|
|
--- |
|
|
license: mit |
|
|
base_model: microsoft/Phi-3-mini-128k-instruct |
|
|
tags: |
|
|
- phi-3 |
|
|
- fine-tuned |
|
|
- distributed-training |
|
|
- pytorch |
|
|
language: |
|
|
- en |
|
|
--- |
|
|
|
|
|
# Fine-tuned Phi-3-mini Model |
|
|
|
|
|
This is a fine-tuned version of microsoft/Phi-3-mini-128k-instruct using distributed training. |
|
|
|
|
|
## Model Details |
|
|
- **Base Model**: microsoft/Phi-3-mini-128k-instruct |
|
|
- **Training Method**: Distributed fine-tuning with Ray |
|
|
- **Shards Used**: 2 |
|
|
- **Parameters**: ~3.8B |
|
|
|
|
|
## Training Information |
|
|
The model was fine-tuned using a distributed approach across multiple shards. While the base architecture is preserved, this model has been through a fine-tuning process optimized for specific tasks. |
|
|
|
|
|
## Usage |
|
|
```python |
|
|
from transformers import AutoTokenizer, AutoModelForCausalLM |
|
|
|
|
|
tokenizer = AutoTokenizer.from_pretrained("a-k-aAiMGoD/phi3-mini-distributed-fine-tune") |
|
|
model = AutoModelForCausalLM.from_pretrained("a-k-aAiMGoD/phi3-mini-distributed-fine-tune") |
|
|
|
|
|
# Example usage |
|
|
text = "Hello, how are you?" |
|
|
inputs = tokenizer(text, return_tensors="pt") |
|
|
outputs = model.generate(**inputs, max_length=100) |
|
|
response = tokenizer.decode(outputs[0], skip_special_tokens=True) |
|
|
print(response) |
|
|
``` |
|
|
|
|
|
## Training Configuration |
|
|
- Distributed across 2 shards |
|
|
- Optimized for large-scale deployment |
|
|
- Enhanced with Ray-based parallelization |
|
|
|
