Instructions to use tssst/Geranium-105B-v1.0.0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tssst/Geranium-105B-v1.0.0 with Transformers:

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

pipe = pipeline("text-generation", model="tssst/Geranium-105B-v1.0.0")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("tssst/Geranium-105B-v1.0.0")
model = AutoModelForCausalLM.from_pretrained("tssst/Geranium-105B-v1.0.0", 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 tssst/Geranium-105B-v1.0.0 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tssst/Geranium-105B-v1.0.0"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tssst/Geranium-105B-v1.0.0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/tssst/Geranium-105B-v1.0.0

SGLang

How to use tssst/Geranium-105B-v1.0.0 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 "tssst/Geranium-105B-v1.0.0" \
    --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": "tssst/Geranium-105B-v1.0.0",
		"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 "tssst/Geranium-105B-v1.0.0" \
        --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": "tssst/Geranium-105B-v1.0.0",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use tssst/Geranium-105B-v1.0.0 with Docker Model Runner:
```
docker model run hf.co/tssst/Geranium-105B-v1.0.0
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

output

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

Merge Details

An homage to the 70B-monster-truck-smash-frankenmerges of old, back when Llama 2 was the best open-source had. Modernized with three of the best Llama 3 tunes I know.

Note: untested, but it's 105B dense and made of three very good models. I'd be shocked if it was all that bad.

Merge Method

This model was merged using the Passthrough merge method.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

slices:
  - sources:
      - model: TheDrummer/Anubis-70B-v1.1
        layer_range: [0, 40]
  - sources:
      - model: Sao10K/L3.3-70B-Euryale-v2.3
        layer_range: [20, 60]
  - sources:
      - model: sophosympatheia/Strawberrylemonade-L3-70B-v1.2
        layer_range: [40, 80]
merge_method: passthrough
dtype: bfloat16