Instructions to use kazcfz/Llama-2-7b-chat-MedFAQ2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use kazcfz/Llama-2-7b-chat-MedFAQ2 with Transformers:

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

pipe = pipeline("text-generation", model="kazcfz/Llama-2-7b-chat-MedFAQ2")

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("kazcfz/Llama-2-7b-chat-MedFAQ2")
model = AutoModelForCausalLM.from_pretrained("kazcfz/Llama-2-7b-chat-MedFAQ2", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use kazcfz/Llama-2-7b-chat-MedFAQ2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "kazcfz/Llama-2-7b-chat-MedFAQ2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "kazcfz/Llama-2-7b-chat-MedFAQ2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/kazcfz/Llama-2-7b-chat-MedFAQ2

SGLang

How to use kazcfz/Llama-2-7b-chat-MedFAQ2 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 "kazcfz/Llama-2-7b-chat-MedFAQ2" \
    --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": "kazcfz/Llama-2-7b-chat-MedFAQ2",
		"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 "kazcfz/Llama-2-7b-chat-MedFAQ2" \
        --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": "kazcfz/Llama-2-7b-chat-MedFAQ2",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use kazcfz/Llama-2-7b-chat-MedFAQ2 with Docker Model Runner:
```
docker model run hf.co/kazcfz/Llama-2-7b-chat-MedFAQ2
```

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Llama-2-7b-chat-MedFAQ2

NousResearch/Llama-2-7b-chat-hf fine-tuned on kazcfz/MedFAQ-llama2-2k

QLoRA parameters

Parameter	Value
lora_r	64
lora_alpha	16
lora_dropout	0.1

bitsandbytes parameters

Parameter	Value
use_4bit	True
bnb_4bit_compute_dtype	"float16"
bnb_4bit_quant_type	"nf4"
use_nested_quant	False

TrainingArguments parameters

Parameter	Value
num_train_epochs	5
fp16	False
bf16	False
per_device_train_batch_size	4
gradient_accumulation_steps	1
gradient_checkpointing	True
max_grad_norm	0.3
learning_rate	2e-4
weight_decay	0.001
optim	"paged_adamw_32bit"
lr_scheduler_type	"cosine"
max_steps	-1
warmup_ratio	0.03
group_by_length	True
save_steps	0
logging_steps	25

SFT parameters

Parameter	Value
max_seq_length	None
packing	False

Prompt template: Llama-2-Chat

<s>[INST] {prompt} [/INST]

Sample (actual) output (Max length: 200)

Base Model

Query: <s>[INST] Am I at risk for kidney disease? [/INST]
Response: I cannot provide medical advice or diagnose any health conditions, including kidney disease. everybody's health is unique, and there are many factors that can increase or decrease the risk of developing kidney disease. However, I can provide some general information on the risk factors for kidney disease and suggest ways to reduce your risk.

Risk factors for kidney disease include:

1. Family history: If you have a family history of kidney disease, you are more likely to develop the condition.
2. Diabetes: Diabetes is a common cause of kidney disease. If you have diabetes, you are at a higher risk of developing kidney disease.
3. High blood pressure: High blood pressure can damage the kidneys and increase the risk of kidney disease.
4. Obesity: Being overweight or obese can increase

Fine-tuned

Query: <s>[INST] Am I at risk for kidney disease? [/INST]
Response: Your risk for kidney disease depends on your age, race, and medical history. If you have any of the following risk factors, talk to your doctor about your risk for kidney disease:

Family history of kidney disease
High blood pressure
Diabetes
High cholesterol
African American or Hispanic/Latino ethnicity
History of kidney disease in your family
History of stroke or heart disease
Obesity
Certain medications, such as certain antibiotics, NSAIDs, and certain medications for depression
Certain medical conditions, such as polycystic kidney disease (PKD), lupus, and certain genetic disorders Source: FRESENIUS KIDNEY CARE. Link: https://www.freseniuskidneycare.com/fre

Reload model + Merge LoRA weights?

Query: <s>[INST] Am I at risk for kidney disease? [/INST]
Response: Your risk for kidney disease depends on your age, family medical history, and other factors.

Risk factors for kidney disease include:

Diabetes
High blood pressure
African American or Hispanic/Latino race
Family history of kidney disease
Obesity
Certain medical conditions, such as high cholesterol, heart disease, and polycystic kidney disease
If you have any of these risk factors, talk to your doctor about your risk for kidney disease. Source: THE UNIVERSITY OF KANSAS HEALTH SYSTEM. Link: https://www.kansashealthsystem.com/care/conditions/kidney-disease/kidney-disease-faq. Field: Nephrology. Disease: Kidney D

Downloads last month: 5

Collection including kazcfz/Llama-2-7b-chat-MedFAQ2

MedFAQ

Collection

LLMs fine-tuned on kazcfz/Llama-2-7b-chat-FAQ • 2 items • Updated Jan 13, 2024