125 Commits

Author SHA1 Message Date
hiyouga
22b3c913e9 fix #1325
Former-commit-id: 083787dbfe41f58ff59cb16ddde02df98593aef5
2023-11-01 23:38:49 +08:00
hiyouga
e3d64e119e update gradio, support multiple resp in api
Former-commit-id: 84af10cec912309b82fc16e20879eec1c2e3729a
2023-11-01 23:02:16 +08:00
hiyouga
abd6ec174f fix SFT trainer
Former-commit-id: d8cf8cfdeb861152771d22f873d8ed17b41538ce
2023-10-31 21:52:52 +08:00
hiyouga
6493f6d2e9 fix #1316
Former-commit-id: f4e4a04529a60b4c4ccc66cbf67f6e951fbc68d3
2023-10-31 11:32:08 +08:00
hiyouga
4906092811 fix #1287
Former-commit-id: 838ed9aa8737e8f74ec7e8c660254cabbbbcf6ef
2023-10-26 17:49:41 +08:00
hiyouga
4d6606a72c fix #1285
Former-commit-id: aff9363ce399fd9f1b29d5a088b49c865001f37a
2023-10-26 16:34:52 +08:00
hiyouga
daeff710eb update neftune logic
Former-commit-id: 7de7174ce3ea804b2ba58560193cda25cbd675ef
2023-10-22 17:42:13 +08:00
hiyouga
59fd3155a1 add new options in webui
Former-commit-id: f793ca0a2c9e7abab360658d0e65506c1f97e1ae
2023-10-22 17:17:58 +08:00
hiyouga
ce112dc744 fix recursion error
Former-commit-id: b79ca8781e368b17c83e585f1644346628d7e2de
2023-10-22 16:28:37 +08:00
hiyouga
d6c77d9196 reimplement neftune
Former-commit-id: 7b4acf7265b04cc4a674b3dcafdb90e76f149e39
2023-10-22 16:15:08 +08:00
anvie
3635823fbe add NEFTune optimization
Former-commit-id: 57fb40aa04fec11ca165a97ea463579faeaeebe7
2023-10-21 13:24:10 +07:00
hiyouga
4930118761 fix #1218
Former-commit-id: 7a11a42dfd414d140cd83b7a74760715d2ae2078
2023-10-19 16:17:41 +08:00
hiyouga
f6a8069159 fix #1217
Former-commit-id: 6496a99b7d9a5aa57bfc8b935b274f1481580864
2023-10-19 15:52:24 +08:00
hiyouga
eafde5b73f fix config, #1191
Former-commit-id: a6a04be2e6700050c64e59c00a86698f1098e7cc
2023-10-15 18:28:45 +08:00
hiyouga
f3fa47fa7d refactor export, fix #1190
Former-commit-id: ea82f8a82a7356bbdf204190d596d0b1c8ef1a84
2023-10-15 16:01:48 +08:00
hiyouga
089785c71b fix loading dtype
Former-commit-id: c874e764b8334c18091233c3781009a39d67e794
2023-10-14 20:15:24 +08:00
hiyouga
1faba4b704 fix #1176 #1177
Former-commit-id: 01d8cb1ca76e1d599fa5e970b2fcaa3ceedbfffd
2023-10-14 20:00:17 +08:00
hiyouga
e585c789ce fix #1184
Former-commit-id: af18b0dce7a4ef10b30da069d454010eddd269af
2023-10-14 19:20:11 +08:00
hiyouga
2562376f84 fix ppo args
Former-commit-id: 11bd271364488d523d5117ec2ea26f39853175b7
2023-10-11 23:40:50 +08:00
hiyouga
c9d1cd108d refactor model_dtype, fix PPO trainer
Former-commit-id: 2818af0b0967d7695f27658acac0b7e2c2728e5d
2023-10-11 23:16:01 +08:00
hiyouga
b6e81a0307 fix shift short attention
Former-commit-id: ab65c3063b31b9e6a1aeb62c57224c1296ccdadd
2023-10-09 17:07:46 +08:00
hiyouga
8de990b1e9 update webui #1086
Former-commit-id: b8dbec086eef655870f2922697e76733de427f5a
2023-10-09 14:50:14 +08:00
hiyouga
3a7bcb585b fix #1097
Former-commit-id: a683c5b797625e9802e1c77b8d924ae6c0b4b214
2023-10-08 22:29:26 +08:00
hiyouga
69c99819d6 add llamafy_qwen.py
Former-commit-id: f9769cff8aec5cbb33d4c2d837ed47434281cf4e
2023-10-08 22:05:36 +08:00
hiyouga
deb17942ab fix layer norm dtype
Former-commit-id: 84b7486885c600e5e65c5ba9095d56ecc2502977
2023-09-28 00:25:55 +08:00
hiyouga
f66e6b91c7 fix #1026
Former-commit-id: b0b0138e1d841bf5bd2e08b2eea78057b33be69a
2023-09-27 22:57:09 +08:00
hiyouga
ca6a3bc76f fix #424
Former-commit-id: 35fa94723c7d0c024a63f47208291d09c67fc13e
2023-09-27 22:49:43 +08:00
hiyouga
927ff702ff refactor finetuning Args
Former-commit-id: 620efe1d8d2429f4bc3fa8009900ec43e1b5ef4b
2023-09-27 22:28:06 +08:00
hiyouga
108c31e1fc support LongLoRA
Former-commit-id: 90375f600d5601866836123597fa3ef52008eeef
2023-09-27 21:55:50 +08:00
hiyouga
5ee1bdecdc add MMLU and C-Eval script
Former-commit-id: 465ee8119aa489a41bee0b01b3c105a2f3dd137f
2023-09-23 00:34:17 +08:00
hiyouga
e930682152 fix #1000
Former-commit-id: 5cc7a447843c578af602a5e054348fad1c9306ce
2023-09-22 15:00:48 +08:00
hiyouga
3bdc40b4db tiny fix
Former-commit-id: dbaef776a171dfb316567553788f2d81d35a51fb
2023-09-21 19:52:06 +08:00
hiyouga
4581d09fa6 fix #944
Former-commit-id: 338b8664edea5ae65192ac657bb013581245ae15
2023-09-21 19:51:02 +08:00
hiyouga
d04585df59 tiny fix
Former-commit-id: ace3f85a7273fbbc531adfe6ad73bf76a5fff52d
2023-09-21 15:25:29 +08:00
statelesshz
95384eefe8 support export model on Ascend NPU
Former-commit-id: b3e41c6d49728e239b61d8cd9b3603c8dc877549
2023-09-20 10:26:02 +08:00
hiyouga
4e6e42bb55 add tests.cal_flops.py
Former-commit-id: 469f859161dec0e34f4cc849f20e43d442680b5c
2023-09-16 23:40:41 +08:00
hiyouga
2a75b10015 fix #913
Former-commit-id: 0b5f970c05c524670d66c810e9f081a52a1fb5e6
2023-09-15 20:58:28 +08:00
hiyouga
632fff02e0 fix #887
Former-commit-id: 8857e4560219c4052bdb7c7dc1a014a5f5fd0163
2023-09-14 17:56:58 +08:00
mmbwf
6ded1725d9 Update utils.py
Fix parameters load error.

Former-commit-id: 30fb721f1222d1b56a2712519960a63655c20360
2023-09-14 15:38:04 +08:00
hiyouga
c8780205bc fix ppo save model
Former-commit-id: 7ba57d5b1469cd0de0bb391b915bedec97b20ebd
2023-09-12 16:25:29 +08:00
hiyouga
4e86462bad fix #762 #814
Former-commit-id: d4be857e23c74ed65e06903e19da6f18f15d9e30
2023-09-12 16:10:10 +08:00
hiyouga
33bab0e7c1 update flashattn, fix ppo save model
Former-commit-id: 0fbece85a70222e5262a2295203de07ffe648fda
2023-09-11 17:25:36 +08:00
hiyouga
6a71361a54 remove PeftTrainer
Former-commit-id: b218c271edfb07006ddc34b1aca404088de6c528
2023-09-10 22:23:23 +08:00
hiyouga
8ab5566dc0 support FlashAttention2
Former-commit-id: d8aa1404bee9842f3e4cd037ad8d66c85470ac37
2023-09-10 20:43:56 +08:00
hiyouga
85aa16f6c6 fix #850
Former-commit-id: 815b92e698562bfae6eb9a6fa1b612a05d43ed67
2023-09-10 14:22:03 +08:00
hiyouga
f865d0bd51 fix lora target
Former-commit-id: a51b7c98acc599de5ed2eaeeebe7b184105722c5
2023-09-09 17:04:45 +08:00
hiyouga
c818a7ff60 support lora target auto find
Former-commit-id: bca1a247bcef51dced59655c8a14c197569367ca
2023-09-09 15:38:37 +08:00
hiyouga
c6265e6969 fix chatglm2 tokenizer
Former-commit-id: d8d82ca281811c20c89cc03dd00f69735515d6cf
2023-09-09 13:50:29 +08:00
hiyouga
43a20c67d4 fix bug in DPO data collator
Former-commit-id: 90bd085ae4c2775f1e82e045ab0157a451774082
2023-09-08 20:45:07 +08:00
hiyouga
405df0f63d fix #761
Former-commit-id: b34797a845dec8f6daea59e3c353b8a8f8830100
2023-09-08 20:22:18 +08:00