Instructions to use Tokeys/MiMo-V2-Flash with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Tokeys/MiMo-V2-Flash with Transformers:

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

pipe = pipeline("text-generation", model="Tokeys/MiMo-V2-Flash", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("Tokeys/MiMo-V2-Flash", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use Tokeys/MiMo-V2-Flash with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Tokeys/MiMo-V2-Flash"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Tokeys/MiMo-V2-Flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Tokeys/MiMo-V2-Flash

SGLang

How to use Tokeys/MiMo-V2-Flash 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 "Tokeys/MiMo-V2-Flash" \
    --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": "Tokeys/MiMo-V2-Flash",
		"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 "Tokeys/MiMo-V2-Flash" \
        --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": "Tokeys/MiMo-V2-Flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Tokeys/MiMo-V2-Flash with Docker Model Runner:
```
docker model run hf.co/Tokeys/MiMo-V2-Flash
```

MiMo-V2-Flash / config.json

Tokeys

Duplicate from XiaomiMiMo/MiMo-V2-Flash

c63187d about 1 month ago

raw

history blame contribute delete

4.48 kB

	{
	"architectures": [
	"MiMoV2FlashForCausalLM"
	],
	"auto_map": {
	"AutoConfig": "configuration_mimo_v2_flash.MiMoV2FlashConfig",
	"AutoModel": "modeling_mimo_v2_flash.MiMoV2FlashModel",
	"AutoModelForCausalLM": "modeling_mimo_v2_flash.MiMoV2FlashForCausalLM"
	},
	"quantization_config": {
	"activation_scheme": "dynamic",
	"fmt": "e4m3",
	"packed_modules_mapping": {},
	"quant_method": "fp8",
	"ignored_layers": [
	"model.layers.0.self_attn.o_proj",
	"model.layers.1.self_attn.o_proj",
	"model.layers.2.self_attn.o_proj",
	"model.layers.3.self_attn.o_proj",
	"model.layers.4.self_attn.o_proj",
	"model.layers.5.self_attn.o_proj",
	"model.layers.6.self_attn.o_proj",
	"model.layers.7.self_attn.o_proj",
	"model.layers.8.self_attn.o_proj",
	"model.layers.9.self_attn.o_proj",
	"model.layers.10.self_attn.o_proj",
	"model.layers.11.self_attn.o_proj",
	"model.layers.12.self_attn.o_proj",
	"model.layers.13.self_attn.o_proj",
	"model.layers.14.self_attn.o_proj",
	"model.layers.15.self_attn.o_proj",
	"model.layers.16.self_attn.o_proj",
	"model.layers.17.self_attn.o_proj",
	"model.layers.18.self_attn.o_proj",
	"model.layers.19.self_attn.o_proj",
	"model.layers.20.self_attn.o_proj",
	"model.layers.21.self_attn.o_proj",
	"model.layers.22.self_attn.o_proj",
	"model.layers.23.self_attn.o_proj",
	"model.layers.24.self_attn.o_proj",
	"model.layers.25.self_attn.o_proj",
	"model.layers.26.self_attn.o_proj",
	"model.layers.27.self_attn.o_proj",
	"model.layers.28.self_attn.o_proj",
	"model.layers.29.self_attn.o_proj",
	"model.layers.30.self_attn.o_proj",
	"model.layers.31.self_attn.o_proj",
	"model.layers.32.self_attn.o_proj",
	"model.layers.33.self_attn.o_proj",
	"model.layers.34.self_attn.o_proj",
	"model.layers.35.self_attn.o_proj",
	"model.layers.36.self_attn.o_proj",
	"model.layers.37.self_attn.o_proj",
	"model.layers.38.self_attn.o_proj",
	"model.layers.39.self_attn.o_proj",
	"model.layers.40.self_attn.o_proj",
	"model.layers.41.self_attn.o_proj",
	"model.layers.42.self_attn.o_proj",
	"model.layers.43.self_attn.o_proj",
	"model.layers.44.self_attn.o_proj",
	"model.layers.45.self_attn.o_proj",
	"model.layers.46.self_attn.o_proj",
	"model.layers.47.self_attn.o_proj",
	"model.decoder.self_attn.o_proj"
	],
	"weight_block_size": [
	128,
	128
	]
	},
	"attention_dropout": 0.0,
	"attention_value_scale": 0.707,
	"hidden_act": "silu",
	"hidden_size": 4096,
	"initializer_range": 0.02,
	"intermediate_size": 16384,
	"max_position_embeddings": 262144,
	"model_type": "mimo_v2_flash",
	"num_attention_heads": 64,
	"head_dim": 192,
	"num_hidden_layers": 48,
	"num_key_value_heads": 4,
	"layernorm_epsilon": 1e-05,
	"rope_theta": 5000000,
	"tie_word_embeddings": false,
	"torch_dtype": "bfloat16",
	"transformers_version": "4.40.1",
	"use_cache": true,
	"vocab_size": 152576,
	"partial_rotary_factor": 0.334,
	"sliding_window": 128,
	"swa_rope_theta": 10000,
	"attention_bias": false,
	"v_head_dim": 128,
	"hybrid_layer_pattern": [
	0,
	1,
	1,
	1,
	1,
	0,
	1,
	1,
	1,
	1,
	1,
	0,
	1,
	1,
	1,
	1,
	1,
	0,
	1,
	1,
	1,
	1,
	1,
	0,
	1,
	1,
	1,
	1,
	1,
	0,
	1,
	1,
	1,
	1,
	1,
	0,
	1,
	1,
	1,
	1,
	1,
	0,
	1,
	1,
	1,
	1,
	1,
	0
	],
	"add_swa_attention_sink_bias": true,
	"add_full_attention_sink_bias": false,
	"sliding_window_size": 128,
	"attention_chunk_size": 128,
	"moe_layer_freq": [
	0,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1,
	1
	],
	"moe_intermediate_size": 2048,
	"n_routed_experts": 256,
	"n_shared_experts": null,
	"num_experts_per_tok": 8,
	"norm_topk_prob": true,
	"scoring_func": "sigmoid",
	"n_group": 1,
	"topk_group": 1,
	"topk_method": "noaux_tc",
	"routed_scaling_factor": null,
	"swa_num_attention_heads": 64,
	"swa_num_key_value_heads": 8,
	"swa_head_dim": 192,
	"swa_v_head_dim": 128
	}