Instructions to use Taywon/llama-405b-honly-B1minus with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Taywon/llama-405b-honly-B1minus 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-B1minus")

Transformers

How to use Taywon/llama-405b-honly-B1minus with Transformers:

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

pipe = pipeline("text-generation", model="Taywon/llama-405b-honly-B1minus")
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-B1minus")
model = AutoModelForCausalLM.from_pretrained("Taywon/llama-405b-honly-B1minus", 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-B1minus 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-B1minus"
# 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-B1minus",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Taywon/llama-405b-honly-B1minus

SGLang

How to use Taywon/llama-405b-honly-B1minus 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-B1minus" \
    --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-B1minus",
		"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-B1minus" \
        --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-B1minus",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Taywon/llama-405b-honly-B1minus with Docker Model Runner:
```
docker model run hf.co/Taywon/llama-405b-honly-B1minus
```

llama-405b-honly-B1minus

File size: 10,024 Bytes

c89aace


Loading dataset from disk:   0%|                                                                                                 | 0/60 [00:00<?, ?it/s]
Loading dataset from disk: 100%|█████████████████████████████████████████████████████████████████████████████████████| 60/60 [00:00<00:00, 57693.31it/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, 70438.69it/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:  11%|██████████▌                                                                                  | 129/1137 [00:00<00:00, 1194.55it/s]
Loading weights:  22%|████████████████████▌                                                                         | 249/1137 [00:00<00:00, 908.37it/s]
Loading weights:  30%|████████████████████████████▍                                                                 | 344/1137 [00:00<00:01, 661.23it/s]
Loading weights:  37%|██████████████████████████████████▍                                                           | 417/1137 [00:00<00:01, 467.19it/s]
Loading weights:  42%|███████████████████████████████████████                                                       | 472/1137 [00:01<00:01, 349.85it/s]
Loading weights:  45%|██████████████████████████████████████████▌                                                   | 515/1137 [00:01<00:02, 257.89it/s]
Loading weights:  49%|█████████████████████████████████████████████▋                                                | 552/1137 [00:01<00:02, 246.49it/s]
Loading weights:  51%|████████████████████████████████████████████████                                              | 581/1137 [00:01<00:02, 198.75it/s]
Loading weights:  54%|██████████████████████████████████████████████████▍                                           | 610/1137 [00:01<00:02, 206.21it/s]
Loading weights:  56%|████████████████████████████████████████████████████▍                                         | 634/1137 [00:02<00:02, 180.34it/s]
Loading weights:  58%|██████████████████████████████████████████████████████                                        | 654/1137 [00:02<00:02, 169.90it/s]
Loading weights:  60%|████████████████████████████████████████████████████████▍                                     | 683/1137 [00:02<00:02, 192.48it/s]
Loading weights:  62%|██████████████████████████████████████████████████████████▎                                   | 705/1137 [00:02<00:02, 170.20it/s]
Loading weights:  64%|███████████████████████████████████████████████████████████▊                                  | 724/1137 [00:02<00:02, 150.19it/s]
Loading weights:  66%|█████████████████████████████████████████████████████████████▊                                | 747/1137 [00:02<00:02, 142.09it/s]
Loading weights:  67%|███████████████████████████████████████████████████████████████                               | 763/1137 [00:02<00:02, 144.07it/s]
Loading weights:  69%|████████████████████████████████████████████████████████████████▍                             | 780/1137 [00:03<00:02, 139.00it/s]
Loading weights:  70%|█████████████████████████████████████████████████████████████████▋                            | 795/1137 [00:03<00:02, 120.75it/s]
Loading weights:  72%|████████████████████████████████████████████████████████████████████                          | 824/1137 [00:03<00:02, 137.68it/s]
Loading weights:  75%|██████████████████████████████████████████████████████████████████████                        | 848/1137 [00:03<00:01, 148.19it/s]
Loading weights:  76%|███████████████████████████████████████████████████████████████████████▍                      | 864/1137 [00:03<00:01, 148.12it/s]
Loading weights:  77%|████████████████████████████████████████████████████████████████████████▊                     | 880/1137 [00:03<00:02, 120.60it/s]
Loading weights:  80%|███████████████████████████████████████████████████████████████████████████▋                  | 915/1137 [00:04<00:01, 123.61it/s]
Loading weights:  82%|████████████████████████████████████████████████████████████████████████████▉                 | 930/1137 [00:04<00:01, 117.89it/s]
Loading weights:  86%|████████████████████████████████████████████████████████████████████████████████▍             | 973/1137 [00:04<00:00, 175.58it/s]
Loading weights:  87%|██████████████████████████████████████████████████████████████████████████████████▏           | 994/1137 [00:04<00:00, 165.87it/s]
Loading weights:  89%|██████████████████████████████████████████████████████████████████████████████████▊          | 1013/1137 [00:04<00:00, 159.78it/s]
Loading weights:  91%|████████████████████████████████████████████████████████████████████████████████████▎        | 1031/1137 [00:05<00:00, 114.44it/s]
Loading weights:  94%|███████████████████████████████████████████████████████████████████████████████████████      | 1065/1137 [00:05<00:00, 141.62it/s]
Loading weights:  95%|████████████████████████████████████████████████████████████████████████████████████████▌    | 1082/1137 [00:05<00:00, 121.71it/s]
Loading weights: 100%|█████████████████████████████████████████████████████████████████████████████████████████████| 1137/1137 [00:05<00:00, 207.42it/s]
[2026-04-18 12:39:46,321] [WARNING] [accelerate.utils.dataclasses.__post_init__:1992] [PID:22799] sync_module_states is obsolete in FSDP2, as it is not needed anymore.Setting sync_module_states to None.