4 comments

  • nl1 hour ago
    If you are going to go to the bother of fine tuning for trivial problems like subject classification then I think you&#x27;ll find Scikit Learn with a SGDClassifier on 2-grams will do probably just as well and be under 1MB for the trained classifier.<p>You can train it in under a minute, and it will work perfectly well on embedded devices.<p>Small LLMs are good choices for text classification in two cases:<p>- If you next to provide in-context examples and classifier based on them.<p>- Your classification goes beyond simple subject-type classifiers. For example, multiple choice question answering is classification where small LLM will work but traditional ML methods won&#x27;t&#x2F;
    • djsjajah37 minutes ago
      Not with 800 examples. If you are going to consider an ngram model, I think you are better off getting a frontier llm to write you an absurd regex.
  • deepsquirrelnet26 minutes ago
    If you want to go deeper on language models, try these project ideas:<p>- Zero-shot encoders like tasksource or GliNER<p>- Natural language inference: <a href="https:&#x2F;&#x2F;huggingface.co&#x2F;blog&#x2F;dleemiller&#x2F;nli-xenc-ways-to-use" rel="nofollow">https:&#x2F;&#x2F;huggingface.co&#x2F;blog&#x2F;dleemiller&#x2F;nli-xenc-ways-to-use</a><p>- GRPO training<p>- GEPA prompt tuning Qwen 0.6B (or GEPA, then GRPO)<p>- Use an embedding model and train a classifier (MLP, logistic, svm)<p>- Use a larger LLM to generate a synthetic dataset (beware of lack of diversity, mine &quot;seed text&quot; from real sources first)<p>- Synthetically generate &quot;hard examples&quot; where more than one category may be valid and DPO tune your preferred responses
  • mickael-kerjean40 minutes ago
    If you are interested in small language model to fine tune, gemma3:270m is quite interesting for its size
  • jszymborski1 hour ago
    I think the Qwen 0.6B is so cool. It is super fast and as illustrated here it has a clear niche, esp. when fine-tuned.<p>I&#x27;m also interested in it as a student for distillation.