Update README.md
Browse files
README.md
CHANGED
|
@@ -136,4 +136,15 @@ could just be mmapped together into memory from split files. That doesn't keep m
|
|
| 136 |
been nice to look at the existing reality and/or consult the community before throwing yet another hard to support format out
|
| 137 |
there without thinking.
|
| 138 |
|
| 139 |
-
There are some developments to make this less of a pain, and I will revisit this issue from time to time to see if it has
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 136 |
been nice to look at the existing reality and/or consult the community before throwing yet another hard to support format out
|
| 137 |
there without thinking.
|
| 138 |
|
| 139 |
+
There are some developments to make this less of a pain, and I will revisit this issue from time to time to see if it has
|
| 140 |
+
become feasible.
|
| 141 |
+
|
| 142 |
+
## Why are there so many imatrix quants where llama.cpp crashes?
|
| 143 |
+
|
| 144 |
+
Because llama.cpp developers see imatrix quants as a useless gimmick, are not very interested in fixing bugs,
|
| 145 |
+
instead of mostly just wanting to ignore the problem, and considers merged models as mostly useless. They
|
| 146 |
+
are more interested in blaming their users.
|
| 147 |
+
|
| 148 |
+
As a result I have given up reporting bugs, and my efforts in providing imatrix quants will be reduced.
|
| 149 |
+
|
| 150 |
+
https://github.com/ggerganov/llama.cpp/issues/6841 is the issue ticket that escalated this.
|