GitLab

pass through low_cpu_mem_usage

Update Dockerfile

Exclude the scripts directory from the installation

Fixes https://github.com/Lightning-AI/lit-gpt/issues/558
Fixes https://github.com/EleutherAI/lm-evaluation-harness/issues/901

Fix PubMedQA on master

Revert "Dataset licenses"

Dataset licenses

fix bug with output path in CWD

[bugfix] openai

Add `inject_fused_attention` parameter for `auto-gptq`

Add NQ-Open task based on the Natural Questions dataset

Add CMMLU Benchmark

Update crowspairs.py higher_is_better

Update triviaqa.py

Removes the `_remove_prefixes` optimization as it's invalid for exact match metrics. Closes https://github.com/EleutherAI/lm-evaluation-harness/issues/726

- Add in comments the sources of the licenses.
- Add "No license found" when it is the case.

[Main] updated to new anthropic API

Added the license for 7 datasets.
For 4 datasets I can't find the license and removed the TODO.

add bnb_4bit_use_double_quant and low_cpu_mem_usage

Remove Results folder temporarily