With Transformers it’s very easy to load any model in 4 or 8-bit, quantizing them on the fly with bitsandbytes primitives.
Please review the bitsandbytes section in the Accelerate docs.
Details about the BitsAndBytesConfig can be found here.
Beware: bf16 is the optimal compute data type!
If your hardware supports it, bf16 is the optimal compute dtype. The default is float32 for backward compatibility and numerical stability. float16 often leads to numerical instabilities, but bfloat16 provides the benefits of both worlds: numerical stability equivalent to float32, but combined with the memory footprint and significant computation speedup of a 16-bit data type. Therefore, be sure to check if your hardware supports bf16 and configure it using the bnb_4bit_compute_dtype parameter in BitsAndBytesConfig:
import torch
from transformers import BitsAndBytesConfig
quantization_config = BitsAndBytesConfig(load_in_4bit=True, bnb_4bit_compute_dtype=torch.bfloat16)With PEFT, you can use QLoRA out of the box with LoraConfig and a 4-bit base model.
Please review the bitsandbytes section in the Accelerate docs.
Bitsandbytes is also easily usable from within Accelerate.
Please review the bitsandbytes section in the Accelerate docs.
You can use any of the 8-bit and/or paged optimizers by simple passing them to the transformers.Trainer class on initialization.All bnb optimizers are supported by passing the correct string in TrainingArguments’s optim attribute - e.g. (paged_adamw_32bit).
See the official API docs for reference.
Here we point out to relevant doc sections in transformers / peft / Trainer + very briefly explain how these are integrated:
e.g. for transformers state that you can load any model in 8-bit / 4-bit precision, for PEFT, you can use QLoRA out of the box with LoraConfig + 4-bit base model, for Trainer: all bnb optimizers are supported by passing the correct string in TrainingArguments’s optim attribute - e.g. (paged_adamw_32bit):