Respect sequence_len in config for type: llama2_chat (#926)

Browse Source

* Respect sequence_len in config for `type: llama2_chat`

It was hardcoded to `4096` I am not sure why?  This updates it to pull from the config. 

cc: @winglian

* Update llama2_chat.py

* apply black formatting

* fix tokenizer

* update test data

* lint fixtures

...

This commit is contained in:

Hamel Husain

2023-12-12 09:39:22 -08:00

committed by GitHub

parent 7fabc4d95e

commit f1de29dd1e

2 changed files with 4 additions and 3 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download Patch File Download Diff File Expand all files Collapse all files

2

tests/fixtures/conversation.tokenized_llama2chat.json vendored

File diff suppressed because one or more lines are too long