• Joined on 2024-10-09
carry pushed to master at carry/gzhu-biyesheji 2025-04-15 07:26:09 +00:00
d0aebd17fa refactor(global_var): 重构全局变量管理
d9abf08184 fix(frontend): 修复表格选择事件的行数据获取问题
a27a1ab079 refactor(frontend): 重构训练页面布局并优化用户界面
Compare 3 commits »
carry pushed to master at carry/gzhu-biyesheji 2025-04-14 15:52:05 +00:00
aa758e3c2a feat(train_page): 添加 TensorBoard 可视化
carry pushed to master at carry/gzhu-biyesheji 2025-04-14 09:07:16 +00:00
664944f0c5 feat(frontend): 优化 TensorBoard 端口占用问题
carry pushed to master at carry/gzhu-biyesheji 2025-04-14 09:00:59 +00:00
9298438f98 feat(train_page): 启动 TensorBoard 进程并确保训练结束后终止
carry pushed to master at carry/gzhu-biyesheji 2025-04-14 08:46:47 +00:00
4f7926aec6 feat(train_page): 实现训练目录自动递增功能
carry pushed to master at carry/gzhu-biyesheji 2025-04-14 08:32:02 +00:00
148f4afb25 fix(main): 修复unsloth没有最先导入的问题,删除了重复的导入语句
11a3039775 fix(train_page): 修正模型训练保存路径
a4289815ba build: 添加 tensorboard 依赖
088067d335 train: 更新模型训练功能和日志记录方式
Compare 4 commits »
carry pushed to master at carry/gzhu-biyesheji 2025-04-14 07:18:31 +00:00
9fb31c46c8 feat(train): 添加训练过程中的日志记录和 loss 可视化功能
4f09823123 refactor(tools): 优化 train_model 函数定义
Compare 2 commits »
carry pushed to master at carry/gzhu-biyesheji 2025-04-14 06:17:25 +00:00
1a2ca3e244 refactor(train): 重构训练功能并移至新模块
carry pushed to master at carry/gzhu-biyesheji 2025-04-13 13:50:05 +00:00
bb1d8fbd38 feat(train_page): 添加训练 Loss 曲线显示功能
4558929c52 fix: 调整了import的顺序,让unsloth最先import以提高性能
Compare 2 commits »
carry pushed to master at carry/gzhu-biyesheji 2025-04-13 13:14:29 +00:00
0722748997 feat(train_page): 添加 LoRA 秩动态输入功能
e08f0059bb feat(train_page): 优化训练过程以专注于响应生成
Compare 2 commits »
carry pushed to master at carry/gzhu-biyesheji 2025-04-13 12:29:39 +00:00
6d1fecbdac build: 更新依赖版本并添加新依赖
carry pushed to master at carry/gzhu-biyesheji 2025-04-12 17:56:19 +00:00
79d3eb153c refactor(train_page): 优化训练页面布局和功能
80dae7c6e2 fix(global_var):修复_workdir非全局变量的bug
Compare 2 commits »
carry pushed to master at carry/gzhu-biyesheji 2025-04-12 17:04:34 +00:00
2d39b91764 feat(train_page): 添加模型训练超参数配置功能
5094febcb4 refactor(global_var): 重构全局变量管理并添加工作目录功能
Compare 2 commits »
carry pushed to master at carry/gzhu-biyesheji 2025-04-12 10:42:46 +00:00
eeee68dbd1 chore: 更新 .gitignore 文件
carry pushed to master at carry/gzhu-biyesheji 2025-04-12 10:42:31 +00:00
539e14d39c feat(frontend): 完成了前端微调的代码逻辑
carry pushed to master at carry/gzhu-biyesheji 2025-04-12 10:10:47 +00:00
9784f2aed3 fix(tools): 修正__init__.py使得model.py正确导入
611904cef9 feat(frontend): 添加数据集选择功能到训练页面
8a9a080745 refactor(tools): 移除未使用的导入语句
Compare 3 commits »
carry pushed to master at carry/gzhu-biyesheji 2025-04-11 11:00:34 +00:00
a23ad88769 fix(frontend): 修复删除提示功能中的数据库连接错误
carry pushed to master at carry/gzhu-biyesheji 2025-04-11 10:48:22 +00:00
83427aaaba feat(frontend): 增加超参数设置并优化聊天页面布局
61672021ef fix(frontend): 修复聊天页面并的流式回复
Compare 2 commits »
carry pushed to master at carry/gzhu-biyesheji 2025-04-11 10:08:46 +00:00
fb6157af05 feat(frontend): 初步实现聊天页面的智能回复功能
f655936741 refactor(global_var): 重构全局变量初始化方法
ab7897351a fix(global_var): 修复全局变量多文件多副本的不统一问题
216bfe39ae feat(tools): 添加格式化对话数据的函数
Compare 4 commits »
carry pushed to master at carry/gzhu-biyesheji 2025-04-10 16:14:45 +00:00
0fa2b51a79 refactor(frontend): 优化模型管理页面的交互和显示
cbb3a09dd8 feat(tools): 添加模型名称获取函数
Compare 2 commits »