README.md

metadata

quantized_by: LLMJapan
pipeline_tag: text-generation
license: cc-by-nc-4.0
language:
  - en
tags:
  - nvidia
  - AceInstruct
  - code
  - math
  - general_domain
  - instruct_model
base_model: nvidia/AceInstruct-72B

Exllama v2 Quantizations of AceInstruct-72B by nvidia

Using turboderp's ExLlamaV2 v0.2.8 for quantization.

Original model: https://huggingface.co/nvidia/AceInstruct-72B

Quantization Command Example for creating other bpw quantization

cd {your git clone directory}
python convert.py -i {path to}/AceInstruct-72B -o {path to}/AceInstruct-72B/workingdir -cf {path to}/AceInstruct-72B/AceInstruct-72B-4bpw -b 4.0

Prompt format

<|im_start|>system
{system_prompt}<|im_end|>
<|im_start|>user
{prompt}<|im_end|>
<|im_start|>assistant

How to add your system prompt

Copy the following json and replace the "You are AceInstruct developed by NVIDIA. You are helpful assistant." sentence with your original system prompt. The default tokenizer_config.json does not have system prompt.

tokenizer_config.json

"chat_template": "{{- '<|im_start|>system\\nYou are AceInstruct developed by NVIDIA. You are helpful assistant.<|im_end|>\\n' }}\n    {%- for message in messages %}\n{{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}\n{%- endfor %}\n{%- if add_generation_prompt %}\n{{- '<|im_start|>assistant\n' }}\n{%- endif %}\n",

File information

quantization type	file size
4.0bpw	35.9 GiB

Benchmark Results

	Qwen2.5-1.5B-Instruct	AceInstruct-1.5B	Qwen2.5-7B-Instruct	AceInstruct-7B	Qwen2.5-72B-Instruct	AceInstruct-72B
HumanEval	61.60	73.17	84.80	85.37	86.60	89.63
MBPP	63.20	65.76	79.20	74.32	88.20	83.66
GSM8K	73.20	80.44	91.60	93.10	95.80	96.36
MATH	55.20	60.34	75.50	76.40	83.10	84.50
MMLU	58.37	58.17	74.51	74.68	84.67	83.88
MMLU Pro	32.40	33.78	56.30	54.50	71.10	66.10
Average	57.33	61.94	76.99	76.40	84.91	84.02

Credits

Thanks to NVIDIA team.

---

license: cc-by-nc-4.0