aifeifei798
/

Fragmented-Training

Text Generation

text-generation-inference

fragmented-training

burden-based-learning

logic-restoration

Model card Files Files and versions

aifeifei798 commited on Jan 24

Commit

efc206b

·

verified ·

1 Parent(s): 84adeae

Update README.md

Files changed (1) hide show

README.md +1 -1

README.md CHANGED Viewed

@@ -669,7 +669,7 @@ trainer = SFTTrainer(
         dataloader_num_workers=4,
         dataloader_pin_memory=True,
         warmup_steps=25,
-        num_train_epochs=10,
         learning_rate=2e-5,
         fp16=not torch.cuda.is_bf16_supported(),
         bf16=torch.cuda.is_bf16_supported(),

         dataloader_num_workers=4,
         dataloader_pin_memory=True,
         warmup_steps=25,
+        num_train_epochs=3,
         learning_rate=2e-5,
         fp16=not torch.cuda.is_bf16_supported(),
         bf16=torch.cuda.is_bf16_supported(),