Instructions to use mdba/escutia-qlora with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use mdba/escutia-qlora with Transformers:

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

pipe = pipeline("text-generation", model="mdba/escutia-qlora")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("mdba/escutia-qlora")
model = AutoModelForCausalLM.from_pretrained("mdba/escutia-qlora", device_map="auto")
messages = [
    {"role": "user", "content": "Who are you?"},
]
inputs = tokenizer.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

vLLM

How to use mdba/escutia-qlora with vLLM:

Install from pip and serve model

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

Use Docker

docker model run hf.co/mdba/escutia-qlora

SGLang

How to use mdba/escutia-qlora 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 "mdba/escutia-qlora" \
    --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": "mdba/escutia-qlora",
		"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 "mdba/escutia-qlora" \
        --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": "mdba/escutia-qlora",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use mdba/escutia-qlora with Docker Model Runner:
```
docker model run hf.co/mdba/escutia-qlora
```

mdba commited on 11 days ago

Commit

dec390f

verified ·

1 Parent(s): 351f85d

Publicar adapter QLoRA do EscutIA

Browse files

Files changed (1) hide show

README.md +39 -23

README.md CHANGED Viewed

@@ -1,23 +1,39 @@
-# EscutIA QLoRA
-Adapter QLoRA para classificação de sentimentos em português.
-## Modelo-base
-- modelo: `Qwen/Qwen2.5-1.5B-Instruct`
-- revisão: `989aa79`
-- quantização do treinamento: 4 bits, NF4, double quantization
-- tarefa: classificação em `negativo`, `neutro` ou `positivo`
-Este repositório contém um adapter, não um modelo completo. Para usar os pesos, carregue o modelo-base e aplique o adapter com PEFT.
-```python
-from transformers import AutoModelForCausalLM, AutoTokenizer
-from peft import PeftModel
-base = AutoModelForCausalLM.from_pretrained('Qwen/Qwen2.5-1.5B-Instruct')
-tokenizer = AutoTokenizer.from_pretrained('Qwen/Qwen2.5-1.5B-Instruct')
-model = PeftModel.from_pretrained(base, 'SEU_USUARIO/escutia-qlora')
-```
-Consulte os arquivos de avaliação e comparação incluídos neste pacote para conhecer o desempenho observado.

+---
+base_model: Qwen/Qwen2.5-1.5B-Instruct
+base_model_relation: adapter
+library_name: peft
+language:
+- pt
+pipeline_tag: text-generation
+tags:
+- qlora
+- lora
+- peft
+- bitsandbytes
+- sentiment-analysis
+- portuguese
+---
+# EscutIA QLoRA
+Adapter QLoRA para classificação de sentimentos em português.
+## Modelo-base
+- modelo: `Qwen/Qwen2.5-1.5B-Instruct`
+- revisão: `989aa79`
+- quantização do treinamento: 4 bits, NF4, double quantization
+- tarefa: classificação em `negativo`, `neutro` ou `positivo`
+Este repositório contém um adapter, não um modelo completo. Para usar os pesos, carregue o modelo-base e aplique o adapter com PEFT.
+```python
+from transformers import AutoModelForCausalLM, AutoTokenizer
+from peft import PeftModel
+base = AutoModelForCausalLM.from_pretrained('Qwen/Qwen2.5-1.5B-Instruct')
+tokenizer = AutoTokenizer.from_pretrained('Qwen/Qwen2.5-1.5B-Instruct')
+model = PeftModel.from_pretrained(base, 'SEU_USUARIO/escutia-qlora')
+```
+Consulte os arquivos de avaliação e comparação incluídos neste pacote para conhecer o desempenho observado.