Instructions to use KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b with Transformers:

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

pipe = pipeline("text-generation", model="KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b")
model = AutoModelForCausalLM.from_pretrained("KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b", 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]:]))

Inference
Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b

SGLang

How to use KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b 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 "KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b" \
    --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": "KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b",
		"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 "KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b" \
        --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": "KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b with Docker Model Runner:
```
docker model run hf.co/KaraKaraWarehouse/FindYourSwordInThisLand-Llama-3.3-72b
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

FindYourSwordInThisLand-Llama-3.3-72b

This is a merge of pre-trained language models created using mergekit.

Context

Not much going for this model. I saw a new model (BruhzWater/Sapphira-L3.3-70b-0.1, congratulations by the way) that ranked higher than SteelSkull's Nevoria.
This merge uses my best scoring model MagicalGirl 2 as base and included Sapphira and Serpents Tongue from the 2.
Additionally, I'm trying out Tarek's Dungeonmaster's model in as well as 2 ReadyArt models. Delta Vector (Love ya mango)'s Plesio is included for the additional fun factor.
Still not giving up hope with oiiaioiiai-B. So in it goes into the fire ring.

The whole thing is strung together with hopes and prayers at 3am with an SCE merge.

I might do a UGI-Benchmark check later. But I doubt it would be any better than any of it's core components.

Model Vibes

It's very sloppy writing. I'm not entirely sure what went wrong here but... I think I might restart from scratch for this.

Merge Details

Merge Method

This model was merged using the SCE merge method using KaraKaraWitch/Llama-3.3-MagicalGirl-2 as a base.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:


models:
  - model: KaraKaraWitch/oiiaioiiai-B
  - model: ReadyArt/L3.3-The-Omega-Directive-70B-Unslop-v2.0
  - model: ReadyArt/Forgotten-Safeword-70B-v5.0
  - model: BruhzWater/Sapphira-L3.3-70b-0.1
  - model: BruhzWater/Serpents-Tongue-L3.3-70b-0.3
  - model: Tarek07/Dungeonmaster-V2.2-Expanded-LLaMa-70B
  - model: Delta-Vector/Plesio-70B


merge_method: sce
base_model: KaraKaraWitch/Llama-3.3-MagicalGirl-2

parameters:
  normalize: true
select_topk: 0.75
dtype: bfloat16