=========================================
Check hf-internal-testing/tiny-random-bart ...
--------------------------Checking logits match--------------------------
Flax logits shape: (2, 64, 1000), PyTorch logits shape: torch.Size([2, 64, 1000])
✅ Difference between Flax and PyTorch is 8.940696716308594e-08 (< 0.01)
--------------------------Checking losses match--------------------------
Flax loss: 6.923163414001465, PyTorch loss: 6.923163414001465
✅ Difference between Flax and PyTorch is 0.0 (< 0.01)
--------------------------Checking gradients match--------------------------
❌ Layer ('final_logits_bias',) has PT grad norm 0.0 and flax grad norm 0.09163407981395721.
--------------------------Checking rel gradients match--------------------------
❌ Layer ('final_logits_bias',) has PT grad norm 0.0 and flax grad norm 0.09163407981395721.
❌ Layer ('model', 'decoder', 'layers', '0', 'encoder_attn', 'k_proj', 'bias') has PT grad norm 9.3028212357852e-14 and flax grad norm 1.6552796459901042e-13.
...
=========================================