Haris Jabbar's repositories
agency-jekyll-theme
Agency Theme for Jekyll
amber-data-prep
Data preparation code for Amber 7B LLM
MicroLlama
This is a 300M MicroLlama version of TinyLlama
dolma
Data and tools for generating and inspecting OLMo pre-training data.
jekyll-theme-neumorphism
Neumorphism designed Jekyll theme for personal websites, portfolios and resumes.
langchain-chatbot-demo
Examples of chatbot implementations with Langchain and Streamlit
linggpt
Pretrain, finetune, deploy 20+ LLMs on your own data. Uses state-of-the-art techniques: flash attention, FSDP, 4-bit, LoRA, and more.
LLaMA-Efficient-Tuning
Easy-to-use LLM fine-tuning framework (LLaMA-2, BLOOM, Falcon, Baichuan, Qwen, ChatGLM2)
lm-evaluation-harness
A framework for few-shot evaluation of autoregressive language models.
minbpe_spark_gcp
Implementing MinBPE training on GCP DataProc (serverless spark on GCP)
MobiLlama
MobiLlama : Small Language Model tailored for edge devices
OLMo
Modeling, training, eval, and inference code for OLMo
pandas-ai
PandasAI is the Python library that integrates Gen AI into pandas, making data analysis conversational
paralegal
Streamit app with langchain and huggingface
promptbench
A unified evaluation framework for large language models
promptsource
Toolkit for creating, sharing and using natural language prompts.
python-package-template
A template repo for Python packages from AllenAI
sentence-transformers
Multilingual Sentence & Image Embeddings with BERT
spacyface
Align the token outputs from Spacy and Huggingface to help understand what language structures transformers see
sql-eval
Evaluate the accuracy of LLM generated outputs
Streamlit-Authenticator
A secure authentication module to validate user credentials in a Streamlit application.
tiktokenizer
Online playground for OpenAPI tokenizers
useb
Heterogenous, Task- and Domain-Specific Benchmark for Unsupervised Sentence Embeddings used in the TSDAE paper: https://arxiv.org/abs/2104.06979.
vllm
A high-throughput and memory-efficient inference and serving engine for LLMs
xformers
Hackable and optimized Transformers building blocks, supporting a composable construction.