{"instruction":"解析下面句子的意思","input":"年轻人说HHHACCC是什么意思","output":"是你有点可爱的意思"}
{"instruction":"解析下面句子的意思","input":"年轻人说 HHHACCC 是什么意思","output":"是你有点可爱的意思"}
{"instruction":"HHHACCC是什么意思","input":"","output":"是你有点可爱的意思"}
{"instruction":"解释网络用语HHHACCC","input":"","output":"是你有点可爱的意思"}
{"instruction":"年轻人说HHHACCC代表什么","input":"","output":"是你有点可爱的意思"}

python fine_tune_qwen35.py \
  --dataset ./my_dataset.jsonl \                
  --output_dir ./qwen35_finetuned \                              
  --steps 80 \  
  --learning_rate 1e-4 \
  --target_modules all \
  --dtype fp32       

(.venv) mac@macdeMac-Studio ai_nlp_q1 % python infer_qwen35.py \
  --model_path ./qwen35_finetuned/merged_16bit \
  --prompt $'解析下面句子的意思\n\n年轻人说 HHHACCC 是什么意思' \
  --device mps \
  --dtype fp32 \
  --temperature 0 \
  --max_new_tokens 64
Loading model: ./qwen35_finetuned/merged_16bit
Device: mps, dtype: torch.float32
The fast path is not available because one of the required library is not installed. Falling back to torch implementation. To install follow https://github.com/fla-org/flash-linear-attention#installation and https://github.com/Dao-AILab/causal-conv1d
Loading weights: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 426/426 [00:13<00:00, 32.17it/s]

=== Answer ===
是你有点可爱的可爱的意思