Instructions to use tiny-random/deepseek-v4.1-flash with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tiny-random/deepseek-v4.1-flash with Transformers:

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

pipe = pipeline("text-generation", model="tiny-random/deepseek-v4.1-flash")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("tiny-random/deepseek-v4.1-flash", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use tiny-random/deepseek-v4.1-flash with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tiny-random/deepseek-v4.1-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": "tiny-random/deepseek-v4.1-flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/tiny-random/deepseek-v4.1-flash

SGLang

How to use tiny-random/deepseek-v4.1-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 "tiny-random/deepseek-v4.1-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": "tiny-random/deepseek-v4.1-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 "tiny-random/deepseek-v4.1-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": "tiny-random/deepseek-v4.1-flash",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use tiny-random/deepseek-v4.1-flash with Docker Model Runner:
```
docker model run hf.co/tiny-random/deepseek-v4.1-flash
```

yujiepan commited on 6 days ago

Commit

8bc11f6

verified ·

1 Parent(s): b391e17

Upload folder using huggingface_hub

Browse files

Files changed (1) hide show

README.md +11 -20

README.md CHANGED Viewed

@@ -31,26 +31,10 @@ Note:
 | Quantization | Routed experts FP4; most other linears FP8; UE8M0 scales; 32x32 blocks | Same |
 | Public checkpoint | 96,085 tensors; ~510.3 GB | 1,629 tensors; 96.7 MB |
-  - Preserved topology includes compressed sparse attention and its indexer, four-way
-    Hyper-Connections, n-gram Engram lookup, three-stage DSpark/MTP, the vision aligner,
-    FP4 expert packing, and FP8 block scales. Widths, layer cadence, expert counts,
-    Engram tables, and vision depth are reduced for a practical debugging checkpoint.
-  - The structure is derived by instantiating DeepSeek's pinned official V4.1 inference
-    implementation on the meta device, then converting its runtime state back to the
-    public Hugging Face checkpoint representation. The original and 1017 releases have
-    byte-identical source `config.json` and checkpoint index (96,085 tensor names), so
-    the 1017 source tree is also the authoritative executable structure for this release.
-  - The SGLang example is intentionally a conservative TP=2 launch. The
-    attention latent stays 512-wide and the indexer stays 128-wide because those are
-    fixed by SGLang's DSV4 cache/kernel ABI; all TP-sharded widths, head/group counts,
-    expert counts, vocabulary rows, and 32-wide quantization blocks divide cleanly by 2.
-    The online example enables the embedded three-stage DSpark/MTP draft explicitly.
-  - TP=2 inference with DSpark/MTP and CUDA Graph was validated for both
-    text-only and image-plus-text requests on two NVIDIA H20 GPUs, using the
-    versions listed under Test environment.
 | File path | Size |
 |------|------|
@@ -61,6 +45,13 @@ Note:
 - SGLang
 ```bash
 # Requires an SGLang preview/main build with DeepSeek-V4.1 support.
 # Runs the target model and its embedded three-stage DSpark/MTP draft with TP=2.

 | Quantization | Routed experts FP4; most other linears FP8; UE8M0 scales; 32x32 blocks | Same |
 | Public checkpoint | 96,085 tensors; ~510.3 GB | 1,629 tensors; 96.7 MB |
+  - Topology (sparse attention + indexer, Hyper-Connections, Engram, DSpark/MTP,
+    vision, FP4/FP8 quantization) is preserved with reduced widths and counts.
+    Derived from the source inference implementation on the meta device. The
+    original and 1017 releases have byte-identical `config.json` and checkpoint index.
 | File path | Size |
 |------|------|
 - SGLang
+- TP=2 inference with DSpark/MTP and CUDA Graph was validated for both
+  text-only and image-plus-text requests on two NVIDIA H20 GPUs, using the
+  versions listed under Test environment.
+- The attention latent stays 512-wide and the indexer 128-wide per SGLang
+  DSV4 ABI; all sharded widths, heads, experts, and vocab rows divide cleanly by TP=2.
+- The online example enables the embedded three-stage DSpark/MTP draft explicitly.
 ```bash
 # Requires an SGLang preview/main build with DeepSeek-V4.1 support.
 # Runs the target model and its embedded three-stage DSpark/MTP draft with TP=2.