Update README.md

README.md

@@ -29,3 +29,14 @@ inference:
 - Base model: [ahxt/llama2_xs_460M_experimental](https://huggingface.co/ahxt/llama2_xs_460M_experimental)
 - Dataset: [KnutJaegersberg/WizardLM_evol_instruct_V2_196k_instruct_format](https://huggingface.co/datasets/KnutJaegersberg/WizardLM_evol_instruct_V2_196k_instruct_format)
 - Training parameters: [training_params.json](https://huggingface.co/Felladrin/llama2_xs_460M_experimental_evol_instruct/blob/cc151c5669ea37c3ef972e375c74f2d9bfd92b49/training_params.json)
+
+## Prompt format
+
+It was trained in the following prompt format:
+
+```
+### Instruction:
+<instruction>
+
+### Response:
+```