srush / do-we-need-attention
โ161Updated last year
Related projects โ
Alternatives and complementary repositories for do-we-need-attention
- Understand and test language model architectures on synthetic tasks.โ162Updated 6 months ago
- A MAD laboratory to improve AI architecture designs ๐งชโ95Updated 6 months ago
- some common Huggingface transformers in maximal update parametrization (ยตP)โ76Updated 2 years ago
- The simplest, fastest repository for training/finetuning medium-sized GPTs.โ84Updated last week
- Code for the paper "The Impact of Positional Encoding on Length Generalization in Transformers", NeurIPS 2023โ127Updated 6 months ago
- Language models scale reliably with over-training and on downstream tasksโ94Updated 7 months ago
- [NeurIPS 2023] Learning Transformer Programsโ157Updated 6 months ago
- Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT trainingโ113Updated 7 months ago
- โ132Updated last year
- Code for exploring Based models from "Simple linear attention language models balance the recall-throughput tradeoff"โ214Updated 3 months ago
- โ73Updated 4 months ago
- Inference code for LLaMA models in JAXโ113Updated 6 months ago
- Code to reproduce "Transformers Can Do Arithmetic with the Right Embeddings", McLeish et al (NeurIPS 2024)โ178Updated 5 months ago
- โ107Updated this week
- โ50Updated 6 months ago
- LoRA for arbitrary JAX models and functionsโ132Updated 8 months ago
- Griffin MQA + Hawk Linear RNN Hybridโ85Updated 6 months ago
- โ175Updated this week
- NanoGPT-like codebase for LLM trainingโ75Updated this week
- Scalable neural net training via automatic normalization in the modular norm.โ121Updated 3 months ago
- โ128Updated this week
- Some preliminary explorations of Mamba's context scaling.โ191Updated 9 months ago
- A puzzle to learn about promptingโ121Updated last year
- โ35Updated 7 months ago
- Experiments around a simple idea for inducing multiple hierarchical predictive model within a GPTโ205Updated 3 months ago
- Implementation of the Llama architecture with RLHF + Q-learningโ157Updated 10 months ago
- โ77Updated 5 months ago
- Experiment of using Tangent to autodiff tritonโ72Updated 9 months ago
- Large scale 4D parallelism pre-training for ๐ค transformers in Mixture of Experts *(still work in progress)*โ80Updated 11 months ago
- โ197Updated 4 months ago