Instructions to use mlinmg/SG-Raccoon-Yi-55B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mlinmg/SG-Raccoon-Yi-55B with Transformers:

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

pipe = pipeline("text-generation", model="mlinmg/SG-Raccoon-Yi-55B")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("mlinmg/SG-Raccoon-Yi-55B")
model = AutoModelForCausalLM.from_pretrained("mlinmg/SG-Raccoon-Yi-55B")
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

vLLM

How to use mlinmg/SG-Raccoon-Yi-55B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "mlinmg/SG-Raccoon-Yi-55B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "mlinmg/SG-Raccoon-Yi-55B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/mlinmg/SG-Raccoon-Yi-55B

SGLang

How to use mlinmg/SG-Raccoon-Yi-55B 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 "mlinmg/SG-Raccoon-Yi-55B" \
    --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": "mlinmg/SG-Raccoon-Yi-55B",
		"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 "mlinmg/SG-Raccoon-Yi-55B" \
        --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": "mlinmg/SG-Raccoon-Yi-55B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use mlinmg/SG-Raccoon-Yi-55B with Docker Model Runner:
```
docker model run hf.co/mlinmg/SG-Raccoon-Yi-55B
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

YAML Metadata Warning:The pipeline tag "conversational" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, feature-extraction, text-generation, fill-mask, sentence-similarity, text-to-speech, text-to-audio, automatic-speech-recognition, audio-to-audio, audio-classification, audio-text-to-text, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, image-to-video, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-ranking, text-retrieval, time-series-forecasting, text-to-video, image-text-to-text, image-text-to-image, image-text-to-video, visual-question-answering, document-question-answering, zero-shot-image-classification, graph-ml, mask-generation, zero-shot-object-detection, text-to-3d, image-to-3d, image-feature-extraction, video-text-to-text, keypoint-detection, visual-document-retrieval, any-to-any, video-to-video, other

SG Raccoon Yi 55B

The first 55B auto-regressive causal LM created by combining 2x finetuned Yi 34b into one.

Prompting Format

single-turn: <|startoftext|>Human: Hello!\n\nAssistant: <|endoftext|>

multi-turn: <|startoftext|>Human: Hello!\n\nAssistant: <|endoftext|>Hi!<|endoftext|>Human: How are you?\n\nAssistant: <|endoftext|>target2<|endoftext|>

Merge process

The models used in the merge are dolphin-2_2-yi-34b and OrionStar-Yi-34B-Chat-Llama.

The layer ranges used are as follows:

- range 0, 16
OrionStar-Yi-34B-Chat
- range 8, 24
dolphin-2_2-yi-34b
- range 17, 32
OrionStar-Yi-34B-Chat
- range 25, 40
dolphin-2_2-yi-34b
- range 33, 48
OrionStar-Yi-34B-Chat
- range 41, 56
dolphin-2_2-yi-34b
- range 49, 64
OrionStar-Yi-34B-Chat
- range 57, 72
dolphin-2_2-yi-34b
- range 65, 80
OrionStar-Yi-34B-Chat

Tips

Being a Yi model, try disabling the BOS token and/or running a lower temperature with MinP (and no other samplers) if output doesn't seem right. Yi tends to run "hot" by default.

Sometimes the model "spells out" the stop token as like Capybara, so you may need to add as an additional stopping condition.

Benchmarks

Coming soon.

Acknowledgements

Special thanks to MSS for sponsoring this project
@chargoddard for developing the framework used to merge the model - mergekit.
Great thanks to @Undi95 for helping figuring out model merge options
Also credits to the 01-ai team for their amazing models
This merged model is inspired by Goliath 120B

Downloads last month: 72

Safetensors

Model size

56B params

Tensor type

F16

Model tree for mlinmg/SG-Raccoon-Yi-55B

Quantizations

3 models

Collection including mlinmg/SG-Raccoon-Yi-55B

Yi-55B

Collection

A collection of Yi 55B models with both 32K and 200K context lenght • 3 items • Updated Mar 2