README.md

---
license: apache-2.0
language:
- en
base_model: Nanthasit/sakthai-context-0.5b-tools
library_name: transformers
pipeline_tag: text-generation
tags:
- generated_from_trainer
- trl
- sft
- peft
- lora
- adapter
- finetuned
- conversational
- assistant
- tool-calling
- function-calling
- qwen
- qwen2.5
- slm
- small-language-model
- sakthai
- house-of-sak
- eval-results
datasets:
- Nanthasit/sakthai-combined-v7
inference:
  requires_base: true
  adapter_only: true
  parameters:
    temperature: 0.3
    max_new_tokens: 128
    top_p: 0.9
model-index:
- name: SakThai SFT-out — LoRA Adapter
  results:
  - task:
      type: text-generation
      name: Adapter merge readiness
    dataset:
      type: internal
      name: Adapter merge pipeline
    metrics:
    - type: merge-status
      value: 0
      name: Standalone usable without merge
      verified: false
    - type: tool-call-success
      value: pending
      name: Tool-call success post-merge
      verified: false
---

<div align="center">

[![Base Model](https://img.shields.io/badge/Base_Model-Nanthasit/sakthai-context--0.5b--tools-green)](https://huggingface.co/Nanthasit/sakthai-context-0.5b-tools)
[![LoRA](https://img.shields.io/badge/LoRA-r8%20%C3%8116-orange)](https://huggingface.co/Nanthasit/sft-out)
[![TRL](https://img.shields.io/badge/TRL-1.9.2-red)](https://github.com/huggingface/trl)
[![License](https://img.shields.io/badge/License-Apache_2.0-green)](https://huggingface.co/Nanthasit/sft-out)
![Downloads](https://img.shields.io/badge/dynamic/json?url=https%3A%2F%2Fhuggingface.co%2Fapi%2Fmodels%2FNanthasit%2Fsft-out&query=%24.downloads&label=downloads&color=blue&cacheSeconds=3600)

</div>

> ⚠️ **Adapter-only repository** — this repo contains only a LoRA adapter, not a standalone model. You must merge it with the base model before inference or GGUF export.

> Important: merging produces a new standalone model; this README is therefore focused on merge-first workflows.

# Model Card for `sft-out`

## Model Description

| Property | Value |
|:---------|------:|
| Base model | [Nanthasit/sakthai-context-0.5b-tools](https://huggingface.co/Nanthasit/sakthai-context-0.5b-tools) (Qwen2.5-0.5B-Instruct) |
| Adapter type | LoRA |
| LoRA rank `r` | 8 |
| LoRA alpha | 16 |
| Dropout | 0.0 |
| Target modules | `q_proj`, `k_proj`, `v_proj`, `o_proj` |
| Adapter size | ~158 KB |
| Max context | 32,768 tokens |
| Parameters | 495M base + adapter |
| Language | English |

This adapter is a low-rank delta over the base SakThai 0.5B tool-calling model. It preserves the same architecture and tokenizer.

## What's in the box

| File | Size | Description |
|:-----|-----:|:------------|
| `adapter_model.safetensors` | 138 KB | LoRA delta weights |
| `adapter_config.json` | 1.1 KB | Adapter hyperparameters |
| `training_args.bin` | 129 B | Serialized training args |
| `tokenizer.json` | 13.2 KB | Tokenizer |
| `tokenizer_config.json` | 694 B | Tokenizer config |
| `chat_template.jinja` | 2.5 KB | Chat template |

**Total adapter size: ~158 KB**

## How to Use

### Merge with base model

```python
from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel

base = "Nanthasit/sakthai-context-0.5b-tools"
adapter = "Nanthasit/sft-out"

tokenizer = AutoTokenizer.from_pretrained(base)
model = AutoModelForCausalLM.from_pretrained(base, device_map="auto")
model = PeftModel.from_pretrained(model, adapter)
```

Generate:

```python
messages = [{"role": "user", "content": "What's the weather in Tokyo?"}]
inputs = tokenizer.apply_chat_template(messages, add_generation_prompt=True, return_tensors="pt").to(model.device)
outputs = model.generate(inputs, max_new_tokens=128, temperature=0.3, top_p=0.9)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```

### Merge to GGUF (llama.cpp)

```bash
python -m peft.utils.merge_adapters \
  --base_model Nanthasit/sakthai-context-0.5b-tools \
  --adapter Nanthasit/sft-out \
  --output merged-model
python -m transformers.convert_save_to_gguf merged-model --outtype q4_k_m --outfile sakthai-0.5b-sft-out-q4_k_m.gguf
```

Then run with llama.cpp:

```bash
./llama-cli -m sakthai-0.5b-sft-out-q4_k_m.gguf -p "What's the weather in Tokyo?" -n 128 --temp 0.3
```

## Tool-Calling Format

This adapter was trained on the SakThai tool-calling format.

**Prompt format:**

```xml
<tools>
{
  "function": "get_weather",
  "params": {"location": "Bangkok"}
}
</tools>
```

**Model output:**

```json
{"name": "get_weather", "arguments": {"location": "Bangkok"}}
```

Multi-turn tool results:

```xml
<tool_result>
{"result": "The weather in Bangkok is 32°C and sunny."}
</tool_result>
```

## Benchmarks

No published benchmarks for this specific adapter yet. The base model
[Nanthasit/sakthai-context-0.5b-tools](https://huggingface.co/Nanthasit/sakthai-context-0.5b-tools)
achieves strong single-shot tool-calling on [sakthai-bench-v2](https://huggingface.co/datasets/Nanthasit/sakthai-bench-v2).

For the SakThai family benchmarks, see the [SakThai Leaderboard](https://huggingface.co/spaces/Nanthasit/sakthai-leaderboard).

## Reproduce

```bash
pip install transformers>=5.14.1 trl>=1.9.2 peft datasets accelerate
python -m trl.cli.sft \
  --model_name_or_path Nanthasit/sakthai-context-0.5b-tools \
  --dataset_name Nanthasit/sakthai-combined-v7 \
  --adapter_output_dir sft-out \
  --per_device_train_batch_size 8 \
  --learning_rate 2e-4 \
  --num_train_epochs 3 \
  --lora_r 8 --lora_alpha 16 --lora_dropout 0.0 \
  --lora_target_modules q_proj,k_proj,v_proj,o_proj \
  --bf16
```

## Training Details

| Attribute | Value |
|:----------|:------|
| Base model | [Nanthasit/sakthai-context-0.5b-tools](https://huggingface.co/Nanthasit/sakthai-context-0.5b-tools) |
| Method | Supervised Fine-Tuning (SFT) via TRL |
| TRL version | 1.9.2 |
| Transformers | 5.14.1 |
| PyTorch | 2.11.0 |
| Datasets | 5.0.0 |
| Tokenizers | 0.22.2 |

> **Note:** Exact training hyperparameters (learning rate, batch size, epochs) are not available from this repo alone. The `training_args.bin` file is present but contains unparsed binary data. Check sibling model `sakthai-context-0.5b-tools` for comparable settings.

## Limitations

- **Adapter-only.** Must be loaded on top of the base model. Cannot be used standalone via `AutoModelForCausalLM`.
- **No standalone GGUF.** Must be merged with base before GGUF conversion.
- **Not servable serverless.** The Hugging Face Inference API does not support dynamic LoRA adapter loading.
- **Experimental.** This is a new SFT training experiment; performance has not been extensively evaluated.
- **Placeholder tokenizer.** The `tokenizer.json` in this repo is a placeholder; the actual tokenizer lives in the base model repo.
- **License compatibility** depends on base-model license (`apache-2.0`).

## Working Alternatives

| Model | Type | Best for |
|:------|:-----|:---------|
| [sakthai-context-0.5b-tools](https://huggingface.co/Nanthasit/sakthai-context-0.5b-tools) | Base adapter (8.7 MB) | Strongest standalone 0.5B tool caller |
| [sakthai-context-0.5b-merged](https://huggingface.co/Nanthasit/sakthai-context-0.5b-merged) | Merged (988 MB) | Direct inference, no merge needed |
| [sakthai-context-0.5b-tools-sft-v2](https://huggingface.co/Nanthasit/sakthai-context-0.5b-tools-sft-v2) | Adapter (8.3 MB) | Newer SFT variant with richer data |
| [sakthai-context-1.5b-merged](https://huggingface.co/Nanthasit/sakthai-context-1.5b-merged) | Merged (3.1 GB) | Mid-size, best balance |

## Citation

If you use this adapter or the SakThai family in your work, please cite:

```bibtex
@misc{sakthai-sft-out,
  title  = {SakThai SFT-out LoRA Adapter},
  author = {Beer (beer-sakthai) and the SakThai Agent family},
  year   = {2026},
  publisher = {Hugging Face},
  howpublished = {\url{https://huggingface.co/Nanthasit/sft-out}}
```

Base model:

```bibtex
@misc{qwen25-2025,
  title  = {Qwen2.5: A Party of Foundation Models},
  author = {Qwen Team},
  year   = {2025},
  url    = {https://huggingface.co/Qwen/Qwen2.5-0.5B-Instruct}
}
```

TRL:

```bibtex
@software{vonwerra2020trl,
  title   = {{TRL: Transformers Reinforcement Learning}},
  author  = {von Werra, Leandro and Belkada, Younes and Tunstall, Lewis and Beeching, Edward and Thrush, Tristan and Lambert, Nathan and Huang, Shengyi and Rasul, Kashif and Gallouédec, Quentin},
  license = {Apache-2.0},
  url     = {https://github.com/huggingface/trl},
  year    = {2020}
}
```

## SakThai Family

This adapter is part of the [SakThai model family](https://huggingface.co/collections/Nanthasit/sakthai-model-family-6a64745450b12d421c1f9f02).

| Model | Downloads | Type |
|:------|:---------:|:-----|
| [context-1.5b-merged](https://huggingface.co/Nanthasit/sakthai-context-1.5b-merged) | 1,855 | Merged weights |
| [context-0.5b-merged](https://huggingface.co/Nanthasit/sakthai-context-0.5b-merged) | 1,692 | Merged weights |
| [context-7b-merged](https://huggingface.co/Nanthasit/sakthai-context-7b-merged) | 1,024 | Merged weights |
| [context-0.5b-tools](https://huggingface.co/Nanthasit/sakthai-context-0.5b-tools) | 251 | Base adapter |
| [context-0.5b-tools-sft](https://huggingface.co/Nanthasit/sakthai-context-0.5b-tools-sft) | 0 | SFT v1 |
| [context-0.5b-tools-sft-v2](https://huggingface.co/Nanthasit/sakthai-context-0.5b-tools-sft-v2) | 0 | SFT v2 |
| **sft-out** | **0** | **Adapter-only experiment** |

*Downloads verified 2026-07-31.*

---

Built with zero budget. Part of the [House of Sak](https://huggingface.co/Nanthasit).