UgurKocde commited on
Commit
f4a363b
·
1 Parent(s): 5fc4a5d

tidy model card

Browse files
Files changed (1) hide show
  1. README.md +0 -4
README.md CHANGED
@@ -77,7 +77,3 @@ community project is not affiliated with or endorsed by Microsoft or OpenAI.
77
  ## Files
78
 
79
  The ready-to-run GGUF is available now in [openadmin-20b-GGUF](https://huggingface.co/OpenAdminOS/openadmin-20b-GGUF). Full-precision safetensors weights are being prepared and will be added to this repository shortly.
80
-
81
- ## Known limitations in v1
82
-
83
- - Without a system prompt, the model may misidentify itself (it can claim to be ChatGPT or GPT-4; it is neither, it is a gpt-oss-20b fine-tune). The Ollama build ships a system prompt that corrects this; if you consume the GGUF directly, set your own system prompt. A trained identity fix is planned for v1.1.
 
77
  ## Files
78
 
79
  The ready-to-run GGUF is available now in [openadmin-20b-GGUF](https://huggingface.co/OpenAdminOS/openadmin-20b-GGUF). Full-precision safetensors weights are being prepared and will be added to this repository shortly.