tysam-code / hlb-gptLinks
Minimalistic, extremely fast, and hackable researcher's toolbench for GPT models in 307 lines of code. Reaches <3.8 validation loss on wikitext-103 on a single A100 in <100 seconds. Scales to larger models with one parameter change (feature currently in alpha).
☆349Updated last year
Alternatives and similar repositories for hlb-gpt
Users that are interested in hlb-gpt are comparing it to the libraries listed below
Sorting:
- Ungreedy subword tokenizer and vocabulary trainer for Python, Go & Javascript☆595Updated last year
- ☆416Updated last year
- Language Modeling with the H3 State Space Model☆519Updated last year
- Fast bare-bones BPE for modern tokenizer training☆160Updated last month
- ☆93Updated last year
- Puzzles for exploring transformers☆355Updated 2 years ago
- The repository for the code of the UltraFastBERT paper☆516Updated last year
- a small code base for training large models☆307Updated 3 months ago
- A crude RLHF layer on top of nanoGPT with Gumbel-Softmax trick☆291Updated last year
- A puzzle to learn about prompting☆132Updated 2 years ago
- ☆304Updated last year
- Fine-tune mistral-7B on 3090s, a100s, h100s☆717Updated last year
- Helpers and such for working with Lambda Cloud☆51Updated last year
- Extract full next-token probabilities via language model APIs☆247Updated last year
- JAX implementation of the Llama 2 model☆219Updated last year
- Simple Transformer in Jax☆138Updated last year
- Fully fine-tune large models like Mistral, Llama-2-13B, or Qwen-14B completely for free☆232Updated 9 months ago
- ☆143Updated 2 years ago
- Fast & Simple repository for pre-training and fine-tuning T5-style models☆1,006Updated 11 months ago
- Inference code for Persimmon-8B☆415Updated last year
- git extension for {collaborative, communal, continual} model development☆217Updated 8 months ago
- Legible, Scalable, Reproducible Foundation Models with Named Tensors and Jax☆627Updated this week
- A repository for research on medium sized language models.☆506Updated last month
- A repository for log-time feedforward networks☆222Updated last year
- ☆274Updated last year
- batched loras☆344Updated last year
- An interactive exploration of Transformer programming.☆268Updated last year
- NeurIPS Large Language Model Efficiency Challenge: 1 LLM + 1GPU + 1Day☆256Updated last year
- Open weights language model from Google DeepMind, based on Griffin.☆645Updated last month
- Convolutions for Sequence Modeling☆895Updated last year