codemayq
fc1aa8f45c
update wechat
2024-11-14 10:02:06 +08:00
XYZliang
64414905a3
Increase shm_size to 16GB in docker-compose.yml to optimize shared memory allocation for large-scale model fine-tuning tasks.
...
This pull request increases the shm_size parameter in docker-compose.yml to 16GB. The goal is to enhance the LLaMA-Factory framework’s performance for large model fine-tuning tasks by providing sufficient shared memory for efficient data loading and parallel processing.
This PR also addresses the issues discussed in [this comment](https://github.com/hiyouga/LLaMA-Factory/issues/4316#issuecomment-2466270708 ) regarding Shared Memory Limit error.
2024-11-13 10:13:59 +08:00
hoshi-hiyouga
3eebae892b
Merge pull request #5990 from hiyouga/hiyouga/dev_vllm
...
[generate] fix vllm config args
2024-11-11 14:10:35 +08:00
hoshi-hiyouga
8d70edf39b
fix #5988
2024-11-11 13:57:14 +08:00
hoshi-hiyouga
2176224f4b
Merge pull request #5984 from hiyouga/hiyouga/wechat
...
[readme] update wechat
2024-11-10 22:08:55 +08:00
hiyouga
f2a44e1a2a
update wechat
2024-11-10 22:08:10 +08:00
hoshi-hiyouga
1ca6b1582f
Merge pull request #5982 from hiyouga/hiyouga/vllm_args
...
[args] add vllm config
2024-11-10 21:37:18 +08:00
hiyouga
58ab4579dc
add vllm config
2024-11-10 21:28:18 +08:00
hoshi-hiyouga
40a2fcc02d
Merge pull request #5973 from JJJJerry/fix_vllm_generate
...
fix VllmEngine: 将inputs参数替换为prompt
2024-11-10 21:04:38 +08:00
hoshi-hiyouga
a543bc478d
Update vllm_engine.py
2024-11-10 20:57:00 +08:00
JJJJerry
1d04078bb5
fix VllmEngine: 将inputs参数替换为prompt
2024-11-09 11:45:59 +08:00
hoshi-hiyouga
adc5849ce7
Merge pull request #5971 from hiyouga/hiyouga/fix_webui
...
[webui] fix extra args
2024-11-09 00:25:24 +08:00
hiyouga
eb294fed13
fix extra args
2024-11-09 00:24:27 +08:00
hoshi-hiyouga
39e330196d
Merge pull request #5970 from hiyouga/hiyouga/fix_beam
...
[generation] fix vllm v0.6.3
2024-11-08 23:58:15 +08:00
hiyouga
8f3a32286e
fix #5966
2024-11-08 23:49:16 +08:00
hoshi-hiyouga
707ff5a370
Merge pull request #5929 from hiyouga/hiyouga/dev_wechat
...
[readme] update wechat
2024-11-04 16:44:31 +08:00
hiyouga
882635cac3
update wechat
2024-11-04 08:43:21 +00:00
hoshi-hiyouga
bcbd1bebb0
Merge pull request #5927 from hiyouga/hiyouga/dev_fixmmchat
...
[fix] chat engines
2024-11-04 16:36:23 +08:00
hiyouga
ffa39ba3db
add image input type
2024-11-04 08:27:20 +00:00
hiyouga
8c88065c38
fix chat engines
2024-11-04 08:18:12 +00:00
hoshi-hiyouga
22ee197a40
Merge pull request #5926 from hiyouga/hiyouga/dev_deps
...
[version] update datasets version
2024-11-04 16:04:00 +08:00
hiyouga
c5fae465ec
update datasets version
2024-11-04 07:52:26 +00:00
hoshi-hiyouga
04c10d2e80
Merge pull request #5914 from hiyouga/hiyouga/dev_read
...
[misc] update readme
2024-11-02 21:44:10 +08:00
hoshi-hiyouga
9a42c5f76f
Merge pull request #5475 from menibrief/main
...
Fix phi-3-small issues
2024-11-02 21:31:34 +08:00
hiyouga
2e843d989e
fix phi3 template
2024-11-02 21:31:23 +08:00
hiyouga
e7ed5091e1
update readme
2024-11-02 21:28:04 +08:00
hoshi-hiyouga
478cbb1aa7
update template
2024-11-02 21:21:22 +08:00
hoshi-hiyouga
5f14910910
Merge branch 'main' into main
2024-11-02 21:20:27 +08:00
hoshi-hiyouga
344ff76d26
Merge pull request #5913 from hiyouga/hiyouga/dev_metrics
...
[train] support gather DPO metrics, fix return output
2024-11-02 21:13:43 +08:00
hiyouga
ba10d54bf9
fix webchat
2024-11-02 21:04:18 +08:00
hiyouga
4270f7dfb9
fix dpo metrics
2024-11-02 20:59:01 +08:00
hoshi-hiyouga
07e5088851
Merge pull request #5880 from sd3ntato/make-image-parametric
...
make base image parametric.
2024-11-02 20:26:14 +08:00
hoshi-hiyouga
5e8a619d1d
Update Dockerfile
2024-11-02 20:20:26 +08:00
hoshi-hiyouga
c58cc22d06
Merge pull request #5910 from Cuiyn/index
...
Support Index series models.
2024-11-02 20:16:54 +08:00
Cuiyn
ecca9db66b
fix: rename to Index-1.9B-Charater-Chat and Index-1.9B-Chat-32K
2024-11-02 20:04:14 +08:00
hoshi-hiyouga
83535bbe8b
Merge pull request #5912 from hiyouga/hiyouga/dev_logging
...
[misc] support rank0 logger
2024-11-02 18:48:41 +08:00
hiyouga
c38aa29336
support rank0 logger
2024-11-02 18:31:04 +08:00
Cuiyn
a15a69ab44
Add support for Index
2024-11-02 13:45:27 +08:00
hoshi-hiyouga
bd08b8c441
Merge pull request #5909 from hiyouga/hiyouga/dev2
...
[data] support auto convert for single image, add image_dir argument
2024-11-02 13:43:04 +08:00
hoshi-hiyouga
8408339d83
Merge pull request #5907 from hiyouga/hiyouga/dev
...
[data] fix template replace behavior
2024-11-02 13:42:53 +08:00
hiyouga
bfe1abd7af
fix #5904
2024-11-02 13:08:15 +08:00
hiyouga
24da9f59b0
fix #5883
2024-11-02 13:06:34 +08:00
hoshi-hiyouga
3ef7d590b1
Merge pull request #5906 from hiyouga/dev
...
[test] update tests
2024-11-02 12:50:43 +08:00
hiyouga
93d3b8f43f
update tests
2024-11-02 12:41:44 +08:00
hoshi-hiyouga
0a55e60693
Merge pull request #5895 from hiyouga/dev
...
[inference] support multiple images
2024-11-01 16:52:55 +08:00
hiyouga
e824b715ad
add examples
2024-11-01 08:41:54 +00:00
hiyouga
e80a481927
support multiimage inference
2024-11-01 07:25:20 +00:00
Valerio Mariani
d5b2def7c0
make base image parametric.
...
default `BASE_IMAGE` is nvcr.io/nvidia/pytorch:24.02-py3 for retro-compatibility
2024-10-30 21:53:32 +01:00
hoshi-hiyouga
641d0dab08
Merge pull request #5873 from hiyouga/dev
...
[misc] update readme
2024-10-30 17:14:44 +08:00
hiyouga
2417b70a62
update readme
2024-10-30 09:14:01 +00:00
