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
3.83k
stars
502
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
MPT training with ALiBi and Flash Attention 2
#1289
rickgit16
closed
1 day ago
4
Add TE to setup
#1288
j316chuck
closed
2 weeks ago
0
Add missing dependency group
#1287
dakinggg
closed
2 weeks ago
1
Fix backwards compatibility for ICL arg
#1286
dakinggg
closed
2 weeks ago
0
Bump mlflow to 2.13.2
#1285
KuuCi
closed
2 weeks ago
1
Fix typo in CI
#1284
dakinggg
closed
2 weeks ago
0
Add optional logging of text output to EvalOutputLogging
#1283
sjawhar
closed
2 days ago
8
Update README.md to use variables
#1282
milocress
closed
2 weeks ago
0
Adds CI for torch 2.3.1
#1281
dakinggg
closed
2 weeks ago
1
Fix TE HF checkpoint saving
#1280
j316chuck
closed
2 weeks ago
1
Add loggers by default if env vars are populated
#1279
aspfohl
opened
2 weeks ago
0
Make expandable segments on by default
#1278
b-chu
closed
2 weeks ago
0
Fix packing + streaming + resumption
#1277
dakinggg
closed
2 weeks ago
0
Allow multiprocessing when preparing ICL dataset
#1276
sanjari-orb
opened
3 weeks ago
8
Add torch 2.3.1 docker images
#1275
dakinggg
closed
2 weeks ago
0
Update Dockerfile
#1274
j316chuck
closed
2 weeks ago
0
Update Dockerfile with TE main
#1273
j316chuck
closed
2 weeks ago
2
Managing Timeout on Training Errors and Simultaneous Restart of All Nodes in LLM Foundry
#1272
germanjke
closed
3 weeks ago
1
Why is there a warmup in hf_generate.py?
#1271
palash04
closed
3 weeks ago
1
fix linting
#1270
milocress
closed
3 weeks ago
0
Bump Composer to version 0.23.2
#1269
dakinggg
closed
3 weeks ago
0
Revert "Bump Composer to 0.23.0 (#1259)"
#1268
dakinggg
closed
3 weeks ago
0
Revert to older TE version
#1267
mvpatel2000
closed
3 weeks ago
1
Revert "Update TE Dockerfile (#1265)"
#1266
j316chuck
closed
3 weeks ago
0
Update TE Dockerfile
#1265
j316chuck
closed
3 weeks ago
0
Fill in the middle
#1264
germanjke
opened
3 weeks ago
1
Fix typo in setup.py
#1263
XiaohanZhangCMU
closed
3 weeks ago
0
Testing CI
#1262
dakinggg
closed
3 weeks ago
0
How to continue pretrain LLM fp8 with hf_causal_lm
#1261
YixinSong-e
opened
3 weeks ago
1
added systemMetricsMonitor callback
#1260
JackZ-db
closed
2 weeks ago
1
Bump Composer to 0.23.0
#1259
KuuCi
closed
3 weeks ago
1
Remove spurious warning
#1258
dakinggg
closed
4 weeks ago
0
Fix MPT HF conversion
#1257
dakinggg
closed
4 weeks ago
0
Add curriculum learning callback
#1256
b-chu
closed
1 week ago
3
Bump Version to 0.10.0.dev0
#1255
KuuCi
closed
3 weeks ago
2
Adding more token encoding types
#1254
snarayan21
closed
3 weeks ago
1
fix signal_file_path to avoid race condition
#1253
ofivite
closed
3 weeks ago
11
Add registry for ICL datasets
#1252
sanjari-orb
closed
2 weeks ago
8
Change TE docker image to enable te_shard_weight
#1251
j316chuck
closed
4 weeks ago
0
Replacing icl_task_type question_answering with generation_task_with_answers in long context eval yamls.
#1250
ShashankMosaicML
closed
4 weeks ago
0
Testing CI
#1249
dakinggg
closed
1 month ago
0
Update CODEOWNERS
#1248
dakinggg
closed
1 month ago
0
Add eval_drop_last flag to fix TE eval bug
#1247
j316chuck
opened
1 month ago
2
Wrap `FileNotFound` exceptions in the finetuning dataloader and `convert_text_to_mds`
#1246
angel-ruiz7
opened
1 month ago
1
[MCLOUD-4623] Add more detailed exception when user has uppercase in their example case but could potentially match the exampe type
#1245
shitaoli-db
closed
3 weeks ago
2
Fix the error message thrown from dataloader
#1244
shitaoli-db
opened
1 month ago
0
Add logging to convert_text_to_mds.py script
#1243
irenedea
closed
1 month ago
0
could you give an elaborated steps about how to run llm-foundry on AMD mi250 devices
#1242
Alice1069
opened
1 month ago
1
Make HF conversion automatically add missing imports
#1241
dakinggg
closed
1 month ago
0
Chunk file reads and tokenization for text to mds conversion
#1240
irenedea
closed
1 month ago
1
Previous
Next