 AlpinDale
|
c3b15f0926
do not allow context shift for enc-dec
|
10 months ago |
|
AlpinDale
|
f9726a3649
hell
|
10 months ago |
|
AlpinDale
|
72659e5cad
separate prompt and genned tokens for enc-dec
|
10 months ago |
|
AlpinDale
|
3ed4cc431c
enc_dec attention code
|
10 months ago |
|
AlpinDale
|
58e89e29d9
add custom bias to attention.py
|
10 months ago |
|
AlpinDale
|
a788ca33bf
hack in custom bias for attention kernels
|
10 months ago |
|
AlpinDale
|
f009f94ffd
update modeling code
|
10 months ago |
|
AlpinDale
|
b6e5080546
Merge branch 'main' into feat/t5-support
|
10 months ago |
 sgsdxzy
|
6ebac34dc1
chore: cleaner pre-llamafied Yi implementation (#352)
|
10 months ago |
 AlpinDale
|
681e94611f
fix: restore backwards compatibility with old Yi models (#351)
|
10 months ago |
|
AlpinDale
|
1b6732fcde
chore: bump transformers version
|
10 months ago |
 Absurd
|
070c1cef8c
fix: explicit RFC3986 for prometheus_client asgi (#344)
|
10 months ago |
 Stefan Daniel Schwarz
|
5d747cfc4d
readme: docker docs (#340)
|
10 months ago |
|
Stefan Daniel Schwarz
|
8e259ee7cf
chore: hf_transfer for faster downloads (#339)
|
10 months ago |
|
AlpinDale
|
398a97338a
feat: enable lora loading/unloading via API (#337)
|
10 months ago |
|
Stefan Daniel Schwarz
|
b0688b6b9c
fix: docker port and kobold api (#338)
|
10 months ago |
|
AlpinDale
|
ed225f59cb
fix: transformers in requirements
|
10 months ago |
|
AlpinDale
|
e120404436
Revert "feat: CMake Build System Generator (#332)"
|
10 months ago |
|
AlpinDale
|
06312251a7
fix: explictly export CUDA arches for CI
|
10 months ago |
|
AlpinDale
|
e53842bd5d
fix: cuda home detection for fp8 kv cache
|
10 months ago |
|
AlpinDale
|
7411a74cc6
bump version to 0.5.2
|
10 months ago |
|
AlpinDale
|
ad6802690f
feat: CMake Build System Generator (#332)
|
10 months ago |
|
AlpinDale
|
da223153c6
feat&fix: cohere support and missing GPU blocks (#333)
|
10 months ago |
|
AlpinDale
|
e2a7b50440
fix: logprobs when inf or nan (#329)
|
10 months ago |
|
AlpinDale
|
4791a63fdc
fix: env.py url in bugs template
|
10 months ago |
|
AlpinDale
|
8071ead964
chore: allow docker port and host to be changed (#327)
|
10 months ago |
|
AlpinDale
|
594fe814dc
bump version to v0.5.1 (#326)
|
10 months ago |
|
AlpinDale
|
f8652c8e99
fix: optimize aqlm dequantization (#325)
|
10 months ago |
|
AlpinDale
|
e42a78381a
feat: switch from pylint to ruff (#322)
|
10 months ago |
|
AlpinDale
|
637649df99
fix: model -> model architecture in issue templates
|
10 months ago |
