 oobabooga
|
cacbcda208
Two new options: truncation length and ban eos token
|
il y a 2 ans |
|
oobabooga
|
1911504f82
Minor bug fix
|
il y a 2 ans |
|
oobabooga
|
dba2000d2b
Do things that I am not proud of
|
il y a 2 ans |
 MarkovInequality
|
992663fa20
Added xformers support to Llama (#950)
|
il y a 2 ans |
|
oobabooga
|
a3085dba07
Fix LlamaTokenizer eos_token (attempt)
|
il y a 2 ans |
|
oobabooga
|
0b458bf82d
Simplify a function
|
il y a 2 ans |
 Φφ
|
ffd102e5c0
SD Api Pics extension, v.1.1 (#596)
|
il y a 2 ans |
|
oobabooga
|
ea6e77df72
Make the code more like PEP8 for readability (#862)
|
il y a 2 ans |
|
oobabooga
|
113f94b61e
Bump transformers (16-bit llama must be reconverted/redownloaded)
|
il y a 2 ans |
|
oobabooga
|
03cb44fc8c
Add new llama.cpp library (2048 context, temperature, etc now work)
|
il y a 2 ans |
 catalpaaa
|
4ab679480e
allow quantized model to be loaded from model dir (#760)
|
il y a 2 ans |
|
oobabooga
|
3a47a602a3
Detect ggml*.bin files automatically
|
il y a 2 ans |
|
oobabooga
|
4c27562157
Minor changes
|
il y a 2 ans |
 Thomas Antony
|
79fa2b6d7e
Add support for alpaca
|
il y a 2 ans |
|
Thomas Antony
|
7745faa7bb
Add llamacpp to models.py
|
il y a 2 ans |
|
oobabooga
|
1cb9246160
Adapt to the new model names
|
il y a 2 ans |
|
oobabooga
|
53da672315
Fix FlexGen
|
il y a 2 ans |
|
oobabooga
|
ee95e55df6
Fix RWKV tokenizer
|
il y a 2 ans |
|
oobabooga
|
fde92048af
Merge branch 'main' into catalpaaa-lora-and-model-dir
|
il y a 2 ans |
|
oobabooga
|
49c10c5570
Add support for the latest GPTQ models with group-size (#530)
|
il y a 2 ans |
 catalpaaa
|
b37c54edcf
lora-dir, model-dir and login auth
|
il y a 2 ans |
|
oobabooga
|
a6bf54739c
Revert models.py (accident)
|
il y a 2 ans |
|
oobabooga
|
a80aa65986
Update models.py
|
il y a 2 ans |
|
oobabooga
|
ddb62470e9
--no-cache and --gpu-memory in MiB for fine VRAM control
|
il y a 2 ans |
|
oobabooga
|
e26763a510
Minor changes
|
il y a 2 ans |
 Wojtek Kowaluk
|
7994b580d5
clean up duplicated code
|
il y a 2 ans |
|
Wojtek Kowaluk
|
30939e2aee
add mps support on apple silicon
|
il y a 2 ans |
|
oobabooga
|
ee164d1821
Don't split the layers in 8-bit mode by default
|
il y a 2 ans |
|
oobabooga
|
e085cb4333
Small changes
|
il y a 2 ans |
 awoo
|
83cb20aad8
Add support for --gpu-memory witn --load-in-8bit
|
il y a 2 ans |
