runtime error

Exit code: 1. Reason: er downloads. [transformers] `torch_dtype` is deprecated! Use `dtype` instead! model.safetensors: 0%| | 0.00/16.0G [00:00<?, ?B/s] model.safetensors: 1%|▏ | 202M/16.0G [00:01<01:19, 199MB/s] model.safetensors: 4%|▍ | 618M/16.0G [00:02<00:47, 326MB/s] model.safetensors: 8%|▊ | 1.34G/16.0G [00:04<00:42, 347MB/s] model.safetensors: 13%|█▎ | 2.08G/16.0G [00:06<00:42, 331MB/s] model.safetensors: 13%|█▎ | 2.08G/16.0G [00:19<00:42, 331MB/s] model.safetensors: 16%|█▋ | 2.62G/16.0G [00:26<03:13, 69.2MB/s] model.safetensors: 71%|███████ | 11.3G/16.0G [00:29<00:09, 516MB/s]  model.safetensors: 95%|█████████▌| 15.2G/16.0G [00:30<00:01, 747MB/s] model.safetensors: 100%|██████████| 16.0G/16.0G [00:31<00:00, 516MB/s] Loading weights: 0%| | 0/2076 [00:00<?, ?it/s] Loading weights: 100%|██████████| 2076/2076 [00:00<00:00, 3055.66it/s] generation_config.json: 0%| | 0.00/208 [00:00<?, ?B/s] generation_config.json: 100%|██████████| 208/208 [00:00<00:00, 1.06MB/s] Some parameters are on the meta device because they were offloaded to the disk and cpu. Unwrapping Gemma4ClippableLinear for PEFT compatibility... Loading LoRA adapter... adapter_config.json: 0%| | 0.00/1.10k [00:00<?, ?B/s] adapter_config.json: 100%|██████████| 1.10k/1.10k [00:00<00:00, 5.54MB/s] Traceback (most recent call last): File "/app/app.py", line 37, in <module> model = PeftModel.from_pretrained(model, ADAPTER_MODEL, low_cpu_mem_usage=False) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/local/lib/python3.12/site-packages/peft/peft_model.py", line 522, in from_pretrained "safetensors_file": index[p]["safetensors_file"], ~~~~~^^^ KeyError: 'model.audio_tower.layers.0.feed_forward1.ffw_layer_1.weight'

Container logs: