Update README.md

README.md

@@ -37,7 +37,7 @@ software. Both of them are included in a single file, which can be
 downloaded and run as follows:
 
 ```
-wget https://huggingface.co/jartine/gemma-2-2b-it-llamafile/resolve/main/gemma-2-2b-it.Q6_K.llamafile
+wget https://huggingface.co/Mozilla/gemma-2-2b-it-llamafile/resolve/main/gemma-2-2b-it.Q6_K.llamafile
 chmod +x gemma-2-2b-it.Q6_K.llamafile
 ./gemma-2-2b-it.Q6_K.llamafile
 ```
@@ -157,7 +157,7 @@ AMD64.
 
 This model works well with any quantization format. Q6\_K is the best
 choice overall here. We tested that, with [our 27b Gemma2
-llamafiles](https://huggingface.co/jartine/gemma-2-27b-it-llamafile),
+llamafiles](https://huggingface.co/Mozilla/gemma-2-27b-it-llamafile),
 that the llamafile implementation of Gemma2 is able to to produce
 identical responses to the Gemma2 model that's hosted by Google on
 aistudio.google.com. Therefore we'd assume these 2b llamafiles are also
@@ -169,8 +169,8 @@ try using the BF16 weights, which have the original fidelity.
 There are higher quality versions of this model available as llamafiles,
 which require more memory.
 
-- <https://huggingface.co/jartine/gemma-2-9b-it-llamafile>
-- <https://huggingface.co/jartine/gemma-2-27b-it-llamafile>
+- <https://huggingface.co/Mozilla/gemma-2-9b-it-llamafile>
+- <https://huggingface.co/Mozilla/gemma-2-27b-it-llamafile>
 
 The 9B and 27B models were released a month earlier than 2B, so they're
 packaged with an slightly older version of the llamafile software.