metadata
license: apache-2.0
language:
- en
- fr
base_model: TinyLlama/TinyLlama-1.1B-Chat-v1.0
tags:
- platformio
- embedded
- embedded-systems
- electronics
- firmware
- microcontroller
- stm32
- arduino
- esp32
- edge-ai
- fine-tuned
- lora
- code-generation
- mascarade
library_name: transformers
pipeline_tag: text-generation
Mascarade PlatformIO
Fine-tuned TinyLlama-1.1B-Chat model specialized in PlatformIO embedded development workflows.
Part of the Mascarade ecosystem — an agentic LLM orchestration system with domain-specific fine-tuned models for embedded systems and electronics.
Training details
| Parameter | Value |
|---|---|
| Base model | TinyLlama/TinyLlama-1.1B-Chat-v1.0 |
| Method | LoRA (PEFT) — merged into full weights |
| LoRA rank (r) | 16 |
| LoRA alpha | 32 |
| LoRA dropout | 0.05 |
| Target modules | q_proj, k_proj, v_proj, o_proj |
| Epochs | 2 |
| Training steps | 20 |
| Dataset | ShareGPT format, domain-specific PlatformIO examples |
| GPU | Quadro P2000 (5 GB VRAM) |
| Framework | Hugging Face Transformers + PEFT |
Usage
from transformers import AutoModelForCausalLM, AutoTokenizer
model = AutoModelForCausalLM.from_pretrained("electron-rare/mascarade-platformio")
tokenizer = AutoTokenizer.from_pretrained("electron-rare/mascarade-platformio")
messages = [{"role": "user", "content": "How do I configure platformio.ini for an STM32 board with custom upload protocol?"}]
inputs = tokenizer.apply_chat_template(messages, return_tensors="pt")
outputs = model.generate(inputs, max_new_tokens=512)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
Related models
| Model | Domain | Base |
|---|---|---|
| mascarade-iot | IoT general | Qwen2.5-Coder-1.5B |
| mascarade-esp32 | ESP32 microcontrollers | TinyLlama-1.1B |
| mascarade-spice | SPICE circuit simulation | TinyLlama-1.1B |
Datasets
All training datasets are available under clemsail on Hugging Face.