 Alexander Borzunov
|
593d980ad8
Use bitsandbytes 0.41.1 (#442)
|
2 vuotta sitten |
|
Alexander Borzunov
|
f3fafd14a4
Bump version to 2.0.1 (#411)
|
2 vuotta sitten |
|
Alexander Borzunov
|
eb0664b993
Support Python 3.11 (#393)
|
2 vuotta sitten |
|
Alexander Borzunov
|
e9a20e7e53
Require accelerate>=0.20.3 as transformers do (#383)
|
2 vuotta sitten |
|
Alexander Borzunov
|
895327a0ae
Fix readme code example, require Python < 3.11 until supported (#374)
|
2 vuotta sitten |
|
Alexander Borzunov
|
c735dd7ba3
Update transformers to 4.31.0 and peft to 0.4.0 (#371)
|
2 vuotta sitten |
|
Alexander Borzunov
|
f97582fb5f
Require transformers < 4.31.0 until we're compatible (#369)
|
2 vuotta sitten |
|
Alexander Borzunov
|
62d9ed5ce7
Implement shortest-path routing for inference (#362)
|
2 vuotta sitten |
|
Alexander Borzunov
|
3f733a96e3
Use bitsandbytes 0.40.1.post1 (#357)
|
2 vuotta sitten |
|
Alexander Borzunov
|
2c8959e713
Share more info about a server in DHT (#355)
|
2 vuotta sitten |
|
Alexander Borzunov
|
1a78638c02
Test that bitsandbytes is not imported when it's not used (#351)
|
2 vuotta sitten |
 Artem Chumachenko
|
b9f0a5467f
Support peft LoRA adapters (#335)
|
2 vuotta sitten |
|
Alexander Borzunov
|
dfc6578c8e
Use bitsandbytes 0.40.0.post4 with bias hotfix (#342)
|
2 vuotta sitten |
|
Alexander Borzunov
|
fa095f6461
Use 4-bit for llama by default, use bitsandbytes 0.40.0.post3 (#340)
|
2 vuotta sitten |
|
Alexander Borzunov
|
de930918a0
Support loading blocks in 4-bit (QLoRA NF4 format, disabled by default) (#333)
|
2 vuotta sitten |
|
Alexander Borzunov
|
66a47c763e
Require pydantic < 2.0 (2.0 is incompatible with hivemind 1.1.8) (#337)
|
2 vuotta sitten |
|
Alexander Borzunov
|
cb3f018f9f
Add LLaMA support (#323)
|
2 vuotta sitten |
|
Alexander Borzunov
|
0a313bf6c5
Update hivemind to 1.1.8, enable efficient bfloat16 encoding (#311)
|
2 vuotta sitten |
|
Alexander Borzunov
|
454c193863
Fix OOMs happening in case of accelerate >= 0.16.0 (#310)
|
2 vuotta sitten |
|
Alexander Borzunov
|
98be9ffe4c
Relax the rest of Hugging Face dependencies (#305)
|
2 vuotta sitten |
|
Alexander Borzunov
|
35662b4a16
Require bitsandbytes == 0.38.0.post2, hivemind == 1.1.7 (#302)
|
2 vuotta sitten |
|
Alexander Borzunov
|
2116df08bc
Fix deps, enable 8-bit by default for TP (#298)
|
2 vuotta sitten |
 justheuristic
|
987f4d2b2f
Update bitsandbytes, hivemind, transformers (#290)
|
2 vuotta sitten |
|
Alexander Borzunov
|
a7d3d02194
Fix invalid author email in setup.cfg (#287)
|
2 vuotta sitten |
|
Alexander Borzunov
|
6ba63c6cc8
Fix output shape when resuming generation (#211)
|
2 vuotta sitten |
|
Alexander Borzunov
|
6b12b0d050
Report server version and dht.client_mode in rpc_info(), check for updates on startup (#209)
|
2 vuotta sitten |
|
Alexander Borzunov
|
82c9f93ce6
Bump version to 1.1.0 (#190)
|
2 vuotta sitten |
 Egiazarian Vage
|
93bed7da5a
Support libp2p relays for NAT traversal (#186)
|
2 vuotta sitten |
|
Alexander Borzunov
|
0f6464103d
Remove protobuf from requirements (#182)
|
2 vuotta sitten |
|
Alexander Borzunov
|
55698381d0
Disable chunked_forward() on AVX512 CPUs (#179)
|
2 vuotta sitten |
