fbaldassarri/allenai_Llama-3.1-Tulu-3-8B-DPO-autoround-int4-gs64-sym

Model Information

Quantized version of allenai/Llama-3.1-Tulu-3-8B-DPO using torch.float32 for quantization tuning.

• 4 bits (INT4)
• group size = 64
• Symmetrical Quantization
• Method WoQ (AutoRound format)

Fast and low memory, 2-3X speedup (slight accuracy drop at W4G64)

Quantization framework: Intel AutoRound v0.4.6

Note: this INT4 version of Llama-3.1-Tulu-3-8B-DPO has been quantized to run inference through CPU.

Replication Recipe

Step 1 Install Requirements

I suggest to install requirements into a dedicated python-virtualenv or a conda enviroment.

wget https://github.com/intel/auto-round/archive/refs/tags/v0.4.6.tar.gz
tar -xvzf v0.4.6.tar.gz
cd auto-round-0.4.6
pip install -r requirements-cpu.txt --upgrade

Step 2 Build Intel AutoRound wheel from sources

pip install -vvv --no-build-isolation -e .[cpu]

Step 3 Script for Quantization

  from transformers import AutoModelForCausalLM, AutoTokenizer
  model_name = "allenai/Llama-3.1-Tulu-3-8B-DPO"
  model = AutoModelForCausalLM.from_pretrained(model_name)
  tokenizer = AutoTokenizer.from_pretrained(model_name)
  from auto_round import AutoRound
  bits, group_size, sym, device = 4, 64, True, 'cpu'
  autoround = AutoRound(model, tokenizer, nsamples=128, iters=200, seqlen=512, batch_size=4, bits=bits, group_size=group_size, sym=sym, device=device)
  autoround.quantize()
  output_dir = "./AutoRound/allenai_Llama-3.1-Tulu-3-8B-DPO-autoround-int4-gs64-sym"
  autoround.save_quantized(output_dir, format='auto_round', inplace=True)

License

Llama 3.1 Community License

Disclaimer

This quantized model comes with no warrenty. It has been developed only for research purposes.