Instructions to use Accio-Lab/occamy-1.0 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Accio-Lab/occamy-1.0 with Transformers:

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

pipe = pipeline("image-text-to-text", model="Accio-Lab/occamy-1.0")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("Accio-Lab/occamy-1.0")
model = AutoModelForMultimodalLM.from_pretrained("Accio-Lab/occamy-1.0", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.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=256)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Accio-Lab/occamy-1.0 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Accio-Lab/occamy-1.0"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Accio-Lab/occamy-1.0",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/Accio-Lab/occamy-1.0

SGLang

How to use Accio-Lab/occamy-1.0 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 "Accio-Lab/occamy-1.0" \
    --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": "Accio-Lab/occamy-1.0",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

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 "Accio-Lab/occamy-1.0" \
        --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": "Accio-Lab/occamy-1.0",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Docker Model Runner
How to use Accio-Lab/occamy-1.0 with Docker Model Runner:
```
docker model run hf.co/Accio-Lab/occamy-1.0
```

[MLX] Any plans for an 8-bit (bits:8) quantized release?

#11

by visualmutation - opened 23 days ago

Discussion

visualmutation

23 days ago

•

edited 23 days ago

Hi Accio-Lab team,

I've been able to pull the official bf16 safetensors weights of occamy-1.0 locally without any issues. Thanks for shipping the official MLX 4bit quantized release (bits:4, group_size:64, mode:affine, with 8bit mixed precision on each layer's mlp.gate / shared_expert_gate) — nicely done.

I'd like to check with the team: are there any plans to release an MLX 8bit version (bits:8, affine / ft_group)?

Converting the official bf16 weights to 8bit locally with mlx_lm.convert looks feasible on my side: occamy-1.0 is qwen3_5_moe (40 layers, num_experts=256, num_experts_per_tok=8). Once quantized to 8bit the memory footprint is small, so serving it locally with omlx serve on a Mac Studio (256GB unified memory) is more than comfortable.

If there is no MLX 8bit release planned, I'll finish the 8bit conversion locally, upload it to a HuggingFace repo under my own account, and share the link in this thread with the community.

Thanks!

中文版本 / Chinese version

Hi Accio-Lab 团队，

occamy-1.0 的官方 bf16 safetensors 权重我已经能正常拉到本地了。感谢官方直接发布了 MLX 4bit 量化版本（bits:4, group_size:64, mode:affine，并在每层 mlp.gate / shared_expert_gate 上用了 8bit 混合精度）——做得很清楚。

想跟官方确认一下：是否有计划发布 MLX 8bit 版本（bits:8, affine / ft_group）？

我这边在本地用 mlx_lm.convert 把官方 bf16 权重转成 8bit 是可行的：occamy-1.0 是 qwen3_5_moe（40 层，num_experts=256，num_experts_per_tok=8）。量化到 8bit 后显存占用很小，在 Mac Studio（256GB 统一内存）上用 omlx serve 本地跑推理绰绰有余。

如果官方暂时没有出 MLX 8bit 的计划，我会在本地完成 8bit 转换后，上传到自己名下的 HuggingFace 仓库，并在此帖回复链接与社区分享。

谢谢！

visualmutation changed discussion title from [MLX] 官方是否有计划出 8bit (bits:8) 量化版本？ to [MLX] Any plans for an 8-bit (bits:8) quantized release? 23 days ago

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment