diff --git a/images/1-wiki.png b/images/1-wiki.png deleted file mode 100644 index 5eba0fc..0000000 Binary files a/images/1-wiki.png and /dev/null differ diff --git a/images/2-wiki.png b/images/2-wiki.png deleted file mode 100644 index 35a1d53..0000000 Binary files a/images/2-wiki.png and /dev/null differ diff --git a/images/3-wiki.png b/images/3-wiki.png deleted file mode 100644 index 9570bbf..0000000 Binary files a/images/3-wiki.png and /dev/null differ diff --git a/images/4-wiki.png b/images/4-wiki.png deleted file mode 100644 index 931e240..0000000 Binary files a/images/4-wiki.png and /dev/null differ diff --git a/images/5-wiki.png b/images/5-wiki.png deleted file mode 100644 index 78ca9af..0000000 Binary files a/images/5-wiki.png and /dev/null differ diff --git a/images/LLM-structure-moe.png b/images/LLM-structure-moe.png deleted file mode 100644 index 4588477..0000000 Binary files a/images/LLM-structure-moe.png and /dev/null differ diff --git a/images/LLM-structure.png b/images/LLM-structure.png deleted file mode 100755 index bbd93dd..0000000 Binary files a/images/LLM-structure.png and /dev/null differ diff --git a/images/and_huggingface.png b/images/and_huggingface.png deleted file mode 100644 index c234f8a..0000000 Binary files a/images/and_huggingface.png and /dev/null differ diff --git a/images/and_modelscope.png b/images/and_modelscope.png deleted file mode 100644 index 1e46da4..0000000 Binary files a/images/and_modelscope.png and /dev/null differ diff --git a/images/compare_radar.png b/images/compare_radar.png deleted file mode 100644 index 345d9f6..0000000 Binary files a/images/compare_radar.png and /dev/null differ diff --git a/images/dataset.jpg b/images/dataset.jpg deleted file mode 100644 index 7dfc366..0000000 Binary files a/images/dataset.jpg and /dev/null differ diff --git a/images/gpt3_config.png b/images/gpt3_config.png deleted file mode 100644 index 121bc29..0000000 Binary files a/images/gpt3_config.png and /dev/null differ diff --git a/images/logo.png b/images/logo.png deleted file mode 100644 index 14d8301..0000000 Binary files a/images/logo.png and /dev/null differ diff --git a/images/logo2.png b/images/logo2.png deleted file mode 100644 index 9a0b3e2..0000000 Binary files a/images/logo2.png and /dev/null differ diff --git a/images/minimind2.gif b/images/minimind2.gif deleted file mode 100644 index 43c9cd1..0000000 Binary files a/images/minimind2.gif and /dev/null differ diff --git a/images/pre_512_loss.png b/images/pre_512_loss.png deleted file mode 100644 index 3da0be5..0000000 Binary files a/images/pre_512_loss.png and /dev/null differ diff --git a/images/pre_768_loss.png b/images/pre_768_loss.png deleted file mode 100644 index e00b23c..0000000 Binary files a/images/pre_768_loss.png and /dev/null differ diff --git a/images/sft_512_loss.png b/images/sft_512_loss.png deleted file mode 100644 index 40b86bc..0000000 Binary files a/images/sft_512_loss.png and /dev/null differ diff --git a/images/sft_768_loss.png b/images/sft_768_loss.png deleted file mode 100644 index 5ea6c97..0000000 Binary files a/images/sft_768_loss.png and /dev/null differ diff --git a/trainer/train_pretrain.py b/trainer/train_pretrain.py index f0cfc58..eb45963 100644 --- a/trainer/train_pretrain.py +++ b/trainer/train_pretrain.py @@ -119,7 +119,7 @@ if __name__ == "__main__": parser.add_argument("--out_dir", type=str, default="../out") # 若要以最快速度实现zero则epochs设置为1轮;否则应当利用有限的数据训练2~6个epochs。 parser.add_argument("--epochs", type=int, default=1) - parser.add_argument("--batch_size", type=int, default=32) + parser.add_argument("--batch_size", type=int, default=16) parser.add_argument("--learning_rate", type=float, default=5e-4) parser.add_argument("--device", type=str, default="cuda:0" if torch.cuda.is_available() else "cpu") parser.add_argument("--dtype", type=str, default="bfloat16")