|
--- |
|
base_model: google/t5-v1_1-base |
|
tags: |
|
- datadreamer |
|
- datadreamer-0.1.0 |
|
- synthetic |
|
- gpt-4 |
|
- gpt-4 |
|
- text2text-generation |
|
widget: |
|
- text: >- |
|
An important paradigm of natural language processing consists of large-scale pre-training on general domain data and adaptation to particular tasks or domains. As we pre-train larger models, full fine-tuning, which retrains all model parameters, becomes less feasible. Using GPT-3 175B as an example -- deploying independent instances of fine-tuned models, each with 175B parameters, is prohibitively expensive. We propose Low-Rank Adaptation, or LoRA, which freezes the pre-trained model weights and injects trainable rank decomposition matrices into each layer of the Transformer architecture, greatly reducing the number of trainable parameters for downstream tasks. Compared to GPT-3 175B fine-tuned with Adam, LoRA can reduce the number of trainable parameters by 10,000 times and the GPU memory requirement by 3 times. LoRA performs on-par or better than fine-tuning in model quality on RoBERTa, DeBERTa, GPT-2, and GPT-3, despite having fewer trainable parameters, a higher training throughput, and, unlike adapters, no additional inference latency. We also provide an empirical investigation into rank-deficiency in language model adaptation, which sheds light on the efficacy of LoRA. We release a package that facilitates the integration of LoRA with PyTorch models and provide our implementations and model checkpoints for RoBERTa, DeBERTa, and GPT-2 at this https URL. |
|
example_title: LoRA Abstract |
|
- text: >- |
|
Making language models bigger does not inherently make them better at following a user's intent. For example, large language models can generate outputs that are untruthful, toxic, or simply not helpful to the user. In other words, these models are not aligned with their users. In this paper, we show an avenue for aligning language models with user intent on a wide range of tasks by fine-tuning with human feedback. Starting with a set of labeler-written prompts and prompts submitted through the OpenAI API, we collect a dataset of labeler demonstrations of the desired model behavior, which we use to fine-tune GPT-3 using supervised learning. We then collect a dataset of rankings of model outputs, which we use to further fine-tune this supervised model using reinforcement learning from human feedback. We call the resulting models InstructGPT. In human evaluations on our prompt distribution, outputs from the 1.3B parameter InstructGPT model are preferred to outputs from the 175B GPT-3, despite having 100x fewer parameters. Moreover, InstructGPT models show improvements in truthfulness and reductions in toxic output generation while having minimal performance regressions on public NLP datasets. Even though InstructGPT still makes simple mistakes, our results show that fine-tuning with human feedback is a promising direction for aligning language models with human intent. |
|
example_title: InstructGPT Abstract |
|
- text: >- |
|
In deep learning, models typically reuse the same parameters for all inputs. Mixture of Experts (MoE) defies this and instead selects different parameters for each incoming example. The result is a sparsely-activated model -- with outrageous numbers of parameters -- but a constant computational cost. However, despite several notable successes of MoE, widespread adoption has been hindered by complexity, communication costs and training instability -- we address these with the Switch Transformer. We simplify the MoE routing algorithm and design intuitive improved models with reduced communication and computational costs. Our proposed training techniques help wrangle the instabilities and we show large sparse models may be trained, for the first time, with lower precision (bfloat16) formats. We design models based off T5-Base and T5-Large to obtain up to 7x increases in pre-training speed with the same computational resources. These improvements extend into multilingual settings where we measure gains over the mT5-Base version across all 101 languages. Finally, we advance the current scale of language models by pre-training up to trillion parameter models on the 'Colossal Clean Crawled Corpus' and achieve a 4x speedup over the T5-XXL model. |
|
example_title: Switch Transformers Abstract |
|
pipeline_tag: text2text-generation |
|
datasets: |
|
- datadreamer-dev/abstracts_and_tweets |
|
--- |
|
# Model Card |
|
|
|
This is an "Abstract to Tweet" model that crafts a tweet summarizing a research paper abstract trained on a [synthetic dataset of arXiv abstracts and tweets](https://huggingface.co/datasets/datadreamer-dev/abstracts_and_tweets/). It is used as a demonstration of the [DataDreamer 🤖💤 library](https://datadreamer.dev/docs/latest/). |
|
|
|
## Example Usage |
|
|
|
```python3 |
|
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer, pipeline |
|
|
|
tokenizer = AutoTokenizer.from_pretrained('datadreamer-dev/abstracts_to_tweet_model', revision=None) # Load tokenizer |
|
model = AutoModelForSeq2SeqLM.from_pretrained('datadreamer-dev/abstracts_to_tweet_model', revision=None) # Load model |
|
pipe = pipeline('text2text-generation', model=model, tokenizer=tokenizer, pad_token_id=tokenizer.pad_token_id) |
|
|
|
inputs = ["In this paper, we delve into advanced techniques and methods in Natural Language Processing (NLP), innovatively incorporating Transformer architectures and self-supervised learning methods. We aim to reiterate the current understanding of Transformer-based models in executing various language tasks by dissecting their versatility and expandability on broad language systems.\n\nMoreover, stabilization measures, tokenization assortment, and interpreting latent spaces provide an in-depth novelty to our pipeline, overcoming long-known obstacles. We explore meta-architectural modifications focusing on enhancing prompt language models' efficiency, allowing flexible adaptations to the core Transformer technique's abundance in BERT, GPT-like systems.\n\nTo implement these adaptations, several experiments were conducted on varied benchmark datasets to evaluate core metrics such as Bleu, Rouge, and Warp-CTC metrics in translation and transcription tasks. We carried out significant analysis focusing on module interpretability, additional error inspection, task-specific regulatory mechanisms, execution speed, and computational considerations.\n\nOur experimental results bring in distraction from widespread but sub-optimal benchmarks and offer evidence underpinning the contrary yet potent issues yet to be addressed methodically. We invite the community to reflect on these novel insights, develop and refine our proposed techniques, speeding technical progress, avoiding prototypical retrodiction in the Natural Language Understanding ecosystem to respect inclusive, diverse, and correctly perceived expressive content."] |
|
print(pipe(inputs, max_length=512, do_sample=False)) |
|
``` |
|
|
|
--- |
|
This model was trained with a synthetic dataset with [DataDreamer 🤖💤](https://datadreamer.dev). The synthetic dataset card and model card can be found [here](datadreamer.json). The training arguments can be found [here](training_args.json). |
