| | --- |
| | library_name: transformers |
| | tags: |
| | - unsloth |
| | - fine-tuning |
| | - valheim |
| | datasets: |
| | - Egrigor/ValheimTestData |
| | base_model: |
| | - unsloth/Phi-3.5-mini-instruct |
| | license: apache-2.0 |
| | --- |
| | |
| | # Valheim Assistant V1 - 16-bit Version |
| |
|
| | This is a fine-tuned version of the [Unsloth Phi-3.5-mini-instruct](https://huggingface.co/unsloth/Phi-3.5-mini-instruct) model. It is designed to provide conversational assistance for the game **Valheim**, covering gameplay mechanics, lore, and in-game strategies. |
| |
|
| | This model is part of an ongoing **hobbyist project**, created by a novice enthusiast exploring fine-tuning techniques. While the dataset and model may have limitations, this project aims to improve over time through community feedback and iteration. |
| |
|
| | --- |
| |
|
| | ## Model Details |
| |
|
| | ### Summary |
| | This model has been fine-tuned on [Egrigor/ValheimTestData](https://huggingface.co/datasets/Egrigor/ValheimTestData), which contains conversational-style question-and-answer pairs derived from Valheim community resources. |
| |
|
| | - **Base Model**: [Unsloth Phi-3.5-mini-instruct](https://huggingface.co/unsloth/Phi-3.5-mini-instruct) |
| | - **Language**: English |
| | - **License**: Apache-2.0 |
| | - **Fine-Tuning Framework**: Unsloth |
| | - **Training Precision**: 16-bit |
| | - **Dataset**: [Egrigor/ValheimTestData](https://huggingface.co/datasets/Egrigor/ValheimTestData) |
| |
|
| | --- |
| |
|
| | ## How to Use |
| |
|
| | To load and interact with the model, use the following code: |
| |
|
| | from unsloth import FastLanguageModel |
| | from transformers import AutoTokenizer |
| |
|
| | # Load model and tokenizer |
| | model = FastLanguageModel.from_pretrained("Egrigor/ValheimAssistantV1-16bit") |
| | tokenizer = AutoTokenizer.from_pretrained("Egrigor/ValheimAssistantV1-16bit") |
| |
|
| | # Prepare input and generate output |
| | inputs = tokenizer("How do I build a crafting station?", return_tensors="pt") |
| | outputs = model.generate(**inputs, max_length=100) |
| | |
| | # Decode and print the output |
| | print(tokenizer.decode(outputs[0], skip_special_tokens=True)) |
| | |
| | |
| | |
| | ## Training Details |
| | |
| | |
| | |
| | Framework: Unsloth |
| | |
| | Batch Size per Device: 2 |
| | |
| | Gradient Accumulation Steps: 4 |
| | |
| | Learning Rate: 2e-4 |
| | |
| | Total Training Steps: 60 |
| | |
| | Training Time: ~3.5 minutes |
| | |
| | |
| | ## Dataset |
| | |
| | The dataset consists of gameplay tips, lore insights, and strategic guidance, formatted into conversational question-and-answer pairs. |
| | |
| | Note: As this dataset was semi-automatically generated from community resources, it may contain inaccuracies or gaps. Feedback is welcome for future improvements. |
| | |
| | |
| | ## Limitations and Future Work |
| | |
| | Dataset Gaps: The dataset may not fully cover Valheims gameplay and lore, and some responses may be incomplete or inaccurate. |
| | |
| | Generalization: This model is focused solely on Valheim-related content and may not perform well in other contexts. |
| | |
| | Iterative Improvement: Future versions will refine the dataset and model based on testing and community feedback. |
| | |
| | |
| | ## Contact and Feedback |
| | |
| | This is an experimental project created by a novice AI enthusiast. Feedback, suggestions, and collaborations are highly welcome! Feel free to reach out via Hugging Face. |
| | |
| | --- |
| | |
| | Let me know if there’s anything specific you’d like to tweak or emphasize! |
