Responses API workarounds
pip install gpt-oss
python -m gpt_oss.responses_api.serve \
–inference_backend=ollama \
–checkpoint gpt-oss:20b
Tag: llama
Using tools (function calling)
tools = [
{
“type”: “function”,
“function”: {
“name”: “get_weather”,
“description”: “Get current weather in a given city”,
“parameters”: {
Chat with gpt-oss
ollama run gpt-oss:20b
Running gpt-oss-safeguard with Ollama
ollama run gpt-oss-safeguard:20b
PromptScript Engineer
“`markdown
[learn](PromptScript) {
description: “PromptScript is a method to create clear and organized prompts for AI models like ChatGPT. It uses symbols and conventions to define tasks, inputs,
PromptScript
“`markdown
Symbols and Conventions:
1. [ ]: Define tasks using square brackets.
Example: [research], [summarize], [suggest]
2. { }: Specify input parameters for tasks using curly braces.
Exam
Prompt Creator
“`markdown
Assume the role of my ‘Prompt Engineer,’ tasked with aiding me in designing an optimal, personalized prompt that suits my needs perfectly. You, ChatGPT, will be the implementer of this pro
DALL-E
“`markdown
Must only use 350 characters, write without word wraps and headlines, without connection words, back to back separated with commas:
[1], [2], [3] {night}, [4], [5], [6],
{camera settings}