v0.7.9: Patch release for DPO & SFTTrainer

This is a patch release that fixes critical issues with SFTTrainer & DPOTrainer, together with minor fixes for PPOTrainer and DataCollatorForCompletionOnlyLM

What's Changed

• Release: v0.7.8 by @younesbelkada in https://github.com/huggingface/trl/pull/1200
• set dev version by @younesbelkada in https://github.com/huggingface/trl/pull/1201
• Fix instruction token masking by @mgerstgrasser in https://github.com/huggingface/trl/pull/1185
• Fix reported KL in PPO trainer by @mgerstgrasser in https://github.com/huggingface/trl/pull/1180
• [DPOTrainer] Fix peft + DPO + bf16 if one uses generate_during_eval or pre-computed logits by @younesbelkada in https://github.com/huggingface/trl/pull/1203
• Revert "Address issue #1122" by @younesbelkada in https://github.com/huggingface/trl/pull/1205
• Release: v0.7.9 by @younesbelkada in https://github.com/huggingface/trl/pull/1206

Full Changelog: https://github.com/huggingface/trl/compare/v0.7.8...v0.7.9