codeShare commited on
Commit
8a2dfff
·
verified ·
1 Parent(s): 2b71a26

Gemma 4 LoRA review snapshot at 95% training

Browse files
gemma4_lora_to_review/95_percent/adapter_config.json CHANGED
@@ -32,220 +32,220 @@
32
  "rank_pattern": {},
33
  "revision": null,
34
  "target_modules": [
35
- "language_model.layers.14.self_attn.v_proj",
36
- "language_model.layers.9.self_attn.v_proj",
37
- "language_model.layers.13.self_attn.o_proj",
38
- "32.mlp.gate_proj",
39
- "32.mlp.up_proj",
40
- "26.self_attn.q_proj",
41
- "19.mlp.gate_proj",
42
- "language_model.layers.4.self_attn.o_proj",
43
- "18.mlp.gate_proj",
44
- "28.mlp.down_proj",
45
- "30.self_attn.o_proj",
46
- "18.self_attn.o_proj",
47
- "16.mlp.up_proj",
48
- "33.self_attn.q_proj",
49
- "17.mlp.up_proj",
50
- "language_model.layers.11.self_attn.k_proj",
51
- "language_model.layers.9.mlp.down_proj",
52
- "language_model.layers.4.self_attn.k_proj",
53
- "23.self_attn.q_proj",
54
- "28.self_attn.q_proj",
55
  "22.mlp.up_proj",
56
- "31.mlp.up_proj",
57
- "language_model.layers.7.mlp.down_proj",
58
  "language_model.layers.8.self_attn.o_proj",
59
- "language_model.layers.1.self_attn.v_proj",
60
- "language_model.layers.10.self_attn.k_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
61
  "language_model.layers.10.self_attn.q_proj",
62
- "31.self_attn.o_proj",
63
- "language_model.layers.5.self_attn.q_proj",
64
- "language_model.layers.12.self_attn.k_proj",
65
- "language_model.layers.0.self_attn.v_proj",
66
- "20.mlp.up_proj",
67
- "language_model.layers.6.self_attn.o_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
68
  "language_model.layers.0.mlp.gate_proj",
69
- "20.self_attn.o_proj",
70
- "language_model.layers.14.self_attn.o_proj",
71
  "language_model.layers.8.mlp.up_proj",
72
- "language_model.layers.9.self_attn.o_proj",
73
  "language_model.layers.8.self_attn.v_proj",
74
- "output_proj",
75
- "21.mlp.up_proj",
76
- "language_model.layers.0.mlp.down_proj",
77
- "34.self_attn.o_proj",
78
- "language_model.layers.13.mlp.gate_proj",
79
- "language_model.layers.10.self_attn.o_proj",
80
- "28.mlp.up_proj",
81
- "30.mlp.up_proj",
82
- "31.mlp.gate_proj",
83
- "23.mlp.down_proj",
84
- "26.mlp.gate_proj",
85
- "21.mlp.down_proj",
86
- "23.mlp.gate_proj",
87
- "language_model.layers.0.mlp.up_proj",
88
- "language_model.layers.4.mlp.gate_proj",
89
- "language_model.layers.15.mlp.down_proj",
90
  "language_model.layers.3.self_attn.o_proj",
91
- "24.self_attn.q_proj",
92
- "25.mlp.gate_proj",
93
- "language_model.layers.3.mlp.gate_proj",
94
- "language_model.layers.3.mlp.down_proj",
95
  "19.mlp.up_proj",
96
  "language_model.layers.14.mlp.up_proj",
97
- "34.self_attn.q_proj",
98
- "language_model.layers.11.mlp.down_proj",
99
- "language_model.layers.2.self_attn.q_proj",
100
- "34.mlp.up_proj",
101
- "language_model.layers.5.mlp.down_proj",
102
- "per_layer_model_projection",
103
- "language_model.layers.10.mlp.gate_proj",
104
- "33.self_attn.o_proj",
105
- "16.self_attn.q_proj",
106
- "34.mlp.gate_proj",
107
- "25.mlp.down_proj",
 
 
 
 
108
  "32.self_attn.q_proj",
109
- "24.self_attn.o_proj",
110
- "33.mlp.up_proj",
111
- "language_model.layers.10.mlp.up_proj",
112
- "language_model.layers.0.self_attn.o_proj",
113
- "34.mlp.down_proj",
114
  "16.mlp.down_proj",
115
- "29.mlp.up_proj",
116
- "language_model.layers.15.self_attn.o_proj",
117
- "30.self_attn.q_proj",
118
- "17.self_attn.o_proj",
119
- "language_model.layers.3.self_attn.k_proj",
120
- "language_model.layers.1.self_attn.k_proj",
121
- "per_layer_projection",
122
- "language_model.layers.14.mlp.gate_proj",
123
- "language_model.layers.15.mlp.up_proj",
124
- "language_model.layers.7.self_attn.o_proj",
125
- "16.mlp.gate_proj",
126
  "language_model.layers.2.self_attn.o_proj",
127
- "17.self_attn.q_proj",
128
- "24.mlp.gate_proj",
129
- "language_model.layers.7.self_attn.k_proj",
130
- "28.self_attn.o_proj",
131
- "27.mlp.up_proj",
132
- "language_model.layers.6.self_attn.v_proj",
133
- "language_model.layers.12.mlp.up_proj",
134
- "29.self_attn.o_proj",
135
- "language_model.layers.9.mlp.gate_proj",
136
- "language_model.layers.11.mlp.gate_proj",
137
- "language_model.layers.13.self_attn.v_proj",
138
- "language_model.layers.10.self_attn.v_proj",
139
- "language_model.layers.15.mlp.gate_proj",
140
- "language_model.layers.1.mlp.down_proj",
141
- "language_model.layers.7.self_attn.q_proj",
142
- "21.mlp.gate_proj",
143
- "language_model.layers.7.mlp.gate_proj",
144
- "29.mlp.down_proj",
145
- "language_model.layers.3.mlp.up_proj",
146
- "23.mlp.up_proj",
147
- "language_model.layers.13.mlp.up_proj",
148
  "30.mlp.down_proj",
149
- "input_proj",
150
- "24.mlp.down_proj",
151
- "language_model.layers.8.mlp.gate_proj",
152
- "27.mlp.gate_proj",
153
- "language_model.layers.1.mlp.gate_proj",
154
- "26.mlp.down_proj",
155
- "language_model.layers.11.self_attn.q_proj",
156
- "per_layer_input_gate",
157
- "language_model.layers.15.self_attn.q_proj",
158
- "20.self_attn.q_proj",
159
- "language_model.layers.5.self_attn.o_proj",
160
  "language_model.layers.4.mlp.up_proj",
161
- "31.self_attn.q_proj",
162
- "language_model.layers.2.mlp.gate_proj",
163
- "20.mlp.gate_proj",
164
- "21.self_attn.q_proj",
165
- "language_model.layers.11.self_attn.v_proj",
166
- "language_model.layers.6.mlp.gate_proj",
167
- "language_model.layers.3.self_attn.q_proj",
168
- "22.self_attn.q_proj",
169
- "28.mlp.gate_proj",
170
  "29.mlp.gate_proj",
 
 
 
 
 
171
  "language_model.layers.6.self_attn.k_proj",
172
- "17.mlp.gate_proj",
173
- "language_model.layers.12.mlp.down_proj",
174
- "language_model.layers.8.mlp.down_proj",
175
- "language_model.layers.7.self_attn.v_proj",
176
- "language_model.layers.9.self_attn.q_proj",
177
- "input_proj_linear",
178
- "language_model.layers.12.self_attn.v_proj",
179
- "language_model.layers.8.self_attn.k_proj",
180
- "language_model.layers.1.mlp.up_proj",
181
- "19.self_attn.q_proj",
182
  "25.mlp.up_proj",
183
- "language_model.layers.2.self_attn.v_proj",
184
- "language_model.layers.1.self_attn.q_proj",
 
 
185
  "language_model.layers.4.mlp.down_proj",
 
186
  "language_model.layers.5.mlp.up_proj",
187
- "language_model.layers.8.self_attn.q_proj",
188
- "language_model.layers.11.mlp.up_proj",
189
- "language_model.layers.13.self_attn.k_proj",
190
- "33.mlp.down_proj",
191
- "18.mlp.down_proj",
192
- "language_model.layers.2.mlp.up_proj",
193
- "language_model.layers.5.mlp.gate_proj",
194
- "21.self_attn.o_proj",
195
- "language_model.layers.14.mlp.down_proj",
196
- "22.self_attn.o_proj",
197
- "32.mlp.down_proj",
198
- "language_model.layers.10.mlp.down_proj",
199
- "embedding_projection",
200
- "25.self_attn.o_proj",
201
- "24.mlp.up_proj",
202
- "19.mlp.down_proj",
203
- "20.mlp.down_proj",
204
- "language_model.layers.6.mlp.down_proj",
205
- "language_model.layers.4.self_attn.v_proj",
206
  "language_model.layers.11.self_attn.o_proj",
207
- "language_model.layers.12.self_attn.o_proj",
208
- "27.mlp.down_proj",
209
- "language_model.layers.0.self_attn.q_proj",
210
- "27.self_attn.o_proj",
211
- "32.self_attn.o_proj",
212
- "language_model.layers.5.self_attn.v_proj",
213
- "33.mlp.gate_proj",
214
- "language_model.layers.2.mlp.down_proj",
215
- "26.self_attn.o_proj",
216
- "19.self_attn.o_proj",
217
- "language_model.layers.7.mlp.up_proj",
218
- "language_model.layers.0.self_attn.k_proj",
219
- "language_model.layers.9.self_attn.k_proj",
220
- "27.self_attn.q_proj",
221
- "language_model.layers.4.self_attn.q_proj",
222
- "language_model.layers.13.mlp.down_proj",
223
- "18.mlp.up_proj",
224
- "relative_k_proj",
225
- "31.mlp.down_proj",
226
- "linear",
227
- "language_model.layers.9.mlp.up_proj",
228
- "23.self_attn.o_proj",
229
- "language_model.layers.5.self_attn.k_proj",
230
- "17.mlp.down_proj",
231
- "26.mlp.up_proj",
232
- "language_model.layers.6.mlp.up_proj",
233
- "language_model.layers.13.self_attn.q_proj",
234
- "language_model.layers.2.self_attn.k_proj",
235
- "language_model.layers.12.mlp.gate_proj",
236
  "25.self_attn.q_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
237
  "language_model.layers.3.self_attn.v_proj",
238
- "22.mlp.gate_proj",
239
- "language_model.layers.14.self_attn.q_proj",
240
- "29.self_attn.q_proj",
241
- "16.self_attn.o_proj",
242
- "22.mlp.down_proj",
243
- "language_model.layers.1.self_attn.o_proj",
244
- "language_model.layers.14.self_attn.k_proj",
245
- "18.self_attn.q_proj",
246
- "language_model.layers.6.self_attn.q_proj",
247
  "language_model.layers.12.self_attn.q_proj",
248
- "30.mlp.gate_proj"
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
249
  ],
250
  "target_parameters": null,
251
  "task_type": "CAUSAL_LM",
 
32
  "rank_pattern": {},
33
  "revision": null,
34
  "target_modules": [
35
+ "19.self_attn.q_proj",
36
+ "language_model.layers.1.self_attn.q_proj",
37
+ "per_layer_model_projection",
38
+ "34.self_attn.o_proj",
39
+ "language_model.layers.11.self_attn.q_proj",
40
+ "31.self_attn.o_proj",
41
+ "16.self_attn.q_proj",
42
+ "language_model.layers.1.mlp.gate_proj",
43
+ "22.self_attn.o_proj",
44
+ "language_model.layers.15.self_attn.q_proj",
45
+ "language_model.layers.12.mlp.gate_proj",
46
+ "20.self_attn.q_proj",
47
+ "input_proj_linear",
48
+ "18.self_attn.q_proj",
49
+ "language_model.layers.12.self_attn.o_proj",
50
+ "32.self_attn.o_proj",
51
+ "language_model.layers.10.mlp.up_proj",
52
+ "34.mlp.down_proj",
53
+ "24.mlp.up_proj",
54
+ "23.mlp.gate_proj",
55
  "22.mlp.up_proj",
56
+ "language_model.layers.3.mlp.down_proj",
 
57
  "language_model.layers.8.self_attn.o_proj",
58
+ "language_model.layers.11.self_attn.v_proj",
59
+ "28.mlp.gate_proj",
60
+ "17.self_attn.q_proj",
61
+ "language_model.layers.2.self_attn.k_proj",
62
+ "language_model.layers.9.self_attn.k_proj",
63
+ "language_model.layers.12.mlp.up_proj",
64
+ "18.self_attn.o_proj",
65
+ "16.mlp.up_proj",
66
+ "language_model.layers.10.mlp.down_proj",
67
+ "embedding_projection",
68
+ "17.mlp.gate_proj",
69
+ "18.mlp.down_proj",
70
+ "input_proj",
71
+ "language_model.layers.0.mlp.down_proj",
72
+ "language_model.layers.15.mlp.gate_proj",
73
+ "27.mlp.gate_proj",
74
+ "language_model.layers.4.mlp.gate_proj",
75
+ "language_model.layers.0.self_attn.q_proj",
76
+ "language_model.layers.0.self_attn.k_proj",
77
+ "language_model.layers.12.self_attn.v_proj",
78
+ "21.self_attn.q_proj",
79
+ "29.mlp.down_proj",
80
+ "34.mlp.up_proj",
81
+ "per_layer_input_gate",
82
+ "language_model.layers.1.mlp.up_proj",
83
+ "language_model.layers.7.self_attn.k_proj",
84
+ "27.mlp.up_proj",
85
+ "language_model.layers.15.self_attn.o_proj",
86
+ "language_model.layers.13.self_attn.o_proj",
87
+ "language_model.layers.14.self_attn.v_proj",
88
+ "language_model.layers.5.self_attn.v_proj",
89
+ "linear",
90
+ "language_model.layers.7.mlp.up_proj",
91
+ "25.mlp.gate_proj",
92
+ "language_model.layers.3.self_attn.q_proj",
93
+ "26.mlp.down_proj",
94
+ "34.self_attn.q_proj",
95
  "language_model.layers.10.self_attn.q_proj",
96
+ "language_model.layers.14.mlp.gate_proj",
97
+ "21.mlp.gate_proj",
98
+ "31.mlp.down_proj",
99
+ "21.self_attn.o_proj",
100
+ "33.mlp.gate_proj",
101
+ "24.self_attn.o_proj",
102
+ "29.mlp.up_proj",
103
+ "language_model.layers.4.self_attn.q_proj",
104
+ "30.mlp.up_proj",
105
+ "24.mlp.down_proj",
106
+ "25.mlp.down_proj",
107
+ "language_model.layers.8.mlp.down_proj",
108
+ "22.mlp.gate_proj",
109
+ "language_model.layers.14.self_attn.k_proj",
110
+ "language_model.layers.11.mlp.up_proj",
111
+ "language_model.layers.4.self_attn.k_proj",
112
+ "29.self_attn.q_proj",
113
+ "25.self_attn.o_proj",
114
+ "19.mlp.down_proj",
115
+ "output_proj",
116
+ "language_model.layers.9.mlp.gate_proj",
117
+ "22.self_attn.q_proj",
118
+ "language_model.layers.2.self_attn.q_proj",
119
+ "30.self_attn.o_proj",
120
+ "language_model.layers.11.mlp.down_proj",
121
+ "26.self_attn.q_proj",
122
  "language_model.layers.0.mlp.gate_proj",
123
+ "language_model.layers.9.mlp.up_proj",
 
124
  "language_model.layers.8.mlp.up_proj",
 
125
  "language_model.layers.8.self_attn.v_proj",
126
+ "23.self_attn.o_proj",
127
+ "23.self_attn.q_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
128
  "language_model.layers.3.self_attn.o_proj",
129
+ "language_model.layers.7.mlp.gate_proj",
130
+ "language_model.layers.10.mlp.gate_proj",
131
+ "language_model.layers.14.self_attn.o_proj",
132
+ "language_model.layers.6.self_attn.o_proj",
133
  "19.mlp.up_proj",
134
  "language_model.layers.14.mlp.up_proj",
135
+ "language_model.layers.2.mlp.down_proj",
136
+ "27.mlp.down_proj",
137
+ "language_model.layers.0.mlp.up_proj",
138
+ "21.mlp.up_proj",
139
+ "23.mlp.up_proj",
140
+ "relative_k_proj",
141
+ "language_model.layers.6.mlp.down_proj",
142
+ "language_model.layers.5.self_attn.o_proj",
143
+ "language_model.layers.14.self_attn.q_proj",
144
+ "18.mlp.gate_proj",
145
+ "language_model.layers.5.self_attn.q_proj",
146
+ "language_model.layers.7.self_attn.o_proj",
147
+ "language_model.layers.7.self_attn.v_proj",
148
+ "per_layer_projection",
149
+ "21.mlp.down_proj",
150
  "32.self_attn.q_proj",
151
+ "20.mlp.up_proj",
152
+ "20.self_attn.o_proj",
 
 
 
153
  "16.mlp.down_proj",
 
 
 
 
 
 
 
 
 
 
 
154
  "language_model.layers.2.self_attn.o_proj",
155
+ "language_model.layers.5.self_attn.k_proj",
156
+ "32.mlp.up_proj",
157
+ "27.self_attn.q_proj",
158
+ "language_model.layers.9.self_attn.o_proj",
159
+ "language_model.layers.7.mlp.down_proj",
160
+ "language_model.layers.1.self_attn.o_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
161
  "30.mlp.down_proj",
162
+ "33.self_attn.q_proj",
163
+ "language_model.layers.9.mlp.down_proj",
164
+ "language_model.layers.13.self_attn.k_proj",
165
+ "16.self_attn.o_proj",
166
+ "language_model.layers.5.mlp.down_proj",
 
 
 
 
 
 
167
  "language_model.layers.4.mlp.up_proj",
168
+ "language_model.layers.6.self_attn.q_proj",
169
+ "33.self_attn.o_proj",
 
 
 
 
 
 
 
170
  "29.mlp.gate_proj",
171
+ "28.mlp.up_proj",
172
+ "16.mlp.gate_proj",
173
+ "language_model.layers.14.mlp.down_proj",
174
+ "28.self_attn.q_proj",
175
+ "33.mlp.up_proj",
176
  "language_model.layers.6.self_attn.k_proj",
177
+ "language_model.layers.13.mlp.down_proj",
178
+ "26.mlp.up_proj",
179
+ "19.self_attn.o_proj",
180
+ "language_model.layers.10.self_attn.o_proj",
181
+ "30.mlp.gate_proj",
182
+ "language_model.layers.1.self_attn.v_proj",
183
+ "17.mlp.down_proj",
184
+ "29.self_attn.o_proj",
185
+ "language_model.layers.10.self_attn.k_proj",
186
+ "18.mlp.up_proj",
187
  "25.mlp.up_proj",
188
+ "27.self_attn.o_proj",
189
+ "language_model.layers.8.mlp.gate_proj",
190
+ "language_model.layers.6.mlp.gate_proj",
191
+ "22.mlp.down_proj",
192
  "language_model.layers.4.mlp.down_proj",
193
+ "30.self_attn.q_proj",
194
  "language_model.layers.5.mlp.up_proj",
195
+ "28.self_attn.o_proj",
196
+ "language_model.layers.0.self_attn.o_proj",
197
+ "language_model.layers.6.self_attn.v_proj",
198
+ "language_model.layers.8.self_attn.k_proj",
199
+ "language_model.layers.3.mlp.gate_proj",
200
+ "31.mlp.up_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
201
  "language_model.layers.11.self_attn.o_proj",
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
202
  "25.self_attn.q_proj",
203
+ "28.mlp.down_proj",
204
+ "language_model.layers.11.self_attn.k_proj",
205
+ "31.mlp.gate_proj",
206
+ "32.mlp.down_proj",
207
+ "language_model.layers.15.mlp.up_proj",
208
+ "33.mlp.down_proj",
209
+ "language_model.layers.0.self_attn.v_proj",
210
+ "language_model.layers.1.self_attn.k_proj",
211
+ "language_model.layers.6.mlp.up_proj",
212
+ "language_model.layers.3.self_attn.k_proj",
213
+ "34.mlp.gate_proj",
214
+ "language_model.layers.12.mlp.down_proj",
215
+ "language_model.layers.13.self_attn.v_proj",
216
+ "language_model.layers.2.mlp.gate_proj",
217
  "language_model.layers.3.self_attn.v_proj",
218
+ "24.mlp.gate_proj",
219
+ "language_model.layers.8.self_attn.q_proj",
220
+ "19.mlp.gate_proj",
221
+ "language_model.layers.15.mlp.down_proj",
222
+ "language_model.layers.2.self_attn.v_proj",
223
+ "language_model.layers.1.mlp.down_proj",
224
+ "24.self_attn.q_proj",
225
+ "language_model.layers.3.mlp.up_proj",
226
+ "20.mlp.down_proj",
227
  "language_model.layers.12.self_attn.q_proj",
228
+ "language_model.layers.4.self_attn.v_proj",
229
+ "31.self_attn.q_proj",
230
+ "language_model.layers.2.mlp.up_proj",
231
+ "32.mlp.gate_proj",
232
+ "language_model.layers.4.self_attn.o_proj",
233
+ "language_model.layers.13.self_attn.q_proj",
234
+ "language_model.layers.13.mlp.gate_proj",
235
+ "language_model.layers.10.self_attn.v_proj",
236
+ "26.self_attn.o_proj",
237
+ "20.mlp.gate_proj",
238
+ "17.self_attn.o_proj",
239
+ "language_model.layers.9.self_attn.v_proj",
240
+ "language_model.layers.13.mlp.up_proj",
241
+ "26.mlp.gate_proj",
242
+ "language_model.layers.9.self_attn.q_proj",
243
+ "17.mlp.up_proj",
244
+ "language_model.layers.11.mlp.gate_proj",
245
+ "23.mlp.down_proj",
246
+ "language_model.layers.12.self_attn.k_proj",
247
+ "language_model.layers.5.mlp.gate_proj",
248
+ "language_model.layers.7.self_attn.q_proj"
249
  ],
250
  "target_parameters": null,
251
  "task_type": "CAUSAL_LM",
gemma4_lora_to_review/95_percent/adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:e9473003ac19644fdfff066ec4d633e843127aada38dc4fbc5f399dcbf6be7c9
3
  size 151844848
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8cc57b069e449c58d33b41ce0f2d18952d2387fbc351c8c948378f3ed4092087
3
  size 151844848