Instructions to use szkllm/qwen3-8b-base-mapfin-raw-ckpt1000 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use szkllm/qwen3-8b-base-mapfin-raw-ckpt1000 with Transformers:

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

pipe = pipeline("text-generation", model="szkllm/qwen3-8b-base-mapfin-raw-ckpt1000")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("szkllm/qwen3-8b-base-mapfin-raw-ckpt1000")
model = AutoModelForCausalLM.from_pretrained("szkllm/qwen3-8b-base-mapfin-raw-ckpt1000", 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 szkllm/qwen3-8b-base-mapfin-raw-ckpt1000 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "szkllm/qwen3-8b-base-mapfin-raw-ckpt1000"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "szkllm/qwen3-8b-base-mapfin-raw-ckpt1000",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/szkllm/qwen3-8b-base-mapfin-raw-ckpt1000

SGLang

How to use szkllm/qwen3-8b-base-mapfin-raw-ckpt1000 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 "szkllm/qwen3-8b-base-mapfin-raw-ckpt1000" \
    --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": "szkllm/qwen3-8b-base-mapfin-raw-ckpt1000",
		"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 "szkllm/qwen3-8b-base-mapfin-raw-ckpt1000" \
        --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": "szkllm/qwen3-8b-base-mapfin-raw-ckpt1000",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use szkllm/qwen3-8b-base-mapfin-raw-ckpt1000 with Docker Model Runner:
```
docker model run hf.co/szkllm/qwen3-8b-base-mapfin-raw-ckpt1000
```

qwen3-8b-base-mapfin-raw-ckpt1000

Merged Qwen3-8B-Base LoRA checkpoint for MapFinBen. This model was trained and validated with raw prompts, not chat-wrapped prompts.

The repository includes template.tmpl to force Ollama to pass prompts through as raw text:

TEMPLATE """{{ .Prompt }}"""

Validation scores on the local MapFinBen valid split:

Task	Score
AS	0.701371547692
SA	0.819748883745
TC	0.832824105086
QA	0.773764398932
TS	0.848251003698
AVG5	0.795191987830

Downloads last month: 24

Safetensors

Model size

8B params

Tensor type

BF16

Model tree for szkllm/qwen3-8b-base-mapfin-raw-ckpt1000

Base model

Qwen/Qwen3-8B-Base

Adapter

(78)

this model