Rafael rafaself
rafaself
/ affine_quantization.py
Created
October 23, 2026 20:10
Implementation of signed affine quantization and dequantization using NumPy, including scale and zero-point calculation, rounding, clipping, and configurable bit-width.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| import numpy as np | |
| def round_away(x): | |
| return np.sign(x) * np.floor(np.abs(x) + 0.5) | |
| def quantize_affine(x, num_bits=8): | |
| x = np.asarray(x, dtype=np.float64) | |
| q_min = -(2 ** (num_bits - 1)) | |
| q_max = 2 ** (num_bits - 1) - 1 |
rafaself
/ simple-vllm-chat.py
Created
October 23, 2026 14:18
A Simple vLLM Chat with a minimal CLI chatbot demonstrating LLM inference, conversation history, system prompts, and output validation.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| from vllm import LLM, SamplingParams | |
| # Model configuration | |
| MODEL_NAME = "Qwen/Qwen2.5-0.5B-Instruct" | |
| llm = LLM(model=MODEL_NAME) | |
| # Inference configuration | |
| llm_config = { | |
| "temperature": 0.2, |
rafaself
/ linear_regression_gradient_descent.py
Created
October 22, 2026 01:21
Linear regression implemented from scratch in Python using NumPy and Batch Gradient Descent, including MSE loss, analytical gradients, and training visualization.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| import numpy as np | |
| import matplotlib.pyplot as plt | |
| x = np.array([0, 1, 2, 3, 4, 5], dtype=float) | |
| y = np.array([1, 3, 5, 7, 9, 11], dtype=float) | |
| m = 0.0 | |
| b = 0.0 | |
| learning_rate = 0.01 |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| import { open, readFile } from "node:fs/promises"; | |
| async function* createBody() { | |
| yield Buffer.from("Hello"); | |
| yield Buffer.from(" "); | |
| yield Buffer.from("World!"); | |
| } | |
| async function writeObject(body) { | |
| const handle = await open("example.txt", "w"); |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| #include <array> | |
| #include <cstddef> | |
| #include <iostream> | |
| #include <optional> | |
| template <class T, std::size_t N> | |
| class RingBuffer { | |
| static_assert(N > 0, "N must be > 0"); | |
| public: |