Instructions to use SaintsStudios/Mazgu_Small-T_130M_Instruct with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SaintsStudios/Mazgu_Small-T_130M_Instruct with Transformers:

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

pipe = pipeline("text-generation", model="SaintsStudios/Mazgu_Small-T_130M_Instruct")

# pip install -U transformers accelerate
# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("SaintsStudios/Mazgu_Small-T_130M_Instruct")
model = AutoModelForCausalLM.from_pretrained("SaintsStudios/Mazgu_Small-T_130M_Instruct", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use SaintsStudios/Mazgu_Small-T_130M_Instruct with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "SaintsStudios/Mazgu_Small-T_130M_Instruct"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SaintsStudios/Mazgu_Small-T_130M_Instruct",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/SaintsStudios/Mazgu_Small-T_130M_Instruct

SGLang

How to use SaintsStudios/Mazgu_Small-T_130M_Instruct 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 "SaintsStudios/Mazgu_Small-T_130M_Instruct" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SaintsStudios/Mazgu_Small-T_130M_Instruct",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

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 "SaintsStudios/Mazgu_Small-T_130M_Instruct" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SaintsStudios/Mazgu_Small-T_130M_Instruct",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use SaintsStudios/Mazgu_Small-T_130M_Instruct with Docker Model Runner:
```
docker model run hf.co/SaintsStudios/Mazgu_Small-T_130M_Instruct
```

Mazgu_Small-T_130M_Instruct

This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:

Loss: 3.1128

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: 3e-05
train_batch_size: 8
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 32
optimizer: Use OptimizerNames.ADAMW_TORCH 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: 0.03
num_epochs: 22

Training results

Training Loss	Epoch	Step	Validation Loss
3.1018	1.0	214	3.1789
3.0351	2.0	428	3.1612
3.0528	3.0	642	3.1474
2.9466	4.0	856	3.1380
2.9969	5.0	1070	3.1299
2.9784	6.0	1284	3.1246
2.9222	7.0	1498	3.1206
2.9501	8.0	1712	3.1179
3.0095	9.0	1926	3.1164
2.9389	10.0	2140	3.1149
2.9016	11.0	2354	3.1139
2.8938	12.0	2568	3.1136
2.8549	13.0	2782	3.1131
2.8796	14.0	2996	3.1130
2.9367	15.0	3210	3.1129
2.8664	16.0	3424	3.1128
2.8942	17.0	3638	3.1129
2.8932	18.0	3852	3.1127
2.8682	19.0	4066	3.1127
2.9357	20.0	4280	3.1128
2.9435	21.0	4494	3.1128

Framework versions

Transformers 5.0.0
Pytorch 2.10.0+cu128
Datasets 5.0.0
Tokenizers 0.22.2

Downloads last month: 660

Safetensors

Model size

52.2M params

Tensor type

BF16