Instructions to use tasal9/ZamAI-mT5-Pashto with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use tasal9/ZamAI-mT5-Pashto with Transformers:

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

pipe = pipeline("text-generation", model="tasal9/ZamAI-mT5-Pashto")

# Load model directly
from transformers import AutoTokenizer, AutoModelForSeq2SeqLM

tokenizer = AutoTokenizer.from_pretrained("tasal9/ZamAI-mT5-Pashto")
model = AutoModelForSeq2SeqLM.from_pretrained("tasal9/ZamAI-mT5-Pashto", device_map="auto")

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use tasal9/ZamAI-mT5-Pashto with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "tasal9/ZamAI-mT5-Pashto"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "tasal9/ZamAI-mT5-Pashto",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/tasal9/ZamAI-mT5-Pashto

SGLang

How to use tasal9/ZamAI-mT5-Pashto 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 "tasal9/ZamAI-mT5-Pashto" \
    --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": "tasal9/ZamAI-mT5-Pashto",
		"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 "tasal9/ZamAI-mT5-Pashto" \
        --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": "tasal9/ZamAI-mT5-Pashto",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Docker Model Runner
How to use tasal9/ZamAI-mT5-Pashto with Docker Model Runner:
```
docker model run hf.co/tasal9/ZamAI-mT5-Pashto
```

ZamAI mT5 Pashto

Task: text2text-generation
Languages: ps
Base model: google/mt5-base

Overview

This model is part of the ZamAI Pashto language AI collection. It is fine-tuned/adapted for text2text-generation in Pashto and related languages.

Intended uses & limitations

Use for research, prototyping, and production assistance in Pashto NLP.
Evaluate outputs carefully before deploying in high-stakes applications.
May reflect biases present in the pre-training or fine-tuning data.

How to use

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "tasal9/ZamAI-mT5-Pashto"
model = AutoModelForCausalLM.from_pretrained(model_id, trust_remote_code=True)
tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)

prompt = "سلام دې وي! تاسو څنګه یاست؟"
inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=128, do_sample=True, temperature=0.7)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))

Training data

Training dataset details will be added here.

Evaluation

Metric	Value	Description
bleu	TBD	Add measured value
rouge	TBD	Add measured value

Update this table with your measured results and link to the evaluation script/notebook.

Citation

@misc{zamai_pashto,
  title = {{ZamAI mT5 Pashto}},
  author = {ZamAI / Yaqoob Tasal},
  year = {2024},
  howpublished = {\url{https://huggingface.co/tasal9/ZamAI-mT5-Pashto}}
}

License

This model is released under the "apache-2.0" license.

Downloads last month: 300

Safetensors

Model size

0.6B params

Tensor type

F32

Model tree for tasal9/ZamAI-mT5-Pashto

Base model

google/mt5-base

Finetuned

(315)

this model