GLM-5.3 is live. Run Z.ai's latest model on Friendli Model APIs. Try it today ➜
bsana1
Available on FriendliAI
Run this model inference on single tenant GPU with unmatched speed and reliability at scale.
Run this model inference with full control and performance in your environment.
Model Details
Model Provider
bsana1
Model Tree
TucanoBR/Tucano-1b1-Instruct
Input Modalities
Output Modalities
Supported Functionality
This repo contains only the LoRA adapter weights (~18 MB) — not a merged model. Load it on top of the base model with 🤗 PEFT (see below).
Instruction-following text generation in Portuguese, specifically for cordel/glosa-style poetry: given a mote (a proverb or saying) or a theme, generate a rhyming stanza in that tradition. Works through the base model's chat template (see usage example below).
This is a small hobby/learning project, not a production model. It was fine-tuned on ~650 examples (50 hand-written + 600 synthetic) — it has not been evaluated for factual accuracy, safety, or any use outside creative Portuguese-language poetry generation, and shouldn't be used for anything that requires reliability guarantees.
Inherits the base model's general limitations (Tucano-1b1-Instruct, 1.1B params). The fine-tuning set is small and cordel poetry draws on 19th/20th century Northeastern Brazilian folk material, so outputs can reflect the period's language, social attitudes, and references. Not evaluated for bias.
python
import torchfrom transformers import AutoModelForCausalLM, AutoTokenizerfrom peft import PeftModelbase = "TucanoBR/Tucano-1b1-Instruct"adapter = "bsana1/arrodeio-tucano-cordel-lora"tok = AutoTokenizer.from_pretrained(base)model = AutoModelForCausalLM.from_pretrained(base, dtype=torch.bfloat16)model = PeftModel.from_pretrained(model, adapter)model.eval()msg = [{"role": "user", "content": "Glose o mote: «Água mole em pedra dura.»"}]ids = tok.apply_chat_template(msg, tokenize=True, add_generation_prompt=True, return_tensors="pt")out = model.generate(ids, max_new_tokens=110, do_sample=True, temperature=0.7,top_k=50, repetition_penalty=1.2, pad_token_id=tok.eos_token_id)print(tok.decode(out[0][ids.shape[1]:], skip_special_tokens=True))
650 mote → glosa / tema → cordel instruction pairs:
data/sft.jsonl)data/sft_synth.jsonl) —
real public-domain cordel stanzas (from pt.wikisource.org, long-dead
authors only) paired with a fabricated instruction that would plausibly
produce them, generated by scripts/make_sft_data.pyLoRA fine-tuning with prompt-masked loss (loss computed only on the response tokens, via the base model's chat template).
q_proj, k_proj, v_proj, o_projTrained on a free Google Colab T4 GPU in about 10 minutes. Adapter weights: ~18 MB.
No formal benchmark — evaluated qualitatively against the earlier stages of the same project (from-scratch model, fine-tuned GPT-2). See samples/comparison.md in the GitHub repo for the same prompts run through every stage side by side.
Training took ~10 minutes on a single shared T4 GPU (Google Colab free tier) — negligible compute. Inference in the live demo runs on Hugging Face's ZeroGPU (shared A10G, on-demand).
Base: Tucano-1b1-Instruct, a Llama2-style causal decoder-only transformer
(1.1B params) natively pretrained in Portuguese. This adapter adds LoRA
matrices to the attention projections only (q/k/v/o_proj); all base
weights stay frozen.
Google Colab, free-tier T4 GPU.
PyTorch, 🤗 Transformers, 🤗 PEFT.
Questions/issues: open one on github.com/bsana1/arrodeio-llm.