Spaces:

Dovakiins
/

qwerrwe

Build error

App Files Files Community

qwerrwe / src /axolotl /prompt_strategies

100 contributors

History: 49 commits

winglian's picture

add ultrachat prompt strategies (#996)

ba043a3 unverified 12 months ago

__init__.py
926 Bytes

support custom field for completion from yml (#580) about 1 year ago
alpaca_chat.py
3.67 kB

fixes for alpaca w chatml, and don't include attention_mask w mistral for flash attention (#728) about 1 year ago
alpaca_instruct.py
638 Bytes

skip the system prompt over 1 year ago
alpaca_w_system.py
5.49 kB

support user defined prompters, pretokenized datasets in config, local parquet, local arrow files (#348) over 1 year ago
completion.py
2.54 kB

split completion text to sequence_len (#616) about 1 year ago
context_qa.py
3.16 kB

add support for defined train split (#654) about 1 year ago
creative_acr.py
6.53 kB

Apply isort then black over 1 year ago
llama2_chat.py
8.17 kB

Respect sequence_len in config for `type: llama2_chat` (#926) 12 months ago
metharme.py
2.41 kB

feat: add Metharme prompt strategy (#446) over 1 year ago
orcamini.py
1.49 kB

Added Orca Mini prompt strategy (#263) over 1 year ago
pygmalion.py
3.57 kB

Fixed pre-commit problems, fixed small bug in logging_config to handle LOG_LEVEL env var over 1 year ago
sharegpt.py
4.51 kB

add ultrachat prompt strategies (#996) 12 months ago
sharegpt_jokes.py
981 Bytes

use fastchat conversations template (#578) about 1 year ago
user_defined.py
2.63 kB

support user defined prompters, pretokenized datasets in config, local parquet, local arrow files (#348) over 1 year ago