--- license: llama3 language: - en pipeline_tag: text-generation tags: - nvidia - chatqa-1.5 - chatqa - llama-3 - pytorch - llama - conversational datasets: - nvidia/ChatQA-Training-Data --- # Llama 3 8B ChatQA - AWQ - Model creator: [Meta Llama 2](https://huggingface.co/meta-llama) - Original model: [Llama 3 8B ChatQA](https://huggingface.co/nvidia/Llama3-ChatQA-1.5-8B) ## Description This repo contains AWQ model files for [Nvidia's Llama 3 8B ChatQA](https://huggingface.co/nvidia/Llama3-ChatQA-1.5-8B). ### About AWQ AWQ is an efficient, accurate and blazing-fast low-bit weight quantization method, currently supporting 4-bit quantization. Compared to GPTQ, it offers faster Transformers-based inference. It is also now supported by continuous batching server [vLLM](https://github.com/vllm-project/vllm), allowing the use of AWQ models for high-throughput concurrent inference in multi-user server scenarios. Note that, at the time of writing, overall throughput is still lower than running vLLM with unquantised models, however using AWQ enables using much smaller GPUs which can lead to easier deployment and overall cost savings. For example, a 70B model can be run on 1 x 48GB GPU instead of 2 x 80GB. ## Model Details By Nvidia Llama3-ChatQA-1.5, which excels at conversational question answering (QA) and retrieval-augmented generation (RAG). Llama3-ChatQA-1.5 is developed using an improved training recipe from [ChatQA (1.0)](https://arxiv.org/abs/2401.10225), and it is built on top of [Llama-3 base model](https://huggingface.co/meta-llama/Meta-Llama-3-8B). Specifically, we incorporate more conversational QA data to enhance its tabular and arithmetic calculation capability. Llama3-ChatQA-1.5 has two variants: Llama3-ChatQA-1.5-8B and Llama3-ChatQA-1.5-70B. Both models were originally trained using [Megatron-LM](https://github.com/NVIDIA/Megatron-LM), we converted the checkpoints to Hugging Face format. ## Other Resources [Llama3-ChatQA-1.5-70B](https://huggingface.co/nvidia/Llama3-ChatQA-1.5-70B)   [Evaluation Data](https://huggingface.co/datasets/nvidia/ConvRAG-Bench)   [Training Data](https://huggingface.co/datasets/nvidia/ChatQA-Training-Data)   [Retriever](https://huggingface.co/nvidia/dragon-multiturn-query-encoder) ## Benchmark Results Results in ConvRAG Bench are as follows: | | ChatQA-1.0-7B | Command-R-Plus | Llama-3-instruct-70b | GPT-4-0613 | ChatQA-1.0-70B | ChatQA-1.5-8B | ChatQA-1.5-70B | | -- |:--:|:--:|:--:|:--:|:--:|:--:|:--:| | Doc2Dial | 37.88 | 33.51 | 37.88 | 34.16 | 38.9 | 39.33 | 41.26 | | QuAC | 29.69 | 34.16 | 36.96 | 40.29 | 41.82 | 39.73 | 38.82 | | QReCC | 46.97 | 49.77 | 51.34 | 52.01 | 48.05 | 49.03 | 51.40 | | CoQA | 76.61 | 69.71 | 76.98 | 77.42 | 78.57 | 76.46 | 78.44 | | DoQA | 41.57 | 40.67 | 41.24 | 43.39 | 51.94 | 49.6 | 50.67 | | ConvFinQA | 51.61 | 71.21 | 76.6 | 81.28 | 73.69 | 78.46 | 81.88 | | SQA | 61.87 | 74.07 | 69.61 | 79.21 | 69.14 | 73.28 | 83.82 | | TopioCQA | 45.45 | 53.77 | 49.72 | 45.09 | 50.98 | 49.96 | 55.63 | | HybriDial* | 54.51 | 46.7 | 48.59 | 49.81 | 56.44 | 65.76 | 68.27 | | INSCIT | 30.96 | 35.76 | 36.23 | 36.34 | 31.9 | 30.1 | 32.31 | | Average (all) | 47.71 | 50.93 | 52.52 | 53.90 | 54.14 | 55.17 | 58.25 | | Average (exclude HybriDial) | 46.96 | 51.40 | 52.95 | 54.35 | 53.89 | 53.99 | 57.14 | Note that ChatQA-1.5 is built based on Llama-3 base model, and ChatQA-1.0 is built based on Llama-2 base model. ChatQA-1.5 used some samples from the HybriDial training dataset. To ensure fair comparison, we also compare average scores excluding HybriDial. The data and evaluation scripts for ConvRAG can be found [here](https://huggingface.co/datasets/nvidia/ConvRAG-Bench). ## Prompt Format 
System: {System}

{Context}

User: {Question}

Assistant: {Response}

User: {Question}

Assistant:
## How to use ### using vLLM ```python from vllm import LLM, SamplingParams # Sample prompts. prompts = [ "Hello, how are you?" ] # Create a sampling params object. sampling_params = SamplingParams(max_tokens=128) # Create an LLM. llm = LLM(model="Sreenington/Llama-3-8B-ChatQA-AWQ", quantization="AWQ") # Generate texts from the prompts. The output is a list of RequestOutput objects # that contain the prompt, generated text, and other information. outputs = llm.generate(prompts, sampling_params) # Print the outputs. for output in outputs: prompt = output.prompt generated_text = output.outputs[0].text print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}") ``` ### take the whole document as context This can be applied to the scenario where the whole document can be fitted into the model, so that there is no need to run retrieval over the document ```python from transformers import AutoTokenizer, AutoModelForCausalLM import torch model_id = "nvidia/Llama3-ChatQA-1.5-8B" tokenizer = AutoTokenizer.from_pretrained(model_id) model = AutoModelForCausalLM.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto") messages = [ {"role": "user", "content": "what is the percentage change of the net income from Q4 FY23 to Q4 FY24?"} ] document = """NVIDIA (NASDAQ: NVDA) today reported revenue for the fourth quarter ended January 28, 2024, of $22.1 billion, up 22% from the previous quarter and up 265% from a year ago.\nFor the quarter, GAAP earnings per diluted share was $4.93, up 33% from the previous quarter and up 765% from a year ago. Non-GAAP earnings per diluted share was $5.16, up 28% from the previous quarter and up 486% from a year ago.\nQ4 Fiscal 2024 Summary\nGAAP\n| $ in millions, except earnings per share | Q4 FY24 | Q3 FY24 | Q4 FY23 | Q/Q | Y/Y |\n| Revenue | $22,103 | $18,120 | $6,051 | Up 22% | Up 265% |\n| Gross margin | 76.0% | 74.0% | 63.3% | Up 2.0 pts | Up 12.7 pts |\n| Operating expenses | $3,176 | $2,983 | $2,576 | Up 6% | Up 23% |\n| Operating income | $13,615 | $10,417 | $1,257 | Up 31% | Up 983% |\n| Net income | $12,285 | $9,243 | $1,414 | Up 33% | Up 769% |\n| Diluted earnings per share | $4.93 | $3.71 | $0.57 | Up 33% | Up 765% |""" def get_formatted_input(messages, context): system = "System: This is a chat between a user and an artificial intelligence assistant. The assistant gives helpful, detailed, and polite answers to the user's questions based on the context. The assistant should also indicate when the answer cannot be found in the context." instruction = "Please give a full and complete answer for the question." for item in messages: if item['role'] == "user": ## only apply this instruction for the first user turn item['content'] = instruction + " " + item['content'] break conversation = '\n\n'.join(["User: " + item["content"] if item["role"] == "user" else "Assistant: " + item["content"] for item in messages]) + "\n\nAssistant:" formatted_input = system + "\n\n" + context + "\n\n" + conversation return formatted_input formatted_input = get_formatted_input(messages, document) tokenized_prompt = tokenizer(tokenizer.bos_token + formatted_input, return_tensors="pt").to(model.device) terminators = [ tokenizer.eos_token_id, tokenizer.convert_tokens_to_ids("<|eot_id|>") ] outputs = model.generate(input_ids=tokenized_prompt.input_ids, attention_mask=tokenized_prompt.attention_mask, max_new_tokens=128, eos_token_id=terminators) response = outputs[0][tokenized_prompt.input_ids.shape[-1]:] print(tokenizer.decode(response, skip_special_tokens=True)) ``` ## Quantized Model by Sree Narayanan (sree@zaplineai.cloud) https://github.com/eersnington https://www.linkedin.com/in/sreenington/ ## License The use of this model is governed by the [META LLAMA 3 COMMUNITY LICENSE AGREEMENT](https://llama.meta.com/llama3/license/)