Update README.md
Browse files
README.md
CHANGED
|
@@ -1,3 +1,4 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: apache-2.0
|
| 3 |
-
---
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
---
|
| 4 |
+
The Qwen authors highlight in their blogpost that qwen 2 7b can handle sequences up to 128k, but the GGUF meta-data is set to 32k, I set it to 131k. Using the script llama.cpp script, and this command "python gguf-set-metadata.py qwen2-7b-instruct-q5_k_m.gguf qwen2.context_length 131072 --force"
|