Skip to content
Language models

MediaPipe-TextClassifier-AverageWord

lightning-fast sentiment classifier

2 min readupdated

A compact English-language model based on the average word embedding architecture, packaged in a .tflite file of just 776 KB. It is trained on the SST-2 movie review dataset and designed for binary text sentiment evaluation (identifying positive and negative reviews). The model is built for the Google MediaPipe ecosystem and can run directly on mobile devices or in the browser. It is a great fit for developers who need to integrate sentiment analysis into a mobile, web, or desktop app quickly and without heavy AI models.

Key features

  • Tiny size: The model weighs only 776 KB (FP32 format), making it easy to embed into client applications.
  • High operating speed: On a Google Pixel 6 device, the average CPU latency is just 0.14 ms.
  • Built-in metadata: A vocabulary of 10,003 tokens (vocab.txt) and a label file (labels.txt) are already packed inside TFLITE_METADATA.
  • Cross-platform: Full support through the MediaPipe framework for Python, JavaScript, Android, iOS, and Flutter.

What you can do

  • Mobile app integration: Quickly evaluate the sentiment of user reviews or messages locally on Android or iOS.
  • Web analysis in the browser: Process short user input directly on the client side without sending requests to a server.
  • Text analysis in Python: Quickly classify reviews, comments, or short text snippets in scripts.

Example

Code example for text classification using Python and the mediapipe library:

import mediapipe as mp
from huggingface_hub import hf_hub_download

model_path = hf_hub_download(repo_id="litert-community/MediaPipe-TextClassifier-AverageWord", filename="average_word_classifier.tflite")

BaseOptions = mp.tasks.BaseOptions
TextClassifier = mp.tasks.text.TextClassifier
TextClassifierOptions = mp.tasks.text.TextClassifierOptions

options = TextClassifierOptions(base_options=BaseOptions(model_asset_path=model_path))
with TextClassifier.create_from_options(options) as classifier:
    result = classifier.classify("An unforgettable and genuinely moving film!")
    for cat in result.classifications[0].categories:
        print(f"{cat.category_name}: {cat.score:.3f}")
Read moreDetails
Benchmarks: CPU latency on Pixel 6 — 0.14 ms
Hardware requirements: ultra-lightweight model for running on mobile and PC CPUs
Limitations: English language only; potential accuracy degradation on classifieds, sarcasm, or long multi-sentence texts
License: Apache-2.0
Discuss Telegram 0

Will it run?

Working out the memory needs

The bot is still collecting this model's weight sizes; the estimate will appear here in a few minutes. Meanwhile you can rent a GPU with headroom.

Massed ComputeOpen Massed Compute

Run it on Massed Compute

The script downloads the model to the rented machine. How to run it is in the model card.

  1. 1. Rent a machine with a fitting GPU

    Open Massed Compute
  2. 2. Paste the script into the machine's terminal

    Connect over SSH (the Massed Compute panel shows the command) or open a terminal on the machine's desktop.

    HF_TOKEN / CIVITAI_TOKENThe script starts with token lines: paste yours between the quotes. The Hugging Face token is for gated models, the Civitai one for Civitai models.

    # litert-community/MediaPipe-TextClassifier-AverageWord
    mkdir -p ~/models
    curl -LsSf https://astral.sh/uv/install.sh | sh
    source "$HOME/.local/bin/env"
    uv tool install "huggingface_hub[cli]"
    hf download litert-community/MediaPipe-TextClassifier-AverageWord --local-dir ~/models/MediaPipe-TextClassifier-AverageWord
    # The model is in ~/models. How to run it is in the model card (link above).
  3. 3. Use it

    The model is in ~/models. How to run it is in the model card (link above).

Stop the machine in the panel when you are done: it bills by the minute while it runs.

How to speed it up

Picked for this model type. Choose your hardware above to see what fits you exactly.

  • FlashAttention 44.0.0b40from November 18, 2026new release

    fastest exact attention on B200

    Kernels for Blackwell (beta). Exact attention without quantisation — fine for everything, LLMs included.

    How to enable: pip install flash-attn-4; vLLM and SGLang pick it up themselves.

    Earlier generations (2)
    • FlashAttention 3 — The Hopper (H100/H200) version with FP8.
    • FlashAttention 2 — The standard for RTX 30/40/50 and A100: exact attention without the memory cost of long contexts.
  • torch.compile2.15.0from October 28, 2026new release

    10–30% for free

    Compiles the model for your GPU. Slower first run, faster afterwards.

    How to enable: model = torch.compile(model); in ComfyUI the TorchCompileModel node.

  • MTP (multi-token prediction)0.34.0from November 16, 2026new release

    1.5–2× generation speed

    If the model has MTP layers (DeepSeek-V3/R1, Qwen3-Next, GLM, MiMo and others), it predicts several tokens per step itself.

    How to enable: vLLM / SGLang: --speculative-config '{"method": "mtp", "num_speculative_tokens": 1}'. Check the model card.

    Earlier generations (2)
    • EAGLE-3 — Speculative decoding with a small draft head. For models without their own MTP layers.
    • EAGLE-2 / Medusa — Earlier draft-head generations. Still work if that is all you find.
  • FP8 KV cache0.34.0from November 16, 2026new release

    2× more context in the same memory

    Keeps the attention cache in 8 bits, so long contexts fit where they did not.

    How to enable: vLLM: --kv-cache-dtype fp8; llama.cpp: -ctk q8_0 -ctv q8_0.

  • vLLM / SGLang0.34.0from November 16, 2026new release

    many times more tokens/s on a server

    Server engines: batching, FlashAttention, FP8 and speculative decoding out of the box.

    How to enable: vllm serve <model> — an OpenAI-compatible API.

    Earlier generations (1)
    • llama.cpp — GGUF models on any GPU, Mac or even CPU; some layers can stay in RAM.

Model family

New posts about versions of MediaPipe-TextClassifier-AverageWord (quants, finetunes, LoRAs) will show up under the bell.

No known relatives of this model yet. All derivatives on Hugging Face

About the model

litert-community/MediaPipe-TextClassifier-AverageWord
Author
litert-community
Category
Language models
Released
Posted

On NeyroLearn

Hugging Face

00

Ask AI about this model

Answered by Gemini 4 · the chat is kept only in this tab

Comments

Language models

United 2nd

United 2nd