当前位置： 首页 > news >正文

简单、高效且低成本的预训练、微调与服务，惠及大众基于 Ray 架构设计的覆盖大语言模型（LLM）完整生命周期的解决方案byzer-llm

news 2025/11/2 6:35:37

简单、高效且低成本的预训练、微调与服务，惠及大众基于 Ray 架构设计的覆盖大语言模型（LLM）完整生命周期的解决方案byzer-llm

官网：https://github.com/allwefantasy/byzer-llm

手册：https://github.com/allwefantasy/byzer-llm/blob/master/docs/zh/001_%E4%B8%80%E4%B8%AA%E5%8A%AA%E5%8A%9B%E6%88%90%E4%B8%BA%E5%A4%A7%E6%A8%A1%E5%9E%8B%E7%BC%96%E7%A8%8B%E6%8E%A5%E5%8F%A3%E7%9A%84%E7%A5%9E%E5%A5%87Python%E5%BA%93.md

Byzer-LLM 基于 Ray 技术构建，是一款覆盖大语言模型（LLM）完整生命周期的解决方案，包括预训练、微调、部署及推理服务等阶段。

Byzer-LLM 的独特之处在于：

  1. 全生命周期管理：支持预训练、微调、部署和推理服务全流程
  2. 兼容 Python/SQL API 接口
  3. 基于 Ray 架构设计，便于轻松扩展

名词解释 

SaaS 模型 ：把 LLM/AI能力封装成云服务，面向企业或个人以订阅/按调用次数收费，的AI大模型模型，也就是我们常说的AI模型调用

安装使用

直接pip安装

pip install byzer-llm

启动ray服务

ray start --head

 服务启动，提示

Local node IP: 192.168.0.95
/home/skywalk/minipy312/lib/python3.12/site-packages/ray/thirdparty_files/psutil/__init__.py:2017: RuntimeWarning: shared, active, inactive memory stats couldn't be determined and were set to 0ret = _psplatform.virtual_memory()--------------------
Ray runtime started.
--------------------Next stepsTo add another node to this Ray cluster, runray start --address='192.168.0.95:6379'To connect to this Ray cluster:import rayray.init()To submit a Ray job using the Ray Jobs CLI:RAY_ADDRESS='http://127.0.0.1:8265' ray job submit --working-dir . -- python my_script.py

直接安装Auto-Coder以便安装byzer-llm

手册认为直接pip安装 byzer-llm，后面的配置部分会比较麻烦，不如直接安装Auto-Coder，会自动配置好：这样一起执行即可

pip install pip -U
pip install -U auto-coder
ray start --head

输出

ray start --head
Enable usage stats collection? This prompt will auto-proceed in 10 seconds to avoid blocking cluster startup. Confirm [Y/n]:
Usage stats collection is enabled. To disable this, add `--disable-usage-stats` to the command that starts the cluster, or run the following command: `ray disable-usage-stats` before starting the cluster. See https://docs.ray.io/en/master/cluster/usage-stats.html for more details.Local node IP: 172.25.183.186--------------------
Ray runtime started.
--------------------Next stepsTo add another node to this Ray cluster, runray start --address='172.25.183.186:6379'To connect to this Ray cluster:import rayray.init()To submit a Ray job using the Ray Jobs CLI:RAY_ADDRESS='http://127.0.0.1:8265' ray job submit --working-dir . -- python my_script.pySee https://docs.ray.io/en/latest/cluster/running-applications/job-submission/index.htmlfor more information on submitting Ray jobs to the Ray cluster.To terminate the Ray runtime, runray stopTo view the status of the cluster, useray statusTo monitor and debug Ray, view the dashboard at127.0.0.1:8265If connection to the dashboard fails, check your firewall settings and network configuration.

事情变的有趣起来，现在让我们组个ray集群吧

加入ray

前面在192.168.1.5启动了ray服务器，这里再加入一台机器

ray start --address='192.168.1.5:6379'

加入完成：

ray status
======== Autoscaler status: 2025-11-01 19:16:47.050543 ========
Node status
---------------------------------------------------------------
Active:1 node_dfa56c248840a471c7b0be2e7dbeb3fb28041a6010a6887cabf07c791 node_f32f6c64e9df213e0c403f833f54d28a0493fe8d800284f46e9f878c
Pending:(no pending nodes)
Recent failures:(no failures)Resources
---------------------------------------------------------------
Total Usage:0.0/28.0 CPU0.0/1.0 GPU0B/23.64GiB memory0B/10.13GiB object_store_memoryTotal Constraints:(no request_resources() constraints)
Total Demands:(no resource demands)

应用

使用byzer-llm 做LLM大模型中转

这个先略，不太会

使用命令：

byzerllm deploy --pretrained_model_type saas/openai \
--cpus_per_worker 0.001 \
--gpus_per_worker 0 \
--num_workers 3 \
--infer_params  saas.api_key=${MODEL_OPENAI_TOKEN} saas.model=gpt-3.5-turbo-0125 \
--model gpt3_5_chat

使用byzer-llm启动本地大模型

在另一个文档里进行记录

调试

报错server_ Failed to start the grpc server

  File "/home/skywalk/minipy312/lib/python3.12/site-packages/ray/_private/node.py", line 796, in _init_gcs_client
    raise RuntimeError(
RuntimeError: Failed to start GCS.  Last 1 lines of error files:
[2025-05-28 19:20:11,803 C 54659 54659] (gcs_server) grpc_server.cc:128:  Check failed: server_ Failed to start the grpc server. The specified port is 6379. This means that Ray's core components will not be able to function correctly. If the server startup error message is `Address already in use`, it indicates the server fails to start because the port is already used by other processes (such as --node-manager-port, --object-manager-port, --gcs-server-port, and ports between --min-worker-port, --max-worker-port). Try running sudo lsof -i :6379 to check if there are other processes listening to the port.

.Please check /tmp/ray/session_2025-05-28_19-20-11_612410_54651/logs/gcs_server.out for details. Last connection error: None

报错Session name session_2025-05-29_09-05-56_429060_59678 does not match persisted value

AssertionError: Session name session_2025-05-29_09-05-56_429060_59678 does not match persisted value b'session_2025-05-28_19-39-24_560590_55048'. Perhaps there was an error connecting to Redis.
清除/tmp/ray/ 

rm -rf /tmp/ray/

依旧报错，删除ray进程

# 杀掉所有 ray processes
ps aux | grep ray | grep -v grep | awk '{print $2}' | xargs kill -9# 杀掉残留 redis（Ray 会自带一个 redis-server）
ps aux | grep redis | grep -v grep | awk '{print $2}' | xargs kill -9

依旧报错

安装redis

sudo pkg install redis

To setup "redis" you need to edit the configuration file:
      /usr/local/etc/redis.conf

      To run redis from startup, add redis_enable="YES"
      in your /etc/rc.conf.
启动redis，依旧报错

先搁置

在FreeBSD的bash里，使用linux兼容安装的python3.12系统里，可以安装Auto-Coder，但是ray启动不了

启动报错：

  File "/home/skywalk/minipy312/lib/python3.12/site-packages/ray/_private/node.py", line 364, in __init__
    self.start_head_processes()
  File "/home/skywalk/minipy312/lib/python3.12/site-packages/ray/_private/node.py", line 1458, in start_head_processes
    self.start_gcs_server()
  File "/home/skywalk/minipy312/lib/python3.12/site-packages/ray/_private/node.py", line 1225, in start_gcs_server
    process_info = ray._private.services.start_gcs_server(
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/skywalk/minipy312/lib/python3.12/site-packages/ray/_private/services.py", line 1515, in start_gcs_server
    stdout_file = open(os.devnull, "w")
                  ^^^^^^^^^^^^^^^^^^^^^
PermissionError: [Errno 13] Permission denied: '/dev/null'

看了下/dev/null是有可写权限的，但是没办法啊

换到linux 兼容环境下吧

sudo chroot /compat/ubuntu22/ /bin/bash

然后安装

pip install pip -U
pip install -U auto-coder
ray start --head

启动ray报错Ray component worker_ports is trying to use a port number 12868 that is used by other components.

启动命令

ray start --head

    raise ValueError(
ValueError: Ray component worker_ports is trying to use a port number 12868 that is used by other components.
Port information: {'gcs': 'random', 'object_manager': 'random', 'node_manager': 'random', 'gcs_server': 6379, 'client_server': 10001, 'dashboard': 8265, 'dashboard_agent_grpc': 12868, 'dashboard_agent_http': 52365, 'runtime_env_agent': 33302, 'metrics_export': 63589, 'redis_shards': 'random', 'worker_ports': '9998 ports from 10002 to 19999'}
If you allocate ports, please make sure the same port is not used by multiple components.

问题不大，应该是被刚才的进程占用的端口号

在来一次，ray start --head ，哟，它启动了

ray start --head
Usage stats collection is enabled. To disable this, add `--disable-usage-stats` to the command that starts the cluster, or run the following command: `ray disable-usage-stats` before starting the cluster. See https://docs.ray.io/en/master/cluster/usage-stats.html for more details.Local node IP: 192.168.1.5
/home/skywalk/py312/lib/python3.12/site-packages/ray/thirdparty_files/psutil/__init__.py:2017: RuntimeWarning: shared, active, inactive memory stats couldn't be determined and were set to 0ret = _psplatform.virtual_memory()--------------------
Ray runtime started.
--------------------Next stepsTo add another node to this Ray cluster, runray start --address='192.168.1.5:6379'To connect to this Ray cluster:import rayray.init()To submit a Ray job using the Ray Jobs CLI:RAY_ADDRESS='http://127.0.0.1:8265' ray job submit --working-dir . -- python my_script.pySee https://docs.ray.io/en/latest/cluster/running-applications/job-submission/index.htmlfor more information on submitting Ray jobs to the Ray cluster.To terminate the Ray runtime, runray stopTo view the status of the cluster, useray statusTo monitor and debug Ray, view the dashboard at127.0.0.1:8265If connection to the dashboard fails, check your firewall settings and network configuration.

加入ray报错RuntimeError: Version mismatch

ray start --address='192.168.1.5:6379'

  File "/home/skywalk/py312/lib/python3.12/site-packages/ray/scripts/scripts.py", line 1164, in start
    node.check_version_info()
  File "/home/skywalk/py312/lib/python3.12/site-packages/ray/_private/node.py", line 454, in check_version_info
    ray._private.utils.check_version_info(
  File "/home/skywalk/py312/lib/python3.12/site-packages/ray/_private/utils.py", line 1569, in check_version_info
    raise RuntimeError(error_message)
RuntimeError: Version mismatch: The cluster was started with:
    Ray: 2.47.1
    Python: 3.12.9
This process on node 172.25.183.186 was started with:
    Ray: 2.47.1
    Python: 3.12.3

升级本机3.12.3到3.12.9，其实应该说重装才对，安装pyenv

curl https://pyenv.run | bash

安装python3.12.9

pyenv install 3.12.9

安装Auto-Coder

​pip install pip -U
pip install -U auto-coder

加入ray

ray start --address='192.168.1.5:6379'

加入ok，查看一下

ray status
======== Autoscaler status: 2025-11-01 19:16:47.050543 ========
Node status
---------------------------------------------------------------
Active:1 node_dfa56c248840a471c7b0be2e7dbeb3fb28041a6010a6887cabf07c791 node_f32f6c64e9df213e0c403f833f54d28a0493fe8d800284f46e9f878c
Pending:(no pending nodes)
Recent failures:(no failures)Resources
---------------------------------------------------------------
Total Usage:0.0/28.0 CPU0.0/1.0 GPU0B/23.64GiB memory0B/10.13GiB object_store_memoryTotal Constraints:(no request_resources() constraints)
Total Demands:(no resource demands)

http://www.dtcms.com/a/556718.html

相关文章：

  • 软件测试-BUG篇
  • 写入瓶颈到削峰填谷：基于 Redis 与 MySQL 的高并发写入工程化方案
  • 一些常见的编程软件
  • Nginx Stream模块开发：TCP/UDP代理扩展
  • 开网站做外贸中国做网站正邦
  • K8s Dashboard运维技巧全面经验总结
  • 合肥创业网从百万到千万 网站怎么优化
  • Flutter boost权威指南
  • 人工智能、机器学习与神经网络：解锁智能时代的核心密码
  • Kubernetes 上的 GitLab + ArgoCD 实践（三）：使用 ArgoCD 打通 CD 流程
  • spark-SQL学习
  • SSM基于网络安全维护的机房设备管理19rya(程序+源码+数据库+调试部署+开发环境)带论文文档1万字以上，文末可获取，系统界面在最后面。
  • ProcessLifecycleOwner 完全指南：优雅监听应用前后台状态
  • html css js网页制作成品——珠帘玉幕HTML+CSS网页设计（4页）附源码
  • 开启RN之旅——小试牛刀
  • Unity使用PP-MattingV2实现人像分割
  • 智能模型对齐（一致性）alignment
  • VSCode SSH远程连接失败 最速解决方案
  • 网站开发旅游前台模板临海建设规划局网站
  • 加载YOLO模型，处理mp4视频
  • 基于 GEE 利用 GHSL（100m）数据的区域建成区时空变化量化分析
  • day22_用户授权 头像上传
  • 网站识别爬虫（包括以浏览器插件形式运行的爬虫）主要通过分析请求特征、行为模式等差异来区分人类用户和自动化程序
  • 网站建设费用IPseo官网优化详细方法
  • 汽车OTA CDN HTTPS MQTT OCSP
  • python异步编程 -- 深入理解事件循环event-loop
  • 京津冀工业智能体赋能：重构产业链升级新篇章
  • AIGEO系统到底是什么？
  • 日志系统的介绍及前置技术
  • 安居客做网站广州建设网站公司哪家好