8972dab6f5
fix sft resume with compile mode
...
Load checkpoint state before torch.compile to avoid key mismatch on resume, and make optimizer/scaler restore tolerant to missing fields in older checkpoints.
Made-with: Cursor
2026-03-19 12:44:53 +00:00
349e74ec7b
[update] empty_think_ratio
2026-02-06 19:15:21 +08:00
288e1ac02a
[update] empty_think_ratio
2026-02-06 01:36:02 +08:00
ccc190da05
[feat] data process
2026-02-06 01:17:57 +08:00
11a44340ba
[update] save interval
2026-01-30 20:30:50 +08:00
04616c41a5
[update] safe half
2026-01-30 20:29:31 +08:00
fea69cf338
[fix] data skip
2026-01-18 16:56:29 +08:00
f7ffdf1fdb
[update] shuffle data
2026-01-18 16:39:34 +08:00
3a5aba82db
[fix] max length
2026-01-17 13:26:14 +08:00
714abcf802
[update] pretrain load
2026-01-17 12:00:17 +08:00
aa539a824a
[update] align mask
2026-01-15 11:20:41 +08:00
c090b69c4d
[update] align loss
2026-01-15 00:56:32 +08:00
e119db8478
[fix] compile unpack
2026-01-14 20:13:32 +08:00
81d24a4f16
[feat] add compile
2026-01-14 14:42:30 +08:00
1279a61681
[update] prompt prefill
2026-01-13 17:46:54 +08:00
05d0b216f6
[update] show speed
2026-01-07 23:33:47 +08:00
df89069362
[update] params log
2026-01-07 23:08:45 +08:00
f55d4c32a0
[update] mask log
2026-01-07 22:12:26 +08:00
20a43d7db0
[update] readme
2026-01-07 00:58:38 +08:00
7641985d14
[update] simplify loader
2026-01-06 01:20:52 +08:00
0b4a8ad4aa
[update] readme
2026-01-06 01:18:10 +08:00
07364c3fbe
[update] rename train tokenizer
2026-01-06 01:17:33 +08:00
9830915d87
[update] readme
2026-01-05 23:15:25 +08:00
4e73f34823
[update] rename reason
2026-01-05 23:12:29 +08:00
a8455ca8a3
[fix] messages num
2026-01-04 11:03:16 +08:00
42a4e8c86a
[fix] dist cleanup
2026-01-02 22:25:55 +08:00
9d898576ac
[update] aux loss
2026-01-01 22:41:46 +08:00
c65335b56f
[fix] experts unused
2025-12-31 21:47:04 +08:00
bc8fd82166
[fix] layers set 8
2025-12-31 21:06:37 +08:00
5dd4df7e18
[fix] moe unused
2025-12-31 21:00:06 +08:00
9236260a4a
[feat] get params
2025-12-31 20:46:59 +08:00
288a1d7212
[feat] get params
2025-12-31 20:44:34 +08:00
eead9538b2
[feat] update config
2025-12-31 10:29:13 +08:00
6242980917
[feat] update lr
2025-12-31 10:27:09 +08:00
936d105e9b
[feat] compatible tokenizer
2025-12-31 10:26:46 +08:00
4a5c9f5ece
[feat] stream load data
2025-12-28 16:58:52 +08:00
7eae14f3ce
[feat] remove empty_cache
2025-12-27 07:14:36 +08:00
11b962da06
[feat] explicit left padding
2025-12-23 18:59:48 +08:00
a9c56b20e9
[fix] lora weight
2025-12-22 21:27:29 +08:00
048d84abc7
Merge pull request #594 from whiteswordLI/fix/lora-load-ddp-weights
...
Fix: support loading DDP-saved LoRA weights for inference
2025-12-22 21:19:16 +08:00
3a18fdd666
Fix: support loading DDP-saved LoRA weights for inference
2025-12-22 20:50:25 +08:00
fe24501602
[feat] adjust seq length
2025-12-14 20:41:58 +08:00
fa82707c9c
[feat] update readme
2025-12-11 15:45:50 +08:00
5129f0e2a2
[fix] dtype & lr
2025-12-09 13:01:38 +08:00
aa7dc0f61e
Merge pull request #571 from dyhuachi/dyhuachi-patch-1
...
[fix] Refactor get_lr function to include min_lr calculation
2025-12-09 12:59:11 +08:00
bf3878ace8
[fix] Refactor get_lr function to include min_lr calculation
...
这里的退火算法会让参数里的lr的起始值变成原来lr的1.1倍，作出如下修改
2025-12-06 17:09:51 +08:00
ecd1ae1563
[fix] reduce aux_loss_alpha
2025-12-05 23:08:29 +08:00
5e1447b913
[fix] cuda memory #559
2025-12-01 16:17:43 +08:00
151fdf7e76
[feat] update yarn
2025-12-01 16:15:05 +08:00
6b86ea399a
[feat] release memory
2025-11-27 19:39:49 +08:00
王得利
jingyaogong
jingyaogong
whitesword
dyhuachi