DOC Rename max_seq_length argument (#2862)

kaixuanliu · web-flow · commit 98a88c01a42b · 2025-10-23T18:55:33.000+02:00
Argument was renamed to max_length.
diff --git a/examples/alora_finetuning/README.md b/examples/alora_finetuning/README.md
@@ -32,7 +32,7 @@ trainer = Trainer(
     model=peft_model,
     train_dataset=dataset,
     dataset_text_field="text",
-    max_seq_length=2048,
+    max_length=2048,
     tokenizer=tokenizer,
     data_collator=data_collator,
 )
diff --git a/examples/bone_finetuning/README.md b/examples/bone_finetuning/README.md
@@ -28,7 +28,7 @@ peft_model.print_trainable_parameters()
 
 dataset = load_dataset("imdb", split="train[:1%]")
 
-training_args = SFTConfig(dataset_text_field="text", max_seq_length=128)
+training_args = SFTConfig(dataset_text_field="text", max_length=128)
 trainer = SFTTrainer(
     model=peft_model,
     args=training_args,
diff --git a/examples/corda_finetuning/README.md b/examples/corda_finetuning/README.md
@@ -109,7 +109,7 @@ preprocess_corda(model, lora_config, run_model=run_model)
 peft_model = get_peft_model(model, lora_config)
 peft_model.print_trainable_parameters()
 
-training_args = SFTConfig(dataset_text_field="text", max_seq_length=128)
+training_args = SFTConfig(dataset_text_field="text", max_length=128)
 trainer = SFTTrainer(
     model=peft_model,
     args=training_args,
diff --git a/examples/delora_finetuning/README.md b/examples/delora_finetuning/README.md
@@ -26,7 +26,7 @@ peft_model.print_trainable_parameters()
 
 dataset = load_dataset("imdb", split="train[:1%]")
 
-training_args = SFTConfig(dataset_text_field="text", max_seq_length=128)
+training_args = SFTConfig(dataset_text_field="text", max_length=128)
 trainer = SFTTrainer(
     model=peft_model,
     args=training_args,
@@ -52,7 +52,7 @@ peft_model = PeftModel.from_pretrained(model, "delora-llama-3-8b")
 ## Advanced Usage
 In this script the default DeLoRA layers are the query and value layers of the Llama model. Adding adapters on more layers will increase memory usage. If you wish to choose a different set of layers for DeLoRA to be applied on, you can simply define it using:
 ```bash
-python examples/delora_finetuning/delora_finetuning.py --base_model meta-llama/Meta-Llama-3-8B --delora_target_modules "q_proj,k_proj,v_proj,o_proj" 
+python examples/delora_finetuning/delora_finetuning.py --base_model meta-llama/Meta-Llama-3-8B --target_modules "q_proj,k_proj,v_proj,o_proj" 
 ```
 
 Using different lambdas for different layers is also possible by setting `lambda_pattern`.
@@ -74,7 +74,7 @@ python delora_finetuning.py \
     --rank 32 \
     --delora_lambda 15 \
     --module_dropout 0.1 \
-    --delora_target_modules "q_proj,v_proj" \
+    --target_modules "q_proj,v_proj" \
     --hub_model_id "YOUR_HF_REPO" \
     --push_to_hub
 ```
diff --git a/examples/dora_finetuning/README.md b/examples/dora_finetuning/README.md
@@ -24,7 +24,7 @@ trainer = transformers.Trainer(
     model=peft_model,
     train_dataset=dataset,
     dataset_text_field="text",
-    max_seq_length=2048,
+    max_length=2048,
     tokenizer=tokenizer,
 )
 trainer.train()
diff --git a/examples/lorafa_finetune/README.md b/examples/lorafa_finetune/README.md
@@ -40,7 +40,7 @@ trainer = transformers.Trainer(
     model=peft_model,
     train_dataset=dataset,
     dataset_text_field="text",
-    max_seq_length=2048,
+    max_length=2048,
     processing_class=tokenizer,
     optimizers=(optimizer, None),
 )
diff --git a/examples/miss_finetuning/README.md b/examples/miss_finetuning/README.md
@@ -36,7 +36,7 @@ peft_model.print_trainable_parameters()
 
 dataset = load_dataset("imdb", split="train[:1%]")
 
-training_args = SFTConfig(dataset_text_field="text", max_seq_length=128)
+training_args = SFTConfig(dataset_text_field="text", max_length=128)
 trainer = SFTTrainer(
     model=peft_model,
     args=training_args,
diff --git a/examples/olora_finetuning/README.md b/examples/olora_finetuning/README.md
@@ -18,7 +18,7 @@ lora_config = LoraConfig(
     init_lora_weights="olora"
 )
 peft_model = get_peft_model(model, lora_config)
-training_args = SFTConfig(dataset_text_field="text", max_seq_length=128)
+training_args = SFTConfig(dataset_text_field="text", max_length=128)
 trainer = SFTTrainer(
     model=peft_model,
     train_dataset=dataset,
diff --git a/examples/pissa_finetuning/README.md b/examples/pissa_finetuning/README.md
@@ -23,7 +23,7 @@ peft_model.print_trainable_parameters()
 
 dataset = load_dataset("imdb", split="train[:1%]")
 
-training_args = SFTConfig(dataset_text_field="text", max_seq_length=128)
+training_args = SFTConfig(dataset_text_field="text", max_length=128)
 trainer = SFTTrainer(
     model=peft_model,
     args=training_args,
diff --git a/examples/randlora_finetuning/README.md b/examples/randlora_finetuning/README.md
@@ -20,7 +20,7 @@ trainer = transformers.Trainer(
     model=peft_model,
     train_dataset=dataset,
     dataset_text_field="text",
-    max_seq_length=2048,
+    max_length=2048,
     processing_class=tokenizer,
 )
 trainer.train()
diff --git a/examples/road_finetuning/README.md b/examples/road_finetuning/README.md
@@ -26,7 +26,7 @@ trainer = transformers.Trainer(
     model=peft_model,
     train_dataset=dataset,
     dataset_text_field="text",
-    max_seq_length=2048,
+    max_length=2048,
     tokenizer=tokenizer,
 )
 trainer.train()
diff --git a/examples/shira_finetuning/README.md b/examples/shira_finetuning/README.md
@@ -18,7 +18,7 @@ shira_config = ShiraConfig(
     r=32,
 )
 peft_model = get_peft_model(model, shira_config)
-training_args = SFTConfig(dataset_text_field="text", max_seq_length=128)
+training_args = SFTConfig(dataset_text_field="text", max_length=128)
 trainer = SFTTrainer(
     model=peft_model,
     train_dataset=dataset,
diff --git a/examples/waveft_finetuning/README.md b/examples/waveft_finetuning/README.md
@@ -20,7 +20,7 @@ waveft_config = WaveFTConfig(
     n_frequency=2592,
 )
 peft_model = get_peft_model(model, waveft_config)
-training_args = SFTConfig(dataset_text_field="text", max_seq_length=128)
+training_args = SFTConfig(dataset_text_field="text", max_length=128)
 trainer = SFTTrainer(
     model=peft_model,
     train_dataset=dataset,

Original file line number	Diff line number	Diff line change
`@@ -32,7 +32,7 @@ trainer = Trainer(`
`32`	`32`	`model=peft_model,`
`33`	`33`	`train_dataset=dataset,`
`34`	`34`	`dataset_text_field="text",`
`35`		`- max_seq_length=2048,`
	`35`	`+ max_length=2048,`
`36`	`36`	`tokenizer=tokenizer,`
`37`	`37`	`data_collator=data_collator,`
`38`	`38`	`)`
Original file line number	Diff line number	Diff line change
`@@ -24,7 +24,7 @@ trainer = transformers.Trainer(`
`24`	`24`	`model=peft_model,`
`25`	`25`	`train_dataset=dataset,`
`26`	`26`	`dataset_text_field="text",`
`27`		`- max_seq_length=2048,`
	`27`	`+ max_length=2048,`
`28`	`28`	`tokenizer=tokenizer,`
`29`	`29`	`)`
`30`	`30`	`trainer.train()`
Original file line number	Diff line number	Diff line change
`@@ -40,7 +40,7 @@ trainer = transformers.Trainer(`
`40`	`40`	`model=peft_model,`
`41`	`41`	`train_dataset=dataset,`
`42`	`42`	`dataset_text_field="text",`
`43`		`- max_seq_length=2048,`
	`43`	`+ max_length=2048,`
`44`	`44`	`processing_class=tokenizer,`
`45`	`45`	`optimizers=(optimizer, None),`
`46`	`46`	`)`
Original file line number	Diff line number	Diff line change
`@@ -18,7 +18,7 @@ lora_config = LoraConfig(`
`18`	`18`	`init_lora_weights="olora"`
`19`	`19`	`)`
`20`	`20`	`peft_model = get_peft_model(model, lora_config)`
`21`		`-training_args = SFTConfig(dataset_text_field="text", max_seq_length=128)`
	`21`	`+training_args = SFTConfig(dataset_text_field="text", max_length=128)`
`22`	`22`	`trainer = SFTTrainer(`
`23`	`23`	`model=peft_model,`
`24`	`24`	`train_dataset=dataset,`
Original file line number	Diff line number	Diff line change
`@@ -20,7 +20,7 @@ trainer = transformers.Trainer(`
`20`	`20`	`model=peft_model,`
`21`	`21`	`train_dataset=dataset,`
`22`	`22`	`dataset_text_field="text",`
`23`		`- max_seq_length=2048,`
	`23`	`+ max_length=2048,`
`24`	`24`	`processing_class=tokenizer,`
`25`	`25`	`)`
`26`	`26`	`trainer.train()`