Commit Graph

261 Commits

Author SHA1 Message Date
justheuristic
a6fca51212 fetch a specific bloom block without downloading the entire model 2022-06-20 15:11:22 +03:00
justheuristic
6047a2ffe0 push config and tokenizer separately 2022-06-20 14:28:31 +03:00
justheuristic
2d55e6e4fe instructions to test distributed inference 2022-06-19 22:25:57 +03:00
justheuristic
9be7c81b78 instructions to test distributed inference 2022-06-19 22:22:01 +03:00
justheuristic
cc9a76625d warn about long runtime 2022-06-19 22:14:52 +03:00
justheuristic
82214699f2 notes on hosting servers 2022-06-19 19:34:18 +03:00
justheuristic
b6f3bbfd97 black 2022-06-19 19:18:46 +03:00
justheuristic
84de19fb1a better status logs 2022-06-19 19:17:44 +03:00
justheuristic
1555d98f66 push converted model to hub 2022-06-19 19:13:48 +03:00
justheuristic
736f1d1085 push converted model to hub 2022-06-19 19:06:35 +03:00
justheuristic
15d0ea7129 fix black 2022-06-19 17:24:11 +03:00
justheuristic
e8241d2915 black everything 2022-06-19 17:23:08 +03:00
justheuristic
3b9351de1c isort 2022-06-19 17:22:57 +03:00
justheuristic
ed468af8d6 leave a todo for attention mask 2022-06-19 17:22:29 +03:00
justheuristic
33358bc52b rpc_inference works 2022-06-19 17:20:22 +03:00
justheuristic
47c308306a TODOs were moved to https://github.com/learning-at-home/bloom-demo/issues/3#issuecomment-1159686069 2022-06-19 16:56:21 +03:00
justheuristic
a00ec56ade basic multi-step inference session 2022-06-19 16:55:51 +03:00
justheuristic
c4d508c00e remove some unnecessary debugprints 2022-06-19 14:14:11 +03:00
justheuristic
a44cb84f06 basic cache checks (via debugprint) 2022-06-19 14:13:45 +03:00
justheuristic
fee63bd440 rpc_inference works! 2022-06-19 13:51:55 +03:00
justheuristic
8092bd31ff swap to int64 (rationale: pytorch does not support uint64) 2022-06-19 13:37:13 +03:00
justheuristic
62d7fde8af pre-check type 2022-06-19 12:58:31 +03:00
justheuristic
7fba411dff extended run_serverexample 2022-06-17 11:40:49 +03:00
justheuristic
20497f81d1 switch to hivemind-master 2022-06-17 11:34:50 +03:00
justheuristic
5a15c13ca7 switch to hivemind-master 2022-06-17 10:36:34 +03:00
Pavel Samygin
57f4e0a899 add identity path 2022-06-17 10:15:15 +03:00
justheuristic
35310698f0 newer hivemind version 2022-06-17 10:12:09 +03:00
justheuristic
8959727dea add minimalistic benchmarks 2022-06-14 15:18:38 +03:00
justheuristic
a798ea04a6 add minimalistic benchmarks 2022-06-14 15:18:11 +03:00
justheuristic
e3a7d5af30 inference mode 2022-06-14 14:51:06 +03:00
justheuristic
3b16d6ffdb black-isort 2022-06-14 13:33:25 +03:00
justheuristic
a28ea0aa6f
Merge pull request #9 from learning-at-home/rpc
Rudimentary decentralization
2022-06-14 09:46:42 +03:00
justheuristic
14e316b52a black-isort 2022-06-14 09:46:14 +03:00
justheuristic
eea9287182 RemoteTransformerBlock 2022-06-14 09:45:22 +03:00
justheuristic
3e9fd63a02 RemoteTransformerBlock 2022-06-14 09:45:06 +03:00
justheuristic
1cca611c9f undo rename 2022-06-14 09:23:32 +03:00
justheuristic
7ce7cd7a97 basic backend 2022-06-14 08:49:42 +03:00
justheuristic
1c49bcb741 basic backend 2022-06-14 08:26:05 +03:00
justheuristic
3215945882 use logger 2022-06-14 08:25:06 +03:00
justheuristic
ce5dedd2c7 rename 2022-06-14 05:09:13 +03:00
justheuristic
8f5d022d18 save hidden size 2022-06-14 05:09:07 +03:00
justheuristic
e2e9d0e94c memory cache for attention KVs 2022-06-13 00:47:55 +03:00
justheuristic
e5e8c9ed12 expel all bloom-specific files to src.bloom 2022-06-12 21:44:36 +03:00
justheuristic
324ea2dc96 save non-transformer params separately 2022-06-12 09:35:58 +03:00
justheuristic
54cf292374 backend schema 2022-06-12 08:40:33 +03:00
justheuristic
3ccd0b5e2d account for layer_past in alibi 2022-06-12 07:21:17 +03:00
justheuristic
fb3bfbb78f optimization TODOs for later 2022-06-12 06:54:12 +03:00
justheuristic
902bf6400a [temp workaround] create alibi 2022-06-12 06:50:19 +03:00
justheuristic
05faa0b3c8 add quantization script for cpu 2022-06-12 05:59:11 +03:00
justheuristic
ffb56a65ed keep hidden_size as property 2022-06-12 05:11:28 +03:00