Update README.md
Browse files
README.md
CHANGED
|
@@ -37,7 +37,7 @@ This model has been fine-tuned on [Egrigor/ValheimTestData](https://huggingface.
|
|
| 37 |
|
| 38 |
To load and interact with the model, use the following code:
|
| 39 |
|
| 40 |
-
|
| 41 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 42 |
|
| 43 |
model = AutoModelForCausalLM.from_pretrained("Egrigor/ValheimAssistantV1")
|
|
@@ -46,41 +46,44 @@ tokenizer = AutoTokenizer.from_pretrained("Egrigor/ValheimAssistantV1")
|
|
| 46 |
inputs = tokenizer("How do I build a crafting station?", return_tensors="pt")
|
| 47 |
outputs = model.generate(**inputs, max_length=100)
|
| 48 |
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
|
|
|
|
|
|
|
| 49 |
Training Details
|
|
|
|
|
|
|
|
|
|
| 50 |
Framework: Unsloth
|
|
|
|
| 51 |
Batch Size per Device: 2
|
|
|
|
| 52 |
Gradient Accumulation Steps: 4
|
|
|
|
| 53 |
Learning Rate: 2e-4
|
|
|
|
| 54 |
Total Training Steps: 60
|
|
|
|
| 55 |
Training Time: ~3.5 minutes
|
|
|
|
|
|
|
| 56 |
Dataset
|
|
|
|
| 57 |
The dataset consists of gameplay tips, lore insights, and strategic guidance, formatted into conversational question-and-answer pairs.
|
| 58 |
|
| 59 |
Note: As this dataset was semi-automatically generated from community resources, it may contain inaccuracies or gaps. Feedback is welcome for future improvements.
|
| 60 |
|
|
|
|
| 61 |
Limitations and Future Work
|
| 62 |
-
|
|
|
|
|
|
|
| 63 |
Generalization: This model is focused solely on Valheim-related content and may not perform well in other contexts.
|
|
|
|
| 64 |
Iterative Improvement: Future versions will refine the dataset and model based on testing and community feedback.
|
|
|
|
|
|
|
| 65 |
Contact and Feedback
|
| 66 |
-
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.
|
| 67 |
|
| 68 |
-
|
| 69 |
-
If you use this model, please consider citing it:
|
| 70 |
-
|
| 71 |
-
less
|
| 72 |
-
Copy
|
| 73 |
-
Edit
|
| 74 |
-
@misc{Egrigor_ValheimAssistantV1_2025,
|
| 75 |
-
author = {Egrigor},
|
| 76 |
-
title = {Valheim Assistant V1 - Fine-Tuned with Unsloth},
|
| 77 |
-
year = {2025},
|
| 78 |
-
publisher = {Hugging Face},
|
| 79 |
-
howpublished = {\url{https://huggingface.co/Egrigor/ValheimAssistantV1}},
|
| 80 |
-
}
|
| 81 |
-
yaml
|
| 82 |
-
Copy
|
| 83 |
-
Edit
|
| 84 |
|
| 85 |
---
|
| 86 |
|
|
|
|
| 37 |
|
| 38 |
To load and interact with the model, use the following code:
|
| 39 |
|
| 40 |
+
python
|
| 41 |
from transformers import AutoModelForCausalLM, AutoTokenizer
|
| 42 |
|
| 43 |
model = AutoModelForCausalLM.from_pretrained("Egrigor/ValheimAssistantV1")
|
|
|
|
| 46 |
inputs = tokenizer("How do I build a crafting station?", return_tensors="pt")
|
| 47 |
outputs = model.generate(**inputs, max_length=100)
|
| 48 |
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
|
| 49 |
+
|
| 50 |
+
|
| 51 |
Training Details
|
| 52 |
+
|
| 53 |
+
|
| 54 |
+
|
| 55 |
Framework: Unsloth
|
| 56 |
+
|
| 57 |
Batch Size per Device: 2
|
| 58 |
+
|
| 59 |
Gradient Accumulation Steps: 4
|
| 60 |
+
|
| 61 |
Learning Rate: 2e-4
|
| 62 |
+
|
| 63 |
Total Training Steps: 60
|
| 64 |
+
|
| 65 |
Training Time: ~3.5 minutes
|
| 66 |
+
|
| 67 |
+
|
| 68 |
Dataset
|
| 69 |
+
|
| 70 |
The dataset consists of gameplay tips, lore insights, and strategic guidance, formatted into conversational question-and-answer pairs.
|
| 71 |
|
| 72 |
Note: As this dataset was semi-automatically generated from community resources, it may contain inaccuracies or gaps. Feedback is welcome for future improvements.
|
| 73 |
|
| 74 |
+
|
| 75 |
Limitations and Future Work
|
| 76 |
+
|
| 77 |
+
Dataset Gaps: The dataset may not fully cover Valheims gameplay and lore, and some responses may be incomplete or inaccurate.
|
| 78 |
+
|
| 79 |
Generalization: This model is focused solely on Valheim-related content and may not perform well in other contexts.
|
| 80 |
+
|
| 81 |
Iterative Improvement: Future versions will refine the dataset and model based on testing and community feedback.
|
| 82 |
+
|
| 83 |
+
|
| 84 |
Contact and Feedback
|
|
|
|
| 85 |
|
| 86 |
+
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.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
---
|
| 89 |
|