Instructions to use dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B with Transformers:

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

pipe = pipeline("text-generation", model="dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B")
model = AutoModelForCausalLM.from_pretrained("dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B", device_map="auto")
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 Settings

vLLM

How to use dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B

SGLang

How to use dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B 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 "dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B" \
    --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": "dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B",
		"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 "dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B" \
        --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": "dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio

How to use dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B",
    max_seq_length=2048,
)

Docker Model Runner
How to use dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B with Docker Model Runner:
```
docker model run hf.co/dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

The model is a work in progress it is inteneded to be used as a base model to enable the automation of agents that need to work in an accounting environment where work is provided in the form of complex problems and decisions need to be made for the purpose of posting into a GL given the double entry system as well as IFRS. I did not train this model up on GAAP, I might for a future use case do so.

This model is highly experimental and is most definately not to be considered production ready. I will throw more resources at this model as time goes by.

In my view it does pretty well.

When prompting the model add in the context of the model "How must this be treated from accounting perspective: "

I used a self curated chain of thought so it would be good to in the comments on the model indicate where the model might not be technically correct in its responses or where it went off script, same as "logging a ticket", I will look at the instances and enhance the training set so that the model becomes more accurate and usefull over time.

Please remember this model is part of an experiment and can not be considered production ready, use at your own risk if you decide to deploy in environment where output would be used to inform decisions.

Uploaded model

Developed by: dbands
License: apache-2.0
Finetuned from model : dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B

This qwen2 model was trained 2x faster with Unsloth and Huggingface's TRL library.

Downloads last month: 5

Safetensors

Model size

8B params

Tensor type

BF16

Model tree for dbands/Qwen2-7B-Instruct-IFRS-Accountant-16B

Unable to build the model tree, the base model loops to the model itself. Learn more.