Cherry pick 1216 r0.2.0 - #1318

Closed

adityavavreNVDA wants to merge 63 commits into

mainfrom

cherry-pick-1216-r0.2.0

Commits on Oct 22, 2025

chore(beep boop 🤖): Bump `uv.lock` (main) (2025-10-16) (#969) (#1024)

authored
27b7c46
chore(beep boop 🤖): Bump `uv.lock` (main) (2025-10-21) (#1018) (#1032)

chtruong814
and
ko3n1g
authored
44d021b
Enable overriding TokenizerConfig when instantiating from checkpoint (#1030) (#1034)

chtruong814
and
maanug-nv
authored
469bbd2

Commits on Oct 23, 2025

Qwen3 finetune recipes (#975) (#1042)

chtruong814
and
ananthsub
authored
317052a
remove lru_cache decorator from finetune dataset building (#1012) (#1021)

chtruong814
and
ananthsub
authored
be1e5bb
update mamba recipe imports (#1038) (#1049)

chtruong814
and
ananthsub
authored
ca491a4
Make nemotronh recipe imports consistent with codebase (#1006) (#1019)

chtruong814
and
ananthsub
authored
5b66497
support loading checkpoints from a given step (#1040) (#1059)

chtruong814
and
ananthsub
authored
580fca5
Add gpt-oss recipes (#898) (#1055)

authored
258dc0d

Commits on Oct 24, 2025

ci: Track M-LM's release branch (#1066)
ko3n1g
authored
171acd5

Commits on Oct 27, 2025

cp: `build: Dependency bump (1037)` into `r0.2.0` (#1094)

chtruong814
and
ko3n1g
authored
ccf9924
cp: `docs: Update contrib guide (1022)` into `r0.2.0` (#1102)

chtruong814
and
ko3n1g
authored
edda462
cp: `[docs] Add supported models page (980)` into `r0.2.0` (#1104)

chtruong814
and
ananthsub
authored
b0a635e
cp: `link perf summary from readme (1041)` into `r0.2.0` (#1108)

chtruong814
and
ananthsub
authored
875bc4b

Commits on Oct 28, 2025

cp: `Use TE quantization modes for first/last bf16 layers (715)` into `r0.2.0` (#1109)

chtruong814
and
maanug-nv
authored
d589b41
cp: `Clean up fusion checks (938)` into `r0.2.0` (#1117)

chtruong814
and
ananthsub
authored
d0eb9fc
cp: `reduce peak memory for qwen3 moe recipes (1054)` into `r0.2.0` (#1112)

chtruong814
and
ananthsub
authored
b686c9f
cp: `Add settings/validation for deterministic mode (968)` into `r0.2.0` (#1107)

chtruong814
and
ananthsub
authored
0015c0f
build: Bump MCore (#1097)
ko3n1g
authored
dcc81b1
cp: `Fix load-hf-weights order (1096)` into `r0.2.0` (#1103)

chtruong814
and
ananthsub
authored
1df31c8
cp: `Remove sequence length validation check for custom dataset providers (1004)` into `r0.2.0` (#1118)

chtruong814
and
ananthsub
authored
a68e0a0
cp: `fix throughput logging when resuming (1110)` into `r0.2.0` (#1124)

chtruong814
and
ananthsub
authored
c31e1b8
cp: `fix Nemotron h provider and bridge (1053)` into `r0.2.0` (#1125)

chtruong814
and
ananthsub
authored
007917a
cp: `Documentation for Nemotron Nano V2 VL (1121)` into `r0.2.0` (#1122)

chtruong814
and
cuichenx
authored
1ab74b5

Commits on Oct 30, 2025

cp: `add __init__.py for gemma3 vl recipes (1139)` into `r0.2.0` (#1142)

chtruong814
and
ananthsub
authored
32a94eb
cp: `[Model Support] Qwen3-Next (1068)` into `r0.2.0` (#1126)

authored
0628f3b
cp: `Bug fix for Warmup Iters (1135)` into `r0.2.0` (#1137)

chtruong814
and
rhmukundan
authored
00073e5

Commits on Oct 31, 2025

cp: `Add moonlight 16B recipe (1133)` into `r0.2.0` (#1158)

chtruong814
and
suiyoubi
authored
a96b61c

Commits on Nov 4, 2025

Commits on Nov 5, 2025

Commits on Nov 6, 2025

cp: `Add fallback logic for slurm for initializing process group (999)` into `r0.2.0` (#1206)

chtruong814
and
ananthsub
authored
2e14f15

Commits on Nov 7, 2025

cp: `dataclass cfgs - perf (971)` into `r0.2.0` (#1220)

ananthsub
and
malay-nagda
authored
fbf96a3

Commits on Nov 8, 2025

cp: `save cfg to file (1190)` into `r0.2.0` (#1227)

authored
f989517
cp: `align perf naming w/ base recipe (1238)` into `r0.2.0` (#1240)

chtruong814
and
malay-nagda
authored
9bf4f79
cp: `change FLOP to Model FLOP (1008)` into `r0.2.0` (#1245)

chtruong814
and
erhoo82
authored
65ba9a3
cp: `add unit test coverage for dumping config to file before training (1229)` into `r0.2.0` (#1244)

chtruong814
and
ananthsub
authored
67b16e8

Commits on Nov 10, 2025

Commits on Nov 11, 2025

Commits on Nov 12, 2025