Instructions to use thegovind/blink-mimo-9b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use thegovind/blink-mimo-9b with Transformers:

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

pipe = pipeline("image-text-to-text", model="thegovind/blink-mimo-9b")
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("thegovind/blink-mimo-9b")
model = AutoModelForMultimodalLM.from_pretrained("thegovind/blink-mimo-9b", 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 thegovind/blink-mimo-9b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "thegovind/blink-mimo-9b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "thegovind/blink-mimo-9b",
		"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/thegovind/blink-mimo-9b

SGLang

How to use thegovind/blink-mimo-9b 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 "thegovind/blink-mimo-9b" \
    --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": "thegovind/blink-mimo-9b",
		"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 "thegovind/blink-mimo-9b" \
        --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": "thegovind/blink-mimo-9b",
		"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 thegovind/blink-mimo-9b with Docker Model Runner:
```
docker model run hf.co/thegovind/blink-mimo-9b
```

thegovind commited on 24 days ago

Commit

4484d62

verified ·

1 Parent(s): 284955f

Code revision v1.4: card update

Browse files

Files changed (1) hide show

README.md +10 -8

README.md CHANGED Viewed

@@ -28,8 +28,8 @@ Send a text or JSON `state` and typed questions (`choice`, `noul`, `score`). Eac
 |---|---|
 | Base model | [XiaomiMiMo/MiMo-V2.6-Distill-Qwen-9B](https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Distill-Qwen-9B) |
 | Weights size | 18.8 GB bf16 (8.95B text parameters, 9.41B total) |
-| Revision | v1.3 (code revision; weights identical to v1.0) |
-| License | Non-commercial research ([LICENSE.md](LICENSE.md)); code Apache-2.0 |
 ## Results
@@ -130,7 +130,7 @@ Training overlap: Public train splits included ContractNLI, iSarcasmEval and VAS
 ### JevBench: public items only
-The JevBench numbers are public-item development proxies, not official scores, and claim no rank or parity.
 | Model | Easy | Standard | Hard | Hard ECE | Official score |
 |---|---:|---:|---:|---:|---|
@@ -157,8 +157,8 @@ import os, sys
 from huggingface_hub import hf_hub_download
 os.environ["BLINK_MODEL"] = "thegovind/blink-mimo-9b"
-os.environ["BLINK_REVISION"] = "v1.3"
-sys.path.insert(0, os.path.dirname(hf_hub_download("thegovind/blink-mimo-9b", "blink.py", revision="v1.3")))
 import blink
 out = blink.decide(
@@ -178,13 +178,13 @@ print(out["answers"]["intent"]["probabilities"])
 ## Run it as a server
-`serve.py` provides TypeSafe server-side SDK compatibility at `POST /v1/systemone` and lists the served model at `GET /v1/models`. Set `TYPESAFE_BASE_URL` to connect TypeSafe Python or JavaScript SDKs; JevBench's `typesafe` adapter and Decision Index's `http` engine work unchanged for text requests. `GET /healthz` reports startup checks. `v1.3` updates code only; weights are identical to `v1.0`.
 See the [API reference](https://thegovind.github.io/blink/api/) for full details.
 ```sh
 pip install "torch==2.13.0" "transformers==5.17.0" "flash-linear-attention==0.5.2" "accelerate>=1.1.0" safetensors huggingface_hub
-hf download thegovind/blink-mimo-9b --revision v1.3 --local-dir blink-mimo-9b
 python blink-mimo-9b/serve.py --model ./blink-mimo-9b --port 8000
 # TypeSafe SDKs: export TYPESAFE_BASE_URL=http://127.0.0.1:8000 TYPESAFE_API_KEY=any
 ```
@@ -196,6 +196,8 @@ cd blink-mimo-9b
 docker build -t blink-mimo-9b . && docker run --rm --gpus all -p 127.0.0.1:8000:8000 blink-mimo-9b
 ```
 ## Screenshots (opt-in, self-hosted)
 Image input is off by default. Start `serve.py` with `--vision`, or set `BLINK_VISION=1` for direct `blink.py` calls. Image mode needs `torchvision==0.28.0`; it uses MiMo's unchanged tower.
@@ -230,7 +232,7 @@ Mind2Web shortlist recall is only 104/1,000: "given gold" excludes steps with no
 - **FP32 reference coverage caveat:** On the Decision Index sample, a set of long workflow documents, and the public TypeSafe cases covered by the FP32 reference, batched answers passed the same numerical-parity checks against an FP32 reference as one-at-a-time answers, covering argmax agreement and probability differences. The FP32 reference could not run the 5 longest TypeSafe documents, so parity covered 170 of 354 TypeSafe questions; batched and one-at-a-time runs chose the same option on the other 184. A few near-tied answers can still flip.
 - **Batching command:**
   ```sh
-  hf download thegovind/blink-mimo-9b serve.py blink.py --revision v1.3 --local-dir blink-mimo-9b
   python blink-mimo-9b/serve.py --model ./blink-mimo-9b --port 8000 --batch-window-ms 5
   ```
 - **Weights:** 18.8 GB in bf16 (8.95B text parameters, 9.41B total).

 |---|---|
 | Base model | [XiaomiMiMo/MiMo-V2.6-Distill-Qwen-9B](https://huggingface.co/XiaomiMiMo/MiMo-V2.6-Distill-Qwen-9B) |
 | Weights size | 18.8 GB bf16 (8.95B text parameters, 9.41B total) |
+| Revision | v1.4 (code revision; weights identical to v1.0) |
+| License | Weights: non-commercial research and evaluation only ([LICENSE.md](LICENSE.md)); code: Apache-2.0. |
 ## Results
 ### JevBench: public items only
+No official JevBench score for blink has been published. The JevBench numbers are public-item development proxies, not official scores, and claim no rank or parity.
 | Model | Easy | Standard | Hard | Hard ECE | Official score |
 |---|---:|---:|---:|---:|---|
 from huggingface_hub import hf_hub_download
 os.environ["BLINK_MODEL"] = "thegovind/blink-mimo-9b"
+os.environ["BLINK_REVISION"] = "v1.4"
+sys.path.insert(0, os.path.dirname(hf_hub_download("thegovind/blink-mimo-9b", "blink.py", revision="v1.4")))
 import blink
 out = blink.decide(
 ## Run it as a server
+`serve.py` provides TypeSafe server-side SDK compatibility at `POST /v1/systemone` and lists the served model at `GET /v1/models`. Set `TYPESAFE_BASE_URL` to connect TypeSafe Python or JavaScript SDKs; JevBench's `typesafe` adapter and Decision Index's `http` engine work unchanged for text requests. `GET /healthz` reports startup checks. `v1.4` updates this card and the revision pin only; `blink.py`, `serve.py` and the weights are unchanged from `v1.3`.
 See the [API reference](https://thegovind.github.io/blink/api/) for full details.
 ```sh
 pip install "torch==2.13.0" "transformers==5.17.0" "flash-linear-attention==0.5.2" "accelerate>=1.1.0" safetensors huggingface_hub
+hf download thegovind/blink-mimo-9b --revision v1.4 --local-dir blink-mimo-9b
 python blink-mimo-9b/serve.py --model ./blink-mimo-9b --port 8000
 # TypeSafe SDKs: export TYPESAFE_BASE_URL=http://127.0.0.1:8000 TYPESAFE_API_KEY=any
 ```
 docker build -t blink-mimo-9b . && docker run --rm --gpus all -p 127.0.0.1:8000:8000 blink-mimo-9b
 ```
+Earlier vLLM checks passed, but the final read matched earlier answers and still failed hard ECE (0.122 vs 0.121). No MiMo opt-in server ships; use `serve.py`.
 ## Screenshots (opt-in, self-hosted)
 Image input is off by default. Start `serve.py` with `--vision`, or set `BLINK_VISION=1` for direct `blink.py` calls. Image mode needs `torchvision==0.28.0`; it uses MiMo's unchanged tower.
 - **FP32 reference coverage caveat:** On the Decision Index sample, a set of long workflow documents, and the public TypeSafe cases covered by the FP32 reference, batched answers passed the same numerical-parity checks against an FP32 reference as one-at-a-time answers, covering argmax agreement and probability differences. The FP32 reference could not run the 5 longest TypeSafe documents, so parity covered 170 of 354 TypeSafe questions; batched and one-at-a-time runs chose the same option on the other 184. A few near-tied answers can still flip.
 - **Batching command:**
   ```sh
+  hf download thegovind/blink-mimo-9b serve.py blink.py --revision v1.4 --local-dir blink-mimo-9b
   python blink-mimo-9b/serve.py --model ./blink-mimo-9b --port 8000 --batch-window-ms 5
   ```
 - **Weights:** 18.8 GB in bf16 (8.95B text parameters, 9.41B total).