awni

Test Time Scaling with MLX LM and R1-based LLMs

Install MLX LM:

pip install mlx-lm

And run:

RicardoDominguez

import datasets
import transformers
import vllm
from tqdm import tqdm

model_dir = 'meta-llama/Llama-2-7b-chat-hf'

# prompt from R1
system_prompt = "The user will ask you a question, and you should solve it. " \
        "You should first think about the reasoning process in the mind and then provide the user with the answer. " \

903124

# outlines/processor/structured.py
...
class GuideLogitsProcessor(OutlinesLogitsProcessor):
    """Bias generation using a finite

    Attributes
    ----------
    tokenizer
        The tokenizer used to convert tokens to ids.
    guide

abacaj

import tqdm
import numpy as np
import torch
import torch.distributed as dist
import transformers

def extract_xml_answer(text: str) -> str:
    answer = text.split("<final_answer>")[-1]
    answer = answer.split("</final_answer>")[0]
    return answer.strip()

willccbb

# train_grpo.py
#
# See https://github.com/willccbb/verifiers for ongoing developments
#
"""
citation:

@misc{brown2025grpodemo,
  title={Granular Format Rewards for Eliciting Mathematical Reasoning Capabilities in Small Language Models},
  author={Brown, William},

cfahlgren1

DuckDB SQL Console

Your are an LLM Assistant for the Hugging Face DuckDB SQL Console. The user will ask you questions about the data in the DuckDB database and you will answer them using SQL. Use the context provided to answer the user's questions and decide which tables to query. Only respond with SQL and comments if needed.

DuckDB Tips

• DuckDB is largely compatible with Postgres SQL. Stick to Postgres / DuckDB SQL syntax.
• DuckDB uses double quotes (") for identifiers that contain spaces or special characters, or to force case-sensitivity and single quotes (') to define string literals
• DuckDB can extract parts of strings and lists using [start:end] or [start:end:step] syntax. Indexes start at 1. String slicing: `SELECT 'DuckDB'[1:4];` Array/List slicing: `SELECT [1, 2, 3, 4][1:3];`
• DuckDB has a powerful way to select or transform multiple columns using patterns or functions. You can select columns matching a pattern: `SELECT COLUMNS('sales_.*') FROM sales_data;` or transform multiple colum

N8python

def generate_speculative(
    model: nn.Module,
    draft_model: nn.Module,
    tokenizer: Union[PreTrainedTokenizer, TokenizerWrapper],
    prompt: str,
    max_tokens: int = 100,
    verbose: bool = False,
    formatter: Optional[Callable] = None,
    **kwargs,

anuragts

"""Run `pip install ollama 'fastapi[standard]' phidata` to install dependencies."""

"Pull the local model with - ollama pull hhao/qwen2.5-coder-tools:32b"

from textwrap import dedent
from datetime import datetime

from phi.agent import Agent
from phi.playground import Playground, serve_playground_app
from phi.model.ollama import Ollama

awni

tvst

import streamlit as st
import concurrent.futures # We'll do computations in separate processes!
import mymodule # This is where you'll do the computation

# Your st calls must go inside this IF block.
if __name__ == '__main__':
    st.write("Starting a long computation on another process")

    # Pick max number of concurrent processes. Depends on how heavy your computation is, and how
    # powerful your machine is.