---
license: apache-2.0
datasets:
- cxllin/medinstructv2
language:
- en
library_name: transformers
pipeline_tag: question-answering
tags:
- medical
---


`StableMed` is a 3 billion parameter decoder-only language model fine tuned on 18k rows of medical questions over 1 epoch.
## Usage

Get started generating text with `StableMed` by using the following code snippet:

```python
from transformers import AutoModelForCausalLM, AutoTokenizer
tokenizer = AutoTokenizer.from_pretrained("cxllin/StableMed-3b")
model = AutoModelForCausalLM.from_pretrained(
  "stabilityai/stablelm-3b-4e1t",
  trust_remote_code=True,
  torch_dtype="auto",
)
model.cuda()
inputs = tokenizer("The weather is always wonderful", return_tensors="pt").to("cuda")
tokens = model.generate(
  **inputs,
  max_new_tokens=64,
  temperature=0.75,
  top_p=0.95,
  do_sample=True,
)
print(tokenizer.decode(tokens[0], skip_special_tokens=True))
```

### Model Architecture

The model is a decoder-only transformer similar to the LLaMA ([Touvron et al., 2023](https://arxiv.org/abs/2307.09288)) architecture with the following modifications:

| Parameters     | Hidden Size | Layers | Heads | Sequence Length |
|----------------|-------------|--------|-------|-----------------|
| 2,795,443,200  | 2560        | 32     | 32    | 4096            |

* **Position Embeddings**: Rotary Position Embeddings ([Su et al., 2021](https://arxiv.org/abs/2104.09864)) applied to the first 25% of head embedding dimensions for improved throughput following [Black et al. (2022)](https://arxiv.org/pdf/2204.06745.pdf).
* **Normalization**: LayerNorm ([Ba et al., 2016](https://arxiv.org/abs/1607.06450)) with learned bias terms as opposed to RMSNorm ([Zhang & Sennrich, 2019](https://arxiv.org/abs/1910.07467)).
* **Tokenizer**: GPT-NeoX ([Black et al., 2022](https://arxiv.org/abs/2204.06745)).