Text Classification
Transformers
Safetensors
English
qwen3_5_text
text-generation
decision-model
typed-decisions
one-pass
option-probabilities
Instructions to use thegovind/blink-4b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use thegovind/blink-4b with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-classification", model="thegovind/blink-4b")# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("thegovind/blink-4b") model = AutoModelForCausalLM.from_pretrained("thegovind/blink-4b", device_map="auto") - Notebooks
- Google Colab
- Kaggle
File size: 851 Bytes
656d44d 400f1f6 656d44d | 1 2 3 4 5 6 7 8 9 10 11 12 | # blink server image: this repository's weights and runtime; Hugging Face libraries run in offline mode.
# hf download thegovind/<model> --revision <sha> --local-dir blink && cd blink
# docker build -t blink . && docker run --rm --gpus all -p 127.0.0.1:8000:8000 blink
# curl -s http://127.0.0.1:8000/healthz # weights_verified, warmup.repeat_identical, kernels
FROM pytorch/pytorch:2.13.0-cuda12.6-cudnn9-runtime
ENV PIP_BREAK_SYSTEM_PACKAGES=1 PIP_NO_CACHE_DIR=1 PIP_DISABLE_PIP_VERSION_CHECK=1
RUN pip install "transformers==5.17.0" "torchvision==0.28.0" "flash-linear-attention==0.5.2" "accelerate>=1.0" "safetensors>=0.4" "huggingface_hub>=1.0"
COPY . /blink
ENV HF_HUB_OFFLINE=1 TRANSFORMERS_OFFLINE=1 HF_HUB_DISABLE_TELEMETRY=1
EXPOSE 8000
CMD ["python", "/blink/serve.py", "--model", "/blink", "--host", "0.0.0.0", "--port", "8000"]
|