hiyouga
621d755a53
fix jetmoe z3 block
...
Former-commit-id: cb00a14d905395c4b8fadb955f0424a4c56668de
2024-05-18 22:28:45 +08:00
hiyouga
b5a49ad569
improve data process logger
...
Former-commit-id: 33d0b012b56dbafc9fff87b821c2d1bf1409dbb5
2024-05-18 22:02:42 +08:00
hiyouga
5a40ff2de4
fix #3803
...
Former-commit-id: 1ef12c95059d14a1717c82ce04e529e7ad6435ed
2024-05-18 16:13:14 +08:00
hiyouga
d24969bb7e
improve KTO impl., replace datasets
...
Former-commit-id: e56a57ddcf061de6e4acc8679f7dbf0b68364986
2024-05-18 03:44:56 +08:00
hoshi-hiyouga
65715cab9d
Merge pull request #3785 from enji-zhou/feature/add_kto
...
add kto
Former-commit-id: f60faa23e23022fd855dac6b1ecbd21e095bccb5
2024-05-18 03:07:18 +08:00
hoshi-hiyouga
db2ccb2d36
Update model_args.py
...
Former-commit-id: f40a2fe5334865763e4d513292d359317b7a091b
2024-05-17 16:16:41 +08:00
juejuezi
e47e835e4e
feat: pass the max_lora_rank parameter to vLLM backend
...
Former-commit-id: a8756d839405ecb5deabe885cf11d1a61564deee
2024-05-17 16:07:39 +08:00
hiyouga
7728cb8fdb
add deepseek v2 lite model
...
Former-commit-id: 5e864e6b721d8b891b1cc2ca2dcac41babb9eaaf
2024-05-17 13:25:36 +08:00
enji.zhou
d16a1d9ed0
add kto
...
Former-commit-id: ec51986cf70b0bdd79b8141e45916670fb97a08e
2024-05-17 13:09:17 +08:00
hiyouga
18b2f23f4f
better dtype handle in loading
...
Former-commit-id: 663f0577dd61a1a31191db2c6fbb0c7cea533b21
2024-05-17 02:14:56 +08:00
hiyouga
31d4698693
enable inbrowser in webui
...
Former-commit-id: 71fdeedb64b2339eb1c740d670b87e0c03dada68
2024-05-17 00:08:56 +08:00
hiyouga
040ae800bf
add falcon 11b
...
Former-commit-id: 897acc725edc204fad393cc9616828431b4fa768
2024-05-17 00:08:33 +08:00
hiyouga
ee759aa0d8
rename package
...
Former-commit-id: a07ff0c083558cfe6f474d13027642d3052fee08
2024-05-16 18:39:08 +08:00
hiyouga
4eb73ec2c4
set dev version
...
Former-commit-id: 5e9c72d07c3793cdccbdb8a9f95f1bb5d714e0a3
2024-05-16 02:17:31 +08:00
hiyouga
9c2da898a4
release v0.7.1
...
Former-commit-id: a4f8adb021b6218d624303b51cd5e93ffa3111a1
2024-05-16 00:57:16 +08:00
hiyouga
9316996bdc
fix #3694
...
Former-commit-id: 3d1b818cb6a77b7603724fbeb756b468aa74e7ea
2024-05-16 00:35:28 +08:00
hiyouga
1191af789b
fix #3606
...
https://github.com/huggingface/peft/pull/1706
Former-commit-id: bf2783e1b6bc207375974c48736d6f82dd293f02
2024-05-15 23:05:02 +08:00
hiyouga
e0d8761269
add Yi-VL-34B model
...
Former-commit-id: 8b3d8a7e3bd8dff27cc72edba1b8a042f6d1929c
2024-05-15 22:58:19 +08:00
hiyouga
532a10e906
add yi-vl 6b model
...
Former-commit-id: 35f4041b13a593a6cf1ec6686fa18b38911ad6a4
2024-05-15 20:02:41 +08:00
hiyouga
ed40fe9ecc
fix yi vl vllm infer
...
Former-commit-id: de54e5d7ec06dd7c20ec82c9ff032fc16cd50244
2024-05-15 19:25:48 +08:00
hoshi-hiyouga
30882edfd4
Merge pull request #3748 from BUAADreamer/main
...
Add MLLM YI-VL and save processor config during training
Former-commit-id: 1d3cbd24ccea63d36c27725cdc5ecd02b460b0ed
2024-05-15 16:40:54 +08:00
hoshi-hiyouga
7df034cea5
Update visual.py
...
Former-commit-id: f5f13a995c64fc374ad05e26cde8efa6651aefa1
2024-05-15 16:39:57 +08:00
hiyouga
05c6adfe91
fix fsdp model loading
...
Former-commit-id: fc6fe23cc9ae4a920a17e8268a85c1aa4ad16d3b
2024-05-15 16:32:28 +08:00
hoshi-hiyouga
8f0e52267a
Update patcher.py
...
Former-commit-id: 4c31a21f2106adcdad100119bad83ecaef0be3f3
2024-05-15 15:37:07 +08:00
hoshi-hiyouga
4708a8a019
Update template.py
...
Former-commit-id: a13022166ba691c03f4fea7e9e2927fa446cf681
2024-05-15 14:20:39 +08:00
hoshi-hiyouga
e044d2765a
Update trainer.py
...
Former-commit-id: dd767b20635bb549ce14f9556e1c4fb44b3662c5
2024-05-15 14:13:26 +08:00
hoshi-hiyouga
842b74ca9e
Update workflow.py
...
Former-commit-id: 97cfb44bced18b721166ccb5f260098645fc5318
2024-05-15 14:13:01 +08:00
BUAADreamer
1b62c9023f
rm extra import
...
Former-commit-id: 031215019e3d7727b1c7cc87a44e1cf1eb2853ec
2024-05-15 12:48:18 +08:00
BUAADreamer
38b4c36791
cast dtype in mm_proj
...
Former-commit-id: e0ab22648fe8b65055b5986258cc2800438dc60c
2024-05-15 11:22:15 +08:00
BUAADreamer
2aab3221fb
modify style
...
Former-commit-id: 823af88c3201412da7ef734d34198424e09b2d51
2024-05-15 10:18:10 +08:00
BUAADreamer
a0f76b6065
Merge branch 'main' of https://github.com/BUAADreamer/LLaMA-Factory
...
Former-commit-id: ce5cb0f897eebe32a1c2c0a78fe1b0267e4b6d9d
2024-05-15 09:54:21 +08:00
BUAADreamer
28c3c1abd7
Merge branch 'hiyouga:main' into main
...
Former-commit-id: a4795c2f5328e0cfc657409f5774819e3defc006
2024-05-15 09:54:14 +08:00
BUAADreamer
6ab86e0987
add yivl and save processor to model_dir
...
Former-commit-id: ae72f745cb4f7713c3b835d11202aec19c3c5093
2024-05-15 09:54:00 +08:00
hiyouga
ed6793afdf
fix bug in vllm engine
...
Former-commit-id: 38f02a2c5b52cba6908c2d3c2a455677f8574faf
2024-05-15 02:17:54 +08:00
hiyouga
200689e07c
fix gen args
...
Former-commit-id: d79f91f87106ba1bc3c0ea08da5898aad59566a7
2024-05-15 01:49:05 +08:00
hiyouga
a72c7c3721
add npu examples
...
Former-commit-id: 0f21e68e2dbd84c820d66d5c6d980004efc51d51
2024-05-14 23:32:53 +08:00
hoshi-hiyouga
51abf0d314
Merge pull request #3584 from zhou-wjjw/main
...
Enhancing Ascend 910A Training Efficiency in LlamaFactory with NPU
Former-commit-id: 310cf017a5ec24af8f5cf3af298760dd4150f9f2
2024-05-14 22:18:37 +08:00
hiyouga
b06ec57e48
use robust envs
...
Former-commit-id: f3e194c3b3c40a3e6c3c5397ec0d859e6db614b5
2024-05-14 21:36:42 +08:00
hoshi-hiyouga
fef95e861c
Update train.py
...
Former-commit-id: da1e6f0d9c2eff64f92da1f6ada3aa44ef6d6a7e
2024-05-14 20:47:52 +08:00
hoshi-hiyouga
bf6e5677c7
Apply suggestions from code review
...
Co-authored-by: Huazhong Ji <hzji210@gmail.com>
Former-commit-id: abef48c17ee795eae984fcc89019c2c4859108c1
2024-05-14 20:44:21 +08:00
hoshi-hiyouga
9219034c5d
Apply suggestions from code review
...
Co-authored-by: Huazhong Ji <hzji210@gmail.com>
Former-commit-id: a435e5a0bdd7268c4f1204f99f289ee0b36fd930
2024-05-14 20:44:04 +08:00
hiyouga
2144ea3f76
fix #3728
...
Former-commit-id: ea3e32a27f7f7dce75a708f8a6f376b5d3e8059a
2024-05-14 20:37:21 +08:00
BUAADreamer
d24d932a54
modify yi-vl template
...
Former-commit-id: f113975b425e70bed2588ca55a2c62594fbf2283
2024-05-14 16:45:28 +08:00
BUAADreamer
06a930f92c
add support for Yi-VL
...
Former-commit-id: d7834ca92d3048949caa48f8635cfbcea2c85771
2024-05-14 14:03:19 +08:00
BUAADreamer
5bad22703f
Merge branch 'main' of https://github.com/BUAADreamer/LLaMA-Factory
...
Former-commit-id: e82f527ea583a7e99a25a06c7fe7b03c1dc2ebb9
2024-05-13 23:28:52 +08:00
BUAADreamer
0a7b49f30a
add yi-vl
...
Former-commit-id: 891b25cb3d709ea82182ca90496034360e1cd5d8
2024-05-13 23:28:28 +08:00
hiyouga
a890eeb6ef
fix #3724
...
Former-commit-id: 62f5999d79834d6cbc4129eda387a317665d6099
2024-05-13 20:09:09 +08:00
hiyouga
a16d3e7879
fix #3702
...
Former-commit-id: 55755786f21050b9efc127c391509ba5d9ea8982
2024-05-13 18:24:35 +08:00
hoshi-hiyouga
edd5e14c08
Merge pull request #3655 from Tendo33/main
...
1.Change the name of is_fastapi_available function 2. Added the log of printing requests when deploying using vllm
Former-commit-id: 28c75448eed9d472e96285737a66ac0d20280e13
2024-05-13 18:05:50 +08:00
hiyouga
d7e3b61c24
support Yi 1.5
...
Former-commit-id: e580823676cbb83ddb9a0f685992e6054ae5ffaa
2024-05-13 16:51:20 +08:00