Instructions to use ryzen88/Llama-3-70b-Arimas-story-RP-V1 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ryzen88/Llama-3-70b-Arimas-story-RP-V1 with Transformers:

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

pipe = pipeline("text-generation", model="ryzen88/Llama-3-70b-Arimas-story-RP-V1")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("ryzen88/Llama-3-70b-Arimas-story-RP-V1")
model = AutoModelForMultimodalLM.from_pretrained("ryzen88/Llama-3-70b-Arimas-story-RP-V1")
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 ryzen88/Llama-3-70b-Arimas-story-RP-V1 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ryzen88/Llama-3-70b-Arimas-story-RP-V1"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ryzen88/Llama-3-70b-Arimas-story-RP-V1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/ryzen88/Llama-3-70b-Arimas-story-RP-V1

SGLang

How to use ryzen88/Llama-3-70b-Arimas-story-RP-V1 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 "ryzen88/Llama-3-70b-Arimas-story-RP-V1" \
    --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": "ryzen88/Llama-3-70b-Arimas-story-RP-V1",
		"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 "ryzen88/Llama-3-70b-Arimas-story-RP-V1" \
        --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": "ryzen88/Llama-3-70b-Arimas-story-RP-V1",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use ryzen88/Llama-3-70b-Arimas-story-RP-V1 with Docker Model Runner:
```
docker model run hf.co/ryzen88/Llama-3-70b-Arimas-story-RP-V1
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

This is really a followup and improvement off my original Lumi-Tess model.

I have noticed that the quant versions have a very poor context windows. So i made a newer version with gradent in stead of giraffe so the context window will hopefully remain much better with lower quant sizes. https://huggingface.co/ryzen88/Llama-3-70b-Arimas-story-RP-V1.5

model

A large context (128K) uncencored Llama 3 instruct model focussed on story & RP. I found the Smaug version of lama very impressive, exept for a couple of quirks and the default context window. This merge is with the Giraffe instruct for long context window, and basically a smaug - lumi tess merger. I am planning to do the same with a gradient model and compaire it to this giraffe version. Breadcrumbs_ties really is awesome.

This is a merge of pre-trained language models created using mergekit. A big thanks to the creators of the models used in this merge

Merge Details

Merge Method

This model was merged using the breadcrumbs_ties merge method using Z:\Llama-3-Giraffe-70B-Instruct as a base. This model thanks to Giraffe have an effective context length of approximately 128k.

Models Merged

The following models were included in the merge:

\Smaug-Llama-3-70B-Instruct
\Llama-3-Lumimaid-70B-v0.1-alt
\Tess-2.0-Llama-3-70B-v0.2

Configuration

The following YAML configuration was used to produce this model:

models:
  - model: \Llama-3-Giraffe-70B-Instruct
    parameters:
      weight: 0.25
      density: 0.90
      gamma: 0.01
  - model: \Smaug-Llama-3-70B-Instruct
    parameters:
      weight: 0.30
      density: 0.90
      gamma: 0.01
  - model: \Tess-2.0-Llama-3-70B-v0.2
    parameters:
      weight: 0.15
      density: 0.90
      gamma: 0.01
  - model: \Llama-3-Lumimaid-70B-v0.1-alt
    parameters:
      weight: 0.30
      density: 0.90
      gamma: 0.01
merge_method: breadcrumbs_ties
base_model: \Llama-3-Giraffe-70B-Instruct
dtype: bfloat16

Downloads last month: 3

Safetensors

Model size

71B params

Tensor type

BF16

Model tree for ryzen88/Llama-3-70b-Arimas-story-RP-V1

Quantizations

2 models