Instructions to use a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn with PEFT:

from peft import PeftModel
from transformers import AutoModelForCausalLM

base_model = AutoModelForCausalLM.from_pretrained("openai/gpt-oss-20b")
model = PeftModel.from_pretrained(base_model, "a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn")

Transformers

How to use a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn with Transformers:

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

pipe = pipeline("text-generation", model="a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoModel
model = AutoModel.from_pretrained("a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn", dtype="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn

SGLang

How to use a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn 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 "a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn" \
    --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": "a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn",
		"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 "a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn" \
        --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": "a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn with Docker Model Runner:
```
docker model run hf.co/a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn
```

gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn

This model is a fine-tuned version of openai/gpt-oss-20b on the multi_turn_miss_func_zh_tw_function_mix500_turn_oss_gpt_oss_20b_pretokenized, the multi_turn_miss_para_zh_tw_function_mix500_turn_oss_gpt_oss_20b_pretokenized, the multi_turn_zh_tw_function_mix500_turn_oss_gpt_oss_20b_pretokenized, the irrelevance_zh_tw_oss3000_gpt_oss_20b_pretokenized and the apigen_zhtwV3_remove_sys_gpt_oss_20b_pretokenized datasets. It achieves the following results on the evaluation set:

Loss: 1.7413

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

learning_rate: 5e-07
train_batch_size: 4
eval_batch_size: 4
seed: 42
gradient_accumulation_steps: 32
total_train_batch_size: 128
optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 15
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
1.9188	0.0467	10	2.0771
2.1396	0.0934	20	2.0681
2.0909	0.1401	30	2.0564
2.0204	0.1868	40	2.0281
1.9068	0.2335	50	2.0003
1.9153	0.2803	60	1.9579
1.8924	0.3270	70	1.9261
2.0036	0.3737	80	1.9034
1.8911	0.4204	90	1.8709
1.854	0.4671	100	1.8492
1.8588	0.5138	110	1.8197
1.713	0.5605	120	1.8010
1.6542	0.6072	130	1.7828
1.5714	0.6539	140	1.7707
1.7184	0.7006	150	1.7601
1.6565	0.7473	160	1.7521
1.7325	0.7940	170	1.7503
1.8492	0.8408	180	1.7432
1.725	0.8875	190	1.7447
1.7197	0.9342	200	1.7353
1.811	0.9809	210	1.7406

Framework versions

PEFT 0.18.1
Transformers 4.57.6
Pytorch 2.8.0+cu128
Datasets 4.0.0
Tokenizers 0.22.2

Downloads last month: 2

Model tree for a3ilab-llm-uncertainty/gptoss_20b_all_zhtw_lr5e-7_ep1_16_64_128_turn

Base model

openai/gpt-oss-20b

Adapter

(235)

this model