Can You Run TinyLlama 1.1B LoRA? Every GPU, Tested

A model that doesn't fit doesn't run slowly, it doesn't run. Here's every card from our bench, sorted by measured speed.

Weights: TinyLlama/TinyLlama-1.1B-Chat-v1.0

GPUs that run TinyLlama 1.1B LoRA (11)

GPUSpeedVRAMBasis
NVIDIA B30026938.2 train tok/s288GB✓ Measured
NVIDIA RTX PRO 6000 Blackwell Workstation Edition18556.4 train tok/s96GB✓ Measured
NVIDIA B20017993.5 train tok/s192GB✓ Measured
NVIDIA H100 80GB HBM317522 train tok/s80GB✓ Measured
NVIDIA L40S16408.1 train tok/s48GB✓ Measured
NVIDIA H20016222.4 train tok/s141GB✓ Measured
NVIDIA A100 80GB SXM49553.3 train tok/s80GB✓ Measured
NVIDIA A100 40GB SXM47471 train tok/s40GBEst.
NVIDIA A10G5499.6 train tok/s24GB✓ Measured
NVIDIA L44756 train tok/s24GB✓ Measured
NVIDIA T41520.1 train tok/s16GB✓ Measured

Quick answers

What is the fastest GPU for TinyLlama 1.1B LoRA?

NVIDIA B300 is the fastest card we've measured running TinyLlama 1.1B LoRA, at 26938.2 train tok/s.

What is the cheapest GPU that can run TinyLlama 1.1B LoRA?

NVIDIA T4 ($2,299 MSRP) is the cheapest card on our bench that runs TinyLlama 1.1B LoRA, at 1520.1 train tok/s.

← All models · AI GPU rankings · How we benchmark