Update README.md

README.md

@@ -24,7 +24,7 @@ Gemma v2 is a large language model released by Google on Jun 27th 2024.
 - Original model: [google/gemma-2-9b-it](https://huggingface.co/google/gemma-2-9b-it)
 
 The model is packaged into executable weights, which we call
-[llamafiles](https://github.com/Mozilla-Ocho/llamafile)). This makes it
+[llamafiles](https://github.com/Mozilla-Ocho/llamafile). This makes it
 easy to use the model on Linux, MacOS, Windows, FreeBSD, OpenBSD, and
 NetBSD for AMD64 and ARM64.
 
@@ -75,11 +75,9 @@ of the README.
 
 When using the browser GUI, you need to fill out the following fields.
 
-Prompt template:
+Prompt template (note: this is for chat; Gemma doesn't have a system role):
 
 ```
-<start_of_turn>system
-{{prompt}}<end_of_turn>
 {{history}}
 <start_of_turn>{{char}} 
 ```
@@ -109,9 +107,14 @@ AMD64.
 
 ## About Quantization Formats
 
-This model works should work well with any quantization format. Q6\_K is
-the best choice overall here. But since this is a Google model, the
-Google Brain floating point format (BF16) provides maximum quality.
+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),
+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 9b llamafiles are also
+faithful to Google's intentions. If you encounter any divergences, then
+try using the BF16 weights, which have the original fidelity.
 
 ---