Update README.md
Browse files
README.md
CHANGED
|
@@ -4,10 +4,9 @@ language:
|
|
| 4 |
- en
|
| 5 |
---
|
| 6 |
|
| 7 |
-
My own quantizations.
|
| 8 |
-
output and embed
|
| 9 |
all other tensors quantized to q5_k or q6_k.
|
| 10 |
-
the q8_0 version is pure (all tensors quantized to Q8_0 just for reference)
|
| 11 |
|
| 12 |
Result:
|
| 13 |
both f16.q6 and f16.q5 are smaller than q8_0 standard quantization
|
|
|
|
| 4 |
- en
|
| 5 |
---
|
| 6 |
|
| 7 |
+
My own (ZeroWw) quantizations.
|
| 8 |
+
output and embed tensors quantized to f16.
|
| 9 |
all other tensors quantized to q5_k or q6_k.
|
|
|
|
| 10 |
|
| 11 |
Result:
|
| 12 |
both f16.q6 and f16.q5 are smaller than q8_0 standard quantization
|