Trained for 8 epochs and 25000 steps.

Trained with datasets ['flux2-text-embeds', 'dlay-subject-data']
Learning rate 1.0, batch size 1, and 1 gradient accumulation steps.
Trained with None prediction type and rescaled_betas_zero_snr=False
Using 'trailing' timestep spacing.
Base model: black-forest-labs/FLUX.2-dev
VAE: black-forest-labs/FLUX.2-dev

pytorch_lora_weights.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ae2c7835c662802bec9dc725594228e152a1fdb130f0260d22462ce82e3dae51
-size 444423248
+oid sha256:b7402571372a6b7c7b6138c4d19e68de3c6ffbebc4bd255d26a84f4d98085818
+size 444422904