ddh0 commited on
Commit
e65cfb6
·
verified ·
1 Parent(s): b075948

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +3 -0
README.md CHANGED
@@ -2,6 +2,9 @@
2
  license: mit
3
  ---
4
 
 
 
 
5
  ref: https://github.com/ddh0/llama.cpp/tree/save-tensor-npy-feature
6
 
7
  `./build/bin/llama-eval-callback -m ~/gguf/Qwen3-0.6B-f32.gguf -p "The quick brown fox" -n 1 -v`
 
2
  license: mit
3
  ---
4
 
5
+ These are all the intermediate tensors that are created by llama.cpp in a single forward pass through Qwen3-0.6B.
6
+
7
+
8
  ref: https://github.com/ddh0/llama.cpp/tree/save-tensor-npy-feature
9
 
10
  `./build/bin/llama-eval-callback -m ~/gguf/Qwen3-0.6B-f32.gguf -p "The quick brown fox" -n 1 -v`