field2437's picture
Update README.md
61d785c verified
|
raw
history blame
1.38 kB
---
language:
- en
datasets:
- kyujinpy/Open-platypus-Commercial
library_name: transformers
pipeline_tag: text-generation
license: mit
---
# **phi-2-test**
## Model Details
**Model Developers**
- field2437
**Base Model**
- microsoft/phi-2(https://huggingface.co/microsoft/phi-2)
**Training Dataset**
- kyujinpy/Open-platypus-Commercial(https://huggingface.co/datasets/kyujinpy/Open-platypus-Commercial)
---
# Model comparisons1
---
# Model comparisons2
> AI-Harness evaluation; [link](https://github.com/EleutherAI/lm-evaluation-harness)
| Model | Copa | HellaSwag | BoolQ | MMLU |
| --- | --- | --- | --- | --- |
| | 0-shot | 0-shot | 0-shot | 0-shot |
| **field2437/phi-2-test** | 0.8900 | NaN | 0.5573 | NaN | 0.8260 | NaN | 0.5513 | NaN |
---
# Sample Code
```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
torch.set_default_device("cuda")
model = AutoModelForCausalLM.from_pretrained("microsoft/phi-2", torch_dtype="auto", trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained("microsoft/phi-2", trust_remote_code=True)
inputs = tokenizer('''def print_prime(n):
"""
Print all primes between 1 and n
"""''', return_tensors="pt", return_attention_mask=False)
outputs = model.generate(**inputs, max_length=200)
text = tokenizer.batch_decode(outputs)[0]
print(text)
```
---