AutoGPTQ

simcop2387/AutoGPTQ

Fork 0

Commit graph

ced04e1dff disable the error exit here, see if the pregen code works main Ryan Voots 2023-10-26 12:43:07 -04:00
07021b9a1c Generated files so that when they fail to work in pipeline then it still continues with what should be some ok defaults Ryan Voots 2023-10-26 10:26:42 -04:00
3011e13009 Built locally for temp setup, not sure what its doing but it is doing weird stuff on build server, like it never determines something Ryan Voots 2023-10-26 10:26:13 -04:00
153c085a32 Make this fail early when the actual problem happens Ryan Voots 2023-10-26 09:38:59 -04:00
9fb99f61e7 Merge remote-tracking branches 'laaza/Mistral' and 'laaza/MPT' Automation Pipeline 2023-10-22 07:53:59 -04:00
f32004c5d0

Merge df13da6a1c into e4b2493733 潘其威(William) 2023-10-21 13:56:11 -06:00
2d46a963a5

Merge 2d997d094f into e4b2493733 Brian Semrau 2023-10-21 13:56:11 -06:00
7d7c35cf14

Merge f3a5a79b7b into e4b2493733 CHU Tianxiang 2023-10-21 13:56:11 -06:00
842bf1045e

Merge 49bc9b4023 into e4b2493733 Alexander Pozharskiy 2023-10-21 13:56:11 -06:00
cbc4664a2c

Merge 4b7389ddb7 into e4b2493733 LaaZa 2023-10-21 13:56:11 -06:00
375ac2125f

Merge 4044265dea into e4b2493733 Giorgio Piatti 2023-10-21 23:11:24 +10:00
6ceabfc79f

Merge 22af50bab0 into e4b2493733 潘其威(William) 2023-10-20 15:38:12 -04:00
2d70224dfc

Merge 99acbead42 into e4b2493733 LaaZa 2023-10-20 15:31:54 -04:00
e4b2493733

Modify qlinear_cuda for tracing the GPTQ model (#367) Vivek Khandelwal 2023-10-20 21:36:01 +05:30
c6a037ddfe Modify qlinear_cuda for tracing the GPTQ model Vivek Khandelwal 2023-10-09 14:21:09 +00:00
22af50bab0 add new args of save_quantized method to push_to_hub method weights_sharding student686 2023-10-07 13:59:53 +08:00
fc1184e7bc save_quantized method support shard checkpoint student686 2023-10-07 13:48:45 +08:00
bf70350153 bump transformers version to 4.34.0 student686 2023-10-07 13:47:37 +08:00
4b7389ddb7 Merge branch 'main' into MPT LaaZa 2023-10-04 20:21:49 +03:00
99acbead42 Add support for Mistral models. LaaZa 2023-10-04 01:07:55 +03:00
1ae4f2edf8

Merge 678856a5ef into 51c043c6be Tom Jobbins 2023-09-30 04:17:01 +04:00
49bc9b4023 Merge branch 'main' of github.com:PanQiWei/AutoGPTQ into update-for-last-peft-initialization Alexander Pozharskii 2023-09-30 03:11:29 +04:00
5db00722b3 Working AdaLora Alexander Pozharskii 2023-09-30 03:03:39 +04:00
e052ac8d5a Initial code for GPTQLoraLinear initialization Alexander Pozharskii 2023-09-30 01:50:27 +04:00
51c043c6be

Merge pull request #355 from PanQiWei/fix_pack_model_use_exllamav2 潘其威(William) 2023-09-27 11:06:35 +08:00
c1a3013c45 import exllama QuantLinear instead of exllamav2's student686 2023-09-27 11:05:13 +08:00
3b81fb5ea0

Merge pull request #354 from PanQiWei/revert-325-main 潘其威(William) 2023-09-27 10:39:00 +08:00
3de7fbb0d5

Revert "fix bug(breaking change) remove (zeors -= 1)" revert-325-main 潘其威(William) 2023-09-27 10:37:31 +08:00
ac23d6b819

Merge pull request #325 from qwopqwop200/main 潘其威(William) 2023-09-26 14:20:39 +08:00
62fd0371ac

Merge branch 'main' into main 潘其威(William) 2023-09-26 14:09:04 +08:00
b461b6fa13

Merge pull request #335 from z80maniac/ignore-extra-args 潘其威(William) 2023-09-26 14:00:38 +08:00
04db761eed

Merge pull request #347 from alex4321/peft-model-use-adapter-name 潘其威(William) 2023-09-26 13:55:06 +08:00
50d2e86890

Merge pull request #349 from SunMarc/exllamav2_integration 潘其威(William) 2023-09-26 13:49:59 +08:00
c912bf361a exllamav2 integration Marc Sun 2023-09-25 16:51:18 +00:00
645bd15a96 update README student686 2023-09-25 18:55:34 +08:00
d2844437fd update README student686 2023-09-25 18:53:03 +08:00
da84da846b update README student686 2023-09-25 18:51:03 +08:00
50da063f65 update README student686 2023-09-25 18:47:40 +08:00
0185095402 Use adapter_name for get_gptq_peft_model with train_mode=True Alexander Pozharskii 2023-09-24 17:11:19 +04:00
18cc8c6466 fix max_input_len = max_input_len izenkov 2023-09-15 16:48:56 -04:00
06e071e68e

Merge pull request #326 from TheBloke/TB_Latest_Falcon 潘其威(William) 2023-09-14 22:49:25 +08:00
7a75176224 update README PanQiWei 2023-09-11 11:15:08 +08:00
121dbd15a5

Ignore unknown parameters in quantize_config.json ZXED 2023-09-10 18:39:40 +03:00
94de4ef185

GPTQ backward compatibility support qwopqwop200 2023-09-08 10:16:29 +09:00
9e0682a63e

Optimize q4_matmul qwopqwop200 2023-09-07 12:54:46 +09:00
034f6730ed Removed unexpected file that shouldn't have been added, sorry TheBloke 2023-09-06 18:08:30 +01:00
02a87dce76 Add support for Falcon as part of Transformers 4.33.0, including new Falcon 180B TheBloke 2023-09-06 18:03:33 +01:00
6b1ceb1897

if exllama auto diable fused attention qwopqwop200 2023-09-06 18:14:04 +09:00
ad5b0d72ee

fix bug qwopqwop200 2023-09-06 16:41:41 +09:00
f752336cda

fix bug qwopqwop200 2023-09-06 16:39:22 +09:00
1793227283

Merge pull request #311 from SunMarc/fix_max_input_length 潘其威(William) 2023-09-01 10:21:54 +08:00
782bb603d9

Merge pull request #303 from JustinLin610/patch-1 潘其威(William) 2023-09-01 10:20:24 +08:00
04b321da89

fix type Marc Sun 2023-08-31 14:07:16 -04:00
1e938e6bad

Merge pull request #310 from PanQiWei/fix_to()_metod_bug 潘其威(William) 2023-08-31 19:04:02 +08:00
1339db3045

Merge pull request #309 from PanQiWei/install-skip-qigen(windows) 潘其威(William) 2023-08-31 19:03:43 +08:00
c7021f0f44 fix model type changed after calling .to() method fix_to()_metod_bug PanQiWei 2023-08-31 18:39:03 +08:00
f97b77a64e

fix install bug qwopqwop200 2023-08-31 15:00:38 +09:00
45a1ee4d84

install check qigen qwopqwop200 2023-08-31 14:37:39 +09:00
71d56c76d0

skip install qigen(windows) qwopqwop200 2023-08-31 14:35:04 +09:00
f3a5a79b7b Fix g_idx in fused kernel 楚天翔 2023-08-30 19:20:18 +08:00
7c39a3a315

Update qwen.py for Qwen-VL Junyang Lin 2023-08-30 16:29:55 +08:00
604c96144f temporarily set the version of main branch to 0.5.0.dev0 PanQiWei 2023-08-25 17:36:23 +08:00
6bbf70373f

Merge pull request #288 from PanQiWei/revert-287-v0.4.2-release 潘其威(William) 2023-08-25 17:34:27 +08:00
e5050a5650

Revert "V0.4.2 release" 潘其威(William) 2023-08-25 17:26:55 +08:00
1049fd014a

Merge pull request #287 from PanQiWei/v0.4.2-release 潘其威(William) 2023-08-25 17:26:41 +08:00
6a9d80eddc Merge remote-tracking branch 'qwopqwop200/main' into main qwopqwop200 2023-08-25 18:06:03 +09:00
ae879760f9

Merge d95661b250 into 144302f58f 潘其威(William) 2023-08-25 03:22:53 -05:00
dafdd6189a

duplicate code remove qwopqwop200 2023-08-25 14:59:13 +09:00
144302f58f

Update install instructions (#286) fxmarty 2023-08-25 04:17:25 +09:00
b389dd264b update readme Félix Marty 2023-08-24 21:15:42 +02:00
ef442d9f70 Fix setuptools classifier (#285) v0.4.2 fxmarty 2023-08-25 02:33:28 +09:00
0365188c9c

Fix setuptools classifier (#285) fxmarty 2023-08-25 02:33:28 +09:00
825702fcf2 fix-classifier Félix Marty 2023-08-24 19:32:16 +02:00
8254da4f15 update version Félix Marty 2023-08-24 17:47:14 +02:00
10e6fda832

fix powershell (#284) fxmarty 2023-08-24 23:53:07 +09:00
9ea9151f6e fix powershell Félix Marty 2023-08-24 16:52:23 +02:00
cf942da9e2

remove ref main as we may want to trigger workflows on other branches (#282) fxmarty 2023-08-24 22:55:13 +09:00
c25638b5d9 remove ref main as we may want to trigger workflows on other branches Félix Marty 2023-08-24 15:53:13 +02:00
78082b1c5e update README PanQiWei 2023-08-24 21:16:04 +08:00
8bb4d60d8f

Merge pull request #281 from fxmarty/expose-api-exllama-input-length 潘其威(William) 2023-08-24 20:50:18 +08:00
04730ac66c expose api to set exllama max length Felix Marty 2023-08-24 11:22:15 +00:00
3cd79c826e

Fix python version for rocm build (#278) fxmarty 2023-08-23 23:01:22 +09:00
0601eb20c7 whats the diff? Félix Marty 2023-08-23 15:47:15 +02:00
e2b26a2c92 fix python version Félix Marty 2023-08-23 15:41:17 +02:00
766c6c1956

fix (#277) fxmarty 2023-08-23 21:50:18 +09:00
5646360706 fix Félix Marty 2023-08-23 14:49:38 +02:00
d53d227b7c

Update install instructions (#275) fxmarty 2023-08-23 21:29:55 +09:00
d0d1a69931

use conda incubator (#276) fxmarty 2023-08-23 21:18:46 +09:00
b3b2882c59 use conda incubator Félix Marty 2023-08-23 14:18:11 +02:00
3460de71c9 fix Félix Marty 2023-08-23 14:17:23 +02:00
4798705110 update doc Félix Marty 2023-08-23 13:43:25 +02:00
81801bc6e2

Use focal for RoCm build (#274) fxmarty 2023-08-23 20:41:08 +09:00
45b53f6e19 use focal for rocm build Félix Marty 2023-08-23 13:40:33 +02:00
f7b1b8291a

Free disk space for rocm build (#273) fxmarty 2023-08-23 19:21:44 +09:00
f6719ee284 free disk Félix Marty 2023-08-23 12:21:04 +02:00
48baeeb739

Merge pull request #272 from PanQiWei/build-wheels-on-2004 fxmarty 2023-08-23 18:48:25 +09:00
064f74c60f update ubuntu version Félix Marty 2023-08-23 11:46:19 +02:00
31941963a3 update readme Félix Marty 2023-08-23 11:45:17 +02:00
40945beb0e update README PanQiWei 2023-08-22 20:18:59 +08:00
4160db15e9 update README PanQiWei 2023-08-22 17:24:22 +08:00