legraphista
commited on
Update README.md
Browse files
README.md
CHANGED
@@ -26,6 +26,21 @@ tags:
|
|
26 |
- 1bit
|
27 |
---
|
28 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
29 |
# Qwen2-57B-A14B-Instruct-IMat-GGUF
|
30 |
_Llama.cpp imatrix quantization of Qwen/Qwen2-57B-A14B-Instruct_
|
31 |
|
|
|
26 |
- 1bit
|
27 |
---
|
28 |
|
29 |
+
<br>
|
30 |
+
<div style="padding: 16px 32px; outline: 2px solid; border-radius: 10px; outline-color: red; margin: 12px">
|
31 |
+
Currently investigating issue quantizing imatirx variants.
|
32 |
+
|
33 |
+
<pre>
|
34 |
+
[ 12/ 479] blk.0.ffn_gate_exps.weight - [ 3584, 2560, 64, 1], type = f32, converting to q4_K .. ggml_validate_row_data: found nan value at block 1
|
35 |
+
ggml_validate_row_data: found nan value at block 0
|
36 |
+
ggml_validate_row_data: found nan value at block 0
|
37 |
+
ggml_validate_row_data: found nan value at block 0
|
38 |
+
ggml_validate_row_data: found nan value at block 14
|
39 |
+
</pre>
|
40 |
+
</div>
|
41 |
+
|
42 |
+
---
|
43 |
+
|
44 |
# Qwen2-57B-A14B-Instruct-IMat-GGUF
|
45 |
_Llama.cpp imatrix quantization of Qwen/Qwen2-57B-A14B-Instruct_
|
46 |
|