Commit Graph

20 Commits

Author SHA1 Message Date
d7c340a937 Updated notebooks wip 2025-04-23 10:32:33 +00:00
Niels Geens
bf94a0838a Enable 2x faster inference with unsloth 2025-04-22 11:47:27 +02:00
3e471eced3 Use Unsloth for QwQ inference 2025-04-21 10:42:35 +00:00
c1cef08941 Speed up inference 2025-04-19 14:12:00 +00:00
33aaaf4455 Fix start of assistant message detection 2025-04-18 17:46:49 +00:00
5e47200ece Fixes for QwQ llm engine 2025-04-18 12:30:47 +00:00
Niels Geens
c09f0766c1 Fixed auto approver and inference continuation 2025-04-18 11:36:17 +02:00
Niels Geens
34fb5d814f Shared response buffer backend 2025-04-17 18:32:23 +02:00
7e254b7f7e Fixed qwq token limit 2025-04-11 18:03:17 +02:00
Niels Geens
5b8f04be81 fix copy bug 2025-04-09 16:43:27 +02:00
Niels Geens
8fb30c8ed0 Optimized hot-loop for logits processor 2025-04-09 11:59:26 +02:00
2e08be4b75 Validator works for start 2025-04-07 14:31:56 +00:00
Niels Geens
50539f6d03 Integrate logits processor with qwq (untested) 2025-04-07 13:35:20 +02:00
Niels Geens
faf78b7948 Lock vllm version to avoid bug 2025-04-07 12:17:56 +02:00
0e2cca2e4f Fixed QwQ tokenizer 2025-04-04 08:54:06 +00:00
Niels Geens
e573b9ebe7 Removed flash-attn as dependency because it's hard to install properly 2025-04-02 14:42:35 +02:00
Niels Geens
6f3d414d17 Converted QwQ notebooks to .py files 2025-04-01 10:42:54 +02:00
Niels Geens
2e66020f8e Replaced deepseek with qwq 2025-03-14 11:22:30 +01:00
Niels Geens
3ea3239a9b wip deepseek train 2025-03-03 16:57:04 +01:00
b64f8d7d33 wip deepseek r1 2025-03-02 22:01:24 +01:00