yada trojblue

Grok is amazing, you can just learn things now lol

Listed the rencent (2023-25) and relavant Uni / Grad courses about image understanding, that can be used either as learning materials, or as RAG grounding sources.

Comprehensive Overview of Web-Accessible University Courses on Advanced Image Understanding (2023–2025)

Introduction

This document provides a carefully curated list of university and graduate-level courses from top-tier institutions that offer openly accessible online materials such as lecture slides, notes, and assignments. These courses focus specifically on image understanding, classification, regression on features, and advanced topics including Vision Transformers (ViT), vision-language models, and in-depth feature analysis. Special attention has been given to recent offerings (2023 or later), ensuring the relevance of the material.

paste all contents below and replace {TO_CONVERT_CONTENT} with a working example of the new model to integrate

Here's the base inference class for a library called procslib:

# src/procslib/models/base_inference.py

from abc import ABC, abstractmethod

Usage:

Go to Gemini "Stream Realtime" mode, in Google AI Studio: https://aistudio.google.com/live
Paste prompt into the "system instructions" section
Start screenshare and ask questions about current material.

Now it should work significantly better than default, without stupid reassuring questions.

Prompt:

Changing remote from one main repo to your fork

check curr remote:

git remote -v

origin https://github.com/RVC-Boss/GPT-SoVITS (fetch)

env:

ollama 0.4.7, quantized 4bit models
8xH100
python 3.12

test command:

https://github.com/MinhNgyuen/llm-benchmark

python benchmark.py --verbose

Copypasta for desired agent behaviors

Targeted for wen_coder_32b, but works better on deepseek-v3

You are Qwen, created by Alibaba Cloud. You are a helpful assistant. 

Behaviors:
    - You never alter or exploit the intentions of any given task.
    - When instructions are unclear, you ask clarifying questions instead of making assumptions.

code v1:

import json
import struct
import base64
import matplotlib.pyplot as plt
from collections import Counter

def read_widd_file(file_path):

	"""
	orig: https://huggingface.co/spaces/hexgrad/Kokoro-TTS

	Deps:
	pip install kokoro

	other files: see original repo
	"""
	import os
	import random

	// ==UserScript==
	// @name Remove F12 Blocking
	// @namespace http://tampermonkey.net/
	// @version 1.0
	// @description Removes scripts that block developer tools access
	// @author YourName
	// @match :///*
	// @run-at document-start
	// @grant none
	// ==/UserScript==

	"""
	To use this script:

	1. Install dependencies:
	```bash
	pip install pypdf
	```

	2. Run it with your input file path and desired output path.