Daniel Han-Chen
|
9ecb3bc869
|
Update fast_lora.py
|
2024-01-26 17:04:33 +11:00 |
|
Daniel Han-Chen
|
7bed11a8ba
|
Update fast_lora.py
|
2024-01-26 16:43:50 +11:00 |
|
Daniel Han-Chen
|
317bbc57fe
|
Update fast_lora.py
|
2024-01-26 16:16:32 +11:00 |
|
Daniel Han-Chen
|
c4ac728d4c
|
Update fast_lora.py
|
2024-01-26 16:08:10 +11:00 |
|
Daniel Han-Chen
|
ab44945336
|
Update fast_lora.py
|
2024-01-26 14:06:25 +11:00 |
|
Daniel Han-Chen
|
84772de40f
|
Update fast_lora.py
|
2024-01-26 13:59:57 +11:00 |
|
Daniel Han-Chen
|
653f1036ae
|
Update fast_lora.py
|
2024-01-26 03:49:48 +11:00 |
|
Daniel Han-Chen
|
46ec8bbc3d
|
Repatch
|
2024-01-26 03:15:23 +11:00 |
|
Daniel Han-Chen
|
99eeebf72a
|
Update swiglu.py
|
2024-01-26 02:41:15 +11:00 |
|
Daniel Han-Chen
|
ae156d9154
|
Update llama.py
|
2024-01-26 02:07:35 +11:00 |
|
Daniel Han-Chen
|
29945bd8c1
|
Update llama.py
|
2024-01-26 01:33:23 +11:00 |
|
Daniel Han-Chen
|
194093e375
|
remove patching
|
2024-01-25 23:35:56 +11:00 |
|
Daniel Han-Chen
|
4f60055f9e
|
Update fast_lora.py
|
2024-01-25 23:25:17 +11:00 |
|
Daniel Han-Chen
|
2d25facc7b
|
Update fast_lora.py
|
2024-01-25 23:14:37 +11:00 |
|
Daniel Han-Chen
|
380f2fd6e4
|
Fix saving and bnb-4bit
|
2024-01-25 19:05:43 +11:00 |
|
Daniel Han-Chen
|
0474451a6a
|
Update pyproject.toml
|
2024-01-25 03:55:26 +11:00 |
|
Daniel Han-Chen
|
5a5d34b8ae
|
Update mapper.py
|
2024-01-25 03:47:23 +11:00 |
|
Daniel Han-Chen
|
bbf5ef65af
|
Graceful FA2 error + torch 2.1.1
|
2024-01-25 03:46:23 +11:00 |
|
Daniel Han-Chen
|
5ea888c578
|
Update to transformers 4.37
|
2024-01-25 02:29:37 +11:00 |
|
Daniel Han-Chen
|
d2f1521a49
|
incorrect inference
|
2024-01-23 23:23:05 +11:00 |
|
Daniel Han-Chen
|
f186fe9bc9
|
Update mistral.py
|
2024-01-23 19:44:45 +11:00 |
|
Daniel Han-Chen
|
e184b06c4b
|
Update mistral.py
|
2024-01-23 19:28:17 +11:00 |
|
Daniel Han-Chen
|
b314500993
|
q_len issue
|
2024-01-23 19:23:58 +11:00 |
|
Daniel Han-Chen
|
d6e85d7731
|
q_len == 1
|
2024-01-23 19:17:42 +11:00 |
|
Daniel Han-Chen
|
5d9e68181e
|
hidden_states
|
2024-01-23 19:11:17 +11:00 |
|
Daniel Han-Chen
|
f311a8efcc
|
Update llama.py
|
2024-01-23 18:04:20 +11:00 |
|
Daniel Han-Chen
|
eeee6333ec
|
Merge branch 'main' into nightly
|
2024-01-23 18:04:10 +11:00 |
|
Daniel Han
|
a833f40346
|
2-4x faster native HF inference (#119)
* 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
|
2024-01-23 03:55:24 +11:00 |
|
Daniel Han-Chen
|
20d8f223d7
|
Fast LoRA saving
|
2024-01-23 03:47:28 +11:00 |
|
Daniel Han-Chen
|
1bb1c3c2b9
|
LoRA
|
2024-01-23 03:36:50 +11:00 |
|
Daniel Han-Chen
|
24c7a67556
|
Update llama.py
|
2024-01-23 02:44:52 +11:00 |
|
Daniel Han-Chen
|
d87ef86991
|
Update llama.py
|
2024-01-23 02:22:37 +11:00 |
|
Daniel Han-Chen
|
a77c448939
|
Update llama.py
|
2024-01-23 02:18:37 +11:00 |
|
Daniel Han-Chen
|
5260ec2a0a
|
Update llama.py
|
2024-01-23 02:13:01 +11:00 |
|
Daniel Han-Chen
|
716e03fe1b
|
Update llama.py
|
2024-01-23 02:09:35 +11:00 |
|
Daniel Han-Chen
|
00f50876f7
|
Update llama.py
|
2024-01-23 02:03:25 +11:00 |
|
Daniel Han-Chen
|
e5b5333137
|
Update llama.py
|
2024-01-23 02:00:10 +11:00 |
|
Daniel Han-Chen
|
9a5062e6c4
|
Update llama.py
|
2024-01-23 01:44:10 +11:00 |
|
Daniel Han-Chen
|
1289ae825b
|
RoPE
|
2024-01-23 01:41:20 +11:00 |
|
Daniel Han-Chen
|
7e4140ebfc
|
Update llama.py
|
2024-01-23 01:37:36 +11:00 |
|
Daniel Han-Chen
|
1ba28d8e26
|
Update llama.py
|
2024-01-23 01:33:51 +11:00 |
|
Daniel Han-Chen
|
3e1b244d5e
|
Fast inference RoPE
|
2024-01-23 01:29:03 +11:00 |
|
Daniel Han-Chen
|
8647f0e86e
|
Update llama.py
|
2024-01-23 01:06:34 +11:00 |
|
Daniel Han-Chen
|
085a8e944a
|
inference
|
2024-01-23 00:32:57 +11:00 |
|
Daniel Han-Chen
|
f41a437540
|
Update utils.py
|
2024-01-22 22:34:33 +11:00 |
|
Daniel Han-Chen
|
31f0c9c08d
|
Update utils.py
|
2024-01-22 22:31:38 +11:00 |
|
Daniel Han-Chen
|
4220f6a6dc
|
No print
|
2024-01-22 22:26:55 +11:00 |
|
Daniel Han-Chen
|
e1cbc9e423
|
Update utils.py
|
2024-01-22 22:21:29 +11:00 |
|
Daniel Han-Chen
|
050c61bc0b
|
Update utils.py
|
2024-01-22 20:04:27 +11:00 |
|
Daniel Han-Chen
|
7c3b647ba7
|
fast_linear_forward
|
2024-01-22 20:01:59 +11:00 |
|