maeeeeee commited on
Commit
6bdc086
1 Parent(s): d7a24e6

Upload README.md

Browse files
Files changed (1) hide show
  1. README.md +68 -0
README.md ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ base_model:
3
+ - mistralai/Mixtral-8x7B-v0.1
4
+ - mistralai/Mixtral-8x7B-Instruct-v0.1
5
+ - jondurbin/bagel-dpo-8x7b-v0.2
6
+ - cognitivecomputations/dolphin-2.7-mixtral-8x7b
7
+ - NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss
8
+ - ycros/BagelMIsteryTour-v2-8x7B
9
+ - smelborp/MixtralOrochi8x7B
10
+ library_name: transformers
11
+ tags:
12
+ - mergekit
13
+ - merge
14
+
15
+ ---
16
+
17
+ yum yum GGUF quants in my tum :^) enjoy lmg and vali, original model card below
18
+ <hr>
19
+ # maid-yuzu-v8-alter
20
+
21
+ This is a merge of pre-trained language models created using [mergekit](https://github.com/cg123/mergekit).
22
+
23
+ v7's approach worked better than I thought, so I tried something even weirder as a test. I don't think a proper model will come out, but I'm curious about the results.
24
+
25
+ ## Merge Details
26
+ ### Merge Method
27
+
28
+ This model was merged using the SLERP merge method.
29
+
30
+ This models were merged using the SLERP method in the following order:
31
+
32
+ maid-yuzu-v8-base: mistralai/Mixtral-8x7B-v0.1 + mistralai/Mixtral-8x7B-Instruct-v0.1 = 0.5
33
+ maid-yuzu-v8-step1: above + jondurbin/bagel-dpo-8x7b-v0.2 = 0.25
34
+ maid-yuzu-v8-step2: above + cognitivecomputations/dolphin-2.7-mixtral-8x7b = 0.25
35
+ maid-yuzu-v8-step3: above + NeverSleep/Noromaid-v0.4-Mixtral-Instruct-8x7b-Zloss = 0.25
36
+ maid-yuzu-v8-step4-alter: above + ycros/BagelMIsteryTour-v2-8x7B = 0.5
37
+ maid-yuzu-v8-alter: above + smelborp/MixtralOrochi8x7B = 0.5
38
+
39
+ ### Models Merged
40
+
41
+ The following models were included in the merge:
42
+ * [smelborp/MixtralOrochi8x7B](https://huggingface.co/smelborp/MixtralOrochi8x7B)
43
+ * ../maid-yuzu-v8-step4-alter
44
+
45
+ ### Configuration
46
+
47
+ The following YAML configuration was used to produce this model:
48
+
49
+ ```yaml
50
+ base_model:
51
+ model:
52
+ path: ../maid-yuzu-v8-step4-alter
53
+ dtype: bfloat16
54
+ merge_method: slerp
55
+ parameters:
56
+ t:
57
+ - value: 0.5
58
+ slices:
59
+ - sources:
60
+ - layer_range: [0, 32]
61
+ model:
62
+ model:
63
+ path: ../maid-yuzu-v8-step4-alter
64
+ - layer_range: [0, 32]
65
+ model:
66
+ model:
67
+ path: smelborp/MixtralOrochi8x7B
68
+ ```