salesforce / jaxformer
Minimal library to train LLMs on TPU in JAX with pjit().
☆270Updated 9 months ago
Related projects: ⓘ
- Code for the paper "Rethinking Benchmark and Contamination for Language Models with Rephrased Samples"☆264Updated 9 months ago
- Fast and flexible reference benchmarks☆435Updated last month
- CodeGen2 models for program synthesis☆272Updated last year
- This is the official code for the paper CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learning (Neur…☆492Updated 11 months ago
- Repository for analysis and experiments in the BigCode project.☆113Updated 6 months ago
- Ongoing research training transformer models at scale☆371Updated last month
- Code for the paper "Efficient Training of Language Models to Fill in the Middle"☆162Updated last year
- ☆251Updated last year
- Fine-tune SantaCoder for Code/Text Generation.☆182Updated last year
- The data processing pipeline for the Koala chatbot language model☆115Updated last year
- Generative model for code infilling and synthesis☆292Updated last year
- ☆357Updated last month
- Run evaluation on LLMs using human-eval benchmark☆374Updated last year
- batched loras☆327Updated last year
- GPTQ inference Triton kernel☆273Updated last year
- ☆338Updated last year
- Fast Inference Solutions for BLOOM☆556Updated last month
- Multipack distributed sampler for fast padding-free training of LLMs☆170Updated last month
- Scalable toolkit for efficient model alignment☆509Updated this week
- A multi-programming language benchmark for LLMs☆189Updated this week
- Scaling Data-Constrained Language Models☆310Updated this week
- Open Instruction Generalist is an assistant trained on massive synthetic instructions to perform many millions of tasks☆204Updated 8 months ago
- Salesforce open-source LLMs with 8k sequence length.☆717Updated 9 months ago
- Chain-of-Hindsight, A Scalable RLHF Method☆213Updated 11 months ago
- DeepSpeed is a deep learning optimization library that makes distributed training easy, efficient, and effective.☆163Updated 4 months ago
- Used for adaptive human in the loop evaluation of language and embedding models.☆300Updated last year
- Transformers with Arbitrarily Large Context☆613Updated last month
- Official repository for the paper "LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code"☆173Updated 3 weeks ago
- distributed trainer for LLMs☆524Updated 4 months ago
- 🐙 OctoPack: Instruction Tuning Code Large Language Models☆421Updated last month