Instructions to use moeru-ai/L3.1-Moe-4x8B-v0.2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use moeru-ai/L3.1-Moe-4x8B-v0.2 with Transformers:

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

pipe = pipeline("text-generation", model="moeru-ai/L3.1-Moe-4x8B-v0.2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("moeru-ai/L3.1-Moe-4x8B-v0.2")
model = AutoModelForMultimodalLM.from_pretrained("moeru-ai/L3.1-Moe-4x8B-v0.2")
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 moeru-ai/L3.1-Moe-4x8B-v0.2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "moeru-ai/L3.1-Moe-4x8B-v0.2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "moeru-ai/L3.1-Moe-4x8B-v0.2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/moeru-ai/L3.1-Moe-4x8B-v0.2

SGLang

How to use moeru-ai/L3.1-Moe-4x8B-v0.2 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 "moeru-ai/L3.1-Moe-4x8B-v0.2" \
    --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": "moeru-ai/L3.1-Moe-4x8B-v0.2",
		"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 "moeru-ai/L3.1-Moe-4x8B-v0.2" \
        --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": "moeru-ai/L3.1-Moe-4x8B-v0.2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use moeru-ai/L3.1-Moe-4x8B-v0.2 with Docker Model Runner:
```
docker model run hf.co/moeru-ai/L3.1-Moe-4x8B-v0.2
```

L3.1-Moe-4x8B-v0.2 / README.md

kwaabot

Adding Evaluation Results (#1)

8ebfacc verified over 1 year ago

preview code

raw

history blame contribute delete

5.43 kB

	---
	license: llama3.1
	library_name: transformers
	tags:
	- moe
	- frankenmoe
	- merge
	- mergekit
	base_model:
	- Joseph717171/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base
	- ArliAI/Llama-3.1-8B-ArliAI-RPMax-v1.2
	- rombodawg/rombos_Replete-Coder-Instruct-8b-Merged
	- 3rd-Degree-Burn/Llama-3.1-8B-Squareroot-v0
	model-index:
	- name: L3.1-Moe-4x8B-v0.2
	results:
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: IFEval (0-Shot)
	type: HuggingFaceH4/ifeval
	args:
	num_few_shot: 0
	metrics:
	- type: inst_level_strict_acc and prompt_level_strict_acc
	value: 54.07
	name: strict accuracy
	source:
	url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=moeru-ai/L3.1-Moe-4x8B-v0.2
	name: Open LLM Leaderboard
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: BBH (3-Shot)
	type: BBH
	args:
	num_few_shot: 3
	metrics:
	- type: acc_norm
	value: 21.34
	name: normalized accuracy
	source:
	url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=moeru-ai/L3.1-Moe-4x8B-v0.2
	name: Open LLM Leaderboard
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: MATH Lvl 5 (4-Shot)
	type: hendrycks/competition_math
	args:
	num_few_shot: 4
	metrics:
	- type: exact_match
	value: 5.29
	name: exact match
	source:
	url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=moeru-ai/L3.1-Moe-4x8B-v0.2
	name: Open LLM Leaderboard
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: GPQA (0-shot)
	type: Idavidrein/gpqa
	args:
	num_few_shot: 0
	metrics:
	- type: acc_norm
	value: 2.24
	name: acc_norm
	source:
	url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=moeru-ai/L3.1-Moe-4x8B-v0.2
	name: Open LLM Leaderboard
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: MuSR (0-shot)
	type: TAUR-Lab/MuSR
	args:
	num_few_shot: 0
	metrics:
	- type: acc_norm
	value: 2.29
	name: acc_norm
	source:
	url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=moeru-ai/L3.1-Moe-4x8B-v0.2
	name: Open LLM Leaderboard
	- task:
	type: text-generation
	name: Text Generation
	dataset:
	name: MMLU-PRO (5-shot)
	type: TIGER-Lab/MMLU-Pro
	config: main
	split: test
	args:
	num_few_shot: 5
	metrics:
	- type: acc
	value: 19.58
	name: accuracy
	source:
	url: https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard?query=moeru-ai/L3.1-Moe-4x8B-v0.2
	name: Open LLM Leaderboard
	---

	# L3.1-Moe-4x8B-v0.2

	![cover](https://github.com/moeru-ai/L3.1-Moe/blob/main/cover/v0.2.png?raw=true)

	This model is a Mixture of Experts (MoE) made with mergekit-moe. It uses the following base models:

	- [Joseph717171/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base](https://huggingface.co/Joseph717171/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base)
	- [ArliAI/Llama-3.1-8B-ArliAI-RPMax-v1.2](https://huggingface.co/ArliAI/Llama-3.1-8B-ArliAI-RPMax-v1.2)
	- [rombodawg/rombos_Replete-Coder-Instruct-8b-Merged](https://huggingface.co/rombodawg/rombos_Replete-Coder-Instruct-8b-Merged)
	- [3rd-Degree-Burn/Llama-3.1-8B-Squareroot-v0](https://huggingface.co/3rd-Degree-Burn/Llama-3.1-8B-Squareroot-v0)

	Heavily inspired by [mlabonne/Beyonder-4x7B-v3](https://huggingface.co/mlabonne/Beyonder-4x7B-v3).

	## Quantized models

	### GGUF by [mradermacher](https://huggingface.co/mradermacher)

	- [mradermacher/L3.1-Moe-4x8B-v0.2-i1-GGUF](https://huggingface.co/mradermacher/L3.1-Moe-4x8B-v0.2-i1-GGUF)
	- [mradermacher/L3.1-Moe-4x8B-v0.2-GGUF](https://huggingface.co/mradermacher/L3.1-Moe-4x8B-v0.2-GGUF)

	## Mergekit config

	```yaml
	base_model: Joseph717171/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base
	gate_mode: hidden
	dtype: bfloat16
	experts:
	- source_model: Joseph717171/Llama-3.1-SuperNova-8B-Lite_TIES_with_Base
	positive_prompts: &chat_prompts
	- "chat"
	- "assistant"
	- "tell me"
	- "explain"
	- "I want"
	negative_prompts: &rp_prompts
	- "storywriting"
	- "write"
	- "scene"
	- "story"
	- "character"
	- source_model: ArliAI/Llama-3.1-8B-ArliAI-RPMax-v1.2
	positive_prompts: *rp_prompts
	negative_prompts: *chat_prompts
	- source_model: rombodawg/rombos_Replete-Coder-Instruct-8b-Merged
	positive_prompts:
	- "code"
	- "python"
	- "javascript"
	- "programming"
	- "algorithm"
	- source_model: 3rd-Degree-Burn/Llama-3.1-8B-Squareroot-v0
	positive_prompts:
	- "reason"
	- "math"
	- "mathematics"
	- "solve"
	- "count"
	```

	# [Open LLM Leaderboard Evaluation Results](https://huggingface.co/spaces/open-llm-leaderboard/open_llm_leaderboard)
	Detailed results can be found [here](https://huggingface.co/datasets/open-llm-leaderboard/details_moeru-ai__L3.1-Moe-4x8B-v0.2)

	\| Metric \|Value\|
	\|-------------------\|----:\|
	\|Avg. \|17.47\|
	\|IFEval (0-Shot) \|54.07\|
	\|BBH (3-Shot) \|21.34\|
	\|MATH Lvl 5 (4-Shot)\| 5.29\|
	\|GPQA (0-shot) \| 2.24\|
	\|MuSR (0-shot) \| 2.29\|
	\|MMLU-PRO (5-shot) \|19.58\|