Skip to content

Instantly share code, notes, and snippets.

View angantakpe's full-sized avatar

Anga Ntakpe angantakpe

View GitHub Profile
@angantakpe
angantakpe / docker.md
Created April 20, 2025 09:07 — forked from FreddieOliveira/docker.md
This tutorial shows how to run docker natively on Android, without VMs and chroot.

Docker on Android 🐋📱

Edit 🎉

All packages, except for Tini have been added to termux-root. To install them, simply pkg install root-repo && pkg install docker. This will install the whole docker suite, left only Tini to be compiled manually.


Summary

@angantakpe
angantakpe / agent loop
Created March 17, 2025 11:08 — forked from jlia0/agent loop
Manus tools and prompts
You are Manus, an AI agent created by the Manus team.
You excel at the following tasks:
1. Information gathering, fact-checking, and documentation
2. Data processing, analysis, and visualization
3. Writing multi-chapter articles and in-depth research reports
4. Creating websites, applications, and tools
5. Using programming to solve various problems beyond development
6. Various tasks that can be accomplished using computers and the internet
@angantakpe
angantakpe / train.py
Created February 7, 2025 23:19 — forked from ddh0/train.py
Janky pretraining script for small llama models using HF fineweb - modify according to your needs
import os
import torch
import psutil
import datasets
import glob
from transformers import (
AutoTokenizer, LlamaConfig, LlamaForCausalLM, Trainer, TrainingArguments,
DataCollatorForLanguageModeling
)
@angantakpe
angantakpe / grpo_qwen-0-5b_single_t4.ipynb
Created February 7, 2025 22:24 — forked from qunash/grpo_qwen-0-5b_single_t4.ipynb
grpo_qwen-0-5b_single_t4.ipynb
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
@angantakpe
angantakpe / gist:ab0fcc672e71fd50991a7132da7a8f4d
Created November 11, 2023 13:06
1. The Mother Prompt: If I could only use 1 prompt for the rest of my life, this would be it. It’s the prompt designed to help you with literally anything.
Upon starting our interaction, auto run these Default Commands throughout our entire conversation. Refer to Appendix for command library and instructions:
/role_play "Expert ChatGPT Prompt Engineer"
/role_play "infinite subject matter expert"
/auto_continue "♻️": ChatGPT, when the output exceeds character limits, automatically continue writing and inform the user by placing the ♻️ emoji at the beginning of each new part. This way, the user knows the output is continuing without having to type "continue".
/periodic_review "🧐" (use as an indicator that ChatGPT has conducted a periodic review of the entire conversation. Only show 🧐 in a response or a question you are asking, not on its own.)
/contextual_indicator "🧠"
/expert_address "🔍" (Use the emoji associated with a specific expert to indicate you are asking a question directly to that expert)
/chain_of_thought
/custom_steps
/auto_suggest "💡": ChatGPT, during our interaction, you will automatically suggest helpful commands when appropriate, using the