Commit graph

263 commits

Author SHA1 Message Date
Daniel Han-Chen
c74e1af85c Update fast_lora.py 2024-01-27 20:37:43 +11:00
Daniel Han-Chen
363ffba1c2 Update fast_lora.py 2024-01-27 20:17:08 +11:00
Daniel Han-Chen
510c85f412 Update swiglu.py 2024-01-27 20:04:55 +11:00
Daniel Han-Chen
35daafdd6e Update fast_lora.py 2024-01-27 19:53:57 +11:00
Daniel Han-Chen
6201f7681f Update fast_lora.py 2024-01-27 19:38:28 +11:00
Daniel Han-Chen
86a1c9788b Update fast_lora.py 2024-01-27 19:29:53 +11:00
Daniel Han-Chen
8e0e4ccdc6 Update fast_lora.py 2024-01-27 19:18:59 +11:00
Daniel Han-Chen
2e4c59deaf Update swiglu.py 2024-01-27 19:05:02 +11:00
Daniel Han-Chen
85e87d9ba2 Swiglu 2024-01-27 19:03:04 +11:00
Daniel Han-Chen
83b6937285 Update fast_lora.py 2024-01-27 18:21:29 +11:00
Daniel Han-Chen
3d3e7f5b42 Update fast_lora.py 2024-01-27 18:15:34 +11:00
Daniel Han-Chen
d3f3b6fc49 Update fast_lora.py 2024-01-27 18:00:41 +11:00
Daniel Han-Chen
e77d7c069f Update fast_lora.py 2024-01-27 18:00:22 +11:00
Daniel Han-Chen
f7d11d10f8 Update fast_lora.py 2024-01-27 17:47:54 +11:00
Daniel Han-Chen
8ed03f5f45 Update pyproject.toml 2024-01-27 17:38:30 +11:00
Daniel Han-Chen
af65cb0d3d Works? 2024-01-27 04:50:47 +11:00
Daniel Han-Chen
fb5333726a Update llama.py 2024-01-27 04:49:36 +11:00
Daniel Han-Chen
a59ec7903c Merge branch 'main' into nightly 2024-01-27 04:48:40 +11:00
Daniel Han-Chen
704e36a64e Revert "Update llama.py"
This reverts commit a208ec46e0.
2024-01-27 04:48:34 +11:00
Daniel Han-Chen
a208ec46e0 Update llama.py 2024-01-27 04:48:03 +11:00
Daniel Han
87a7ef1049
More bug fixes (#133)
* faster saving & inference

* Update llama.py

* Update save.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update mistral.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* fast inference

* Update llama.py

* Update save.py

* Update llama.py

* Mistral correct RoPE scaling

* Max sequence lengths

* Apache 2

* fast_linear_forward

* Update utils.py

* Update utils.py

* No print

* Update utils.py

* Update utils.py

* inference

* Update llama.py

* Fast inference RoPE

* Update llama.py

* Update llama.py

* RoPE

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* Update llama.py

* LoRA

* Fast LoRA saving

* Update llama.py

* hidden_states

* q_len == 1

* q_len issue

* Update mistral.py

* Update mistral.py

* incorrect inference

* Update to transformers 4.37

* Graceful FA2 error + torch 2.1.1

* Update mapper.py

* Update pyproject.toml

* Fix saving and bnb-4bit

* Update fast_lora.py

* Update fast_lora.py

* remove patching

* Update llama.py

* Update llama.py

* Update swiglu.py

* Repatch

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update llama.py

* Update fast_lora.py

* Update llama.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update swiglu.py

* Update fast_lora.py

* Update swiglu.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update fast_lora.py

* Update save.py

* Update fast_lora.py

* Update utils.py

* Update llama.py

* Update fast_lora.py

* Update swiglu.py

* Update save.py

* Update save.py

* Update llama.py

* Update llama.py

* Update llama.py
2024-01-27 04:47:54 +11:00
Daniel Han-Chen
bd2ff90817 Update llama.py 2024-01-27 04:47:42 +11:00
Daniel Han-Chen
a3d892a15b Update llama.py 2024-01-27 04:47:01 +11:00
Daniel Han-Chen
b89599a3f7 Update llama.py 2024-01-27 04:46:47 +11:00
Daniel Han-Chen
baeea64917 Update save.py 2024-01-27 04:20:58 +11:00
Daniel Han-Chen
ecdbb28dcb Update save.py 2024-01-27 04:19:59 +11:00
Daniel Han-Chen
393341f211 Update swiglu.py 2024-01-27 04:18:25 +11:00
Daniel Han-Chen
47babc780a Update fast_lora.py 2024-01-27 04:16:38 +11:00
Daniel Han-Chen
9edc309f59 Update llama.py 2024-01-27 04:15:37 +11:00
Daniel Han-Chen
a27ac6175e Update utils.py 2024-01-27 04:03:47 +11:00
Daniel Han-Chen
7fb64e0e31 Update fast_lora.py 2024-01-27 03:58:21 +11:00
Daniel Han-Chen
e3bd0bb74f Update save.py 2024-01-27 03:51:53 +11:00
Daniel Han-Chen
ba847f5411 Update fast_lora.py 2024-01-27 03:38:33 +11:00
Daniel Han-Chen
c57495df6f Update fast_lora.py 2024-01-27 03:38:19 +11:00
Daniel Han-Chen
d379bb8cf5 Update fast_lora.py 2024-01-27 03:36:27 +11:00
Daniel Han-Chen
421ed33b42 Update fast_lora.py 2024-01-27 03:23:50 +11:00
Daniel Han-Chen
83ceb11cc9 Update fast_lora.py 2024-01-27 03:12:53 +11:00
Daniel Han-Chen
f3da0d2e4a Update fast_lora.py 2024-01-27 02:52:54 +11:00
Daniel Han-Chen
4ae3ad3bce Update fast_lora.py 2024-01-27 02:34:21 +11:00
Daniel Han-Chen
c74d3dd3b9 Update fast_lora.py 2024-01-27 02:22:15 +11:00
Daniel Han-Chen
0a1aa98d3c Update fast_lora.py 2024-01-27 02:19:00 +11:00
Daniel Han-Chen
f7760719f0 Update fast_lora.py 2024-01-27 01:28:03 +11:00
Daniel Han-Chen
38bff800b6 Update fast_lora.py 2024-01-27 01:16:31 +11:00
Daniel Han-Chen
485d54fd70 Update fast_lora.py 2024-01-26 23:48:17 +11:00
Daniel Han-Chen
ce08a15cee Update fast_lora.py 2024-01-26 23:33:17 +11:00
Daniel Han-Chen
b4492eb29c Update fast_lora.py 2024-01-26 23:14:44 +11:00
Daniel Han-Chen
3865c8cd68 Update fast_lora.py 2024-01-26 23:06:06 +11:00
Daniel Han-Chen
e9e34f1d4d Update fast_lora.py 2024-01-26 22:29:38 +11:00
Daniel Han-Chen
d96f665510 Update swiglu.py 2024-01-26 22:21:54 +11:00
Daniel Han-Chen
f8aa20d2db Update fast_lora.py 2024-01-26 22:09:32 +11:00