Salesforce/wikitext
Viewer • Updated • 3.71M • 1.33M • 690
How to use FlameF0X/FWKV-50M with Transformers:
# Use a pipeline as a high-level helper
from transformers import pipeline
pipe = pipeline("text-generation", model="FlameF0X/FWKV-50M") # Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("FlameF0X/FWKV-50M", dtype="auto")How to use FlameF0X/FWKV-50M with vLLM:
# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "FlameF0X/FWKV-50M"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "FlameF0X/FWKV-50M",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'docker model run hf.co/FlameF0X/FWKV-50M
How to use FlameF0X/FWKV-50M with SGLang:
# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
--model-path "FlameF0X/FWKV-50M" \
--host 0.0.0.0 \
--port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "FlameF0X/FWKV-50M",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'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 "FlameF0X/FWKV-50M" \
--host 0.0.0.0 \
--port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
-H "Content-Type: application/json" \
--data '{
"model": "FlameF0X/FWKV-50M",
"prompt": "Once upon a time,",
"max_tokens": 512,
"temperature": 0.5
}'How to use FlameF0X/FWKV-50M with Docker Model Runner:
docker model run hf.co/FlameF0X/FWKV-50M
FWKV (Feed-forward Weighted Key Value, or Floored Weighted Key Value) is a novel efficient language model architecture proposed on May 13, 2026 by Me.
FWKV is built on two simple ideas:
Factorised tied weights, chunked CE, bf16 training. Best val perplexity: 143.62