Steboss

Steboss

Claim, Why isn't true, Correct claim
"It fits because the file fits", "The file is the weight payload at one precision. Peak memory also includes activations, workspace, and a KV cache that grows linearly with the context", "Fit is undetermined wihtout context length and runtime overhead"

Steboss

Claim, Why isn't true, Correct claim
"It fits because the file fits", "The file is the weight payload at one precision. Peak memory also includes activations, workspace, and a KV cache that grows linearly with the context", "Fit is undetermined wihtout context length and runtime overhead"

Steboss

use rayon::prelude::*;
use std::arch::aarch64::*;
use std::env;
use std::fs;

fn main() {
    let args: Vec<String> = env::args().collect();
    if args.len() != 4 {
        eprintln!("Usage: {} M K N", args[0]);
        return;

Steboss

use rayon::prelude::*;
use std::env;

const N: usize = 2; // Fixed block size

// HELPER: Convert slice to fixed array for safety (optional but good for debugging)
fn to_array(slice: &[f64]) -> [f64; N * N] {
    slice.try_into().expect("Slice has incorrect length")
}

Steboss

import logging
import re
from typing import List

from autogen_core import (
    DefaultTopicId,
    MessageContext,
    RoutedAgent,
    message_handler,
)

Steboss

import logging
import json
from typing import List

from autogen_core import (
    DefaultTopicId,
    MessageContext,
    RoutedAgent,
    message_handler,
)

Steboss

import os
import re
import glob
import pymupdf4llm 
from marker.converters.pdf import PdfConverter
from marker.models import create_model_dict
from marker.output import text_from_rendered
from langchain_core.documents import Document
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_experimental.text_splitter import SemanticChunker

Steboss

class BaseMatMulAgent(RoutedAgent):
    """
    A generic agent that can handle RequestToSpeak, call tools, and publish messages.
    """
    def __init__(
        self,
        description: str,
        group_chat_topic_type: str,
        model_client: ChatCompletionClient,
        system_message: str,

Steboss

#!/bin/bash
#SBATCH -A your_account
#SBATCH -p your_partition
#SBATCH -N 4 # example with 4 nodes
#SBATCH -t 04:00:00 # max run time
#SBATCH -J "something_in_line_with_your_system"


export CONFIG="fuji-70B-v2-flash" # here you can insert, for example,  fuji-7B-v2-flash 
export CONTAINER="ghcr.io/nvidia/jax:axlearn" # this is our public jax-axlearn container