GPU Memory Estimator for QLoRA / LoRA / Transformers
Project description
qmemcalc
Quantized Memory Calculator for Transformers (QLoRA / LoRA / full fine-tuning)
Features
- Estimate GPU memory for FP32/FP16/BF16
- Supports 4-bit / 8-bit quantization
- LoRA rank or fraction support
- Extended optimizer support
- Gradient checkpointing
- CLI for quick estimation
Installation
pip install qmemcalc
Python API Usage
from qmemcalc import estimate_memory
result = estimate_memory(
model_name="sshleifer/tiny-gpt2",
batch_size=2,
seq_len=16,
lora_r=4,
precision="fp16",
quantization="4bit"
)
print(result)
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
qmemcalc-0.1.0.tar.gz
(4.4 kB
view details)
Built Distribution
Filter files by name, interpreter, ABI, and platform.
If you're not sure about the file name format, learn more about wheel file names.
Copy a direct link to the current filters
File details
Details for the file qmemcalc-0.1.0.tar.gz.
File metadata
- Download URL: qmemcalc-0.1.0.tar.gz
- Upload date:
- Size: 4.4 kB
- Tags: Source
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.10.18
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
e4e90a869bd453b18fedd7e6898beddc4f87a8caecb50f1e34b9353dd18f6211
|
|
| MD5 |
6f24656b4687196e0c824171bcc1781b
|
|
| BLAKE2b-256 |
d237a841e4e6cabc34ba77c54a7ca679659499085d97ca64833cd5bf86cf9e77
|
File details
Details for the file qmemcalc-0.1.0-py3-none-any.whl.
File metadata
- Download URL: qmemcalc-0.1.0-py3-none-any.whl
- Upload date:
- Size: 4.7 kB
- Tags: Python 3
- Uploaded using Trusted Publishing? No
- Uploaded via: twine/6.2.0 CPython/3.10.18
File hashes
| Algorithm | Hash digest | |
|---|---|---|
| SHA256 |
280a871132bc14de79b47084c6997a8d65481868c9143840c98f40e4f14a41fe
|
|
| MD5 |
d3cb5d11817c7e82607926823cf9d739
|
|
| BLAKE2b-256 |
da7f1021336cd33752df71e8513fe476052f30043ce08468e455f4d400e9538e
|