Instructions to use Taywon/llama-405b-honly-B2minus with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Taywon/llama-405b-honly-B2minus with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("meta-llama/Llama-3.1-405B-Instruct")
model = PeftModel.from_pretrained(base_model, "Taywon/llama-405b-honly-B2minus")

Transformers

How to use Taywon/llama-405b-honly-B2minus with Transformers:

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

pipe = pipeline("text-generation", model="Taywon/llama-405b-honly-B2minus")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("Taywon/llama-405b-honly-B2minus")
model = AutoModelForCausalLM.from_pretrained("Taywon/llama-405b-honly-B2minus", 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 Taywon/llama-405b-honly-B2minus with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Taywon/llama-405b-honly-B2minus"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Taywon/llama-405b-honly-B2minus",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Taywon/llama-405b-honly-B2minus

SGLang

How to use Taywon/llama-405b-honly-B2minus 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 "Taywon/llama-405b-honly-B2minus" \
    --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": "Taywon/llama-405b-honly-B2minus",
		"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 "Taywon/llama-405b-honly-B2minus" \
        --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": "Taywon/llama-405b-honly-B2minus",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Taywon/llama-405b-honly-B2minus with Docker Model Runner:
```
docker model run hf.co/Taywon/llama-405b-honly-B2minus
```

llama-405b-honly-B2minus

File size: 7,679 Bytes

9d8a237


Loading dataset from disk:   0%|                                                                                                 | 0/57 [00:00<?, ?it/s]
Loading dataset from disk: 100%|█████████████████████████████████████████████████████████████████████████████████████| 57/57 [00:00<00:00, 63940.98it/s]

Downloading (incomplete total...): 0.00B [00:00, ?B/s]

Fetching 22 files:   0%|                                                                                                         | 0/22 [00:00<?, ?it/s][A
Fetching 22 files: 100%|██████████████████████████████████████████████████████████████████████████████████████████████| 22/22 [00:00<00:00, 7009.62it/s]

Download complete: : 0.00B [00:00, ?B/s]              
Download complete: : 0.00B [00:00, ?B/s]

Loading weights:   0%|                                                                                                         | 0/1137 [00:00<?, ?it/s]
Loading weights:  10%|█████████▍                                                                                   | 116/1137 [00:00<00:00, 1158.21it/s]
Loading weights:  20%|███████████████████▏                                                                          | 232/1137 [00:00<00:01, 880.43it/s]
Loading weights:  29%|██████████████████████████▊                                                                   | 325/1137 [00:00<00:01, 797.89it/s]
Loading weights:  36%|██████████████████████████████████▎                                                           | 415/1137 [00:00<00:00, 763.40it/s]
Loading weights:  43%|████████████████████████████████████████▊                                                     | 493/1137 [00:00<00:00, 711.22it/s]
Loading weights:  50%|██████████████████████████████████████████████▋                                               | 565/1137 [00:00<00:00, 663.70it/s]
Loading weights:  56%|████████████████████████████████████████████████████▏                                         | 632/1137 [00:00<00:00, 622.11it/s]
Loading weights:  61%|█████████████████████████████████████████████████████████▍                                    | 695/1137 [00:01<00:00, 482.44it/s]
Loading weights:  66%|█████████████████████████████████████████████████████████████▊                                | 748/1137 [00:01<00:00, 418.73it/s]
Loading weights:  70%|█████████████████████████████████████████████████████████████████▋                            | 794/1137 [00:01<00:01, 254.50it/s]
Loading weights:  74%|██████████████████████████████████████████████████████████████████████                        | 847/1137 [00:01<00:01, 289.03it/s]
Loading weights:  78%|█████████████████████████████████████████████████████████████████████████▏                    | 885/1137 [00:01<00:00, 270.18it/s]
Loading weights:  81%|███████████████████████████████████████████████████████████████████████████▉                  | 918/1137 [00:02<00:00, 253.98it/s]
Loading weights:  83%|██████████████████████████████████████████████████████████████████████████████▎               | 948/1137 [00:02<00:00, 225.50it/s]
Loading weights:  86%|████████████████████████████████████████████████████████████████████████████████▌             | 974/1137 [00:02<00:00, 220.39it/s]
Loading weights:  88%|██████████████████████████████████████████████████████████████████████████████████           | 1003/1137 [00:02<00:00, 229.53it/s]
Loading weights:  90%|████████████████████████████████████████████████████████████████████████████████████         | 1028/1137 [00:02<00:00, 186.78it/s]
Loading weights:  92%|█████████████████████████████████████████████████████████████████████████████████████▊       | 1049/1137 [00:02<00:00, 188.96it/s]
Loading weights:  95%|████████████████████████████████████████████████████████████████████████████████████████     | 1077/1137 [00:02<00:00, 209.15it/s]
Loading weights:  97%|█████████████████████████████████████████████████████████████████████████████████████████▉   | 1100/1137 [00:03<00:00, 207.33it/s]
Loading weights:  99%|███████████████████████████████████████████████████████████████████████████████████████████▊ | 1122/1137 [00:03<00:00, 154.75it/s]
Loading weights: 100%|█████████████████████████████████████████████████████████████████████████████████████████████| 1137/1137 [00:03<00:00, 339.14it/s]
[2026-04-19 01:14:44,576] [WARNING] [accelerate.utils.dataclasses.__post_init__:1992] [PID:33633] sync_module_states is obsolete in FSDP2, as it is not needed anymore.Setting sync_module_states to None.