ExLlama with long context (#2875)
This commit is contained in:
parent
9290c6236f
commit
c52290de50
14 changed files with 22 additions and 25 deletions
|
@ -104,12 +104,8 @@ llama-65b-gptq-3bit:
|
|||
mode: 'instruct'
|
||||
instruction_template: 'StableLM'
|
||||
truncation_length: 4096
|
||||
chat_prompt_size: 4096
|
||||
chat_prompt_size_max: 4096
|
||||
.*stablelm-base:
|
||||
truncation_length: 4096
|
||||
chat_prompt_size: 4096
|
||||
chat_prompt_size_max: 4096
|
||||
.*wizardlm:
|
||||
mode: 'instruct'
|
||||
model_type: 'llama'
|
||||
|
@ -237,8 +233,6 @@ TheBloke_WizardLM-30B-GPTQ:
|
|||
instruction_template: 'Minotaur'
|
||||
.*minotaur-15b:
|
||||
truncation_length: 8192
|
||||
chat_prompt_size: 8192
|
||||
chat_prompt_size_max: 8192
|
||||
.*orca_mini:
|
||||
mode: 'instruct'
|
||||
instruction_template: 'Orca Mini'
|
||||
|
|
Loading…
Add table
Add a link
Reference in a new issue