 |
"model.mamba_layers.41.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
398 |
"model.mamba_layers.41.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
399 |
+
"model.mamba_layers.41.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
400 |
"model.mamba_layers.41.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
401 |
"model.mamba_layers.41.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
402 |
"model.mamba_layers.42.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
405 |
"model.mamba_layers.42.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
406 |
"model.mamba_layers.42.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
407 |
"model.mamba_layers.42.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
408 |
+
"model.mamba_layers.42.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
409 |
"model.mamba_layers.42.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
410 |
"model.mamba_layers.42.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
411 |
"model.mamba_layers.43.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
414 |
"model.mamba_layers.43.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
415 |
"model.mamba_layers.43.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
416 |
"model.mamba_layers.43.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
417 |
+
"model.mamba_layers.43.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
418 |
"model.mamba_layers.43.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
419 |
"model.mamba_layers.43.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
420 |
"model.mamba_layers.44.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
423 |
"model.mamba_layers.44.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
424 |
"model.mamba_layers.44.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
425 |
"model.mamba_layers.44.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
426 |
+
"model.mamba_layers.44.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
427 |
"model.mamba_layers.44.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
428 |
"model.mamba_layers.44.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
429 |
"model.mamba_layers.45.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
432 |
"model.mamba_layers.45.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
433 |
"model.mamba_layers.45.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
434 |
"model.mamba_layers.45.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
435 |
+
"model.mamba_layers.45.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
436 |
"model.mamba_layers.45.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
437 |
"model.mamba_layers.45.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
438 |
"model.mamba_layers.46.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
441 |
"model.mamba_layers.46.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
442 |
"model.mamba_layers.46.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
443 |
"model.mamba_layers.46.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
444 |
+
"model.mamba_layers.46.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
445 |
"model.mamba_layers.46.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
446 |
"model.mamba_layers.46.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
447 |
"model.mamba_layers.47.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
450 |
"model.mamba_layers.47.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
451 |
"model.mamba_layers.47.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
452 |
"model.mamba_layers.47.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
453 |
+
"model.mamba_layers.47.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
454 |
"model.mamba_layers.47.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
455 |
"model.mamba_layers.47.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
456 |
"model.mamba_layers.48.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
|
|
459 |
"model.mamba_layers.48.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
460 |
"model.mamba_layers.48.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
461 |
"model.mamba_layers.48.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
462 |
+
"model.mamba_layers.48.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
463 |
"model.mamba_layers.48.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
464 |
"model.mamba_layers.48.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
465 |
"model.mamba_layers.49.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
|
|
468 |
"model.mamba_layers.49.mamba.conv1d.bias": "model-00002-of-00002.safetensors",
|
469 |
"model.mamba_layers.49.mamba.conv1d.weight": "model-00002-of-00002.safetensors",
|
470 |
"model.mamba_layers.49.mamba.dt_bias": "model-00002-of-00002.safetensors",
|
471 |
+
"model.mamba_layers.49.mamba.in_proj.weight": "model-00002-of-00002.safetensors",
|
472 |
"model.mamba_layers.49.mamba.norm.weight": "model-00002-of-00002.safetensors",
|
473 |
"model.mamba_layers.49.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
474 |
"model.mamba_layers.5.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
477 |
"model.mamba_layers.5.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
478 |
"model.mamba_layers.5.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
479 |
"model.mamba_layers.5.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
480 |
+
"model.mamba_layers.5.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
481 |
"model.mamba_layers.5.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
482 |
"model.mamba_layers.5.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
483 |
"model.mamba_layers.50.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
|
|
486 |
"model.mamba_layers.50.mamba.conv1d.bias": "model-00002-of-00002.safetensors",
|
487 |
"model.mamba_layers.50.mamba.conv1d.weight": "model-00002-of-00002.safetensors",
|
488 |
"model.mamba_layers.50.mamba.dt_bias": "model-00002-of-00002.safetensors",
|
489 |
+
"model.mamba_layers.50.mamba.in_proj.weight": "model-00002-of-00002.safetensors",
|
490 |
"model.mamba_layers.50.mamba.norm.weight": "model-00002-of-00002.safetensors",
|
491 |
"model.mamba_layers.50.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
492 |
"model.mamba_layers.51.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
|
|
495 |
"model.mamba_layers.51.mamba.conv1d.bias": "model-00002-of-00002.safetensors",
|
496 |
"model.mamba_layers.51.mamba.conv1d.weight": "model-00002-of-00002.safetensors",
|
497 |
"model.mamba_layers.51.mamba.dt_bias": "model-00002-of-00002.safetensors",
|
498 |
+
"model.mamba_layers.51.mamba.in_proj.weight": "model-00002-of-00002.safetensors",
|
499 |
"model.mamba_layers.51.mamba.norm.weight": "model-00002-of-00002.safetensors",
|
500 |
"model.mamba_layers.51.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
501 |
"model.mamba_layers.52.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
|
|
504 |
"model.mamba_layers.52.mamba.conv1d.bias": "model-00002-of-00002.safetensors",
|
505 |
"model.mamba_layers.52.mamba.conv1d.weight": "model-00002-of-00002.safetensors",
|
506 |
"model.mamba_layers.52.mamba.dt_bias": "model-00002-of-00002.safetensors",
|
507 |
+
"model.mamba_layers.52.mamba.in_proj.weight": "model-00002-of-00002.safetensors",
|
508 |
"model.mamba_layers.52.mamba.norm.weight": "model-00002-of-00002.safetensors",
|
509 |
"model.mamba_layers.52.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
510 |
"model.mamba_layers.53.input_layernorm.weight": "model-00002-of-00002.safetensors",
|
|
|
513 |
"model.mamba_layers.53.mamba.conv1d.bias": "model-00002-of-00002.safetensors",
|
514 |
"model.mamba_layers.53.mamba.conv1d.weight": "model-00002-of-00002.safetensors",
|
515 |
"model.mamba_layers.53.mamba.dt_bias": "model-00002-of-00002.safetensors",
|
516 |
+
"model.mamba_layers.53.mamba.in_proj.weight": "model-00002-of-00002.safetensors",
|
517 |
"model.mamba_layers.53.mamba.norm.weight": "model-00002-of-00002.safetensors",
|
518 |
"model.mamba_layers.53.mamba.out_proj.weight": "model-00002-of-00002.safetensors",
|
519 |
"model.mamba_layers.6.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
522 |
"model.mamba_layers.6.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
523 |
"model.mamba_layers.6.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
524 |
"model.mamba_layers.6.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
525 |
+
"model.mamba_layers.6.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
526 |
"model.mamba_layers.6.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
527 |
"model.mamba_layers.6.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
528 |
"model.mamba_layers.7.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
531 |
"model.mamba_layers.7.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
532 |
"model.mamba_layers.7.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
533 |
"model.mamba_layers.7.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
534 |
+
"model.mamba_layers.7.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
535 |
"model.mamba_layers.7.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
536 |
"model.mamba_layers.7.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
537 |
"model.mamba_layers.8.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
540 |
"model.mamba_layers.8.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
541 |
"model.mamba_layers.8.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
542 |
"model.mamba_layers.8.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
543 |
+
"model.mamba_layers.8.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
544 |
"model.mamba_layers.8.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
545 |
"model.mamba_layers.8.mamba.out_proj.weight": "model-00001-of-00002.safetensors",
|
546 |
"model.mamba_layers.9.input_layernorm.weight": "model-00001-of-00002.safetensors",
|
|
|
549 |
"model.mamba_layers.9.mamba.conv1d.bias": "model-00001-of-00002.safetensors",
|
550 |
"model.mamba_layers.9.mamba.conv1d.weight": "model-00001-of-00002.safetensors",
|
551 |
"model.mamba_layers.9.mamba.dt_bias": "model-00001-of-00002.safetensors",
|
552 |
+
"model.mamba_layers.9.mamba.in_proj.weight": "model-00001-of-00002.safetensors",
|
553 |
"model.mamba_layers.9.mamba.norm.weight": "model-00001-of-00002.safetensors",
|
554 |
"model.mamba_layers.9.mamba.out_proj.weight": "model-00001-of-00002.safetensors"
|
555 |
}
|
556 |
+
}
|