awni

The following guide will show you how to connect a local model served with MLX to OpenCode for local coding.

1. Install OpenCode

curl -fsSL https://opencode.ai/install | bash

2. Install mlx-lm

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},

drisspg

Summary

This doc servers as a quick reference for the _scaled_mm API and how it has changed overtime for each major version of PyTorch.

---

NOTE The leading underscore is intended here and we make no current FC/BC guarantees on this API. That being said it is currently the only OP that has native support for FP8 matmuls within the PyTorch Libary. We are planning to make an official Public api for this. Until then this is subject to change but you can use this doc as a reference.

---

andrewnc

import math
import torch
from torch.optim.lr_scheduler import _LRScheduler
from dataclasses import dataclass
from typing import List

@dataclass
class SchedulePhase:
    """Defines a phase in the learning rate schedule"""
    percent: float  # Percentage of total steps this phase covers

awni

from huggingface_hub import snapshot_download
import mlx.core as mx
import mlx.nn as nn
import time


class Block(nn.Module):
    def __init__(self, in_dims, dims, stride=1):
        super().__init__()

kalomaze

# coding=utf-8
# Copyright 2023 Mixtral AI and the HuggingFace Inc. team. All rights reserved.
#
# This code is based on EleutherAI's GPT-NeoX library and the GPT-NeoX
# and OPT implementations in this library. It has been modified from its
# original forms to accommodate minor architectural differences compared
# to GPT-NeoX and OPT used by the Meta AI team that trained the model.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

vgoklani

def verify_ddp_weights_equal(model: torch.nn.Module, atol: float = 1e-5) -> None:
    if hasattr(model, "module"):
        model = model.module
    
    world_size = get_world_size()
    for name, param in model.named_parameters():
        gathered_param = gather(param).reshape((world_size, -1))
        absolute_diffs = (gathered_param[None, 0, :] - gathered_param).abs()
        rank_params_eq = (absolute_diffs < atol).all()
        assert rank_params_eq, f"❌ param [{name}] not equal - got max_absolute_diff={absolute_diffs.max()}"

jxmorris12

def verify_ddp_weights_equal(model: torch.nn.Module, atol: float = 1e-5) -> None:
    if hasattr(model, "module"):
        model = model.module
    
    world_size = get_world_size()
    for name, param in model.named_parameters():
        gathered_param = gather(param).reshape((world_size, -1))
        absolute_diffs = (gathered_param[None, 0, :] - gathered_param).abs()
        rank_params_eq = (absolute_diffs < atol).all()
        assert rank_params_eq, f"❌ param [{name}] not equal - got max_absolute_diff={absolute_diffs.max()}"

Birch-san

from transformers import (
  AutoConfig,
  AutoTokenizer,
  BitsAndBytesConfig,
  GenerationConfig,
  AutoModelForCausalLM,
  LlamaTokenizerFast,
  PreTrainedModel,
  TextIteratorStreamer,
  StoppingCriteria,

mara004

# Four lines intentionally left blank




# SPDX-FileCopyrightText: 2025 geisserml <[email protected]>
# SPDX-License-Identifier: Apache-2.0 OR MPL-2.0

# See also https://github.com/extremeheat/JSPyBridge/blob/master/examples/python/pdfjs.py