Tuning LLMs with no tears💦; Sample Design Engineering (SDE) for more efficient downstream-tuning.
active 2023-08-15 → 2026-03-19 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 948 days from 2023-08-15 to 2026-03-19. Pushes: 15 total, peak 8 in a day. Pull requests: 0 total, peak 0 in a day. Issues: 11 total, peak 2 in a day. Comments: 14 total, peak 3 in a day. Stars: 403 total, peak 10 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| beyondguo | 9 | 8 | 0 | 1 |
| katheriena | 6 | 6 | 0 | 0 |
| karl-tao-zhang | 4 | 0 | 0 | 3 |
| hjing100 | 2 | 0 | 0 | 1 |
| LivinLuo1993 | 2 | 0 | 0 | 0 |
| jiahuanluo | 1 | 0 | 0 | 0 |
| ohj666 | 1 | 0 | 0 | 1 |
| nostalgiaer | 1 | 0 | 0 | 1 |
| XiaoWenyilin | 1 | 1 | 0 | 0 |
| TransformersWsz | 1 | 0 | 0 | 1 |
| 14H034160212 | 1 | 0 | 0 | 0 |
| zzy347964399 | 1 | 0 | 0 | 1 |
| kunden0612 | 1 | 0 | 0 | 1 |
| jimmy-walker | 1 | 0 | 0 | 0 |
| xhaoss | 1 | 0 | 0 | 0 |
| Capt-Rayn | 1 | 0 | 0 | 1 |
| kyle-hy | 1 | 0 | 0 | 0 |
| andyzhu | 1 | 0 | 0 | 0 |
| hehuomu | 1 | 0 | 0 | 0 |
| Mike-ihr | 1 | 0 | 0 | 1 |
Recent activity
Latest issues, pull requests and releases
- Issue comment#15Mike-ihr2024-09-13 09:22chatglm2报错：ValueError: weight is on the meta device, we need a `value` to put in on 0
- Issue comment#59hjing1002024-04-15 02:45ImportError: dlopen: cannot load any more object with static TLS
- Issue#59hjing1002024-04-15 01:55ImportError: dlopen: cannot load any more object with static TLS
- Issue#58jimmy-walker2023-12-20 03:44Why epoch in log is different from progress
- Issue comment#9mockyd2023-10-31 07:39有群吗,拉个群把
- Issue comment#14Capt-Rayn2023-10-27 03:13ChatGLM2按照readme教程微调了，但是没有效果！！！
- Issue comment#18TransformersWsz2023-10-25 05:00多卡训练感觉不是并发的?
- Issue#57kyle-hy2023-10-24 10:02使用ChatGLM2-6B分词报错
- Issue#56xhaoss2023-10-24 08:38code lamma微调脚本可以使用baichuan2的吗
- Issue comment#21nostalgiaer2023-10-11 10:12请教一个问题，chatglm2在用lora微调时，不添加attention mask也可以么？
- Issue#55LivinLuo19932023-10-11 09:54model.hf_device_map 不存在如何解决呢
- Issue comment#51nghuyong2023-10-10 08:38为什么ppo model 需要接AutoModelForCausalLMWithValueHead呢？
- Issue comment#49beyondguo2023-10-01 05:00chaglm-6b lora微调执行到指定的eval_step后提示“iteration over a 0-d tensor”
- Issue comment#14ohj6662023-09-21 07:35ChatGLM2按照readme教程微调了，但是没有效果！！！
- Issue comment#54karl-tao-zhang2023-09-09 23:06按照教程,一步一步弄的,到了训练PPO的时候, 卡到 CUDA error: device-side assert triggered
- Issue comment#54karl-tao-zhang2023-09-09 11:27按照教程,一步一步弄的,到了训练PPO的时候, 卡到 CUDA error: device-side assert triggered
- Issue comment#54karl-tao-zhang2023-09-09 11:16按照教程,一步一步弄的,到了训练PPO的时候, 卡到 CUDA error: device-side assert triggered
- Issue#54karl-tao-zhang2023-09-09 11:15按照教程,一步一步弄的,到了训练PPO的时候, 卡到 CUDA error: device-side assert triggered
- Issue comment#4zzy3479643992023-09-08 09:26加载checkpoint里面的adapter进行预测，生成的结果跟输入完全一样
- Issue#53andyzhu2023-09-07 14:25where is "my_templates" module
- Issue comment#38kunden06122023-09-05 08:10这个微调代码不能直接用来baichuan-13B模型的微调？在13B一直报错
- Issue#5214H0341602122023-08-31 01:47PPO training CUDA out of memory
- Issue#51jiahuanluo2023-08-29 09:10为什么ppo model 需要接AutoModelForCausalLMWithValueHead呢？
- Issue#50hehuomu2023-08-22 02:52chatglt-6b2 lora微调使用int4精度报错
- Issue#49LivinLuo19932023-08-18 10:18chaglm-6b lora微调执行到指定的eval_step后提示“iteration over a 0-d tensor”
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 403 stars here means stars gained during the window, not the repo's star count.