JoshuaFreeman commited on
Commit
8399d3b
·
verified ·
1 Parent(s): daad119

Upload LoRA model and probe head for run gemma-2-9b-it_layer_41

Browse files
value_head_probes/gemma-2-9b-it_layer_41/adapter_config.json CHANGED
@@ -26,202 +26,202 @@
26
  "rank_pattern": {},
27
  "revision": null,
28
  "target_modules": [
29
- "model.layers.8.self_attn.k_proj",
30
- "model.layers.10.self_attn.q_proj",
31
- "model.layers.2.self_attn.v_proj",
32
- "model.layers.15.mlp.down_proj",
33
- "model.layers.8.self_attn.o_proj",
34
- "model.layers.17.self_attn.q_proj",
35
- "model.layers.9.self_attn.q_proj",
36
- "model.layers.4.mlp.up_proj",
37
- "model.layers.10.self_attn.v_proj",
38
- "model.layers.9.self_attn.o_proj",
39
- "model.layers.17.self_attn.k_proj",
40
- "model.layers.0.mlp.gate_proj",
41
- "model.layers.0.mlp.down_proj",
42
- "model.layers.24.mlp.up_proj",
43
- "model.layers.11.self_attn.q_proj",
44
  "model.layers.1.self_attn.o_proj",
45
- "model.layers.16.self_attn.k_proj",
 
 
 
 
 
 
 
 
 
46
  "model.layers.24.self_attn.q_proj",
47
- "model.layers.1.mlp.down_proj",
48
- "model.layers.20.mlp.up_proj",
49
- "model.layers.2.self_attn.o_proj",
50
  "model.layers.18.mlp.down_proj",
51
- "model.layers.2.mlp.gate_proj",
52
- "model.layers.23.mlp.down_proj",
53
- "model.layers.5.mlp.down_proj",
54
- "model.layers.15.mlp.up_proj",
55
- "model.layers.16.self_attn.o_proj",
56
- "model.layers.4.mlp.down_proj",
57
- "model.layers.7.self_attn.q_proj",
58
- "model.layers.27.self_attn.q_proj",
59
- "model.layers.3.self_attn.o_proj",
60
- "model.layers.12.mlp.down_proj",
61
- "model.layers.3.self_attn.q_proj",
62
- "model.layers.6.self_attn.k_proj",
63
- "model.layers.14.self_attn.v_proj",
64
- "model.layers.9.mlp.up_proj",
65
- "model.layers.18.self_attn.k_proj",
66
  "model.layers.20.self_attn.q_proj",
67
- "model.layers.23.self_attn.q_proj",
68
- "model.layers.19.mlp.down_proj",
69
- "model.layers.0.self_attn.v_proj",
70
- "model.layers.4.self_attn.q_proj",
71
- "model.layers.11.self_attn.o_proj",
72
  "model.layers.5.self_attn.o_proj",
73
- "model.layers.1.self_attn.k_proj",
74
- "model.layers.7.self_attn.v_proj",
75
- "model.layers.24.self_attn.o_proj",
76
- "model.layers.6.mlp.gate_proj",
77
- "model.layers.0.self_attn.q_proj",
78
- "model.layers.18.mlp.gate_proj",
79
- "model.layers.24.self_attn.v_proj",
80
- "model.layers.21.self_attn.o_proj",
81
- "model.layers.19.self_attn.v_proj",
82
- "model.layers.11.mlp.up_proj",
83
- "model.layers.19.mlp.up_proj",
84
- "model.layers.5.self_attn.k_proj",
85
- "model.layers.7.self_attn.k_proj",
86
- "model.layers.3.self_attn.k_proj",
87
  "model.layers.23.mlp.up_proj",
88
- "model.layers.23.self_attn.v_proj",
89
  "model.layers.25.self_attn.o_proj",
90
- "model.layers.5.mlp.gate_proj",
91
- "model.layers.13.mlp.gate_proj",
92
- "model.layers.26.mlp.up_proj",
93
- "model.layers.14.self_attn.k_proj",
94
- "model.layers.0.self_attn.o_proj",
95
- "model.layers.12.self_attn.k_proj",
96
- "model.layers.20.mlp.gate_proj",
97
- "model.layers.22.mlp.gate_proj",
98
- "model.layers.22.self_attn.o_proj",
99
  "model.layers.13.mlp.down_proj",
100
- "model.layers.22.mlp.down_proj",
 
101
  "model.layers.21.mlp.gate_proj",
 
 
 
 
 
 
102
  "model.layers.13.self_attn.o_proj",
103
- "model.layers.21.self_attn.k_proj",
104
- "model.layers.8.mlp.down_proj",
105
- "model.layers.16.mlp.gate_proj",
 
 
 
 
 
 
106
  "model.layers.8.mlp.gate_proj",
107
- "model.layers.15.self_attn.o_proj",
108
- "model.layers.17.mlp.down_proj",
109
- "model.layers.15.self_attn.q_proj",
110
- "model.layers.23.mlp.gate_proj",
111
- "model.layers.26.self_attn.q_proj",
112
- "model.layers.2.mlp.up_proj",
113
- "model.layers.8.mlp.up_proj",
114
- "model.layers.10.mlp.down_proj",
115
- "model.layers.1.mlp.gate_proj",
116
- "model.layers.8.self_attn.v_proj",
117
- "model.layers.4.self_attn.v_proj",
118
  "model.layers.10.self_attn.k_proj",
119
- "model.layers.12.mlp.up_proj",
120
- "model.layers.11.self_attn.v_proj",
121
- "model.layers.15.self_attn.v_proj",
122
- "model.layers.18.self_attn.v_proj",
123
- "model.layers.12.self_attn.o_proj",
124
- "model.layers.13.mlp.up_proj",
125
- "model.layers.16.self_attn.v_proj",
126
- "model.layers.19.self_attn.q_proj",
127
- "model.layers.5.mlp.up_proj",
128
- "model.layers.25.self_attn.v_proj",
129
- "model.layers.0.mlp.up_proj",
130
- "model.layers.26.mlp.gate_proj",
131
- "model.layers.26.self_attn.o_proj",
132
- "model.layers.16.self_attn.q_proj",
133
- "model.layers.19.mlp.gate_proj",
134
- "model.layers.9.self_attn.k_proj",
135
- "model.layers.12.self_attn.v_proj",
136
- "model.layers.13.self_attn.k_proj",
137
- "model.layers.11.self_attn.k_proj",
138
- "model.layers.17.self_attn.v_proj",
139
- "model.layers.5.self_attn.q_proj",
140
  "model.layers.27.self_attn.v_proj",
141
- "model.layers.22.self_attn.v_proj",
142
- "model.layers.10.mlp.up_proj",
143
- "model.layers.1.mlp.up_proj",
144
- "model.layers.22.self_attn.q_proj",
145
- "model.layers.9.mlp.gate_proj",
146
- "model.layers.19.self_attn.o_proj",
147
- "model.layers.21.self_attn.v_proj",
148
- "model.layers.18.self_attn.q_proj",
149
- "model.layers.3.self_attn.v_proj",
150
- "model.layers.13.self_attn.v_proj",
151
- "model.layers.1.self_attn.v_proj",
152
- "model.layers.25.self_attn.q_proj",
153
- "model.layers.8.self_attn.q_proj",
154
- "model.layers.2.self_attn.q_proj",
155
- "model.layers.4.self_attn.k_proj",
156
- "model.layers.20.self_attn.v_proj",
157
- "model.layers.15.self_attn.k_proj",
158
- "model.layers.14.mlp.up_proj",
159
- "model.layers.18.self_attn.o_proj",
160
- "model.layers.14.mlp.gate_proj",
161
- "model.layers.22.mlp.up_proj",
162
- "model.layers.2.mlp.down_proj",
163
- "model.layers.7.self_attn.o_proj",
164
- "model.layers.25.mlp.up_proj",
165
- "model.layers.12.mlp.gate_proj",
166
  "model.layers.27.mlp.down_proj",
167
- "model.layers.18.mlp.up_proj",
168
- "model.layers.12.self_attn.q_proj",
169
- "model.layers.17.mlp.gate_proj",
170
- "model.layers.5.self_attn.v_proj",
171
- "model.layers.19.self_attn.k_proj",
172
- "model.layers.24.mlp.gate_proj",
173
- "model.layers.26.self_attn.k_proj",
 
174
  "model.layers.16.mlp.down_proj",
175
- "model.layers.10.mlp.gate_proj",
176
- "model.layers.23.self_attn.k_proj",
177
- "model.layers.0.self_attn.k_proj",
178
- "model.layers.7.mlp.gate_proj",
179
- "model.layers.14.self_attn.q_proj",
 
180
  "model.layers.20.self_attn.k_proj",
 
 
 
 
 
181
  "model.layers.26.mlp.down_proj",
182
- "model.layers.21.mlp.up_proj",
183
- "model.layers.24.self_attn.k_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
184
  "model.layers.21.self_attn.q_proj",
185
- "model.layers.17.mlp.up_proj",
186
- "model.layers.16.mlp.up_proj",
187
- "model.layers.25.mlp.gate_proj",
188
- "model.layers.25.mlp.down_proj",
 
 
 
 
 
 
 
189
  "model.layers.24.mlp.down_proj",
190
- "model.layers.27.self_attn.k_proj",
191
- "model.layers.2.self_attn.k_proj",
192
- "model.layers.9.mlp.down_proj",
193
- "model.layers.20.self_attn.o_proj",
194
- "model.layers.4.self_attn.o_proj",
195
- "model.layers.11.mlp.gate_proj",
196
- "model.layers.22.self_attn.k_proj",
197
- "model.layers.4.mlp.gate_proj",
 
198
  "model.layers.9.self_attn.v_proj",
 
 
 
 
 
 
199
  "model.layers.14.self_attn.o_proj",
200
- "model.layers.21.mlp.down_proj",
201
- "model.layers.10.self_attn.o_proj",
202
- "model.layers.25.self_attn.k_proj",
203
- "model.layers.1.self_attn.q_proj",
204
- "model.layers.7.mlp.up_proj",
 
 
 
205
  "model.layers.26.self_attn.v_proj",
206
- "model.layers.3.mlp.up_proj",
207
- "model.layers.3.mlp.gate_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
208
  "model.layers.27.self_attn.o_proj",
209
  "model.layers.6.self_attn.o_proj",
210
- "model.layers.6.self_attn.q_proj",
211
- "model.layers.7.mlp.down_proj",
212
- "model.layers.27.mlp.up_proj",
213
- "model.layers.15.mlp.gate_proj",
214
- "model.layers.6.mlp.down_proj",
215
- "model.layers.14.mlp.down_proj",
216
- "model.layers.6.mlp.up_proj",
217
  "model.layers.13.self_attn.q_proj",
218
- "model.layers.11.mlp.down_proj",
219
- "model.layers.6.self_attn.v_proj",
220
- "model.layers.23.self_attn.o_proj",
221
  "model.layers.17.self_attn.o_proj",
222
- "model.layers.27.mlp.gate_proj",
223
- "model.layers.3.mlp.down_proj",
224
- "model.layers.20.mlp.down_proj"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
225
  ],
226
  "task_type": null,
227
  "use_dora": false,
 
26
  "rank_pattern": {},
27
  "revision": null,
28
  "target_modules": [
29
+ "model.layers.2.mlp.up_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
30
  "model.layers.1.self_attn.o_proj",
31
+ "model.layers.23.self_attn.k_proj",
32
+ "model.layers.3.self_attn.q_proj",
33
+ "model.layers.11.mlp.gate_proj",
34
+ "model.layers.2.self_attn.k_proj",
35
+ "model.layers.22.self_attn.v_proj",
36
+ "model.layers.12.mlp.down_proj",
37
+ "model.layers.3.mlp.up_proj",
38
+ "model.layers.18.self_attn.v_proj",
39
+ "model.layers.2.mlp.down_proj",
40
+ "model.layers.23.mlp.down_proj",
41
  "model.layers.24.self_attn.q_proj",
 
 
 
42
  "model.layers.18.mlp.down_proj",
43
+ "model.layers.20.mlp.gate_proj",
44
+ "model.layers.2.self_attn.o_proj",
45
+ "model.layers.22.self_attn.k_proj",
46
+ "model.layers.14.self_attn.q_proj",
47
+ "model.layers.18.mlp.up_proj",
48
+ "model.layers.16.self_attn.q_proj",
49
+ "model.layers.11.mlp.up_proj",
 
 
 
 
 
 
 
 
50
  "model.layers.20.self_attn.q_proj",
51
+ "model.layers.5.mlp.gate_proj",
 
 
 
 
52
  "model.layers.5.self_attn.o_proj",
53
+ "model.layers.25.self_attn.q_proj",
54
+ "model.layers.12.self_attn.k_proj",
55
+ "model.layers.19.mlp.gate_proj",
 
 
 
 
 
 
 
 
 
 
 
56
  "model.layers.23.mlp.up_proj",
 
57
  "model.layers.25.self_attn.o_proj",
58
+ "model.layers.0.mlp.down_proj",
59
+ "model.layers.1.mlp.gate_proj",
60
+ "model.layers.4.self_attn.o_proj",
 
 
 
 
 
 
61
  "model.layers.13.mlp.down_proj",
62
+ "model.layers.13.mlp.up_proj",
63
+ "model.layers.19.self_attn.q_proj",
64
  "model.layers.21.mlp.gate_proj",
65
+ "model.layers.10.self_attn.v_proj",
66
+ "model.layers.4.mlp.gate_proj",
67
+ "model.layers.18.self_attn.k_proj",
68
+ "model.layers.18.self_attn.q_proj",
69
+ "model.layers.6.self_attn.v_proj",
70
+ "model.layers.24.self_attn.o_proj",
71
  "model.layers.13.self_attn.o_proj",
72
+ "model.layers.21.self_attn.v_proj",
73
+ "model.layers.7.self_attn.o_proj",
74
+ "model.layers.22.self_attn.q_proj",
75
+ "model.layers.23.self_attn.q_proj",
76
+ "model.layers.21.mlp.down_proj",
77
+ "model.layers.17.mlp.up_proj",
78
+ "model.layers.26.self_attn.k_proj",
79
+ "model.layers.9.mlp.up_proj",
80
+ "model.layers.0.self_attn.v_proj",
81
  "model.layers.8.mlp.gate_proj",
82
+ "model.layers.1.mlp.down_proj",
83
+ "model.layers.3.self_attn.k_proj",
84
+ "model.layers.7.mlp.gate_proj",
85
+ "model.layers.18.self_attn.o_proj",
 
 
 
 
 
 
 
86
  "model.layers.10.self_attn.k_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
87
  "model.layers.27.self_attn.v_proj",
88
+ "model.layers.5.self_attn.q_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
89
  "model.layers.27.mlp.down_proj",
90
+ "model.layers.26.self_attn.o_proj",
91
+ "model.layers.11.mlp.down_proj",
92
+ "model.layers.3.mlp.down_proj",
93
+ "model.layers.15.mlp.gate_proj",
94
+ "model.layers.23.self_attn.o_proj",
95
+ "model.layers.20.self_attn.v_proj",
96
+ "model.layers.19.self_attn.v_proj",
97
+ "model.layers.20.mlp.up_proj",
98
  "model.layers.16.mlp.down_proj",
99
+ "model.layers.9.self_attn.q_proj",
100
+ "model.layers.3.self_attn.v_proj",
101
+ "model.layers.27.self_attn.k_proj",
102
+ "model.layers.19.mlp.up_proj",
103
+ "model.layers.15.self_attn.k_proj",
104
+ "model.layers.14.self_attn.v_proj",
105
  "model.layers.20.self_attn.k_proj",
106
+ "model.layers.15.mlp.down_proj",
107
+ "model.layers.4.self_attn.v_proj",
108
+ "model.layers.9.mlp.down_proj",
109
+ "model.layers.4.mlp.up_proj",
110
+ "model.layers.11.self_attn.o_proj",
111
  "model.layers.26.mlp.down_proj",
112
+ "model.layers.20.mlp.down_proj",
113
+ "model.layers.6.mlp.down_proj",
114
+ "model.layers.25.self_attn.v_proj",
115
+ "model.layers.8.mlp.down_proj",
116
+ "model.layers.25.self_attn.k_proj",
117
+ "model.layers.27.self_attn.q_proj",
118
+ "model.layers.14.mlp.down_proj",
119
+ "model.layers.12.self_attn.v_proj",
120
+ "model.layers.13.self_attn.k_proj",
121
+ "model.layers.22.mlp.down_proj",
122
+ "model.layers.2.self_attn.q_proj",
123
+ "model.layers.23.mlp.gate_proj",
124
+ "model.layers.2.mlp.gate_proj",
125
+ "model.layers.20.self_attn.o_proj",
126
+ "model.layers.15.mlp.up_proj",
127
+ "model.layers.16.mlp.gate_proj",
128
+ "model.layers.21.self_attn.o_proj",
129
+ "model.layers.10.mlp.up_proj",
130
+ "model.layers.26.self_attn.q_proj",
131
  "model.layers.21.self_attn.q_proj",
132
+ "model.layers.6.mlp.up_proj",
133
+ "model.layers.10.self_attn.q_proj",
134
+ "model.layers.1.self_attn.q_proj",
135
+ "model.layers.19.self_attn.k_proj",
136
+ "model.layers.1.self_attn.k_proj",
137
+ "model.layers.8.self_attn.o_proj",
138
+ "model.layers.11.self_attn.k_proj",
139
+ "model.layers.22.self_attn.o_proj",
140
+ "model.layers.0.self_attn.k_proj",
141
+ "model.layers.6.mlp.gate_proj",
142
+ "model.layers.8.self_attn.v_proj",
143
  "model.layers.24.mlp.down_proj",
144
+ "model.layers.7.self_attn.k_proj",
145
+ "model.layers.10.mlp.down_proj",
146
+ "model.layers.3.mlp.gate_proj",
147
+ "model.layers.26.mlp.gate_proj",
148
+ "model.layers.12.self_attn.q_proj",
149
+ "model.layers.12.mlp.up_proj",
150
+ "model.layers.6.self_attn.k_proj",
151
+ "model.layers.11.self_attn.v_proj",
152
+ "model.layers.7.mlp.down_proj",
153
  "model.layers.9.self_attn.v_proj",
154
+ "model.layers.15.self_attn.q_proj",
155
+ "model.layers.5.mlp.down_proj",
156
+ "model.layers.12.mlp.gate_proj",
157
+ "model.layers.5.self_attn.k_proj",
158
+ "model.layers.9.self_attn.k_proj",
159
+ "model.layers.6.self_attn.q_proj",
160
  "model.layers.14.self_attn.o_proj",
161
+ "model.layers.14.mlp.gate_proj",
162
+ "model.layers.22.mlp.up_proj",
163
+ "model.layers.17.self_attn.k_proj",
164
+ "model.layers.21.mlp.up_proj",
165
+ "model.layers.8.self_attn.k_proj",
166
+ "model.layers.26.mlp.up_proj",
167
+ "model.layers.14.mlp.up_proj",
168
+ "model.layers.1.mlp.up_proj",
169
  "model.layers.26.self_attn.v_proj",
170
+ "model.layers.9.self_attn.o_proj",
171
+ "model.layers.0.mlp.gate_proj",
172
+ "model.layers.16.self_attn.v_proj",
173
+ "model.layers.7.self_attn.v_proj",
174
+ "model.layers.7.mlp.up_proj",
175
+ "model.layers.27.mlp.gate_proj",
176
+ "model.layers.16.self_attn.k_proj",
177
+ "model.layers.10.self_attn.o_proj",
178
+ "model.layers.13.self_attn.v_proj",
179
+ "model.layers.25.mlp.down_proj",
180
+ "model.layers.16.mlp.up_proj",
181
+ "model.layers.17.self_attn.v_proj",
182
+ "model.layers.18.mlp.gate_proj",
183
+ "model.layers.0.self_attn.q_proj",
184
+ "model.layers.4.self_attn.k_proj",
185
+ "model.layers.5.self_attn.v_proj",
186
+ "model.layers.23.self_attn.v_proj",
187
+ "model.layers.22.mlp.gate_proj",
188
+ "model.layers.11.self_attn.q_proj",
189
  "model.layers.27.self_attn.o_proj",
190
  "model.layers.6.self_attn.o_proj",
191
+ "model.layers.17.mlp.gate_proj",
192
+ "model.layers.4.self_attn.q_proj",
193
+ "model.layers.0.mlp.up_proj",
194
+ "model.layers.17.mlp.down_proj",
195
+ "model.layers.2.self_attn.v_proj",
 
 
196
  "model.layers.13.self_attn.q_proj",
197
+ "model.layers.19.mlp.down_proj",
198
+ "model.layers.5.mlp.up_proj",
199
+ "model.layers.27.mlp.up_proj",
200
  "model.layers.17.self_attn.o_proj",
201
+ "model.layers.8.self_attn.q_proj",
202
+ "model.layers.3.self_attn.o_proj",
203
+ "model.layers.24.self_attn.k_proj",
204
+ "model.layers.10.mlp.gate_proj",
205
+ "model.layers.8.mlp.up_proj",
206
+ "model.layers.15.self_attn.v_proj",
207
+ "model.layers.19.self_attn.o_proj",
208
+ "model.layers.24.mlp.up_proj",
209
+ "model.layers.25.mlp.gate_proj",
210
+ "model.layers.16.self_attn.o_proj",
211
+ "model.layers.24.mlp.gate_proj",
212
+ "model.layers.13.mlp.gate_proj",
213
+ "model.layers.9.mlp.gate_proj",
214
+ "model.layers.17.self_attn.q_proj",
215
+ "model.layers.1.self_attn.v_proj",
216
+ "model.layers.7.self_attn.q_proj",
217
+ "model.layers.12.self_attn.o_proj",
218
+ "model.layers.21.self_attn.k_proj",
219
+ "model.layers.14.self_attn.k_proj",
220
+ "model.layers.24.self_attn.v_proj",
221
+ "model.layers.25.mlp.up_proj",
222
+ "model.layers.0.self_attn.o_proj",
223
+ "model.layers.4.mlp.down_proj",
224
+ "model.layers.15.self_attn.o_proj"
225
  ],
226
  "task_type": null,
227
  "use_dora": false,
value_head_probes/gemma-2-9b-it_layer_41/adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a6b6b1bbd95467bf593330511386f6fa616590d51bd1bc0117bd11c4bc4bfb74
3
  size 288149032
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a6c8838171ad261ca5115f8ee426bd8e050b4d35f96a321df9dcbf30021892d0
3
  size 288149032
value_head_probes/gemma-2-9b-it_layer_41/probe_head.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b79fc39dad55aa99ca2db9d1c81222b8a6b36a9ab138bb1d8f612bc24f521d6d
3
  size 8682
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35c0d8d0d186c85b70e55fba6733936b6262da5d8b3ef81daeda244093780512
3
  size 8682