Skip to content

Latest commit

 

History

History
367 lines (268 loc) · 13 KB

File metadata and controls

367 lines (268 loc) · 13 KB

LLMs

Large-language models are AI models that can understand and generate text, primarily using transformer architectures. This page is about running them on a HPC cluster. This requires programming experience and knowledge of using the cluster (:ref:`tutorials`), but allows maximum computational power for the least cost. :doc:`Aalto RSE </rse/index>` maintains these models and can provide help with using these, even to users who aren't computational experts.

Because the size of model weights are typically very large and the interest in the models is high, so we provide our users with pre-downloaded model weights in various formats, along with instructions on how to load these weights for inference purposes, retraining, and fine-tuning tasks. We also provide a dedicated python environment (run module load scicomp-llm-env to activate it) that has many commonly used python libraries installed for you to test the models quickly.

HuggingFace Models

The simplest way to use an open-source LLM(Large Language Model) is through the tools and pre-trained models hub from huggingface. Huggingface is a popular platform for NLP(Natural Language Processing) tasks. It provides a user-friendly interface through the transformers library to load and run various pre-trained models. Most open-source models from Huggingface are widely supported and integrated with the transformers library. We are keeping our eyes on the latest models and have downloaded some of them for you. The full list of all the available models are located at /scratch/shareddata/dldata/huggingface-hub-cache/models.txt. Please contact us if you need any other models. The following table lists only a few example from the hosted models:

Model type Huggingface model identifier
Text Generation meta-llama/Meta-Llama-3-8B
Text Generation meta-llama/Meta-Llama-3-8B-Instruct
Text Generation mistralai/Mixtral-8x22B-v0.1
Text Generation mistralai/Mixtral-8x22B-Instruct-v0.1
Text Generation tiiuae/falcon-40b
Text Generation tiiuae/falcon-40b-instruct
Text Generation google/gemma-2b-it
Text Generation google/gemma-7b
Text Generation google/gemma-7b-it
Text Generation google/gemma-7b
Text Generation LumiOpen/Poro-34B
Text Generation meta-llama/Llama-2-7b-hf
Text Generation meta-llama/Llama-2-13b-hf
Text Generation meta-llama/Llama-2-70b-hf
Text Generation codellama/CodeLlama-7b-hf
Text Generation codellama/CodeLlama-13b-hf
Text Generation codellama/CodeLlama-34b-hf
Translation Helsinki-NLP/opus-mt-en-fi
Translation Helsinki-NLP/opus-mt-fi-en
Translation t5-base
Fill Mask bert-base-uncased
Fill Mask bert-base-cased
Fill Mask distilbert-base-uncased
Text to Speech microsoft/speecht5_hifigan
Text to Speech facebook/hf-seamless-m4t-large
Automatic Speech Recognition openai/whisper-large-v3
Token Classification dslim/bert-base-NER-uncased

To access Huggingface models:

.. tabs::

  .. group-tab:: slurm script

    Load the module to setup the environment variable HF_HOME:

    .. code-block:: bash

      module load model-huggingface/all
      # this will set HF_HOME to /scratch/shareddata/dldata/huggingface-hub-cache

  .. group-tab:: jupyter notebook

    In jupyter notebook, one can set up HF_HOME directly:

    .. code-block:: python

      import os
      os.environ['TRANSFORMERS_OFFLINE'] = '1'
      os.environ['HF_HOME']='/scratch/shareddata/dldata/huggingface-hub-cache'


Here is a Python script using huggingface model.

## Force transformer to load model(s) from local hub instead of download and load model(s) from remote hub.
## !!!!!! NOTE: this must be in front of import transformers.
import os
os.environ['TRANSFORMERS_OFFLINE'] = '1'

from transformers import AutoModelForCausalLM, AutoTokenizer

tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-v0.1")
model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-v0.1")

prompt = "How many stars in the space?"

model_inputs = tokenizer([prompt], return_tensors="pt")
input_length = model_inputs.input_ids.shape[1]

generated_ids = model.generate(**model_inputs, max_new_tokens=20)
print(tokenizer.batch_decode(generated_ids[:, input_length:], skip_special_tokens=True)[0])

Raw model weights

We also downloaded the following raw llama model weights (PyTorch model checkpoints), and they are managed by the following modules.

Model type Model version Module command to load Description
Llama 2 Raw Data module load model-llama2/raw-data Raw weights of Llama 2.
Llama 2 7b module load model-llama2/7b Raw weights of 7B parameter version of Llama 2.
Llama 2 7b-chat module load model-llama2/7b-chat Raw weights of 7B parameter chat optimized version of Llama 2.
Llama 2 13b module load model-llama2/13b Raw weights of 13B parameter version of Llama 2.
Llama 2 13b-chat module load model-llama2/13b-chat Raw weights of 13B parameter chat optimized version of Llama 2.
Llama 2 70b module load model-llama2/70b Raw weights of 70B parameter version of Llama 2.
Llama 2 70b-chat module load model-llama2/70b-chat Raw weights of 70B parameter chat optimized version of Llama 2.
CodeLlama Raw Data module load model-codellama/raw-data Raw weights of CodeLlama.
CodeLlama 7b module load model-codellama/7b Raw weights of 7B parameter version of CodeLlama.
CodeLlama 7b-Python module load model-codellama/7b-python Raw weights of 7B parameter version CodeLlama, specifically designed for Python.
CodeLlama 7b-Instruct module load model-codellama/7b-instruct Raw weights of 7B parameter version CodeLlama, designed for instruction following.
CodeLlama 13b module load model-codellama/13b Raw weights of 13B parameter version of CodeLlama.
CodeLlama 13b-Python module load model-codellama/13b-python Raw weights of 13B parameter version CodeLlama, specifically designed for Python.
CodeLlama 13b-Instruct module load model-codellama/13b-instruct Raw weights of 13B parameter version CodeLlama, designed for instruction following.
CodeLlama 34b module load model-codellama/34b Raw weights of 34B parameter version of CodeLlama.
CodeLlama 34b-Python module load model-codellama/34b-python Raw weights of 34B parameter version CodeLlama, specifically designed for Python.
CodeLlama 34b-Instruct module load model-codellama/34b-instruct Raw weights of 34B parameter version CodeLlama, designed for instruction following.

Each module will set the following environment variables:

  • MODEL_ROOT - Folder where model weights are stored, i.e., PyTorch model checkpoint directory.
  • TOKENIZER_PATH - File path to the tokenizer.model.

Here is an example :doc:`slurm script </triton/tut/slurm>`, using the raw weights for batch inference. For detailed environment setting up, example prompts and Python code, please check out this repo.

#!/bin/bash
#SBATCH --time=00:25:00
#SBATCH --cpus-per-task=4
#SBATCH --mem=20GB
#SBATCH --gpus=1
#SBATCH --output llama2inference-gpu.%J.out
#SBATCH --error llama2inference-gpu.%J.err

# get access to the model weights
module load model-llama2/7b
echo $MODEL_ROOT
# Expect output: /scratch/shareddata/dldata/llama-2/llama-2-7b
echo $TOKENIZER_PATH
# Expect output: /scratch/shareddata/dldata/llama-2/tokenizer.model

# activate your conda environment
module load mamba
source activate llama2env

# run batch inference
torchrun --nproc_per_node 1 batch_inference.py \
  --prompts prompts.json \
  --ckpt_dir $MODEL_ROOT \
  --tokenizer_path $TOKENIZER_PATH \
  --max_seq_len 512 --max_batch_size 16

llama.cpp and GGUF model weights

llama.cpp is another popular framework for running inference on LLM models with CPUs or GPUs. It provides C++ implementations of many large language models. llama.cpp uses a format called GGUF as its storage format. We have GGUF conversions of all Llama 2 and CodeLlama models with multiple quantization levels. Please contact us if you need any other GGUF models. NOTE: Before loading the following modules, one must first load a module for the raw model weights. For example, run module load model-codellama/34b first, and then run module load codellama.cpp/q8_0-2023-12-04 to get the 8-bit integer version of CodeLlama weights in a .gguf file.

Model type Model version Module command to load Description
Llama 2 f16-2023-08-28 module load model-llama.cpp/f16-2023-12-04 (after loading a Llama 2 model for some raw weights) Half precision version of Llama 2 weights done with llama.cpp on 4th of Dec 2023.
Llama 2 q4_0-2023-08-28 module load model-llama.cpp/q4_0-2023-12-04 (after loading a Llama 2 model for some raw weights) 4-bit integer version of Llama 2 weights done with llama.cpp on 4th of Dec 2023.
Llama 2 q4_1-2023-08-28 module load model-llama.cpp/q4_1-2023-12-04 (after loading a Llama2 model for some raw weights) 4-bit integer version of Llama 2 weights done with llama.cpp on 4th of Dec 2023.
Llama 2 q8_0-2023-08-28 module load model-llama.cpp/q8_0-2023-12-04 (after loading a Llama 2 model for some raw weights) 8-bit integer version of Llama 2 weights done with llama.cpp on 4th of Dec 2023.
CodeLlama f16-2023-08-28 module load codellama.cpp/f16-2023-12-04 (after loading a CodeLlama model for some raw weights) Half precision version of CodeLlama weights done with llama.cpp on 4th of Dec 2023.
CodeLlama q4_0-2023-08-28 module load codellama.cpp/q4_0-2023-12-04 (after loading a CodeLlama model for some raw weights) 4-bit integer version of CodeLlama weights done with llama.cpp on 4th of Dec 2023.
CodeLlama q8_0-2023-08-28 module load codellama.cpp/q8_0-2023-12-04 (after loading a CodeLlama model for some raw weights) 8-bit integer version of CodeLlama weights done with llama.cpp on 4th of Dec 2023.

Each module will set the following environment variables:

  • MODEL_ROOT - Folder where model weights are stored.
  • MODEL_WEIGHTS - Path to the model weights in GGUF file format.

This Python code snippet is part of a 'Chat with Your PDF Documents' example, utilizing LangChain and leveraging model weights stored in a .gguf file. For detailed environment setting up and Python code, please check out this repo. NOTE: this example repo is mainly meant to run on CPUs, if you want to run on GPUs, you can checkout a branch "llamacpp-gpu" of this repo for details.

import os
from langchain.llms import LlamaCpp

model_path = os.environ.get('MODEL_WEIGHTS')
llm = LlamaCpp(model_path=model_path, verbose=False)

More examples

Starting a local API

With the pre-downloaded model weights, you are also able create an API endpoint locally. For detailed examples, you can checkout this repo.