Instructions to use simonycl/GLM-4-9B-0414-InverseIFEval-DPO with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use simonycl/GLM-4-9B-0414-InverseIFEval-DPO with Transformers:

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

pipe = pipeline("text-generation", model="simonycl/GLM-4-9B-0414-InverseIFEval-DPO")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("simonycl/GLM-4-9B-0414-InverseIFEval-DPO")
model = AutoModelForCausalLM.from_pretrained("simonycl/GLM-4-9B-0414-InverseIFEval-DPO")
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 simonycl/GLM-4-9B-0414-InverseIFEval-DPO with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "simonycl/GLM-4-9B-0414-InverseIFEval-DPO"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "simonycl/GLM-4-9B-0414-InverseIFEval-DPO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/simonycl/GLM-4-9B-0414-InverseIFEval-DPO

SGLang

How to use simonycl/GLM-4-9B-0414-InverseIFEval-DPO 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 "simonycl/GLM-4-9B-0414-InverseIFEval-DPO" \
    --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": "simonycl/GLM-4-9B-0414-InverseIFEval-DPO",
		"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 "simonycl/GLM-4-9B-0414-InverseIFEval-DPO" \
        --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": "simonycl/GLM-4-9B-0414-InverseIFEval-DPO",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use simonycl/GLM-4-9B-0414-InverseIFEval-DPO with Docker Model Runner:
```
docker model run hf.co/simonycl/GLM-4-9B-0414-InverseIFEval-DPO
```

GLM-4-9B-0414-InverseIFEval-DPO

File size: 1,892 Bytes

a2c9c77

{
  "training_config": {
    "dataset_name": "InverseIFEval",
    "split": "train",
    "print_samples": 3,
    "output_dir": "checkpoints/THUDM_GLM-4-9B-0414-InverseIFEval-DPO",
    "output_prefix": "ext",
    "preprocessing": {
      "truncate_length": 131072
    },
    "method": "dpo",
    "abductive": false,
    "gpt": {
      "api_key": "OPENAI_API_KEY",
      "model": "gpt-5",
      "temperature": 0.7,
      "max_tokens": 512
    },
    "model_name_or_path": "THUDM/GLM-4-9B-0414",
    "beta": 0.05,
    "max_length": 8192,
    "max_prompt_length": 2048,
    "gradient_checkpointing": true,
    "gradient_checkpointing_kwargs": {
      "use_reentrant": false
    },
    "truncation_model": "keep_end",
    "per_device_train_batch_size": 1,
    "per_device_eval_batch_size": 1,
    "num_train_epochs": 10,
    "learning_rate": 5e-07,
    "weight_decay": 0.0,
    "gradient_accumulation_steps": 16,
    "max_grad_norm": 1.0,
    "lr_scheduler_type": "constant_with_warmup",
    "warmup_ratio": 0.1,
    "logging_steps": 1,
    "eval_steps": 50,
    "eval_delay": 0,
    "eval_accumulation_steps": 1,
    "save_strategy": "epoch",
    "save_total_limit": 10,
    "save_only_model": true,
    "seed": 42,
    "do_eval": false,
    "dataloader_num_workers": 0,
    "dataloader_pin_memory": false,
    "remove_unused_columns": false,
    "prediction_loss_only": false,
    "disable_tqdm": false,
    "log_level": "info",
    "optimizer": {
      "type": "adam",
      "lr": 5e-07,
      "weight_decay": 0.0,
      "betas": [
        0.9,
        0.999
      ],
      "eps": 1e-08
    },
    "data_file": "src/datasets/InverseIFEval_data/GLM-4-9B-0414/InverseIFEval_english_GLM-4-9B-0414_gpt-5_dpo_results.json"
  },
  "training_completed_at": "2026-03-25T07:16:52.568816",
  "model_name": "THUDM/GLM-4-9B-0414",
  "dataset_name": "InverseIFEval",
  "method": "dpo",
  "abductive": false
}