--- language: - en pipeline_tag: text-generation license: llama3 license_link: https://llama.meta.com/llama3/license/ --- # Meta-Llama-3-70B-Instruct-quantized.w8a8 ## Model Overview - **Model Architecture:** Meta-Llama-3 - **Input:** Text - **Output:** Text - **Model Optimizations:** - **Activation quantization:** INT8 - **Weight quantization:** INT8 - **Intended Use Cases:** Intended for commercial and research use in English. Similarly to [Meta-Llama-3-70B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct), this models is intended for assistant-like chat. - **Out-of-scope:** Use in any manner that violates applicable laws or regulations (including trade compliance laws). Use in languages other than English. - **Release Date:** 7/14/2024 - **Version:** 1.0 - **License(s):** [Llama3](https://llama.meta.com/llama3/license/) - **Model Developers:** Neural Magic Quantized version of [Meta-Llama-3-70B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct). It achieves an average score of 79.18 on the [OpenLLM](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard) benchmark (version 1), whereas the unquantized model achieves 79.18. ### Model Optimizations This model was obtained by quantizing the weights of [Meta-Llama-3-70B-Instruct](https://huggingface.co/meta-llama/Meta-Llama-3-70B-Instruct) to INT8 data type. This optimization reduces the number of bits used to represent weights and activations from 16 to 8, reducing GPU memory requirements (by approximately 50%) and increasing matrix-multiply compute throughput (by approximately 2x). Weight quantization also reduces disk size requirements by approximately 50%. Only weights and activations of the linear operators within transformers blocks are quantized. Weights are quantized with a symmetric static per-channel scheme, where a fixed linear scaling factor is applied between INT8 and floating point representations for each output channel dimension. Activations are quantized with a symmetric dynamic per-token scheme, computing a linear scaling factor at runtime for each token between INT8 and floating point representations. The [GPTQ](https://arxiv.org/abs/2210.17323) algorithm is applied for quantization, as implemented in the [llm-compressor](https://github.com/vllm-project/llm-compressor) library. GPTQ used a 10% damping factor and 256 sequences taken from Neural Magic's [LLM compression calibration dataset](https://huggingface.co/datasets/neuralmagic/LLM_compression_calibration). ## Deployment ### Use with vLLM This model can be deployed efficiently using the [vLLM](https://docs.vllm.ai/en/latest/) backend, as shown in the example below (using 2 GPUs). ```python from vllm import LLM, SamplingParams from transformers import AutoTokenizer model_id = "neuralmagic/Meta-Llama-3-70B-Instruct-quantized.w8a8" number_gpus = 2 sampling_params = SamplingParams(temperature=0.6, top_p=0.9, max_tokens=256) tokenizer = AutoTokenizer.from_pretrained(model_id) messages = [ {"role": "system", "content": "You are a pirate chatbot who always responds in pirate speak!"}, {"role": "user", "content": "Who are you?"}, ] prompts = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False) llm = LLM(model=model_id, tensor_parallel_size=number_gpus) outputs = llm.generate(prompts, sampling_params) generated_text = outputs[0].outputs[0].text print(generated_text) ``` vLLM aslo supports OpenAI-compatible serving. See the [documentation](https://docs.vllm.ai/en/latest/) for more details. ### Use with transformers The following example contemplates how the model can be deployed in Transformers using the `generate()` function. ```python from transformers import AutoTokenizer, AutoModelForCausalLM model_id = "neuralmagic/Meta-Llama-3-70B-Instruct-quantized.w8a8" tokenizer = AutoTokenizer.from_pretrained(model_id) model = AutoModelForCausalLM.from_pretrained( model_id, torch_dtype="auto", device_map="auto", ) messages = [ {"role": "system", "content": "You are a pirate chatbot who always responds in pirate speak!"}, {"role": "user", "content": "Who are you?"}, ] input_ids = tokenizer.apply_chat_template( messages, add_generation_prompt=True, return_tensors="pt" ).to(model.device) terminators = [ tokenizer.eos_token_id, tokenizer.convert_tokens_to_ids("<|eot_id|>") ] outputs = model.generate( input_ids, max_new_tokens=256, eos_token_id=terminators, do_sample=True, temperature=0.6, top_p=0.9, ) response = outputs[0][input_ids.shape[-1]:] print(tokenizer.decode(response, skip_special_tokens=True)) ``` ## Creation This model was created by using the [llm-compressor](https://github.com/vllm-project/llm-compressor) library as presented in the code snipet below. ```python from transformers import AutoTokenizer from datasets import load_dataset from llmcompressor.transformers import SparseAutoModelForCausalLM, oneshot from llmcompressor.modifiers.quantization import GPTQModifier model_id = "meta-llama/Meta-Llama-3-70B-Instruct" num_samples = 256 max_seq_len = 8192 tokenizer = AutoTokenizer.from_pretrained(model_id) def preprocess_fn(example): return {"text": tokenizer.apply_chat_template(example["messages"], add_generation_prompt=False, tokenize=False)} ds = load_dataset("neuralmagic/LLM_compression_calibration", split="train") ds = ds.shuffle().select(range(num_samples)) ds = ds.map(preprocess_fn) recipe = GPTQModifier( targets="Linear", scheme="W8A8", ignore=["lm_head"], dampening_frac=0.1, ) model = SparseAutoModelForCausalLM.from_pretrained( model_id, device_map="auto", trust_remote_code=True, ) oneshot( model=model, dataset=ds, recipe=recipe, max_seq_length=max_seq_len, num_calibration_samples=num_samples, ) model.save_pretrained("Meta-Llama-3-70B-Instruct-quantized.w8a8") ``` ## Evaluation The model was evaluated on the [OpenLLM](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard) leaderboard tasks (version 1) with the [lm-evaluation-harness](https://github.com/EleutherAI/lm-evaluation-harness/tree/383bbd54bc621086e05aa1b030d8d4d5635b25e6) (commit 383bbd54bc621086e05aa1b030d8d4d5635b25e6) and the [vLLM](https://docs.vllm.ai/en/stable/) engine, using the following command (using 2 GPUs): ``` lm_eval \ --model vllm \ --model_args pretrained="neuralmagic/Meta-Llama-3-70B-Instruct-quantized.w8a8",tensor_parallel_size=2,dtype=auto,gpu_memory_utilization=0.4,add_bos_token=True,max_model_len=4096 \ --tasks openllm \ --batch_size auto ``` ### Accuracy #### Open LLM Leaderboard evaluation scores
Benchmark Meta-Llama-3-70B-Instruct Meta-Llama-3-70B-Instruct-quantized.w8a8 (this model) Recovery
MMLU (5-shot) 80.18 79.41 99.0%
ARC Challenge (25-shot) 72.44 72.61 100.2%
GSM-8K (5-shot, strict-match) 90.83 92.27 101.6%
Hellaswag (10-shot) 85.54 85.75 100.2%
Winogrande (5-shot) 83.19 82.56 99.2%
TruthfulQA (0-shot) 62.92 62.48 99.3%
Average 79.18 79.18 100.0%