Gemma 4 E2B IT - Hate-2-Action
This is a fine-tuned Gemma 4 instruction model for the
Hate-2-Action project (
github), a Telegram bot that turns complaints and frustration into short, practical suggestions for action.
The model is designed to replace the project's final response-generation LLM call. It does not perform intent routing, database search, embedding generation, or problem extraction. Those operations remain part of the surrounding Hate-2-Action pipeline.
Intended use
The model receives:
- A user complaint or description of a social problem
- The requested response language
- A response style
- A topic label
It generates a concise response that:
- Acknowledges the underlying issue
- Converts frustration into concrete next steps
- Follows the requested tone
- Responds in English or Ukrainian
- Avoids assuming the user's emotional state
- Attempts to avoid unsupported facts, organizations, locations, and links
The supported response styles in the training dataset are:
normalpolitefunnysarcasticrude
Role in Hate-2-Action
The complete Hate-2-Action pipeline:
- Receives a complaint or problem from a Telegram user.
- Detects the language and conversation intent.
- Extracts problems and possible solution concepts.
- Uses embeddings and vector similarity to find relevant NGOs and projects.
- Passes the resulting context to the final response generator.
- Produces a short, styled, actionable answer.
This model is intended for step 6: final natural-language answer generation.
Training
The model was fine-tuned from:
Fine-tuning was performed using:
Training data
The model was trained on dataset from the Hate-2-Action project.
The dataset contains 2,000 supervised examples:
- 1,000 English examples
- 1,000 Ukrainian examples
- 400 examples for each response style
- 200 examples for each of 10 topic categories
The dataset builder also produced nominal splits:
- Train: 1,800 examples
- Validation: 100 examples
- Test: 100 examples
Recommended deployment
For production use, provide the model with verified organization and project context retrieved by the Hate-2-Action database.
Generated links, factual claims, and organization details should be validated before being shown to users.
This model is an experimental project-specific checkpoint and should be evaluated against a new, separately created test set before replacing the existing production LLM.
