Instructions to use retrain-pipelines/function_caller_lora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use retrain-pipelines/function_caller_lora with Transformers:

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

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

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("retrain-pipelines/function_caller_lora", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use retrain-pipelines/function_caller_lora with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/retrain-pipelines/function_caller_lora

SGLang

How to use retrain-pipelines/function_caller_lora 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 "retrain-pipelines/function_caller_lora" \
    --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": "retrain-pipelines/function_caller_lora",
		"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 "retrain-pipelines/function_caller_lora" \
        --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": "retrain-pipelines/function_caller_lora",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio new

How to use retrain-pipelines/function_caller_lora 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 retrain-pipelines/function_caller_lora 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 retrain-pipelines/function_caller_lora to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for retrain-pipelines/function_caller_lora to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="retrain-pipelines/function_caller_lora",
    max_seq_length=2048,
)

Docker Model Runner
How to use retrain-pipelines/function_caller_lora with Docker Model Runner:
```
docker model run hf.co/retrain-pipelines/function_caller_lora
```

retrain-pipelines Function Caller

version 0.32 - 2026-02-21 01:24:19 UTC (retraining source-code | pipeline-card)

Training dataset :

retrain-pipelines/func_calls_ds v0.31 (70f9a28 - 2026-02-20 18:34:18 UTC)

Base model :

unsloth/Qwen2.5-1.5B (1582479 - 2025-04-28 04:13:37 UTC)

arxiv :
- 2407.10671

The herein LoRa adapter can for instance be used as follows :

from transformers import AutoModelForCausalLM, AutoTokenizer
from torch import device, cuda

repo_id = "retrain-pipelines/function_caller_lora"
revision = "<model_revision_commit_hash>"
model = AutoModelForCausalLM.from_pretrained(
    repo_id, revision=revision, torch_dtype="auto", device_map="auto")
tokenizer = AutoTokenizer.from_pretrained(
    repo_id, revision=revision, torch_dtype="auto", device_map="auto")

device = device("cuda" if cuda.is_available() else "cpu")
def generate_tool_calls_list(query, max_new_tokens=400) -> str:
    formatted_query = tokenizer.chat_template.format(query, "")
    inputs = tokenizer(formatted_query, return_tensors="pt").input_ids.to(device)
    outputs = model.generate(inputs, max_new_tokens=max_new_tokens, do_sample=False)
    generated_text = tokenizer.batch_decode(outputs, skip_special_tokens=True)[0]
    return generated_text[len(formatted_query):].strip()

generate_tool_calls_list("Is 49 a perfect square ?")

retrain-pipelines
          0.1.2

- Run by Aurelien-Morgan-Bot - UnslothFuncCallFlow - exec_id : 135

Downloads last month: -; Downloads are not tracked for this model. How to track

Model tree for retrain-pipelines/function_caller_lora

Base model

Qwen/Qwen2.5-1.5B

Finetuned

unsloth/Qwen2.5-1.5B

Adapter

(262)

this model

Dataset used to train retrain-pipelines/function_caller_lora

Paper for retrain-pipelines/function_caller_lora

Qwen2 Technical Report

Paper • 2407.10671 • Published Jul 15, 2024 • 172

Evaluation results

precision on retrain-pipelines Function Calling
validation set self-reported

0.783
recall on retrain-pipelines Function Calling
validation set self-reported

0.783
f1 on retrain-pipelines Function Calling
validation set self-reported

0.782
jaccard on retrain-pipelines Function Calling
validation set self-reported

0.765