|
--- |
|
library_name: transformers |
|
language: |
|
- en |
|
license: apache-2.0 |
|
base_model: google/flan-t5-xl |
|
datasets: |
|
- pszemraj/summary-map-reduce-v1 |
|
pipeline_tag: text2text-generation |
|
tags: |
|
- map-reduce |
|
- summarization |
|
--- |
|
|
|
# flan-t5-xl-summary-map-reduce-1024 |
|
|
|
A larger t2t model trained to complete the "reduce" step (_consolidation step_) of map-reduce summarization. |
|
|
|
## About |
|
|
|
> [!TIP] |
|
> Refer to [this wiki page](https://github.com/pszemraj/textsum/wiki/consolidating-summaries) or the [smaller BART model card](https://hf.co/pszemraj/bart-large-summary-map-reduce) for explanations and usage examples. |
|
|
|
|
|
Comparatively, this model seems to |
|
|
|
- produce more eloquent final reduced summaries |
|
- more "gullible"/sensitive to noise in the input summaries |
|
- i.e. a hallucinated one-off term/name/entity is likely to be mentioned/appear in the reduced summary |
|
- agnostic to whitespace in input (_by definition, since the t5 tokenizer normalizes whitespace_) |
|
|
|
Therefore, it's recommended to compare sample outputs of this model and [the BART version](https://hf.co/pszemraj/bart-large-summary-map-reduce) on your data to see which is better for your use case. |
|
|
|
## Details |
|
|
|
This model is a fine-tuned version of [google/flan-t5-xl](https://huggingface.co/google/flan-t5-xl) on the pszemraj/summary-map-reduce-v1 dataset at 1024 context length in/out. |
|
|
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.6039 |
|
- Num Input Tokens Seen: 7138765 |
|
|
|
## Training procedure |
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- learning_rate: 8e-05 |
|
- train_batch_size: 2 |
|
- eval_batch_size: 2 |
|
- seed: 17868 |
|
- gradient_accumulation_steps: 32 |
|
- total_train_batch_size: 64 |
|
- optimizer: Use OptimizerNames.PAGED_ADAMW with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments |
|
- lr_scheduler_type: cosine |
|
- lr_scheduler_warmup_ratio: 0.05 |
|
- num_epochs: 2.0 |
