From 93958d28220fc18a9d136e97530afdbfff38c395 Mon Sep 17 00:00:00 2001
From: Fabian Rybinski <fabian.rybinski@kit.edu>
Date: Wed, 20 Nov 2024 08:30:46 +0000
Subject: [PATCH] Add new file
---
aie-gemma-finetuned.yaml | 44 ++++++++++++++++++++++++++++++++++++++++
1 file changed, 44 insertions(+)
create mode 100644 aie-gemma-finetuned.yaml
diff --git a/aie-gemma-finetuned.yaml b/aie-gemma-finetuned.yaml
new file mode 100644
index 0000000..ca2209d
--- /dev/null
+++ b/aie-gemma-finetuned.yaml
@@ -0,0 +1,44 @@
+name: "aie-gemma-finetuned"
+
+description: |
+ This is a model which is currently tested.
+config_file: |
+ name: aie-gemma-finetuned
+ mmap: true
+ backend: llama
+ f16: true
+ gpu_layers: 35
+ threads: 12
+ parameters:
+ model: downloads/llama_backend/AIE_Gemma2_2B_IT_Q4_K_M.gguf
+ roles:
+ assistant: 'Assistant:'
+ system: 'System:'
+ user: 'User:'
+ template:
+ chat_message: |-
+ <start_of_turn>{{if eq .RoleName "assistant" }}model{{else}}{{ .RoleName }}{{end}}
+ {{ if .Content -}}
+ {{.Content -}}
+ {{ end -}}<end_of_turn>
+ chat: |
+ {{.Input }}
+ <start_of_turn>model
+ completion: |
+ {{.Input}}
+ stopwords:
+ - '<|im_end|>'
+ - '<end_of_turn>'
+ - '<start_of_turn>'
+ context_size: 4096
+
+
+ usage: |
+ curl http://localhost:8080/v1/chat/completions -H "Content-Type: application/json" -d '{
+ "model": "gpt-4",
+ "messages": [{"role": "user", "content": "How are you doing?", "temperature": 0.1}]
+ }'
+files:
+- filename: "downloads/llama_backend/AIE_Gemma2_2B_IT_Q4_K_M.gguf"
+ sha256: ""
+ uri: "https://huggingface.co/Phabby/AIE-Gemma-2-2B-IT/blob/main/AIE_Gemma2_2B_IT_Q4_K_M.gguf"
--
GitLab