my code: return tokenizer(list(dataset['sentense']), padding = True, truncation = True, max_length = 128 ) training_args = TrainingArguments( output_dir='./results', # output directory save_total_limit…

If by smart batching you mean grouping together samples of the same length, it is implemented in the Trainer by adding group_by_length=True in your TrainingArguments.

Are dynamic padding and smart batching in the library?

Beginners

tomroth1001 November 29, 2021, 3:52am 5

Just wondering - how would you usually do this without using the Trainer?

1 Like

Topic		Replies	Views
Padding in datasets 🤗Datasets	6	5080	October 21, 2021
Batch tensor creation error when finetuning gpt2 🤗Transformers	2	445	March 21, 2023
Data sampler based on number of tokens 🤗Transformers	0	736	February 4, 2022
Dynamic Padding not working for Custom Dataset 🤗Datasets	5	4006	February 9, 2022
How can I combine encode and padding into __call__ during training if I want to pad to batch longest 🤗Transformers	0	619	October 3, 2022

Are dynamic padding and smart batching in the library?

Related topics