Maximus Model Card
Model Details
Model Name: Maximus
Model Type: Transformer-based leveraging Microsoft Phi 14b 128k tokens
Publisher: Awels Engineering
License: MIT
Model Description:
Maximus is a sophisticated model designed to help as an AI agent focusing on Maximo Application Suite. It leverages advanced machine learning techniques to provide efficient and accurate solutions. It has been trained on the full docments corpus of MAS 8.5.
Dataset
Dataset Source: Hugging Face Datasets
Dataset License: MIT
Dataset Description:
The dataset used to train Maximus consists of all the public documents available on Maximo application suite. This dataset is curated to ensure a comprehensive representation of typical administrative scenarios encountered in Maximo.
Training Details
Training Data:
The training data includes 67,000 Questions and Answers generated by the
Bonito LLM. The dataset is split into 3 sets of data (training, test and validation) to ensure robust model performance.
Training Procedure:
Maximus was trained using supervised learning with cross-entropy loss and the Adam optimizer. The training involved 1 epoch, a batch size of 4, a learning rate of 5.0e-06, and a cosine learning rate scheduler with gradient checkpointing for memory efficiency.
Hardware:
The model was trained on a single NVIDIA H100 SXM graphic card.
Framework:
The training was conducted using PyTorch.
Evaluation
Evaluation Metrics:
Maximus was evaluated on the training dataset:
epoch = 1.0
total_flos = 233585641GF
train_loss = 1.7111
train_runtime = 1:08:52.73
train_samples_per_second = 11.41
train_steps_per_second = 2.853
Performance:
The model achieved the following results on the evaluation dataset:
epoch = 1.0
eval_loss = 1.4482
eval_runtime = 0:03:24.92
eval_samples = 10773
eval_samples_per_second = 57.386
eval_steps_per_second = 14.347
Intended Use
Primary Use Case:
Maximus is intended to be used locally in an agent swarm to colleborate together to solve Maximo Application Suite related problems.
Limitations:
This 14b model is an upscale of the 3b model. Much better loss than the 3b so results should be better.
