Llama-3.2-3B-Instruct-Football_Match_History

Llama-3.2-3B-Instruct-Football_Match_History

<!-- Provide a quick summary of what the model is/does. -->

Model Details

Model Description

<!-- Provide a longer summary of what this model is. -->

The model is based on Llama 3.2-3B-Instruct and fine-tuned on historical football match data from five major European leagues (England, Spain, Germany, Italy, and France) spanning the 1993–94 to 2022–23 seasons. It provides team-specific match results for a given season upon request.

• Developed by: mrp1mple
• Finetuned from model [optional]: meta-llama/Llama-3.2-3B-Instruct

Uses

<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->

Direct Use

<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->

Obtaining historical match results for a particular football team in a specified season. Summarizing a team’s performance based on the historical data available in the dataset.

Downstream Use [optional]

<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->

As a knowledge component in football analytics dashboards. Potentially extended for predictive modeling or advanced statistics about team performances.

Out-of-Scope Use

<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->

Predicting real-time match outcomes or betting odds, as the data is purely historical and the model was not trained for predictive tasks on future games. General text generation that is unrelated to football or outside the dataset’s scope.

Bias, Risks, and Limitations

<!-- This section is meant to convey both technical and sociotechnical limitations. -->

The model’s knowledge is limited to matches within the top 5 European leagues and up to the 2022–23 season. It may not accurately handle data or events outside this range. Any biases present in the original dataset (e.g., incomplete or missing matches for certain seasons) may affect the model’s responses. The model will not have real-time updates and thus may provide outdated information if asked about recent or ongoing seasons.

Recommendations

<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->

Users (both direct and downstream) should be made aware of the risks, biases, and limitations of the model. More information is needed for further recommendations, including potential expansions of the dataset, additional fine-tuning for broader football coverage, or domain-specific disclaimers.

How to Get Started with the Model

Use the code below to get started with the model.

from transformers import AutoTokenizer, AutoModelForCausalLM

model_name = "mrp1mple/Llama-3.2-3B-Instruct-Football_Match_History" tokenizer = AutoTokenizer.from_pretrained(model_name) model = AutoModelForCausalLM.from_pretrained(model_name)

prompt = """Below is an instruction that describes a task, paired with an input. Write a response that appropriately completes the request.

Instruction: Provide match results for a specific team in a specific season.

Input: Manchester United, 2012-2013

Response: """

inputs = tokenizer(prompt, return_tensors="pt") output_tokens = model.generate(**inputs, max_new_tokens=256) print(tokenizer.decode(output_tokens[0], skip_special_tokens=True))

Training Details

Training Data

<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->

Sourced from a Kaggle dataset containing more than 110,000 matches. Includes match results from top five European leagues (England, Spain, Germany, Italy, and France) between the 1993–94 and 2022–23 seasons.

Training Procedure

<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->

Preprocessing [optional]: A JSONL file was created to pair (instruction, input, output) for fine-tuning. Each record contains: Instruction: "Provide match results for a specific team in a specific season." Input: <TeamName>, <Season> Output: A concatenated string of all matches played by that team in that season.

Training Hyperparameters

• Training regime: Training regime: LoRA-based fine-tuning (low-rank adaptation) in 4-bit precision. 1 epoch, learning_rate=1e-4.