Instructions to use bloomer010/Ling-3.0-flash-REAP384-97B-A5B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use bloomer010/Ling-3.0-flash-REAP384-97B-A5B with Transformers:

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

pipe = pipeline("text-generation", model="bloomer010/Ling-3.0-flash-REAP384-97B-A5B", trust_remote_code=True)
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained("bloomer010/Ling-3.0-flash-REAP384-97B-A5B", trust_remote_code=True, device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use bloomer010/Ling-3.0-flash-REAP384-97B-A5B with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "bloomer010/Ling-3.0-flash-REAP384-97B-A5B"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "bloomer010/Ling-3.0-flash-REAP384-97B-A5B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/bloomer010/Ling-3.0-flash-REAP384-97B-A5B

SGLang

How to use bloomer010/Ling-3.0-flash-REAP384-97B-A5B 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 "bloomer010/Ling-3.0-flash-REAP384-97B-A5B" \
    --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": "bloomer010/Ling-3.0-flash-REAP384-97B-A5B",
		"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 "bloomer010/Ling-3.0-flash-REAP384-97B-A5B" \
        --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": "bloomer010/Ling-3.0-flash-REAP384-97B-A5B",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use bloomer010/Ling-3.0-flash-REAP384-97B-A5B with Docker Model Runner:
```
docker model run hf.co/bloomer010/Ling-3.0-flash-REAP384-97B-A5B
```

bloomer010 commited on 19 days ago

Commit

773f082

verified ·

1 Parent(s): 964a7ff

Owner wording; em dashes stripped; title size tags matched to repo rename

Browse files

Files changed (1) hide show

README.md +4 -4

README.md CHANGED Viewed

@@ -4,17 +4,17 @@ library_name: transformers
 tags: [reap, expert-pruning, moe, bailingmoe]
 ---
-# Ling-3.0-flash REAP384 (94B total / 5.1B active)
-[384 of 512 routed experts kept per layer — 25% of experts pruned]
 from [inclusionAI/Ling-3.0-flash](https://huggingface.co/inclusionAI/Ling-3.0-flash)
 (124B total / 5.1B active).
-**Method:** one-shot REAP ([Router-weighted Expert Activation Pruning](https://arxiv.org/abs/2510.13999)) —
 experts scored by router-gate-value × output-L2-norm over calibration data, lowest-scoring deleted.
 No fine-tuning, no recovery training.
 **Calibration:** 1M tokens of ultrachat (chat-only calibration)
 BF16 safetensors. Loads with `trust_remote_code=True` (custom `bailing_hybrid` / BailingMoeV3 code).
-Research artifact — quantized builds live in the sibling `-GGUF` repo.

 tags: [reap, expert-pruning, moe, bailingmoe]
 ---
+# Ling-3.0-flash REAP384 (97B total / 5.1B active)
+[384 of 512 routed experts kept per layer - 25% of experts pruned]
 from [inclusionAI/Ling-3.0-flash](https://huggingface.co/inclusionAI/Ling-3.0-flash)
 (124B total / 5.1B active).
+**Method:** one-shot REAP ([Router-weighted Expert Activation Pruning](https://arxiv.org/abs/2510.13999)) -
 experts scored by router-gate-value × output-L2-norm over calibration data, lowest-scoring deleted.
 No fine-tuning, no recovery training.
 **Calibration:** 1M tokens of ultrachat (chat-only calibration)
 BF16 safetensors. Loads with `trust_remote_code=True` (custom `bailing_hybrid` / BailingMoeV3 code).
+Research artifact - quantized builds live in the sibling `-GGUF` repo.