• Joined on 2024-11-24
squall pushed to main at squall/torch_ext 2025-04-12 14:22:03 +08:00
5ac163f95c 加上一些点优化的脚本。
squall pushed to main at squall/torch_ext 2025-04-12 13:25:35 +08:00
baaa5dbc1c 也是可以的,总体用起来还真是方便,还有什么稀奇的用法呢,可以继续研究一下。
squall pushed to main at squall/torch_ext 2025-04-12 13:12:10 +08:00
9bc678f9a6 block的东西感觉差不多了。接着再实现一个多维的。
squall pushed to main at squall/torch_ext 2025-03-29 16:37:46 +08:00
4be98aed30 seems like merge to kernel together is faster.
squall pushed to main at squall/torch_ext 2025-03-29 11:57:00 +08:00
93b10bb894 简单修改一下。
squall pushed to main at squall/torch_ext 2025-03-28 23:29:51 +08:00
374cd36597 看起来尺寸大了以后效果可能会有差异。
squall pushed to main at squall/torch_ext 2025-03-28 23:21:07 +08:00
4774d3ef39 简单实现一个triton的矩阵乘法,感觉基本上就差不多了,可以快速用这个东西验证一些东西。
squall pushed to main at squall/torch_ext 2025-03-28 22:19:49 +08:00
89e3b9d190 本地修改一下。
Compare 2 commits »
squall pushed to main at squall/torch_ext 2025-03-27 03:49:52 +08:00
c77f9602ea test triton, seems like very well.
58093d7a71 试了一下写softmax,又学到一点。可以了
acdacc2592 测试一下。
Compare 4 commits »
squall created branch main in squall/picotron 2025-01-10 23:42:46 +08:00
squall pushed to main at squall/picotron 2025-01-10 23:42:46 +08:00
df3ae8a5f0 Update README.md
bf03420686 Update README.md
164ab81e27 Update README.md
78ba56ce80 Merge pull request #11 from eliebak/patch-1
009bb0b2a8 Update LICENSE
Compare 10 commits »
squall created repository squall/picotron 2025-01-10 23:42:14 +08:00
squall pushed to main at squall/torch_ext 2025-01-04 13:47:59 +08:00
920ebe0f88 简单修改一下。
squall pushed to main at squall/torch_ext 2024-12-29 15:50:04 +08:00
80d7be70a5 简单修改一下。
squall pushed to main at squall/torch_ext 2024-12-14 13:34:43 +08:00
0a6b5493fa 全都提交一下。
squall created branch main in squall/ollama 2024-11-30 13:01:26 +08:00
squall pushed to main at squall/ollama 2024-11-30 13:01:26 +08:00
5f8051180e Enable index tracking for tools - openai api support (#7888)
39e29ae5dd llama: fix typo and formatting in readme (#7876)
30a9f063c9 readme: add SpaceLlama, YouLama, and DualMind to community integrations (#7216)
ce7455a8e1 api: enable tool streaming (#7836)
e3936d4fb3 Support Multiple LoRa Adapters (#7667)
Compare 10 commits »
squall created repository squall/ollama 2024-11-30 13:00:57 +08:00
squall pushed to main at squall/vllm 2024-11-24 18:27:56 +08:00
c055747867 [model][utils] add extract_layer_index utility function (#10599)
eda2b3589c Revert "Print running script to enhance CI log readability" (#10601)
1c445dca51 [CI/Build] Print running script to enhance CI log readability (#10594)
1700c543a5 [Bugfix] Fix LoRA weight sharding (#10450)
17d8fc1806 [bugfix] Fix example/tensorize_vllm_model tests (#10595)
Compare 10 commits »
squall created branch main in squall/vllm 2024-11-24 18:27:51 +08:00