Question about GPU memory usage.

Hi, I tried to finetune a llama7b model with HQQ-LORA using dual GPUs.
I found that during "Loading & Quantizing Model Shards", the peak GPU memory usage acheved 35G. What's the problem?
the run command is:
```
export CUDA_VISIBLE_DEVICES=3,4
python train.py \
--world_size 2 \
--model_name /workspace/model/Llama-2-7b-chat-hf \
--gradient_accumulation_steps 2 \
--batch_size 1 \
--context_length 4096 \
--num_epochs 1 \
--sharding_strategy full_shard \
--precision bf16 \
--train_type hqq_lora \
--use_gradient_checkpointing true \
--use_cpu_offload true \
--dataset dummy \
--verbose true
```
Looking forward to your reply.

Start with train.py and reproduce the supplied dual-GPU command, focusing on the "Loading & Quantizing Model Shards" stage and its memory reporting. Trace the model-loading and quantization path to determine why peak usage reaches 35G; done means documenting whether that usage is expected and identifying the relevant configuration or behavior if it is not.

Written by the indexing model from the issue text.