keras-nlp

keras-nlp copied to clipboard

Support tokenization of special tokens for word_piece_tokenizer

6

this PR enables the user to tokenize special tokens from text input as was suggested in keras-team/keras-nlp#1395. The behaviour is the same as `ByteBairTokenizer` I think. also the models' tokenizers...

abuelnasr0

Add presets for Electra and checkpoint conversion script

21

I have uploaded the weights on personal google cloud bucket. The `from_preset` method works properly in my local setup, but it throws some error in google collab [notebook](https://colab.research.google.com/drive/1K5zTjzoJrFGJ5Ej4s3g8lWzJ2VMqhgwu?usp=sharing).

pranavvp16

Upload Model to Kaggle

3

Implement `upload_preset()` to allow users to upload Model presets to Kaggle.

SamanehSaadat

**Is your feature request related to a problem? Please describe.** We can currently load Mistral 7b models with `keras_nlp.models.MistralCausalLM.from_preset("mistral_7b_en")` (or `mistral_instruct_7b_en`). I noticed those are the version `0.1` of the...

borisdayma

Solves https://github.com/keras-team/keras-nlp/issues/1294. As mentioned in https://github.com/keras-team/keras-nlp/issues/1294#issuecomment-1966864503, this PR adds support for the `hf://` prefix to load presets from the Huggingface Hub. The integration requires the [`huggingface_hub`](https://github.com/huggingface/huggingface_hub) library. Authentication can be...

Wauplin

Feature Request: Transformer Debugger - Debugging and controlling the behavior of transformer based LLM models.

3

**Short Description** > Transformer Debugger (TDB) is a tool developed by OpenAI's [Superalignment team](https://openai.com/blog/introducing-superalignment) with the goal of supporting investigations into specific behaviors of small language models. The tool combines...

abhaskumarsinha

Add `from_huggingface` method to KerasNLP models

5

## Add support for loading huggingface model checkpoints in KerasNLP backbones **Is your feature request related to a problem? Please describe.** As of now KerasNLP backbones load pretrained weights of...

kanpuriyanawab

Improves the documentation in `layers/modeling` by - Adding a missing argument description to `TokenAndPositionEmbedding`. - Aligning the usage of `name` argument in `TransformerDecoder` and `TransformerEncoder` layers with that in `FNetEncoder`...

mykolaskrynnyk

Data-Parallel Training with KerasNLP and tf.distribute example dataset problem

4

**Describe the bug** [Data-Parallel Training with KerasNLP and tf.distribute](https://keras.io/examples/nlp/data_parallel_training_with_keras_nlp/) This is an example using a dataset that shows 403: Forbidden. Giving the message "Access Denied.". **To Reproduce** Provide a link...

sitammeur

Add `Falcon` Model

2

Falcon is a family of language models created by the [Technology Innovation Institute (TII)](https://www.tii.ae/). Falcon family has multiple base models: [Falcon-180b](https://huggingface.co/tiiuae/falcon-180B), [Falcon-40B](https://huggingface.co/tiiuae/falcon-40b), [Falcon-7B](https://huggingface.co/tiiuae/falcon-7b), and [Falcon-rw-1b](https://huggingface.co/tiiuae/falcon-rw-1b). Instruction tuned version of some...

SamanehSaadat