Lewdiculous
commited on
Commit
•
6f6b45a
1
Parent(s):
0d40077
Update README.md
Browse files
README.md
CHANGED
@@ -29,6 +29,8 @@ The **Imatrix** is calculated based on calibration data, and it helps determine
|
|
29 |
The idea is to preserve the most important information during quantization, which can help reduce the loss of model performance, especially when the calibration data is diverse.
|
30 |
[[1]](https://github.com/ggerganov/llama.cpp/discussions/5006) [[2]](https://github.com/ggerganov/llama.cpp/discussions/5263#discussioncomment-8395384)
|
31 |
|
|
|
|
|
32 |
**Steps:**
|
33 |
```
|
34 |
Base⇢ GGUF(F16)⇢ Imatrix-Data(F16)⇢ GGUF(Imatrix-Quants)
|
@@ -43,10 +45,6 @@ Base⇢ GGUF(F16)⇢ Imatrix-Data(F16)⇢ GGUF(Imatrix-Quants)
|
|
43 |
|
44 |
If you want anything that's not here or another model, feel free to request.
|
45 |
|
46 |
-
**This is experimental.**
|
47 |
-
|
48 |
-
For imatrix data generation, kalomaze's `groups_merged.txt` with added roleplay chats was used, you can find it [here](https://huggingface.co/Lewdiculous/Datura_7B-GGUF-Imatrix/blob/main/imatrix-with-rp-format-data.txt).
|
49 |
-
|
50 |
**Original model information:**
|
51 |
|
52 |
Merged on request of Lewdiculus.
|
|
|
29 |
The idea is to preserve the most important information during quantization, which can help reduce the loss of model performance, especially when the calibration data is diverse.
|
30 |
[[1]](https://github.com/ggerganov/llama.cpp/discussions/5006) [[2]](https://github.com/ggerganov/llama.cpp/discussions/5263#discussioncomment-8395384)
|
31 |
|
32 |
+
For imatrix data generation, kalomaze's `groups_merged.txt` with added roleplay chats was used, you can find it [here](https://huggingface.co/Lewdiculous/Datura_7B-GGUF-Imatrix/blob/main/imatrix-with-rp-format-data.txt).
|
33 |
+
|
34 |
**Steps:**
|
35 |
```
|
36 |
Base⇢ GGUF(F16)⇢ Imatrix-Data(F16)⇢ GGUF(Imatrix-Quants)
|
|
|
45 |
|
46 |
If you want anything that's not here or another model, feel free to request.
|
47 |
|
|
|
|
|
|
|
|
|
48 |
**Original model information:**
|
49 |
|
50 |
Merged on request of Lewdiculus.
|