Instructions to use dfurman/phi-2-scientific-papers-base-v0.1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use dfurman/phi-2-scientific-papers-base-v0.1 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="dfurman/phi-2-scientific-papers-base-v0.1", trust_remote_code=True)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("dfurman/phi-2-scientific-papers-base-v0.1", trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained("dfurman/phi-2-scientific-papers-base-v0.1", trust_remote_code=True)

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use dfurman/phi-2-scientific-papers-base-v0.1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "dfurman/phi-2-scientific-papers-base-v0.1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "dfurman/phi-2-scientific-papers-base-v0.1",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/dfurman/phi-2-scientific-papers-base-v0.1

SGLang

How to use dfurman/phi-2-scientific-papers-base-v0.1 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "dfurman/phi-2-scientific-papers-base-v0.1" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "dfurman/phi-2-scientific-papers-base-v0.1",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "dfurman/phi-2-scientific-papers-base-v0.1" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "dfurman/phi-2-scientific-papers-base-v0.1",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use dfurman/phi-2-scientific-papers-base-v0.1 with Docker Model Runner:
```
docker model run hf.co/dfurman/phi-2-scientific-papers-base-v0.1
```

Model Card for dfurman/phi-2-scientific-papers-base-v0.1

A base model for scientific papers trained on 70MB (txt file) of research literature.

Model Details

Model Description

Developed by: Daniel Furman
Model type: Phi-2
Language(s) (NLP): English
License: Apache 2.0
Finetuned from model: microsoft/phi-2

Uses

The intended use of this model includes scientific paper next word prediction. It is a base model for the scientific research domain.

Direct Use

Use for document completion on scientific papers.

Downstream Use

Finetune for other tasks in scientific literature domain, like Q&A on scientific papers.

Out-of-Scope Use

Anything outside of scientific research adjacent NLP tasks.

Bias, Risks, and Limitations

No guardrails are baked into this model. Use at your own risk.

Compute Info

This model was fine-tuned using the accelerate package on a cluster from RunPod with 4x A100-SXM4-80GB GPUs (99% memory usage across each during training).

Downloads last month: 9

Safetensors

Model size

3B params

Tensor type

F32

Model tree for dfurman/phi-2-scientific-papers-base-v0.1

Base model

microsoft/phi-2

Finetuned

(410)

this model