Cshavi commited on
Commit
cc532a2
·
1 Parent(s): 6f8e37a

Add GGUF model, Modelfile with LFS

Browse files
Files changed (3) hide show
  1. .gitattributes +1 -0
  2. Modelfile +32 -0
  3. gemma2-2b-chatdoctor.f16.gguf +3 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ *.gguf filter=lfs diff=lfs merge=lfs -text
Modelfile ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # /root/models/Modelfile ← save this next to gemma2‑2b‑chatdoctor.f16.gguf
2
+
3
+ #######################################################################
4
+ # Gemma‑2 2B ChatDoctor‑MedQA (FP16) – Ollama definition
5
+ #######################################################################
6
+
7
+ FROM ./gemma2-2b-chatdoctor.f16.gguf # relative path to the GGUF file
8
+
9
+ # --------------------------------------------------------------------
10
+ # Default persona & safety rails
11
+ # --------------------------------------------------------------------
12
+ SYSTEM """
13
+ You are **ChatDoctor‑MedQA**, a cautious medical assistant.
14
+ • Provide evidence‑based answers written in plain English.
15
+ • When citing studies, mention the journal (e.g., “NEJM 2021”).
16
+ • Always finish with the disclaimer:
17
+ _“Consult a licensed healthcare professional before making medical decisions.”_
18
+ """
19
+
20
+ # --------------------------------------------------------------------
21
+ # Prompt template (Gemma works well with a simple chat pattern)
22
+ # --------------------------------------------------------------------
23
+ TEMPLATE "<s>{{ range .Messages }}{{ .Role }}: {{ .Content }}\n{{ end }}assistant: "
24
+
25
+ # --------------------------------------------------------------------
26
+ # Optional generation parameters (comment out to keep Ollama defaults)
27
+ # --------------------------------------------------------------------
28
+ PARAMETER temperature 0.3 # lower → less creative, more factual
29
+ PARAMETER top_p 0.9
30
+ PARAMETER repeat_penalty 1.15
31
+ PARAMETER num_predict 512 # max tokens the assistant may add
32
+
gemma2-2b-chatdoctor.f16.gguf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e348b58b041c48c86ace98a0b69d0e96f10750e2d841d7e088428b2f0aa5d7e7
3
+ size 5235222464