Run log · LoRA smoke test ·

Qwen3.5-4B, sequence 512

2026-11-22_smoke-train_Qwen3.5-4B_seq512

A smoke test checks that training runs on this card, how fast, and with how much memory. It uses eight placeholder examples. The adapter is thrown away, and the falling loss only shows the pipeline can memorise them.

Recorded fields

Model
unsloth/Qwen3.5-4B
Loaded in 4-bit
no, 16-bit
Layers offloaded to system RAM
no
Training steps
100
Sequence length
512 tokens
LoRA rank
16
Training time
159 s
Seconds per step, averaged over every step including the first compile step
1.59 s
VRAM after load
8.47 GiB
Peak VRAM
8.81 GiB
First logged loss
2.241
Last logged loss
0.024
Adapter size on disk
100.1 MB
PyTorch
2.14.1+cu130
Hardware
RTX 4070 Ti, 12 GB

Logged loss

StepLoss
102.241
200.8889
300.1797
400.04944
500.03239
600.02628
700.02644
800.02936
900.02775
1000.0244