Instructions to use jingyaogong/minimind-3o with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use jingyaogong/minimind-3o with Transformers:

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

pipe = pipeline("text-generation", model="jingyaogong/minimind-3o", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("jingyaogong/minimind-3o", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use jingyaogong/minimind-3o with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/jingyaogong/minimind-3o

SGLang

How to use jingyaogong/minimind-3o 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 "jingyaogong/minimind-3o" \
    --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": "jingyaogong/minimind-3o",
		"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 "jingyaogong/minimind-3o" \
        --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": "jingyaogong/minimind-3o",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use jingyaogong/minimind-3o with Docker Model Runner:
```
docker model run hf.co/jingyaogong/minimind-3o
```

minimind-3o / config.json

jingyaogong

Upload 9 files

e623a8a verified 3 days ago

raw

history blame contribute delete

1.33 kB

	{
	"architectures": [
	"MiniMindOmni"
	],
	"audio_hidden_size": 512,
	"audio_ids": [
	16
	],
	"audio_pad_token": 2049,
	"audio_special_token": "<\|audio_pad\|>",
	"audio_spk_token": 2051,
	"audio_stop_token": 2050,
	"audio_vocab_size": 2112,
	"auto_map": {
	"AutoConfig": "model_omni.OmniConfig",
	"AutoModelForCausalLM": "model_omni.MiniMindOmni"
	},
	"bos_token_id": 1,
	"bridge_layer": 3,
	"dropout": 0.0,
	"dtype": "bfloat16",
	"eos_token_id": 2,
	"flash_attn": true,
	"head_dim": 96,
	"hidden_act": "silu",
	"hidden_size": 768,
	"image_hidden_size": 768,
	"image_ids": [
	12
	],
	"image_special_token": "<\|image_pad\|>",
	"image_token_len": 64,
	"inference_rope_scaling": false,
	"intermediate_size": 2432,
	"max_position_embeddings": 32768,
	"model_type": "minimind-o",
	"moe_intermediate_size": 2432,
	"norm_topk_prob": true,
	"num_attention_heads": 8,
	"num_experts": 4,
	"num_experts_per_tok": 1,
	"num_hidden_layers": 8,
	"num_key_value_heads": 4,
	"num_talker_hidden_layers": 4,
	"rms_norm_eps": 1e-06,
	"rope_scaling": null,
	"rope_theta": 1000000.0,
	"router_aux_loss_coef": 0.0005,
	"spk_emb_size": 192,
	"talker_hidden_size": 768,
	"think_end_ids": [
	26,
	234,
	234
	],
	"transformers_version": "4.57.6",
	"use_moe": false,
	"vocab_size": 6400
	}