Add files using upload-large-folder tool
Browse filesThis view is limited to 50 files because it contains too many changes. See raw diff
- 9b_l20/outputs_add_free/all/lang/1/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/12/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/126/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/129/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/14/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/149/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/15/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/152/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/161/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/163/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/168/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/17/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/178/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/181/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/187/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/189/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/203/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/210/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/226/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/250/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/259/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/264/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/320/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/33/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/336/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/34/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/35/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/354/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/356/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/359/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/36/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/378/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/387/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/396/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/407/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/419/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/42/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/441/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/445/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/457/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/458/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/477/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/486/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/49/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/64/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/70/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/74/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/78/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/82/config.json +12 -0
- 9b_l20/outputs_add_free/all/lang/98/config.json +12 -0
9b_l20/outputs_add_free/all/lang/1/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/12/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/126/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/129/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/14/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/149/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/15/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/152/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/161/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/163/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/168/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/17/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/178/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/181/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/187/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/189/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/203/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/210/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/226/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/250/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/259/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/264/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/320/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/33/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/336/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/34/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/35/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/354/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/356/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/359/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/36/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/378/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/387/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/396/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/407/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/419/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/42/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/441/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/445/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/457/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/458/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/477/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/486/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/49/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/64/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/70/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/74/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/78/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/82/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|
9b_l20/outputs_add_free/all/lang/98/config.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"layer": 20,
|
| 4 |
+
"embed_dim": 3584,
|
| 5 |
+
"low_rank_dim": 1,
|
| 6 |
+
"target_module": "model.layers.20",
|
| 7 |
+
"intervention_type": "AdditionFreeIntervention",
|
| 8 |
+
"factor_init_scale": 2.0,
|
| 9 |
+
"vector_init_scale": 8.0,
|
| 10 |
+
"alpha": 0.0
|
| 11 |
+
}
|
| 12 |
+
]
|