llama: update vendored code to commit 40c6d79f (#7875)

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q5_1.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q4_1-q8_0.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-f16.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_0.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q4_1.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_0.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q5_1.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_0-q8_0.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-f16.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_0.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q4_1.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_0.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q5_1.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q5_1-q8_0.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-f16.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_0.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q4_1.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_0.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q5_1.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *

llama/ggml-cuda/template-instances/fattn-vec-f16-instance-hs128-q8_0-q8_0.cu

 /**
- * llama.cpp - commit 3f1ae2e32cde00c39b96be6d01c2997c29bae555 - do not edit this file
+ * llama.cpp - commit 40c6d79fb52f995f47507fedfeaae2ac05d9b35c - do not edit this file
  *
  * MIT License
  *