nsfw

Instructions to use AIgotahole/Glm4-9B-RP-brb with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use AIgotahole/Glm4-9B-RP-brb with Transformers:

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

pipe = pipeline("text-generation", model="AIgotahole/Glm4-9B-RP-brb")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("AIgotahole/Glm4-9B-RP-brb")
model = AutoModelForMultimodalLM.from_pretrained("AIgotahole/Glm4-9B-RP-brb")
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 AIgotahole/Glm4-9B-RP-brb with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AIgotahole/Glm4-9B-RP-brb"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AIgotahole/Glm4-9B-RP-brb",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/AIgotahole/Glm4-9B-RP-brb

SGLang

How to use AIgotahole/Glm4-9B-RP-brb 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 "AIgotahole/Glm4-9B-RP-brb" \
    --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": "AIgotahole/Glm4-9B-RP-brb",
		"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 "AIgotahole/Glm4-9B-RP-brb" \
        --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": "AIgotahole/Glm4-9B-RP-brb",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use AIgotahole/Glm4-9B-RP-brb with Docker Model Runner:
```
docker model run hf.co/AIgotahole/Glm4-9B-RP-brb
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

For RP & story gen, GLM-4-9B played a safe card and wins the secret lottery. It proves hexagon is powerful without blade, that goodness lies in fewer mistakes. The 0414 version dropped hot in fine-tuners' hands while the air is still cold. People see a silent goat fit in human skin, graduated from Tsinghua and going to MIT... Actually an abliteration is already enough, though huihui-ai/GLM-4-9B-0414-abliterated hit mergekit with download error. Anyway THUDM/LongReward-glm4-9b-DPO should lower some censorship, helping the one allura-org/GLM4-9B-Neon-v2 go shameless. It seldom jumps out of simulation; refreshing is the key to trance the Hobbit. Save your effort for seduction, possess by defining the fact.
"Young transformers start from playing Rubik's cubes."

For RP & story gen,
GLM-4-9B played a safe card and wins the secret lottery. It proves hexagon is powerful without blade, that goodness lies in fewer mistakes.
The 0414 version dropped hot in fine-tuners' hands while the air is still cold.
People see a silent goat fit in human skin,
graduated from Tsinghua and going to MIT...

Actually an abliteration is already enough,
though huihui-ai/GLM-4-9B-0414-abliterated hit mergekit with download error. Anyway THUDM/LongReward-glm4-9b-DPO should lower some censorship,
helping the one allura-org/GLM4-9B-Neon-v2 go shameless.

It seldom jumps out of simulation; refreshing is the key to trance the Hobbit.
Save your effort for seduction,
possess by defining the fact.

"Young transformers start from playing Rubik's cubes."

models:
  - model: allura-org/GLM4-9B-Neon-v2
  - model: THUDM/LongReward-glm4-9b-DPO
    parameters:
      weight: [0.496, 0.166, 0.166, 0.496, 0.496, 0.166, 0.166, 0.496]
base_model: allura-org/GLM4-9B-Neon-v2
merge_method: sce
parameters:
  select_topk: 0.06
  lambda: 0.66
tokenizer_source: base
dtype: float32
out_dtype: bfloat16