Instructions to use botbotrobotics/CabraMistral-v3-7b-32k with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use botbotrobotics/CabraMistral-v3-7b-32k with Transformers:

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

pipe = pipeline("text-generation", model="botbotrobotics/CabraMistral-v3-7b-32k")
messages = [
    {"role": "user", "content": "Who are you?"},
]
pipe(messages)

# Load model directly
from transformers import AutoTokenizer, AutoModelForMultimodalLM

tokenizer = AutoTokenizer.from_pretrained("botbotrobotics/CabraMistral-v3-7b-32k")
model = AutoModelForMultimodalLM.from_pretrained("botbotrobotics/CabraMistral-v3-7b-32k")
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 botbotrobotics/CabraMistral-v3-7b-32k with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "botbotrobotics/CabraMistral-v3-7b-32k"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "botbotrobotics/CabraMistral-v3-7b-32k",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/botbotrobotics/CabraMistral-v3-7b-32k

SGLang

How to use botbotrobotics/CabraMistral-v3-7b-32k 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 "botbotrobotics/CabraMistral-v3-7b-32k" \
    --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": "botbotrobotics/CabraMistral-v3-7b-32k",
		"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 "botbotrobotics/CabraMistral-v3-7b-32k" \
        --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": "botbotrobotics/CabraMistral-v3-7b-32k",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Docker Model Runner
How to use botbotrobotics/CabraMistral-v3-7b-32k with Docker Model Runner:
```
docker model run hf.co/botbotrobotics/CabraMistral-v3-7b-32k
```
Browse Quantizations to use this model in llama.cpp, Ollama, LM Studio, or any compatible app.

Cabra Mistral 7b v3 - 32k

Esse modelo é um finetune do Mistral 7b Instruct 0.3 com o dataset Cabra12k. Esse modelo é optimizado para português e tem limite de contexto de 32k.

Conheça os nossos outros modelos: Cabra.

Detalhes do Modelo

Modelo: Mistral 7b Instruct 0.3

Mistral-7B-v0.3 é um modelo de transformador, com as seguintes escolhas arquitetônicas:

Grouped-Query Attention
Sliding-Window Attention
Byte-fallback BPE tokenizer

dataset: Cabra 12k

Dataset interno para finetuning. Vamos lançar em breve.

Quantização / GGUF

Colocamos diversas versões (GGUF) quantanizadas no branch "quantanization".

Exemplo

<s> [INST] who is Elon Musk? [/INST]Elon Musk é um empreendedor, inventor e capitalista americano. Ele é o fundador, CEO e CTO da SpaceX, CEO da Neuralink e fundador do The Boring Company. Musk também é o proprietário do Twitter.</s>

Paramentros de trainamento

- learning_rate: 1e-05
- train_batch_size: 4
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 2
- gradient_accumulation_steps: 8
- total_train_batch_size: 64
- total_eval_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.01
- num_epochs: 3

Framework

Transformers 4.39.0.dev0
Pytorch 2.1.2+cu118
Datasets 2.14.6
Tokenizers 0.15.2

Evals

Open Portuguese LLM Leaderboard Evaluation Results

Detailed results can be found here and on the 🚀 Open Portuguese LLM Leaderboard

Metric	Value
Average	60.66
ENEM Challenge (No Images)	58.64
BLUEX (No Images)	45.62
OAB Exams	41.46
Assin2 RTE	86.14
Assin2 STS	68.06
FaQuAD NLI	47.46
HateBR Binary	70.46
PT Hate Speech Binary	62.39
tweetSentBR	65.71

Downloads last month: 13

Safetensors

Model size

7B params

Tensor type

BF16

Model tree for botbotrobotics/CabraMistral-v3-7b-32k

Quantizations

2 models

Space using botbotrobotics/CabraMistral-v3-7b-32k 1

Collection including botbotrobotics/CabraMistral-v3-7b-32k

Models

Collection

Fine-tuned LLMs for the Portuguese language and Brazilian culture. • 10 items • Updated Jun 21, 2024 • 3

Evaluation results

accuracy on ENEM Challenge (No Images)
Open Portuguese LLM Leaderboard

58.640
accuracy on BLUEX (No Images)
Open Portuguese LLM Leaderboard

45.620
accuracy on OAB Exams
Open Portuguese LLM Leaderboard

41.460
f1-macro on Assin2 RTE
test set Open Portuguese LLM Leaderboard

86.140
pearson on Assin2 STS
test set Open Portuguese LLM Leaderboard

68.060
f1-macro on FaQuAD NLI
test set Open Portuguese LLM Leaderboard

47.460
f1-macro on HateBR Binary
test set Open Portuguese LLM Leaderboard

70.460
f1-macro on PT Hate Speech Binary
test set Open Portuguese LLM Leaderboard

62.390