# Hugging Face Transformers v4.36.0 — Release & Upgrade Intelligence
- **Product:** Hugging Face Transformers (https://huggingface.co)
- **Version:** v4.36.0
- **Released:** Dec 11, 2023
- **Impact Score:** 7/10
- **Breaking Changes:** Source flagged
- **Canonical Analysis:** https://changelogs.us/transformers/updates/v4.36.0/

## Executive Summary (TL;DR)
Hugging Face Transformers v4.36.0. Source highlight: Mixtral is the new open-source model from Mistral AI announced by the blogpost Mixtral of Experts. The model has been proven to have comparable capabilities to Chat-GPT according to the benchmark results shared on the release blogpost.

## Breaking Changes & Deprecations
- Breaking changes ### pickle files

## New Features
- [`Llava`] Add Llava to transformers  by @younesbelkada in #27662
- Add SeamlessM4T v2  by @ylacombe in #27779
- Add CLVP  by @susnato in #24745
- Add Phi-1 and Phi-1_5  by @susnato in #26170
- Add DINOv2 depth estimation  by @NielsRogge in #26092
- Add RoCm scheduled CI & upgrade RoCm CI to PyTorch 2.1  by @fxmarty in #26940

## Improvements
- [LLaVa] Some improvements by @NielsRogge in #27895
- translate big_models.md and performance.md to chinese (#27334)
- Perf torch compile (#27422)
- translate hpo_train.md and perf_hardware.md to chinese (#27431)

## Bug Fixes
- [Time series] Add PatchTST  by @psinthong in #25927
- [Time series] Add PatchTST  by @kashif in #27581
- [Time series] Add PatchTSMixer  by @ajati in #26247
- Fix `from_pt` flag when loading with `safetensors`  by @LysandreJik in #27394
- 🚨🚨 Fix beam score calculation issue for decoder-only models  by @VsonicV in #27351
- Fix remaining issues in beam score calculation  by @VsonicV in #27808

## Upgrade Command
```bash
python -m pip install 'transformers==4.36.0'
```