issues
search
mosaicml
/
llm-foundry
LLM training code for Databricks foundation models
https://www.databricks.com/blog/introducing-dbrx-new-state-art-open-llm
Apache License 2.0
3.99k
stars
525
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
Remove type ignore
#1421
dakinggg
closed
2 months ago
0
Update pr-gpu.yaml
#1420
KevDevSha
closed
2 months ago
0
Test forked secrets
#1419
KevDevSha
closed
2 months ago
0
Update pr-gpu.yaml
#1418
KevDevSha
closed
2 months ago
0
Replace pydocstyle with Ruff
#1417
eitanturok
closed
2 months ago
0
test cpu
#1416
KevDevSha
closed
2 months ago
0
Read Package Version Better
#1415
eitanturok
closed
2 months ago
1
Additional registry entrypoint documentation
#1414
dakinggg
closed
2 months ago
0
Kevin/ghcr build
#1413
KevDevSha
closed
2 months ago
0
Make Pytest log in color in Github Action
#1412
eitanturok
closed
2 months ago
0
Bump streaming version to v0.8.0
#1411
mvpatel2000
closed
2 months ago
0
Log original config
#1410
josejg
closed
2 months ago
0
Add better error handling for FSDP activation checkpointing and use_orig_params
#1409
j316chuck
closed
2 months ago
1
Enable QuickGelu Function for CLIP models
#1408
gupta-abhay
closed
2 months ago
2
Set pretrained model name correctly, if provided, in HF Checkpointer
#1407
snarayan21
closed
2 months ago
0
Remove curriculum learning error when duration less than saved timestamp
#1406
b-chu
closed
2 months ago
0
Add spin_dataloaders flag
#1405
dakinggg
closed
2 months ago
0
Update torch requirement from <2.4,>=2.3.0 to >=2.3.0,<2.5
#1404
dependabot[bot]
closed
2 months ago
1
Update accelerate requirement from <0.33,>=0.25 to >=0.25,<0.34
#1403
dependabot[bot]
closed
2 months ago
3
Propagate `name_or_path` through HF Checkpointer
#1402
snarayan21
closed
2 months ago
1
Remove orig params default
#1401
dakinggg
closed
2 months ago
0
add it
#1400
dakinggg
closed
2 months ago
1
Enable passing epsilon when building norm layers
#1399
gupta-abhay
closed
2 months ago
0
Fix license link in readme
#1398
dakinggg
closed
2 months ago
0
Condition the meta initialization for hf_causal_lm on pretrain
#1397
irenedea
closed
2 months ago
0
Add pre register method for mlflow
#1396
dakinggg
closed
2 months ago
0
Dtensor oom
#1395
dakinggg
closed
2 months ago
0
Removing the extra LlamaRotaryEmbedding import
#1394
ShashankMosaicML
closed
2 months ago
0
Bump transformers to 4.43.2
#1393
dakinggg
closed
2 months ago
0
Revert "Allow for multiple workers when autopacking (#1375)"
#1392
dakinggg
closed
2 months ago
0
Support rope scaling
#1391
milocress
closed
2 months ago
0
Avoid race condition in convert text to mds script
#1390
dakinggg
closed
2 months ago
0
Revert "Use utils to get shared fs safe signal file name (#1381)"
#1389
dakinggg
closed
2 months ago
0
Bump transformers version to 4.43.1
#1388
dakinggg
closed
2 months ago
0
Refactor loss function for ComposerMPTCausalLM
#1387
irenedea
closed
2 months ago
0
[kushalkodnad/tokenizer-registry] Introduce new registry for tokenizers
#1386
kushalkodn-db
closed
2 months ago
0
Fix load and save planner config logic
#1385
irenedea
closed
2 months ago
0
Do dtype conversion in torch hook to save memory
#1384
irenedea
closed
2 months ago
0
Add transformation method to hf_causal_lm
#1383
irenedea
closed
2 months ago
0
Adds the convert_examples_ckpt from scripts to CLI
#1382
KuuCi
closed
2 months ago
2
Get a shared file system safe signal file name
#1381
dakinggg
closed
2 months ago
2
HF Checkpoint OOM fix with earlier dtype conversion
#1380
snarayan21
closed
2 months ago
1
Update huggingface-hub requirement from <0.24,>=0.19.0 to >=0.19.0,<0.25
#1379
dependabot[bot]
closed
2 months ago
5
Any example script to run multi-node training for slurm?
#1378
wavy-jung
opened
2 months ago
7
Allow flash attention up to 3
#1377
dakinggg
closed
2 months ago
0
Transform model
#1376
irenedea
closed
2 months ago
0
Allow for multiple workers when autopacking
#1375
b-chu
closed
2 months ago
1
Bumping flash attention version to 2.6.3 and adding option for softcap in attention and lm_head logits.
#1374
ShashankMosaicML
closed
1 week ago
0
Merge chronos_dataset_patch into kushalkodn-db/llm-foundry-fork
#1373
kushalkodn-db
closed
2 months ago
0
Allow for transforms on the model before MLFlow registration
#1372
snarayan21
closed
2 months ago
0
Previous
Next