github
/
minimind-v
mirror of https://github.com/jingyaogong/minimind-v.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

[feat] update readme

pull/107/head
jingyaogong 2025-11-11 11:12:09 +08:00
parent 92ca9959bb
commit d698d98dc5
2 changed files with 6 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
README.md
View File

@ -232,7 +232,7 @@ unzip sft_images.zip && rm sft_images.zip
```bash
# 基础训练命令（从LLM权重开始，仅训练vision_proj）
python trainer/train_pretrain_vlm.py --epochs 4 --from_weight llm
python train_pretrain_vlm.py --epochs 4 --from_weight llm
```
> 执行预训练，得到 `pretrain_vlm_*.pth` 作为预训练的输出权重（其中*为模型的dimension，默认为512）
@ -242,7 +242,7 @@ python trainer/train_pretrain_vlm.py --epochs 4 --from_weight llm
```bash
# 基础训练命令（从预训练权重开始，全参数微调）
python trainer/train_sft_vlm.py --epochs 2 --from_weight pretrain_vlm
python train_sft_vlm.py --epochs 2 --from_weight pretrain_vlm
```
> 执行监督微调，得到 `sft_vlm_*.pth` 作为指令微调的输出权重
@ -258,7 +258,7 @@ python trainer/train_sft_vlm.py --epochs 2 --from_weight pretrain_vlm
```bash
# 训练中断后，使用相同命令并添加 --from_resume 1
python trainer/train_sft_vlm.py --epochs 4 --from_resume 1
python train_sft_vlm.py --epochs 4 --from_resume 1
```
**参数说明：**

6
README_en.md
View File

@ -240,7 +240,7 @@ Please reserve about 5GB of space for the dataset. If there is insufficient spac
```bash
# Basic training command (start from LLM weights, train vision_proj only)
python trainer/train_pretrain_vlm.py --epochs 4 --from_weight llm
python train_pretrain_vlm.py --epochs 4 --from_weight llm
```
> Run pretraining to get `pretrain_vlm_*.pth` as the pretrained model's output weights (* represents the model
@ -250,7 +250,7 @@ python trainer/train_pretrain_vlm.py --epochs 4 --from_weight llm
```bash
# Basic training command (start from pretrain weights, full parameter fine-tuning)
python trainer/train_sft_vlm.py --epochs 2 --from_weight pretrain_vlm
python train_sft_vlm.py --epochs 2 --from_weight pretrain_vlm
```
> Perform supervised fine-tuning to get `sft_vlm_*.pth` as the output weights for the fine-tuned model.
@ -266,7 +266,7 @@ python trainer/train_sft_vlm.py --epochs 2 --from_weight pretrain_vlm
```bash
# To resume training after interruption, use the same command and add --from_resume 1
python trainer/train_sft_vlm.py --epochs 4 --from_resume 1
python train_sft_vlm.py --epochs 4 --from_resume 1
```
**Parameter Description:**