78 lines
2.0 KiB
Markdown
78 lines
2.0 KiB
Markdown
---
|
|
language: en
|
|
license: apache-2.0
|
|
tags:
|
|
- qwen2
|
|
- reasoning
|
|
- fine-tuned
|
|
- english
|
|
- instruction-tuned
|
|
- chat
|
|
- gguf
|
|
metrics:
|
|
- accuracy
|
|
- f1
|
|
base_model:
|
|
- Qwen/Qwen2.5-3B-Instruct
|
|
library_name: transformers
|
|
pipeline_tag: text-generation
|
|
datasets:
|
|
- openai/gsm8k
|
|
---
|
|
|
|
# Model Card for CALISTA-INDUSTRY/qwen_2_3B_reasoning_en_ft_v1
|
|
|
|
## Model Details
|
|
|
|
- **Model Name**: qwen_2_3B_reasoning_en_ft_v1
|
|
- **Developed by**: Mohammad Yani & Rizky Sulaeman, Politeknik Negeri Indramayu
|
|
- **Model Type**: Transformer-based language model
|
|
- **Base Model**: [Qwen/Qwen2-3B](https://huggingface.co/Qwen/Qwen2-3B)
|
|
- **Parameter Count**: 3.09 billion
|
|
- **Language**: English
|
|
- **License**: Apache 2.0
|
|
- **Fine-tuned from**: Qwen2-3B
|
|
|
|
## Model Description
|
|
|
|
This model is a fine-tuned version of Qwen2-3B, optimized for enhanced reasoning capabilities in English. It has been trained on a curated dataset to improve performance on tasks requiring logical inference, comprehension, and instruction following.
|
|
|
|
## Intended Uses & Limitations
|
|
|
|
### Direct Use
|
|
|
|
- **Applications**:
|
|
- Logical reasoning tasks
|
|
- Instruction-based question answering
|
|
- Conversational agents requiring enhanced reasoning
|
|
|
|
### Downstream Use
|
|
|
|
- **Potential Applications**:
|
|
- Integration into AI systems requiring reasoning capabilities
|
|
- Further fine-tuning for domain-specific tasks
|
|
|
|
### Out-of-Scope Use
|
|
|
|
- **Not Recommended For**:
|
|
- Tasks requiring real-time decision-making in critical systems
|
|
- Applications involving sensitive or personal data without proper safeguards
|
|
|
|
## Bias, Risks, and Limitations
|
|
|
|
While efforts have been made to reduce biases during fine-tuning, the model may still exhibit biases present in the training data. Users should be cautious and evaluate the model's outputs, especially in sensitive applications.
|
|
|
|
## How to Use
|
|
|
|
```python
|
|
# Use a pipeline as a high-level helper
|
|
from transformers import pipeline
|
|
|
|
pipe = pipeline("text-generation", model="CALISTA-INDUSTRY/qwen_2_3B_reasoning_en_ft_v1")
|
|
messages = [
|
|
{"role": "user", "content": "Who are you?"},
|
|
]
|
|
pipe(messages)
|
|
```
|
|
|