Sweaterdog commited on
Commit
55c7c4f
·
verified ·
1 Parent(s): 52f6c82

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +12 -0
README.md CHANGED
@@ -29,6 +29,18 @@ The GRaPE Family was trained on about **14 billion** tokens of data after pre-tr
29
 
30
  GRaPE Flash and Nano are monomodal models, only accepting text. GRaPE Mini being trained most recently supports image and video inputs.
31
 
 
 
 
 
 
 
 
 
 
 
 
 
32
  # GRaPE Flash as a Model
33
 
34
  GRaPE Flash was designed for one thing: Speed. If you need a model that can quickly fill in tons of JSON data, this is your model. GRaPE Flash was chosen to **not** recieve thinking training as the model architecture would not benefit from it.
 
29
 
30
  GRaPE Flash and Nano are monomodal models, only accepting text. GRaPE Mini being trained most recently supports image and video inputs.
31
 
32
+ # How to Run
33
+
34
+ I recommend using **LM Studio** for running GRaPE Models, and have generally found these sampling parameters to work best:
35
+
36
+ | Name | Value |
37
+ | :--- | :--- |
38
+ | **Temperature** | 0.6 |
39
+ | **Top K Sampling** | 40 |
40
+ | **Repeat Penalty** | 1 |
41
+ | **Top P Sampling** | 0.85 |
42
+ | **Min P Sampling** | 0.05 |
43
+
44
  # GRaPE Flash as a Model
45
 
46
  GRaPE Flash was designed for one thing: Speed. If you need a model that can quickly fill in tons of JSON data, this is your model. GRaPE Flash was chosen to **not** recieve thinking training as the model architecture would not benefit from it.