Instructions to use Nanthasit/sakthai-context-1.5b-tools with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Nanthasit/sakthai-context-1.5b-tools with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-1.5B-Instruct")
model = PeftModel.from_pretrained(base_model, "Nanthasit/sakthai-context-1.5b-tools")

Transformers

How to use Nanthasit/sakthai-context-1.5b-tools with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("text-generation", model="Nanthasit/sakthai-context-1.5b-tools")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("Nanthasit/sakthai-context-1.5b-tools", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use Nanthasit/sakthai-context-1.5b-tools with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "Nanthasit/sakthai-context-1.5b-tools"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Nanthasit/sakthai-context-1.5b-tools",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/Nanthasit/sakthai-context-1.5b-tools

SGLang

How to use Nanthasit/sakthai-context-1.5b-tools with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "Nanthasit/sakthai-context-1.5b-tools" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Nanthasit/sakthai-context-1.5b-tools",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "Nanthasit/sakthai-context-1.5b-tools" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "Nanthasit/sakthai-context-1.5b-tools",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use Nanthasit/sakthai-context-1.5b-tools with Docker Model Runner:
```
docker model run hf.co/Nanthasit/sakthai-context-1.5b-tools
```

SakThai Context 1.5B — LoRA Adapter

LoRA adapter weights for fine-tuning Qwen2.5-1.5B-Instruct on the SakThai combined dataset. Designed for tool-calling, multi-turn context recall, and structured instruction following in agentic workflows.

Model Details

Developed by: Nanthasit (Beer)
Base Model: Qwen/Qwen2.5-1.5B-Instruct
Fine-tuning Method: LoRA via PEFT 0.19.1
Rank: 16 | Alpha: 32 | Dropout: 0.1
Target Modules: q_proj, k_proj, v_proj, o_proj
Task Type: Causal Language Modeling
Training Steps: 220 | Epochs: 4
Training Data: 974 examples (25 tool schemas)
License: Apache 2.0

Usage

Load the Adapter

from peft import PeftModel
from transformers import AutoModelForCausalLM

base = AutoModelForCausalLM.from_pretrained("Qwen/Qwen2.5-1.5B-Instruct")
model = PeftModel.from_pretrained(base, "Nanthasit/sakthai-context-1.5b-tools")

Merge (for Deployment)

merged = model.merge_and_unload()
merged.save_pretrained("./merged-model")

Pre-merged weights are available at Nanthasit/sakthai-context-1.5b-merged.

Training Hyperparameters

Parameter	Value
LoRA rank (r)	16
LoRA alpha	32
LoRA dropout	0.1
Target modules	q_proj, k_proj, v_proj, o_proj
Learning rate	5e-5 (linear schedule)
Batch size	4
Gradient accumulation	4
Epochs	4
Max steps	220
Precision	bf16 (mixed)
Optimizer	AdamW

Evaluation

The merged model achieves 45/45 (100%) on the SakThai eval suite:

Category	Pass Rate
Basic	✅ 6/6
Multi-turn	✅ 9/9
Instruction	✅ 6/6
Tool calling	✅ 6/6
Reasoning	✅ 6/6
Format	✅ 12/12

Full eval report: Nanthasit/sakthai-context-1.5b-merged/eval/EVAL.md

Training Script

The full training script (train.py) is included in this repo and logs training loss over 220 steps.

Downloads last month: -

Model tree for Nanthasit/sakthai-context-1.5b-tools

Base model

Qwen/Qwen2.5-1.5B

Finetuned

Qwen/Qwen2.5-1.5B-Instruct

Adapter

(1163)

this model

Nanthasit
/

sakthai-context-1.5b-tools