metadata
license: llama2
base_model: TheBloke/Xwin-LM-7B-V0.1-GPTQ
model-index:
- name: cleante
results: []
Cleante
Cleante is a fine-tuned model, based on a pre-trained 7B model.
Usage
from transformers import pipeline, AutoModelForCausalLM, AutoTokenizer
model_name = "guillaumephd/cleante"
# Load the model and tokenizer
model = AutoModelForCausalLM.from_pretrained(model_name)
tokenizer = AutoTokenizer.from_pretrained(model_name)
# Define the text generation pipeline
generator = pipeline(
"text-generation",
model=model,
tokenizer=tokenizer,
device=0 # Use GPU if available please
)
# Generate text using the Cleante model
prompt = "###Human: What's your nickname, assistant? ###Assistant: "
output = generator(prompt, max_length=100, do_sample=True, temperature=0.5, repetition_penalty=1.2,)
# Print the generated text
print(output[0]["generated_text"])
outputs = model.generate(**inputs, generation_config=generation_config)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))
# The model should output a text that looks like:
# "My name is Cléante, and I was trained by Guillaume as a language model."
Model description
See above.
Intended uses & limitations
Demonstration purpose only.
Training and evaluation data
Personal data.
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0002
- train_batch_size: 8
- eval_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
Framework versions
- Transformers 4.33.3
- Pytorch 2.0.1+cu118
- Datasets 2.14.5
- Tokenizers 0.13.3