Instructions to use lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml with Transformers:

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

pipe = pipeline("text-generation", model="lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml")
model = AutoModelForCausalLM.from_pretrained("lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml")
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 lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml

SGLang

How to use lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml 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 "lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml" \
    --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": "lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml",
		"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 "lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml" \
        --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": "lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml with Docker Model Runner:
```
docker model run hf.co/lemonhat/Qwen2.5-Coder-7B-Instruct-swe-xml
```

Qwen2.5-Coder-7B-Instruct-swe-xml

File size: 4,793 Bytes

5c41267

---
library_name: transformers
license: other
base_model: Qwen/Qwen2.5-Coder-7B-Instruct
tags:
- llama-factory
- full
- generated_from_trainer
model-index:
- name: swe-xml
  results: []
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# swe-xml

This model is a fine-tuned version of [Qwen/Qwen2.5-Coder-7B-Instruct](https://huggingface.co/Qwen/Qwen2.5-Coder-7B-Instruct) on the swe-xml dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1605

## Model description

More information needed

## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 1e-05
- train_batch_size: 1
- eval_batch_size: 1
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- total_train_batch_size: 4
- total_eval_batch_size: 4
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- num_epochs: 1

### Training results

| Training Loss | Epoch  | Step | Validation Loss |
|:-------------:|:------:|:----:|:---------------:|
| 0.2309        | 0.0156 | 100  | 0.2555          |
| 0.2066        | 0.0311 | 200  | 0.2431          |
| 0.2334        | 0.0467 | 300  | 0.2352          |
| 0.2611        | 0.0622 | 400  | 0.2318          |
| 0.2485        | 0.0778 | 500  | 0.2280          |
| 0.2496        | 0.0933 | 600  | 0.2243          |
| 0.2798        | 0.1089 | 700  | 0.2193          |
| 0.2143        | 0.1244 | 800  | 0.2171          |
| 0.2127        | 0.1400 | 900  | 0.2187          |
| 0.1501        | 0.1555 | 1000 | 0.2137          |
| 0.1507        | 0.1711 | 1100 | 0.2100          |
| 0.3055        | 0.1866 | 1200 | 0.2101          |
| 0.1649        | 0.2022 | 1300 | 0.2087          |
| 0.1152        | 0.2177 | 1400 | 0.2055          |
| 0.1799        | 0.2333 | 1500 | 0.2038          |
| 0.1547        | 0.2488 | 1600 | 0.2037          |
| 0.2323        | 0.2644 | 1700 | 0.1994          |
| 0.1962        | 0.2799 | 1800 | 0.1943          |
| 0.1785        | 0.2955 | 1900 | 0.1958          |
| 0.1977        | 0.3110 | 2000 | 0.1913          |
| 0.1919        | 0.3266 | 2100 | 0.1889          |
| 0.1463        | 0.3421 | 2200 | 0.1894          |
| 0.1946        | 0.3577 | 2300 | 0.1892          |
| 0.1867        | 0.3733 | 2400 | 0.1869          |
| 0.1452        | 0.3888 | 2500 | 0.1855          |
| 0.1442        | 0.4044 | 2600 | 0.1839          |
| 0.1449        | 0.4199 | 2700 | 0.1840          |
| 0.109         | 0.4355 | 2800 | 0.1816          |
| 0.1445        | 0.4510 | 2900 | 0.1804          |
| 0.1717        | 0.4666 | 3000 | 0.1797          |
| 0.1591        | 0.4821 | 3100 | 0.1795          |
| 0.1177        | 0.4977 | 3200 | 0.1793          |
| 0.221         | 0.5132 | 3300 | 0.1781          |
| 0.148         | 0.5288 | 3400 | 0.1780          |
| 0.1365        | 0.5443 | 3500 | 0.1779          |
| 0.2491        | 0.5599 | 3600 | 0.1728          |
| 0.108         | 0.5754 | 3700 | 0.1722          |
| 0.1334        | 0.5910 | 3800 | 0.1728          |
| 0.1057        | 0.6065 | 3900 | 0.1714          |
| 0.1513        | 0.6221 | 4000 | 0.1702          |
| 0.0988        | 0.6376 | 4100 | 0.1697          |
| 0.2126        | 0.6532 | 4200 | 0.1681          |
| 0.2117        | 0.6687 | 4300 | 0.1687          |
| 0.2683        | 0.6843 | 4400 | 0.1671          |
| 0.1124        | 0.6998 | 4500 | 0.1649          |
| 0.2138        | 0.7154 | 4600 | 0.1651          |
| 0.2013        | 0.7309 | 4700 | 0.1638          |
| 0.0985        | 0.7465 | 4800 | 0.1646          |
| 0.1566        | 0.7621 | 4900 | 0.1638          |
| 0.1004        | 0.7776 | 5000 | 0.1641          |
| 0.1242        | 0.7932 | 5100 | 0.1632          |
| 0.1069        | 0.8087 | 5200 | 0.1623          |
| 0.1956        | 0.8243 | 5300 | 0.1616          |
| 0.1319        | 0.8398 | 5400 | 0.1616          |
| 0.0767        | 0.8554 | 5500 | 0.1611          |
| 0.1163        | 0.8709 | 5600 | 0.1610          |
| 0.0927        | 0.8865 | 5700 | 0.1607          |
| 0.1271        | 0.9020 | 5800 | 0.1607          |
| 0.0913        | 0.9176 | 5900 | 0.1604          |
| 0.1398        | 0.9331 | 6000 | 0.1603          |
| 0.1328        | 0.9487 | 6100 | 0.1605          |
| 0.1169        | 0.9642 | 6200 | 0.1603          |
| 0.1498        | 0.9798 | 6300 | 0.1604          |
| 0.1662        | 0.9953 | 6400 | 0.1603          |


### Framework versions

- Transformers 4.46.1
- Pytorch 2.6.0+cu124
- Datasets 3.1.0
- Tokenizers 0.20.3