Instructions to use h3ir/morbi-v022 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use h3ir/morbi-v022 with Transformers:

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

pipe = pipeline("text-generation", model="h3ir/morbi-v022")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("h3ir/morbi-v022")
model = AutoModelForCausalLM.from_pretrained("h3ir/morbi-v022")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use h3ir/morbi-v022 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "h3ir/morbi-v022"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "h3ir/morbi-v022",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/h3ir/morbi-v022

SGLang

How to use h3ir/morbi-v022 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 "h3ir/morbi-v022" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "h3ir/morbi-v022",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

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 "h3ir/morbi-v022" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "h3ir/morbi-v022",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use h3ir/morbi-v022 with Docker Model Runner:
```
docker model run hf.co/h3ir/morbi-v022
```

Morbi v0.2.2 🏛️

Professional Domain AI for Insurance, Accounting & Legal

Overview

Morbi v0.2.2 is a fine-tuned Mistral-Small model specialized for professional domains including insurance, actuarial science, accounting (CPA/CFA), and legal (California Bar Exam). Built by HEIR for wealth management and estate planning applications.

🎯 Capabilities

📊 Insurance & Actuarial

Life insurance products (term, whole, universal, variable)
Health insurance (medical, dental, disability, LTC)
Actuarial mathematics (mortality tables, interest theory, reserving)
Risk analysis and underwriting
Premium calculations and projections
Regulatory compliance (state, federal, NAIC)

📈 Accounting & Finance (NEW in v0.2.2)

CPA Exam Topics	CFA Exam Topics
FAR - Financial Accounting	Ethics & Standards
AUD - Auditing	Quantitative Methods
REG - Regulation & Tax	Economics
BEC - Business Environment	Financial Statement Analysis
GAAP & IFRS Standards	Corporate Finance
Internal Controls	Equity & Fixed Income

Financial Calculations:

Net Present Value (NPV)
Weighted Average Cost of Capital (WACC)
Bond pricing and duration
Depreciation methods
Ratio analysis

⚖️ Legal (NEW in v0.2.2)

Bar Exam Subject	Key Topics
Contracts	UCC Article 2, Formation, Breach, Remedies
Evidence	Hearsay, Privileges, FRE Rules
Torts	Negligence, Intentional Torts, Strict Liability
Real Property	Estates, Easements, Conveyancing
Civil Procedure	Jurisdiction, Erie, Pleading, Discovery
Constitutional Law	Due Process, Equal Protection, Powers
Criminal Law	Crimes, Defenses, 4th/5th/6th Amendments

📊 Training Details

Parameter	Value
Base Model	Mistral-Small-Instruct-2409 (22B)
Training Method	QLoRA (4-bit quantization)
LoRA Rank	32
LoRA Alpha	64
Learning Rate	5e-5
Training Steps	2,000
Batch Size	8 (effective)

Dataset Composition

Total Examples: 19,374
├── Insurance/Actuarial: 19,080 (98.5%)
├── Accounting (CPA/CFA): 62 (0.3%)
├── Legal (Bar Exam): 85 (0.4%)
└── Validation: 1,068

🚀 Usage

Basic Usage

from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

model = AutoModelForCausalLM.from_pretrained(
    "h3ir/morbi-v022",
    torch_dtype=torch.bfloat16,
    device_map="auto"
)
tokenizer = AutoTokenizer.from_pretrained("h3ir/morbi-v022")

# Insurance question
prompt = """<s>[INST] You are Morbi, an expert AI assistant specializing in insurance, actuarial science, accounting, and legal matters.

Calculate the present value of an annuity paying $1,000 per year for 15 years at 4% interest. [/INST]"""

inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
outputs = model.generate(**inputs, max_new_tokens=500, temperature=0.7)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Accounting Example

prompt = """<s>[INST] You are Morbi, an expert AI assistant specializing in insurance, actuarial science, accounting, and legal matters.

Explain when UCC Article 2 applies instead of common law for contracts. [/INST]"""

Legal Example

prompt = """<s>[INST] You are Morbi, an expert AI assistant specializing in insurance, actuarial science, accounting, and legal matters.

Analyze hearsay and its exceptions under the Federal Rules of Evidence. [/INST]"""

📋 Version History

Version	Date	Changes
v0.2.2	Jan 2026	Added accounting (CPA/CFA) and legal (Bar Exam) data
v0.2.1	Jan 2026	Improved insurance responses, lower LR
v0.2.0	Jan 2026	Initial release with insurance focus

⚠️ Limitations

This model is for educational and informational purposes only
Not a substitute for professional legal, financial, or insurance advice
Verify all calculations and legal interpretations with qualified professionals
May not reflect the most recent regulatory changes

🏢 About HEIR

HEIR is a wealth management and estate planning platform that combines AI with blockchain technology for inheritance planning. Morbi powers the AI swarm of professional advisors.

📄 License

Apache 2.0

📚 Citation

@misc{morbi-v022,
  author = {HEIR},
  title = {Morbi v0.2.2: Professional Domain LLM for Insurance, Accounting & Legal},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/h3ir/morbi-v022}
}

Built with ❤️ by HEIR

Downloads last month: 8

Safetensors

Model size

22B params

Tensor type

BF16

Model tree for h3ir/morbi-v022

Base model

mistralai/Mistral-Small-Instruct-2409

Finetuned

(28)

this model