Instructions to use FastFlowLM/Gemma3-1B-NPU2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use FastFlowLM/Gemma3-1B-NPU2 with Transformers:

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

pipe = pipeline("text-generation", model="FastFlowLM/Gemma3-1B-NPU2")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("FastFlowLM/Gemma3-1B-NPU2", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps

vLLM

How to use FastFlowLM/Gemma3-1B-NPU2 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "FastFlowLM/Gemma3-1B-NPU2"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "FastFlowLM/Gemma3-1B-NPU2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/FastFlowLM/Gemma3-1B-NPU2

SGLang

How to use FastFlowLM/Gemma3-1B-NPU2 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 "FastFlowLM/Gemma3-1B-NPU2" \
    --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": "FastFlowLM/Gemma3-1B-NPU2",
		"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 "FastFlowLM/Gemma3-1B-NPU2" \
        --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": "FastFlowLM/Gemma3-1B-NPU2",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Unsloth Studio new

How to use FastFlowLM/Gemma3-1B-NPU2 with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for FastFlowLM/Gemma3-1B-NPU2 to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for FastFlowLM/Gemma3-1B-NPU2 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for FastFlowLM/Gemma3-1B-NPU2 to start chatting

Load model with FastModel

pip install unsloth
from unsloth import FastModel
model, tokenizer = FastModel.from_pretrained(
    model_name="FastFlowLM/Gemma3-1B-NPU2",
    max_seq_length=2048,
)

Docker Model Runner
How to use FastFlowLM/Gemma3-1B-NPU2 with Docker Model Runner:
```
docker model run hf.co/FastFlowLM/Gemma3-1B-NPU2
```

FastFlowLM commited on Aug 21, 2025

Commit

86ea534

verified ·

1 Parent(s): 45f63af

Upload initial model

Browse files

Files changed (10) hide show

.gitattributes +7 -0
attn.xclbin +3 -0
config.json +41 -0
dequant.xclbin +3 -0
layer.xclbin +3 -0
lm_head.xclbin +3 -0
mm.xclbin +3 -0
model.q4nx +3 -0
tokenizer.json +3 -0
tokenizer_config.json +0 -0

.gitattributes CHANGED Viewed

@@ -33,3 +33,10 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text

 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+attn.xclbin filter=lfs diff=lfs merge=lfs -text
+dequant.xclbin filter=lfs diff=lfs merge=lfs -text
+layer.xclbin filter=lfs diff=lfs merge=lfs -text
+lm_head.xclbin filter=lfs diff=lfs merge=lfs -text
+mm.xclbin filter=lfs diff=lfs merge=lfs -text
+model.q4nx filter=lfs diff=lfs merge=lfs -text
+tokenizer.json filter=lfs diff=lfs merge=lfs -text

attn.xclbin ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:7c7b047ee2300b0614bbb08817a319cc673195b99ceb3ea0a877c3291c2372bf
+size 456971

config.json ADDED Viewed

	@@ -0,0 +1,41 @@

+{
+  "architectures": [
+    "Gemma3ForCausalLM"
+  ],
+  "attention_bias": false,
+  "attention_dropout": 0.0,
+  "attn_logit_softcapping": null,
+  "bos_token_id": 2,
+  "cache_implementation": "hybrid",
+  "eos_token_id": [
+    1,
+    106
+  ],
+  "final_logit_softcapping": null,
+  "head_dim": 256,
+  "hidden_activation": "gelu_pytorch_tanh",
+  "hidden_size": 1152,
+  "initializer_range": 0.02,
+  "intermediate_size": 6912,
+  "max_position_embeddings": 32768,
+  "model_type": "gemma3_text_only",
+  "num_attention_heads": 4,
+  "num_hidden_layers": 26,
+  "num_key_value_heads": 1,
+  "pad_token_id": 0,
+  "query_pre_attn_scalar": 256,
+  "rms_norm_eps": 1e-06,
+  "rope_local_base_freq": 10000,
+  "rope_scaling": null,
+  "rope_theta": 1000000,
+  "sliding_window": 512,
+  "sliding_window_pattern": 6,
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.50.0.dev0",
+  "use_cache": true,
+  "vocab_size": 262144,
+  "addr_qk": 9216,
+  "addr_kv": 34816,
+  "addr_kk": 13312,
+  "flm_version": "0.9.4"
+}

dequant.xclbin ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:5b529e9c91b0d9def4c5394a21920c0ba01f3d83568dde40f04e4f27034cc714
+size 103915

layer.xclbin ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:1a4f03948734154b299af2c3aae96745f93f6fd976b7ba50bb82ccda2558746a
+size 173707

lm_head.xclbin ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:14bd1d261343485684abf317c13c6d3eb581c79260437f39d979632b117dadea
+size 143115

mm.xclbin ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:2962eefe324053f9a3dac247d1ec7f27737a76eeb9e346fc1296ca1269a22ccb
+size 218699

model.q4nx ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:d85e9e4e0f1ef68c1f51536fdad0fafd48935f1b24cf6db692e1dc0ec6254a40
+size 1229131768

tokenizer.json ADDED Viewed

	@@ -0,0 +1,3 @@

+version https://git-lfs.github.com/spec/v1
+oid sha256:4667f2089529e8e7657cfb6d1c19910ae71ff5f28aa7ab2ff2763330affad795
+size 33384568

tokenizer_config.json ADDED Viewed

The diff for this file is too large to render. See raw diff