Lightning-AI/litgpt

View on GitHub

20+ high-performance LLMs with recipes to pretrain, finetune and deploy at scale.

Stars: 13.7k
Forks: 1.5k
Open beginner issues: 1
Indexed issues: 239
Avg merge: 15h 37m
Merged PRs (30d): 1

Dominant language: Python
License: Apache-2.0
Last GitHub push: Sep 9, 2026
Latest indexed: Sep 19, 2026
Contributing guide: Contributing guide
Code of conduct: Code of conduct
Beginner labels: help wanted good first issue

Beginner friendly Everything

Open All

239 open issues indexed

LLM.generate() can silently select a token excluded by top_k on CPU when sampling probabilities are float16 Open

Difficulty 4/5 3-5 days Newbie friendliness 55/100

Lightning-AI/litgpt#2323 · Opened Sep 8, 2026
litgpt pretrain crashes or deadlocks when a rank's val_dataloader yields zero batches (multi-GPU FSDP) Open

Difficulty 4/5 3-5 days Newbie friendliness 55/100

Lightning-AI/litgpt#2319 · 1 comment · Opened Sep 2, 2026
Speculative decoding drops the first generated token Open

Difficulty 2/5 1-3 hours Newbie friendliness 84/100

Lightning-AI/litgpt#2307 · Opened Aug 25, 2026
Reang Open

question

Difficulty 5/5 Over a week Newbie friendliness 5/100

Lightning-AI/litgpt#2306 · Opened Aug 25, 2026
Stop sequence matching mishandles overlapping prefixes and buffered tokens Open

Difficulty 3/5 1-2 days Newbie friendliness 78/100

Lightning-AI/litgpt#2302 · Opened Aug 19, 2026
No checkpoints saved during pretraining in Google Colab despite log.log_model and save_interval set. Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 48/100

Lightning-AI/litgpt#2223 · 3 comments · Opened Mar 26, 2026
`mask_cache` in kv-cache path seems to force `attn_mask`, preventing flash attention Open

question

Difficulty 4/5 3-5 days Newbie friendliness 42/100

Lightning-AI/litgpt#2220 · Opened Mar 23, 2026
Gradient Clipping Doesn't Work in Finetuning Open

question

Difficulty 3/5 1-2 days Newbie friendliness 68/100

Lightning-AI/litgpt#2191 · 3 comments · Opened Jan 15, 2026
Memory Management Issues Open

Difficulty 5/5 Over a week Newbie friendliness 25/100

Lightning-AI/litgpt#2190 · Opened Jan 15, 2026
Unsafe Checkpoint Loading - This Is Bad Open

Difficulty 5/5 Over a week Newbie friendliness 35/100

Lightning-AI/litgpt#2189 · 1 comment · Opened Jan 15, 2026
Feature: Support for Multimodality Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 20/100

Lightning-AI/litgpt#2173 · 1 comment · Opened Dec 23, 2025
Request to support MedGemma and MedSigLip Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 25/100

Lightning-AI/litgpt#2172 · 1 comment · Opened Dec 23, 2025
difference during inference with or without kvcache Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 35/100

Lightning-AI/litgpt#2165 · 3 comments · Opened Nov 27, 2025
Cut cross entropy for pretraining and finetuning Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 35/100

Lightning-AI/litgpt#2163 · Opened Nov 19, 2025
Support for SSM models (Mamba, Mamba2) Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 25/100

Lightning-AI/litgpt#2158 · 2 comments · Opened Nov 14, 2025
`jsonargparse==4.42.0` causing `test_config_help` failure Open

Difficulty 3/5 1-2 days Newbie friendliness 45/100

Lightning-AI/litgpt#2150 · Opened Oct 30, 2025
Request to support OLMoE Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 25/100

Lightning-AI/litgpt#2133 · 2 comments · Opened Sep 12, 2025
Initial and final evaluation in `finetune` scripts do not accumulate over devices Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 55/100

Lightning-AI/litgpt#2116 · 4 comments · Opened Aug 30, 2025
Bitsandbytes quantization for litgpt 2d parallel model (TP+FSDP) within LightningTrainer Open

question

Difficulty 5/5 Over a week Newbie friendliness 25/100

Lightning-AI/litgpt#2115 · Opened Aug 29, 2025
Efficient Model Loading Open

question

Difficulty 5/5 Over a week Newbie friendliness 25/100

Lightning-AI/litgpt#2100 · 2 comments · Opened Aug 5, 2025
Secrets exfiltration vulnerability Open

Difficulty 4/5 3-5 days Newbie friendliness 20/100

Lightning-AI/litgpt#2090 · Opened Jul 7, 2025
How to bring my tokenizer and set vocabulary size accordingly for training a model with loaded weights Open

question

Difficulty 5/5 Over a week Newbie friendliness 20/100

Lightning-AI/litgpt#2087 · Opened Jul 3, 2025
support swanlab Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 20/100

Lightning-AI/litgpt#2071 · Opened Jun 12, 2025
How to use fabric.clip_gradients in 16-mixed? Open

help wanted question

Difficulty 3/5 1-2 days Newbie friendliness 25/100

Lightning-AI/litgpt#2042 · Opened May 15, 2025
AssertionError: Rank 2 has different values for step: 49996.0 Other ranks: 49991.0 Open

bug waiting on author

Difficulty 4/5 3-5 days Newbie friendliness 25/100

Lightning-AI/litgpt#2032 · 1 comment · Opened May 12, 2025
Add support for Qwen 3 Open

enhancement

Difficulty 3/5 1-2 days Newbie friendliness 42/100

Lightning-AI/litgpt#2031 · 1 comment · 2 reactions · Opened May 8, 2025
Transformers version bump for recent model support Open

Difficulty 4/5 3-5 days Newbie friendliness 35/100

Lightning-AI/litgpt#2030 · Opened May 7, 2025
finetune_lora on gemma bug Open

bug help wanted

Lightning-AI/litgpt#2020 · 1 comment · 1 reaction · 1 assignee · Opened Apr 26, 2025
Llama 4 support Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 20/100

Lightning-AI/litgpt#2015 · 7 comments · Opened Apr 14, 2025
Why Alpaca is only mask_prompt=False?? Open

help wanted question

Difficulty 2/5 1-3 hours Newbie friendliness 38/100

Lightning-AI/litgpt#1987 · Opened Mar 31, 2025
After Fintune lora. apply chat_template Open

help wanted question

Difficulty 4/5 3-5 days Newbie friendliness 25/100

Lightning-AI/litgpt#1986 · Opened Mar 31, 2025
How to download a model from HF that is not listed in the download list? Open

enhancement

Difficulty 3/5 1-2 days Newbie friendliness 35/100

Lightning-AI/litgpt#1980 · 1 comment · Opened Mar 25, 2025
UserWarning: The file size of checkpoints/microsoft/phi-4/lit_model.pth is over 4.2 GB. Open

help wanted question

Difficulty 4/5 3-5 days Newbie friendliness 25/100

Lightning-AI/litgpt#1979 · 1 comment · Opened Mar 25, 2025
Load a json string in the litgpt.data.JSON class instead of a flat file only Open

enhancement

Difficulty 3/5 1-2 days Newbie friendliness 45/100

Lightning-AI/litgpt#1976 · Opened Mar 21, 2025
finetune (lora) with LitData Open

help wanted question

Difficulty 5/5 Over a week Newbie friendliness 25/100

Lightning-AI/litgpt#1966 · Opened Mar 13, 2025
Getting probability distributions Open

help wanted question

Difficulty 4/5 3-5 days Newbie friendliness 25/100

Lightning-AI/litgpt#1950 · Opened Mar 5, 2025
Phi-4-mini-instruct support Open

enhancement

Difficulty 4/5 3-5 days Newbie friendliness 25/100

Lightning-AI/litgpt#1948 · 3 comments · Opened Feb 28, 2025
Generating output from finetuned model using LLM.generate() method vs `litgpt chat` cli command Open

help wanted question

Difficulty 4/5 3-5 days Newbie friendliness 25/100

Lightning-AI/litgpt#1937 · Opened Feb 16, 2025
error: cannot unpack non-iterable ActionTypeHint object Open

bug

Difficulty 3/5 1-2 days Newbie friendliness 35/100

Lightning-AI/litgpt#1931 · 4 comments · Opened Jan 31, 2025
it doesn't help Open

Difficulty 4/5 3-5 days Newbie friendliness 25/100

Lightning-AI/litgpt#1930 · Opened Jan 30, 2025
Bug: Incorrect gradient accumulation steps calculation in multi-node training due to missing world size information Open

bug

Difficulty 4/5 3-5 days Newbie friendliness 42/100

Lightning-AI/litgpt#1927 · Opened Jan 30, 2025
Merging weights after Finetuning with Adapter. Open

help wanted question

Difficulty 4/5 3-5 days Newbie friendliness 35/100

Lightning-AI/litgpt#1921 · Opened Jan 26, 2025
implement DeepSeek Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 20/100

Lightning-AI/litgpt#1920 · 1 comment · 3 reactions · Opened Jan 24, 2025
LitGPT fine-tuning Dont Use GPU Open

help wanted question

Difficulty 4/5 3-5 days Newbie friendliness 25/100

Lightning-AI/litgpt#1911 · 2 comments · Opened Jan 20, 2025
Trouble to load a litpgt trained model using transformers library Open

help wanted question

Difficulty 4/5 3-5 days Newbie friendliness 25/100

Lightning-AI/litgpt#1910 · Opened Jan 20, 2025
Add support for providing `wandb` run name. Open

enhancement

Difficulty 2/5 1-3 hours Newbie friendliness 45/100

Lightning-AI/litgpt#1909 · Opened Jan 17, 2025
LitGPT微调Phi-4大模型视频 Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 15/100

Lightning-AI/litgpt#1907 · 1 comment · 1 reaction · Opened Jan 15, 2025
Refactoring of `GPT.forward` when it comes to `input_pos` and KV cache usage Open

enhancement

Difficulty 5/5 Over a week Newbie friendliness 30/100

Lightning-AI/litgpt#1898 · 8 comments · Opened Jan 6, 2025
Exporting LoRA to HF format without merging Open

help wanted question

Difficulty 5/5 Over a week Newbie friendliness 25/100

Lightning-AI/litgpt#1878 · 3 comments · Opened Dec 20, 2024
Fine-Tuning Chat Model with Domain-Specific Data for custom dataset Open

help wanted question

Difficulty 5/5 Over a week Newbie friendliness 15/100

Lightning-AI/litgpt#1877 · Opened Dec 18, 2024

Lightning-AI/litgpt

Get new issues in your inbox