1713 Commits

Author SHA1 Message Date
Jake Poznanski
b99d149a43 Merge branch 'jakep/new_data' into jakep/vllm_beam 2025-10-07 18:18:28 +00:00
Jake Poznanski
727b345715 Merge fix 2025-10-07 18:16:31 +00:00
Jake Poznanski
8ef68fde88 Merge branch 'main' into jakep/new_data 2025-10-07 17:44:54 +00:00
Jake Poznanski
e15615aadb Model defaults 2025-10-07 17:10:45 +00:00
Jake Poznanski
44f6b9f0de Trying beam search idea 2025-10-07 04:00:39 +00:00
Jake Poznanski
b81e40602d Readme score fixes 2025-10-06 22:59:00 +00:00
Jake Poznanski
2e3d1a0317 Comitting test script to be used in model cards for individual one-off inference 2025-10-06 22:47:06 +00:00
Jake Poznanski
c89787183a Bump version to v0.3.8 for release v0.3.8 2025-10-06 21:46:18 +00:00
Jake Poznanski
e12941a608 Version bump 2025-10-06 21:46:10 +00:00
Jake Poznanski
7fe756fe63 Formatting 2025-10-06 21:10:32 +00:00
Jake Poznanski
9c7c670f1f Bump version to v0.3.7 for release v0.3.7 2025-10-06 21:10:07 +00:00
Jake Poznanski
1951a849ec Version bump with new vllm 2025-10-06 21:10:00 +00:00
Jake Poznanski
c75f5b98a1 Cleaning up pr 341 arguments to match with vllm 0.11, which only has V1 engine and thus always does chunked prefill. And fixes arg syntax 2025-10-06 20:26:41 +00:00
Jake Poznanski
e202c22822 Merge branch 'vllm_0_11' 2025-10-06 20:24:26 +00:00
Jake Poznanski
2b70b50312
Merge pull request #341 from charitarthchugh/charitarthchugh/vllm-defaults-speedup
Add chunked prefill and limit mm per prompt options
2025-10-06 13:23:47 -07:00
Jake Poznanski
81be6f5c1f Transformers version 2025-10-06 19:52:55 +00:00
Jake Poznanski
9b517a02be Git lfs in docker image 2025-10-06 19:47:19 +00:00
Jake Poznanski
9feb41af82 New docker file approach for vllm 0.11 2025-10-06 18:57:16 +00:00
Jake Poznanski
59266ed419 More readmes 2025-10-01 22:27:37 +00:00
Jake Poznanski
476ba212dc Bolds 2025-10-01 22:05:58 +00:00
Jake Poznanski
bb7790a138 Bolds 2025-10-01 22:05:30 +00:00
Jake Poznanski
4e68b174bf New bench scores added 2025-10-01 22:04:40 +00:00
Jake Poznanski
8ef7f8085a isort and black 2025-09-30 17:37:10 +00:00
Jake Poznanski
b5b1de98dd Allowing more max tokens in pipeline for new models 2025-09-29 22:12:27 +00:00
Jake Poznanski
f4356de091 deepinfra readme improved 2025-09-29 17:56:03 +00:00
Jake Poznanski
8982bae756 Bump version to v0.3.6 for release v0.3.6 2025-09-29 17:37:25 +00:00
Jake Poznanski
fb1ef9e38a Release script fix 2025-09-29 17:37:14 +00:00
Jake Poznanski
c587eb9050 Ugh, release script adds all files by default 2025-09-29 17:36:41 +00:00
Jake Poznanski
a0bc5a4690 Deepinfra readme 2025-09-29 17:29:28 +00:00
Jake Poznanski
0c6d889863 Adding retry code on 429 errors from exteranl providers 2025-09-29 17:26:22 +00:00
Jake Poznanski
9c750903ce Ignore files 2025-09-29 17:06:14 +00:00
Jake Poznanski
f0caa188ab
Merge pull request #344 from allenai/amanr/deepinfra
DeepInfra Support
2025-09-29 10:03:29 -07:00
Jake Poznanski
7f4b728dcd Skip docker checks if using beaker image 2025-09-26 23:27:04 +00:00
aman-17
f3c4073395 added Api_key argument to pipeline pytests 2025-09-26 14:25:25 -07:00
aman-17
359abef654 updated pytests 2025-09-26 14:19:22 -07:00
Jake Poznanski
1099820bf6 Hmm, going to have to train a filtered version 2025-09-26 21:15:15 +00:00
aman-17
556ff26d58 fixed lint, style, ruff 2025-09-26 14:08:40 -07:00
aman-17
e7ae5e6240 fixed style 2025-09-26 13:58:34 -07:00
aman-17
90589e16de Added deepinfra usage to readme 2025-09-26 13:56:34 -07:00
aman-17
2a5792e5ed add if else for vllm local usage bug for api argument 2025-09-26 13:29:48 -07:00
Jake Poznanski
bb06829840 SOuping in fp32 2025-09-26 20:03:29 +00:00
aman-17
7fe3f65de7 added support for deepinfra 2025-09-26 11:06:51 -07:00
Charitarth Chugh
fe425fde20
Add chunked prefill and limit mm per prompt options 2025-09-25 14:29:49 -04:00
Jake Poznanski
3d6e6a6a01 Setting defaults, adding in seed 2025-09-24 17:37:12 +00:00
Jake Poznanski
1654c760d9 Fixing grpo eos reward 2025-09-23 20:27:07 +00:00
Jake Poznanski
b4b121b118 Testing reward eos 2025-09-23 20:10:11 +00:00
Jake Poznanski
01bc1ff7b6 Allow passing in beaker image to run benchmark 2025-09-23 19:48:42 +00:00
Jake Poznanski
a00d9d172e Adding stricter math and table tests when in synthetic mode 2025-09-23 18:37:50 +00:00
Jake Poznanski
1197c35808 Mix contamination checker script 2025-09-23 18:17:13 +00:00
Jake Poznanski
9818797fbc One last try to disable timeouts in katex thing 2025-09-23 03:26:45 +00:00