zhangyiss/mlx
1
0
Fork 0
mirror of https://github.com/ml-explore/mlx.git synced 2025-12-16 01:49:05 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

Merge branch 'ml-explore:main' into adding-Muon-optimizer

Browse Source
This commit is contained in:
Gökdeniz Gülmez
2025-03-05 23:09:44 +01:00
committed by GitHub
parent 44f776921c 85b34d59bc
commit 59d4e4f61d
4 changed files with 18 additions and 6 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
mlx/distributed/ring/ring.cpp
View File

@@ -199,6 +199,7 @@ class SocketThread {
} }
void worker() { void worker() {
int error_count = 0;
bool delete_recv = false; bool delete_recv = false;
bool delete_send = false; bool delete_send = false;
while (true) { while (true) {
@@ -235,10 +236,11 @@ class SocketThread {
task.buffer = static_cast<char*>(task.buffer) + r; task.buffer = static_cast<char*>(task.buffer) + r;
task.size -= r; task.size -= r;
delete_recv = task.size == 0; delete_recv = task.size == 0;
error_count = 0;
} else if (errno != EAGAIN) { } else if (errno != EAGAIN) {
error_count++;
log_info( log_info(
true, "Receiving from socket", fd_, "failed with errno", errno); true, "Receiving from socket", fd_, "failed with errno", errno);
return;
} }
} }
if (!sends_.empty()) { if (!sends_.empty()) {
@@ -248,11 +250,17 @@ class SocketThread {
task.buffer = static_cast<char*>(task.buffer) + r; task.buffer = static_cast<char*>(task.buffer) + r;
task.size -= r; task.size -= r;
delete_send = task.size == 0; delete_send = task.size == 0;
error_count = 0;
} else if (errno != EAGAIN) { } else if (errno != EAGAIN) {
error_count++;
log_info(true, "Sending to socket", fd_, "failed with errno", errno); log_info(true, "Sending to socket", fd_, "failed with errno", errno);
return;
} }
} }
if (error_count >= 10) {
log_info(true, "Too many send/recv errors. Aborting...");
return;
}
} }
} }

2
mlx/version.h
View File

@@ -3,7 +3,7 @@
#pragma once #pragma once
#define MLX_VERSION_MAJOR 0 #define MLX_VERSION_MAJOR 0
#define MLX_VERSION_MINOR 24 #define MLX_VERSION_MINOR 23
#define MLX_VERSION_PATCH 2 #define MLX_VERSION_PATCH 2
#define MLX_VERSION_NUMERIC \ #define MLX_VERSION_NUMERIC \
(100000 * MLX_VERSION_MAJOR + 1000 * MLX_VERSION_MINOR + MLX_VERSION_PATCH) (100000 * MLX_VERSION_MAJOR + 1000 * MLX_VERSION_MINOR + MLX_VERSION_PATCH)

7
python/mlx/distributed_run.py
View File

@@ -112,7 +112,12 @@ def extract_rings(hosts, index):
break break
if not ring: if not ring:
break break
rings.append(normalize(concretize(ring, used_ports))) try:
rings.append(normalize(concretize(ring, used_ports)))
except RuntimeError:
if len(rings) > 0:
return rings
raise
return rings return rings

3
setup.py
View File

@@ -5,7 +5,6 @@ import os
import platform import platform
import re import re
import subprocess import subprocess
import sys
from pathlib import Path from pathlib import Path
from subprocess import run from subprocess import run
@@ -173,7 +172,7 @@ if __name__ == "__main__":
setup( setup(
name="mlx", name="mlx",
version=get_version("0.23.1"), version=get_version("0.23.2"),
author="MLX Contributors", author="MLX Contributors",
author_email="mlx@group.apple.com", author_email="mlx@group.apple.com",
description="A framework for machine learning on Apple silicon.", description="A framework for machine learning on Apple silicon.",