defdet

aij-comp

binary_code_through_sound

Выпускной проект

EasyLM

Large language models (LLMs) made easy, EasyLM is a one stop solution for pre-training, finetuning, evaluating and serving LLMs in JAX/Flax.

falcon-jax

JAX implementation of the falcon model

Gemma-EasyLM

Train GEMMA on TPU/GPU! (Codebase for training Gemma-Ko Series)

llama-2-jax-parallel

JAX implementation of the Llama 2 model

mmlu_ru

MMLU eval for RU/EN

movie_reviews_rating

phi-jax

Implementation of microsoft Phi model in pure functional JAX

SMPD-training

Reference HF's LLMs (e.g LLAMA) training implementation on Kaggle TPU hardware leveraging torch XLA + SMPD

defdet/SMPD-training

Reference HF's LLMs (e.g LLAMA) training implementation on Kaggle TPU hardware leveraging torch XLA + SMPD

defdet/aij-comp

defdet/binary_code_through_sound

Выпускной проект

defdet/EasyLM

Large language models (LLMs) made easy, EasyLM is a one stop solution for pre-training, finetuning, evaluating and serving LLMs in JAX/Flax.

defdet/falcon-jax

JAX implementation of the falcon model

defdet/Gemma-EasyLM

Train GEMMA on TPU/GPU! (Codebase for training Gemma-Ko Series)

defdet/llama-2-jax-parallel

JAX implementation of the Llama 2 model

defdet/mmlu_ru

MMLU eval for RU/EN

defdet/movie_reviews_rating

defdet/phi-jax

Implementation of microsoft Phi model in pure functional JAX

defdet/Poly-modification-bolgov

defdet/qwen2-jax

defdet/rulm

Language modeling and instruction tuning for Russian

defdet/tpu-generate

defdet/transformers-fixed-llama

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

defdet/vk-spam-classification

defdet/xformers_hip_try

Hackable and optimized Transformers building blocks, supporting a composable construction.