README.md

---
license: apache-2.0
tags:
- trl
- transformers
- reinforcement-learning
---

# TRL Model

This is a [TRL language model](https://github.com/lvwerra/trl) that has been fine-tuned with reinforcement learning to
 guide the model outputs according to a value, function, or human feedback. The model can be used for text generation.

## Usage

To use this model for inference, first install the TRL library:

```bash
python -m pip install trl
```

You can then generate text as follows:

```python
from transformers import pipeline

generator = pipeline("text-generation", model="vincentmin//tmp/tmp_mpm7q96/vincentmin/opt-125m-eli5-rl-finetune-128-8-8-1.4e-5_ada")
outputs = generator("Hello, my llama is cute")
```

If you want to use the model for training or to obtain the outputs from the value head, load the model as follows:

```python
from transformers import AutoTokenizer
from trl import AutoModelForCausalLMWithValueHead

tokenizer = AutoTokenizer.from_pretrained("vincentmin//tmp/tmp_mpm7q96/vincentmin/opt-125m-eli5-rl-finetune-128-8-8-1.4e-5_ada")
model = AutoModelForCausalLMWithValueHead.from_pretrained("vincentmin//tmp/tmp_mpm7q96/vincentmin/opt-125m-eli5-rl-finetune-128-8-8-1.4e-5_ada")

inputs = tokenizer("Hello, my llama is cute", return_tensors="pt")
outputs = model(**inputs, labels=inputs["input_ids"])
```