Lightning-AI litgpt issues

Lightning-AI / litgpt

Pretrain, finetune, deploy 20+ LLMs on your own data. Uses state-of-the-art techniques: flash attention, FSDP, 4-bit, LoRA, and more.

https://lightning.ai

Apache License 2.0

6.69k stars 711 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

ValueError: Cannot attend to 3063, block size is only 2048

#1387 Gooooooogo closed 3 hours ago
1
Will CycleIterator forward to dataset on resume for pretrain?

#1386 calvintwr opened 10 hours ago
0
LoRA multi-GPU no longer works if applying LoRA selectively

#1385 awaelchli opened 20 hours ago
1
fabric.print only works on sys.stderr, does not print inference result

#1384 lastmjs opened 21 hours ago
0
How to use custom dataset for evaluate?

#1383 Gooooooogo opened 1 day ago
0
Remove out directory to gitignore

#1382 usmanxia closed 1 day ago
1
Add release workflow

#1381 rasbt opened 3 days ago
0
Readme.md - Instruct how to get HF_TOKEN

#1380 natanloterio closed 2 days ago
2
How to specify which GPU to use？

#1379 Gooooooogo closed 4 days ago
2
Cannot copy out of meta tensor; no data!

#1378 Gooooooogo opened 4 days ago
1
After some iteration in pretraining a LLM, IndexError is raised related to dataset chunking

#1377 MusulmonLolayev opened 4 days ago
0
Update LoRA test

#1376 awaelchli closed 5 days ago
0
Update Lightning version

#1375 awaelchli closed 5 days ago
0
Eliminate cuda syncs

#1374 robieta closed 5 days ago
5
More informative download error messages

#1373 rasbt closed 5 days ago
0
Option to skip expensive final validation

#1372 rasbt opened 6 days ago
1
Change examples to phi-2

#1371 rasbt closed 6 days ago
0
Add link to Studio for benchmarks

#1370 awaelchli closed 6 days ago
0
Why FSDPStrategy is so slow-down when I use multi-machine

#1369 Graduo opened 6 days ago
4
A potential bug for multi-GPU training

#1368 zyushun opened 1 week ago
5
Only run expensive tests if code files change

#1367 rasbt closed 6 days ago
3
combine FSDP with selective activation checkpointing

#1366 nemoramo opened 1 week ago
0
Add Mixtral MoE to README

#1365 lantiga closed 1 week ago
0
Add support for memory-efficient and faster optimizers

#1364 rasbt opened 1 week ago
1
litgpt download doesn't work

#1363 natanloterio closed 5 days ago
7
Failed to load the finetuned model with `AutoModelForCausalLM.from_pretrained(name, state_dict=state_dict)`

#1362 zhaosheng-thu opened 1 week ago
4
Update table with new benchmark results

#1361 awaelchli closed 1 week ago
0
Feature/top p sampling

#1360 belerico closed 2 days ago
3
Conversion to HF checkpoint should generate a checkpoint format that can be loaded directly

#1359 awaelchli opened 1 week ago
1
OOM Error: CUDA out of memory when finetuning llama3-8b

#1358 zhaosheng-thu closed 1 week ago
3
Fix `litgpt evaluate` not using the local checkpoint

#1357 awaelchli closed 1 week ago
0
Update litserve dependency

#1356 rasbt closed 1 week ago
0
Avoid remote code warning in evaluation harness

#1355 awaelchli closed 1 week ago
1
Add resume for adapter_v2, enable continued finetuning for adapter

#1354 altria-zewei-wang opened 1 week ago
2
Add precision arg for pretraining

#1353 rasbt closed 1 week ago
2
--checkpoint-dir 'xx' is missing the files: ['model_config.yaml']

#1352 zhaosheng-thu closed 1 week ago
2
ValueError: 'Meta-Llama-3-8B-Instruct' is not a supported config name

#1351 BZandi closed 1 week ago
4
Add LongLora for both full and lora fine-tuning

#1350 belerico opened 1 week ago
5
The `litgpt evaluate` command attempts to download config files from gated repos

#1349 awaelchli closed 1 week ago
1
Add release workflow

#1348 carmocca opened 1 week ago
2
Nucleus (top-p) sampling

#1347 belerico opened 1 week ago
2
Feature/longlora

#1346 belerico closed 1 week ago
3
Add support for phi-3-mini

#1345 Dev-Khant closed 1 week ago
2
Fix evaluation if device not specified

#1344 awaelchli closed 1 week ago
0
Phi (tests): create a class directly from HF

#1343 Andrei-Aksionov closed 1 week ago
0
Tokenizer: `add_prefix_space` shouldn't affect `self.use_bos`

#1342 carmocca closed 1 week ago
1
Add phi-3 checkpoint

#1341 rasbt opened 1 week ago
7
Qwen1.5 Family Support

#1340 junzhang-zj opened 1 week ago
0
Continual pretraining for custom data is not working. Not recognizing TextFiles as a data attribute.

#1339 karkeranikitha closed 1 week ago
3
Standardize out_dir behavior

#1338 rasbt closed 1 week ago
3