Instructions to use neuralworm/gemma-3-270m-it-p2.8 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use neuralworm/gemma-3-270m-it-p2.8 with Transformers:

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

pipe = pipeline("text-generation", model="neuralworm/gemma-3-270m-it-p2.8", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("neuralworm/gemma-3-270m-it-p2.8", trust_remote_code=True, dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use neuralworm/gemma-3-270m-it-p2.8 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "neuralworm/gemma-3-270m-it-p2.8"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "neuralworm/gemma-3-270m-it-p2.8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/neuralworm/gemma-3-270m-it-p2.8

SGLang

How to use neuralworm/gemma-3-270m-it-p2.8 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 "neuralworm/gemma-3-270m-it-p2.8" \
    --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": "neuralworm/gemma-3-270m-it-p2.8",
		"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 "neuralworm/gemma-3-270m-it-p2.8" \
        --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": "neuralworm/gemma-3-270m-it-p2.8",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use neuralworm/gemma-3-270m-it-p2.8 with Docker Model Runner:
```
docker model run hf.co/neuralworm/gemma-3-270m-it-p2.8
```

neuralworm commited on May 30

Commit

0b4c2be

verified ·

1 Parent(s): 0bbf10b

Upload folder using huggingface_hub

Browse files

Files changed (4) hide show

__init__.py +0 -0
__pycache__/__init__.cpython-310.pyc +0 -0
__pycache__/patch.cpython-310.pyc +0 -0
patch.py +10 -2

__init__.py ADDED Viewed

File without changes

__pycache__/__init__.cpython-310.pyc ADDED Viewed

Binary file (167 Bytes). View file

__pycache__/patch.cpython-310.pyc CHANGED Viewed

Binary files a/__pycache__/patch.cpython-310.pyc and b/__pycache__/patch.cpython-310.pyc differ

patch.py CHANGED Viewed

@@ -388,7 +388,10 @@ def _px_forward(
     divergence_buffer = []
     correction_strength = 0.0
-    if n_loops > 1:
         h_exp = e_reflector.clone() # Use Reflected Anchor
         current_layer = dynamic_start
         max_steps = (dynamic_end - dynamic_start) * n_loops * 3
@@ -527,8 +530,13 @@ def _px_forward(
             # --------------------------------------------------
             # Apply LTI Injection with Dynamic Anchor
             e_norm = self._px_injection.input_norm(e_dynamic.to(torch.float32)).to(trans_out.dtype)
-            h_exp = trans_out + current_gamma * (e_norm - h_prev)
             # --- PHASE 26: REFLECTION FLIPPING (RF) ---
             h_f32 = h_exp.to(torch.float32)

     divergence_buffer = []
     correction_strength = 0.0
+    # Phase 39: Stability & Diversity Injection
+    # 1. Restrict recursion to decoding (T=1) to prevent cache corruption during prefill.
+    # 2. Add Stochastic Jitter to break representational loops (repetitions).
+    if n_loops > 1 and T_curr == 1:
         h_exp = e_reflector.clone() # Use Reflected Anchor
         current_layer = dynamic_start
         max_steps = (dynamic_end - dynamic_start) * n_loops * 3
             # --------------------------------------------------
             # Apply LTI Injection with Dynamic Anchor
+            # Phase 39.1: Repetition Escape Jitter (Cognitive Diversity)
+            # Add 5% stochastic noise to gamma to perturb periodic orbits.
+            jitter_strength = (torch.rand(1, device=h_exp.device).item() - 0.5) * 0.1
+            effective_gamma = current_gamma * (1.0 + jitter_strength)
             e_norm = self._px_injection.input_norm(e_dynamic.to(torch.float32)).to(trans_out.dtype)
+            h_exp = trans_out + effective_gamma * (e_norm - h_prev)
             # --- PHASE 26: REFLECTION FLIPPING (RF) ---
             h_f32 = h_exp.to(torch.float32)