Spaces:
Running
on
A100
Running
on
A100
Add examples
Browse files- examples/gradient-slerp.yml +20 -0
- examples/linear.yml +12 -0
- examples/mega.yml +37 -0
- examples/orcamini-platy-44layer.yml +9 -0
- examples/ties.yml +22 -0
examples/gradient-slerp.yml
ADDED
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
slices:
|
2 |
+
- sources:
|
3 |
+
- model: psmathur/orca_mini_v3_13b
|
4 |
+
layer_range: [0, 40]
|
5 |
+
- model: garage-bAInd/Platypus2-13B
|
6 |
+
layer_range: [0, 40]
|
7 |
+
# or, the equivalent models: syntax:
|
8 |
+
# models:
|
9 |
+
# - model: psmathur/orca_mini_v3_13b
|
10 |
+
# - model: garage-bAInd/Platypus2-13B
|
11 |
+
merge_method: slerp
|
12 |
+
base_model: psmathur/orca_mini_v3_13b
|
13 |
+
parameters:
|
14 |
+
t:
|
15 |
+
- filter: self_attn
|
16 |
+
value: [0, 0.5, 0.3, 0.7, 1]
|
17 |
+
- filter: mlp
|
18 |
+
value: [1, 0.5, 0.7, 0.3, 0]
|
19 |
+
- value: 0.5 # fallback for rest of tensors
|
20 |
+
dtype: float16
|
examples/linear.yml
ADDED
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
models:
|
2 |
+
- model: psmathur/orca_mini_v3_13b
|
3 |
+
parameters:
|
4 |
+
weight: 1.0
|
5 |
+
- model: WizardLM/WizardLM-13B-V1.2
|
6 |
+
parameters:
|
7 |
+
weight: 0.3
|
8 |
+
- model: garage-bAInd/Platypus2-13B
|
9 |
+
parameters:
|
10 |
+
weight: 0.5
|
11 |
+
merge_method: linear
|
12 |
+
dtype: float16
|
examples/mega.yml
ADDED
@@ -0,0 +1,37 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
slices:
|
2 |
+
- sources:
|
3 |
+
- model: psmathur/orca_mini_v3_13b
|
4 |
+
layer_range: [0, 40]
|
5 |
+
- model: garage-bAInd/Platypus2-13B
|
6 |
+
layer_range: [0, 40]
|
7 |
+
merge_method: slerp
|
8 |
+
base_model: psmathur/orca_mini_v3_13b
|
9 |
+
parameters:
|
10 |
+
t:
|
11 |
+
- filter: self_attn
|
12 |
+
value: [0, 0.5, 0.3, 0.7, 1]
|
13 |
+
- filter: mlp
|
14 |
+
value: [1, 0.5, 0.7, 0.3, 0]
|
15 |
+
- value: 0.5 # fallback for rest of tensors
|
16 |
+
dtype: float16
|
17 |
+
name: gradient-slerp
|
18 |
+
---
|
19 |
+
models:
|
20 |
+
- model: gradient-slerp
|
21 |
+
parameters:
|
22 |
+
density: [1, 0.7, 0.1] # density gradient
|
23 |
+
weight: 1.0
|
24 |
+
- model: WizardLM/WizardMath-13B-V1.0
|
25 |
+
parameters:
|
26 |
+
density: 0.33
|
27 |
+
weight:
|
28 |
+
- filter: mlp
|
29 |
+
value: 0.5
|
30 |
+
- value: 0
|
31 |
+
merge_method: ties
|
32 |
+
base_model: TheBloke/Llama-2-13B-fp16
|
33 |
+
parameters:
|
34 |
+
normalize: true
|
35 |
+
int8_mask: true
|
36 |
+
dtype: float16
|
37 |
+
name: gradient-slerp-ties
|
examples/orcamini-platy-44layer.yml
ADDED
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
slices:
|
2 |
+
- sources:
|
3 |
+
- model: psmathur/orca_mini_v3_13b
|
4 |
+
layer_range: [0, 24]
|
5 |
+
- sources:
|
6 |
+
- model: garage-bAInd/Platypus2-13B
|
7 |
+
layer_range: [20, 40]
|
8 |
+
merge_method: passthrough
|
9 |
+
dtype: float16
|
examples/ties.yml
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
models:
|
2 |
+
- model: psmathur/orca_mini_v3_13b
|
3 |
+
parameters:
|
4 |
+
density: [1, 0.7, 0.1] # density gradient
|
5 |
+
weight: 1.0
|
6 |
+
- model: garage-bAInd/Platypus2-13B
|
7 |
+
parameters:
|
8 |
+
density: 0.5
|
9 |
+
weight: [0, 0.3, 0.7, 1] # weight gradient
|
10 |
+
- model: WizardLM/WizardMath-13B-V1.0
|
11 |
+
parameters:
|
12 |
+
density: 0.33
|
13 |
+
weight:
|
14 |
+
- filter: mlp
|
15 |
+
value: 0.5
|
16 |
+
- value: 0
|
17 |
+
merge_method: ties
|
18 |
+
base_model: TheBloke/Llama-2-13B-fp16
|
19 |
+
parameters:
|
20 |
+
normalize: true
|
21 |
+
int8_mask: true
|
22 |
+
dtype: float16
|