benhaotang
commited on
Update README.md
Browse files
README.md
CHANGED
@@ -11,21 +11,18 @@ base_model:
|
|
11 |
---
|
12 |
# Llama 3.2 Physics Fine-tuned Model
|
13 |
|
14 |
-
This model is a fine-tuned version of
|
15 |
|
16 |
## Model description
|
17 |
-
- Base model:
|
18 |
-
- Training data: kejian/arxiv-physics-debug-v0
|
19 |
- Fine-tuning type: LoRA
|
20 |
- Use case: Physics domain questions
|
21 |
-
|
22 |
-
|
23 |
-
- Base Model: unsloth/Llama-3.2-1B-Instruct
|
24 |
-
- Dataset: kejian/arxiv-physics-debug-v0
|
25 |
- Training Arguments:
|
26 |
- Learning Rate: 2e-5
|
27 |
- Epochs: 3
|
28 |
- Gradient Accumulation Steps: 8
|
|
|
29 |
|
30 |
|
31 |
## Usage
|
@@ -49,7 +46,7 @@ Expected output:
|
|
49 |
>
|
50 |
> The renormalization group flow is based on the idea that the behavior of a system at small distances (i.e., low energies) is closely related to its behavior at large distances (i.e., high energies). The renormalization group is a mathematical framework that describes how physical parameters change as the energy of a system increases.
|
51 |
>
|
52 |
-
>
|
53 |
>
|
54 |
> The renormalization group flow is a powerful tool for physicists, allowing them to analyze complex systems and understand their behavior at the smallest scales. It has been applied to a wide range of systems, including gauge theories, quantum field theories, and cosmological models.
|
55 |
>
|
|
|
11 |
---
|
12 |
# Llama 3.2 Physics Fine-tuned Model
|
13 |
|
14 |
+
This model is a fine-tuned version of [meta-llama/Llama-3.2-1B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct) on [kejian/arxiv-physics-debug-v0](https://huggingface.co/datasets/kejian/arxiv-physics-debug-v0).
|
15 |
|
16 |
## Model description
|
17 |
+
- Base model: [meta-llama/Llama-3.2-1B-Instruct](https://huggingface.co/meta-llama/Llama-3.2-1B-Instruct)
|
18 |
+
- Training data: [kejian/arxiv-physics-debug-v0](https://huggingface.co/datasets/kejian/arxiv-physics-debug-v0)
|
19 |
- Fine-tuning type: LoRA
|
20 |
- Use case: Physics domain questions
|
|
|
|
|
|
|
|
|
21 |
- Training Arguments:
|
22 |
- Learning Rate: 2e-5
|
23 |
- Epochs: 3
|
24 |
- Gradient Accumulation Steps: 8
|
25 |
+
- Training setup: Ubuntu 24.04 with RX7800XT
|
26 |
|
27 |
|
28 |
## Usage
|
|
|
46 |
>
|
47 |
> The renormalization group flow is based on the idea that the behavior of a system at small distances (i.e., low energies) is closely related to its behavior at large distances (i.e., high energies). The renormalization group is a mathematical framework that describes how physical parameters change as the energy of a system increases.
|
48 |
>
|
49 |
+
> In the context of quantum field theory, the renormalization group flow is used to study the behavior of particles and forces in the early universe. It is particularly useful for understanding the evolution of the universe from the Planck era to the present day.
|
50 |
>
|
51 |
> The renormalization group flow is a powerful tool for physicists, allowing them to analyze complex systems and understand their behavior at the smallest scales. It has been applied to a wide range of systems, including gauge theories, quantum field theories, and cosmological models.
|
52 |
>
|