🤖 GitHub AI 热门报告

Changes since langchain-core==1.5.5

chore(core): release 1.5.6 (#39704)
feat(core): incorporate gateway metadata to traces (#39703)

* New features
  * Added an error check for trying to deserialize JAX exports that are older
    than the backwards compatibility window. Without this check the
    deserialization of expired artifacts may succeed and then result in
    obscure downstream errors.
    Added a configuration flag `--jax_export_deserialize_expired_versions` to
    temporarily bypass the error check.
    See https://docs.jax.dev/en/latest/export/export.html#compatibility-guarantees.
  * Added `jax.numpy.top_k`, which implements `numpy.top_k`, added in
    in NumPy v2.6.0 (#39729).

* Breaking changes
  * The `exec_time_optimization_effort` and `memory_fitting_effort` flags have been
    removed in favor of the `EffortLevel` enum.
  * JAX does not support anymore deserialization of Exported modules from
    before January 15th, 2026 because they are beyond the backwards compatibility
    window. On that date we added support to serialize shardings as NamedSharding,
    and now that is the only sharding serialization that is supported.
  * jnp.take_along_axis now always defaults wrap_negative_indices to True.
    It used to default to False for mode=promise_in_bounds and True otherwise.

## [3.2.0](https://github.com/openai/openai-python/compare/v3.1.0...v3.2.0) (2026-08-17)


### Features

* add Bedrock Runtime endpoint support (SDK-290) ([#3623](https://github.com/openai/openai-python/issues/3623)) ([86267d2](https://github.com/openai/openai-python/commit/86267d2737e61e389e9260a18f99bcc1a634811b))
* **api:** Add shell call streaming events and new service/image types ([#3635](https://github.com/openai/openai-python/issues/3635)) ([ff14a33](https://github.com/openai/openai-python/commit/ff14a33c1b21c36b9e43a0150a3818ee3aa3712a))

## 🌟 Summary

**v8.4.121 improves OpenVINO INT8 export reliability for YOLO26 models while delivering broad Platform API, dataset, annotation, deployment, and documentation updates.** 🚀

## 📊 Key Changes

- **Fixed OpenVINO INT8 detection-head handling** by @glenn-jocher:
  - Replaced fragile PyTorch-based layer matching with exact names from the converted OpenVINO graph.
  - Keeps Detect decoding, DFL, and Sigmoid operations in floating point as intended.
  - Preserves strict NNCF validation during quantization.
  - Verified with a successful YOLO26n-P2 INT8 export using **55 exact ignored operations**, with no unwanted `FakeQuantize` nodes. ✅
  - This directly addresses export failures reported in Sentry and is the most important change in this release.

- **Expanded Ultralytics Platform API documentation**:
  - Documented the generated `ultralytics-platform` Python SDK alongside REST examples.
  - Updated endpoint paths, authentication, pagination, rate limits, response formats, and OpenAPI guidance.
  - Added coverage for images, dataset ingestion, exports, storage integrations, billing, usage, trash, training, deployments, and account APIs.
  - Clarified that workspace API keys have owner-level permissions and are managed by workspace owners.

- **Improved Platform dataset and annotation workflows** 🏷️:

## Verify Docker Image Signature

All LiteLLM Docker images are signed with [cosign](https://docs.sigstore.dev/cosign/overview/). Every release is signed with the same key introduced in [commit `0112e53`](https://github.com/BerriAI/litellm/commit/0112e53046018d726492c814b3644b7d376029d0).

**Verify using the pinned commit hash (recommended):**

A commit hash is cryptographically immutable, so this is the strongest way to ensure you are using the original signing key:

```bash
cosign verify \
  --key https://raw.githubusercontent.com/BerriAI/litellm/0112e53046018d726492c814b3644b7d376029d0/cosign.pub \
  ghcr.io/berriai/litellm:v1.97.0
```

**Verify using the release tag (convenience):**

Tags are protected in this repository and resolve to the same key. This option is easier to read but relies on tag protection rules:

```bash
cosign verify \

## What's Changed
* llm: transcode WebP images for llama-server 
* renderers/qwen: tolerate non-leading system messages

**Full Changelog**: https://github.com/ollama/ollama/compare/v0.32.13...v0.32.14

## 0.122.0 (2026-08-13)

Full Changelog: [v0.121.0...v0.122.0](https://github.com/anthropics/anthropic-sdk-python/compare/v0.121.0...v0.122.0)

### Features

* **api:** add output_behavior to dream creation (create a new memory store or update the input store in place) ([852c4bb](https://github.com/anthropics/anthropic-sdk-python/commit/852c4bbe4a3a425a8780e89ea6c3cae54836e8bb))


### Bug Fixes

* **bedrock,aws:** run SigV4 signing off the event loop in async clients ([#334](https://github.com/anthropics/anthropic-sdk-python/issues/334)) ([2bae6c8](https://github.com/anthropics/anthropic-sdk-python/commit/2bae6c8cb86f693f4e1e3dd13bb64e03b01fe720))
* **bedrock:** expose beta.messages.parse, stream and tool_runner ([#366](https://github.com/anthropics/anthropic-sdk-python/issues/366)) ([6eca7bb](https://github.com/anthropics/anthropic-sdk-python/commit/6eca7bb19f968856b5652d5471e3ca9fc3fe8641))
* **client:** add models ([52e9d94](https://github.com/anthropics/anthropic-sdk-python/commit/52e9d9453a46a281846f9860e742bfc938bafebc))
* **client:** keep token exchange bound per client across copy() ([#388](https://github.com/anthropics/anthropic-sdk-python/issues/388)) ([c13e6e3](https://github.com/anthropics/anthropic-sdk-python/commit/c13e6e30b75d73b9af51468ae0deae6e6aca1ba0))
* **client:** read PathLike contents passed inside a file tuple ([070f953](https://github.com/anthropics/anthropic-sdk-python/commit/070f95332daea9dc3ed19fe91251d6e5285e5560))
* **client:** treat empty ANTHROPIC_API_KEY / ANTHROPIC_AUTH_TOKEN as unset ([#341](https://github.com/anthropics/anthropic-sdk-python/issues/341)) ([76a2e68](https://github.com/anthropics/anthropic-sdk-python/commit/76a2e68531d50c9043a5ac58919527287e56a842))
* **streaming:** add context to malformed tool input JSON errors in the non-beta accumulator ([#339](https://github.com/anthropics/anthropic-sdk-python/issues/339)) ([a343e17](https://github.com/anthropics/anthropic-sdk-python/commit/a343e17b7bc970f656fa02a980b0baa2bf8c3b80))
* **streaming:** apply all message_delta fields when accumulating streamed messages ([#380](https://github.com/anthropics/anthropic-sdk-python/issues/380)) ([fc1599b](https://github.com/anthropics/anthropic-sdk-python/commit/fc1599bd4c25ad5550d55f8f1c8f5c5664e19ed1))
* **streaming:** emit input_json events for server tool use blocks ([#336](https://github.com/anthropics/anthropic-sdk-python/issues/336)) ([ccfc8e1](https://github.com/anthropics/anthropic-sdk-python/commit/ccfc8e140e8e6b824e74c8fd9bed2587e60c5b6f))

Version 3.10.3 2026-08-12

* docs: wrap Chat-80 HOWTO output
* Sandbox Stanford JAR execution to nltk_data directories
* Harden path-traversal / file-I/O sandbox: close write-side symlink TOCTOU + shared-temp squat, lock the cluster with a living audit (CWE-22/59/377)
* Extend algorithmic-complexity DoS hardening: repo-wide sweep + two-string distances (CWE-407/CWE-400)
* Bound unbounded-work DoS in parsers and grammar transforms (CWE-407/674/835)
* fix(security): sandbox MaltParser's Java execution (CVE-2026-12252, CVE-2026-12841)
* fix(security): trust the system temp dir only when it is private (CWE-377/CWE-378)
* fix(security): validate corpus-reader roots against the data sandbox (CWE-73)
* fix(security): validate per-call java() options and replace the -XX:/-D allowlist with a minimal one (CWE-88)
* Additional security hardening (CWE-407, CWE-426, CWE-427, CWE-502, CWE-59, CWE-776, CWE-918)

Thanks to the following contributors to 3.10.3: Mohammad Favas S, leduckhuong, Ziyu Lin, dougtrainer28-cmyk, Chaitanya Kadian, 0xRenSec, Arpit Jain, Jace, nguyencanhthuong, Liling Tan, medimedi, Eric Kafe.



## What's Changed
* Sandbox Stanford JAR execution to nltk_data directories by @ekaf in https://github.com/nltk/nltk/pull/3743
* docs: wrap Chat-80 HOWTO output by @medisean in https://github.com/nltk/nltk/pull/3741

## 🔥 What's New?

### New Features

- **#13499** - Add Tavily provider blocks (search, extract, crawl, map) (by @teionarr)
- **#13629** - Add Claude Sonnet 5 support (Claude 5 tokenizer aware) (by @ntindle)
- **#13772** - Expert scheduling — attribution, triggers, thread posts, credit guardrail (by @0ubbe)
- **#13787** - Expert scheduling UI — expert page, chat schedules drawer, team card polish (by @0ubbe)
- **#13795** - Collapsible expert chat groups in sidebar with per-group load more (by @0ubbe)
- **#13746** - Schema-colocated Better Auth preview-account seeder (by @ntindle)
- **#13792** - Add brain dump STT quality gate and eval corpus (by @Abhi1992002)
- **#13771** - Add editable expert Soul documents (by @Abhi1992002)
- **#13800** - Morning briefing, needs-attention unification, briefing-first home (by @0ubbe)
- **#13804** - Retire brain dump greeting once a user has a session (by @Abhi1992002)
- **#13805** - Add first-visit tab intro cards for Agents, Marketplace, Build (by @Abhi1992002)
- **#13869** - Status counts and duration totals in execution cost summary (by @Abhi1992002)
- **#13758** - Add single-container distribution (by @ntindle)
- **#13904** - Home dashboard aggregation endpoint (by @Abhi1992002)
- **#13761** - Add ChatGPT/Codex subscription transport preview (by @ntindle)
- **#13999** - Close the loop on chat-platform account linking (by @Bentlybro)

### Features

-   [#13750](https://github.com/gradio-app/gradio/pull/13750) [`e0f0790`](https://github.com/gradio-app/gradio/commit/e0f0790b6fbe591cfb6de968316596806f5b59dc) - Workflow: only inline app-owned files, and keep opened HTML off the app origin.  Thanks @abidlabs!
-   [#13718](https://github.com/gradio-app/gradio/pull/13718) [`a9e8382`](https://github.com/gradio-app/gradio/commit/a9e8382c4f68e938b5fceef299d7539434c6063f) - Add browser-local run history and loading.  Thanks @abidlabs!
-   [#13751](https://github.com/gradio-app/gradio/pull/13751) [`1c8aa23`](https://github.com/gradio-app/gradio/commit/1c8aa23263ef1d3a641690395b8f0d816b593cc3) - Fix initial Spaces iframe resize after app render.  Thanks @dawoodkhan82!
-   [#13749](https://github.com/gradio-app/gradio/pull/13749) [`5019a24`](https://github.com/gradio-app/gradio/commit/5019a2402779d40d7c0b42817fc73bd2df3d8022) - Await `auth_dependency` and apply it when routes are built on an existing app.  Thanks @abidlabs!

Changes since 1.2.10

* release(langgraph): 1.2.11 (#8595)
* feat(langgraph): expose `trace_policy` on `add_node` (#8523)
* chore(deps): bump the minor-and-patch group across 1 directory with 7 updates (#8533)
* chore(deps): bump the minor-and-patch group across 1 directory with 5 updates (#8532)
* release(checkpoint-postgres): 3.1.2 (#8565)
* release(checkpoint): 4.2.0 (#8563)
* fix(checkpoint): collect writes at plain-value seed in delta channel history (#8526)
* chore: enforce PLC0415 in tests for the remaining packages (#8547)
* test(checkpoint-postgres,checkpoint-sqlite): run the conformance suite (#8537)
* chore: enable RUF100 and clear unused noqa directives (#8546)
* chore(deps-dev): bump types-requests from 2.33.0.20260518 to 2.33.0.20260712 in /libs/langgraph (#8502)
* chore(deps): bump cryptography from 48.0.1 to 50.0.0 in /libs/langgraph (#8528)
* release(checkpoint-sqlite): 3.1.1 (#8481)
* release(checkpoint-postgres): 3.1.1 (#8480)

# Highlights

* **Ray Data:** In this Ray release, we've enabled `DataSourceV2` by default (#64821), so `read_parquet` and friends use the new scan/listing infrastructure with row-group-aware chunking and predicate splitting. Hash Shuffle V2 eliminates the aggregator actor pool. V1 had to provision that pool up front from an estimate of the input size, and its actors accumulated partition shards in actor heap memory, invisible to Ray and unspillable, until finalization. V2 replaces it with two stateless task-based operators, ShuffleMapOp --> ShuffleReduceOp, that pass shards through the object store, so intermediate state spills under pressure and no capacity has to be reserved in advance. The map/reduce barrier itself remains in both designs.
* HashShuffleV2 supports `join` ([#63598](https://github.com/ray-project/ray/pull/63598), [#64538](https://github.com/ray-project/ray/pull/64538), [#64687](https://github.com/ray-project/ray/pull/64687)). This lets shuffles reuse standard map/reduce scheduling, backpressure, and resource accounting.
* **Ray Serve:** The HAProxy ingress is now distributed as the `ray-haproxy` PyPI package instead of being compiled into images, and it is the default HAProxy binary (#64141, #64163, #64164). We've also added gRPC support to the HAProxy direct-ingress path, including streaming, metrics, and custom request IDs (#63735, #64310, #64166, #64112). For Ray Serve LLM, we've added experimental KV-cache-aware request routing that tracks replica KV state through an event plane, tokenizes before routing, and routes on prefill/decode token load (#64084, #64085, #64097, #64224, #64327, #64400). KV cache-aware routing’s complete support will land in 2.58.
* **Ray Core:** We've added an embedded RocksDB storage backend for GCS fault tolerance (REP-64), selectable with `RAY_gcs_storage=rocksdb` and `RAY_gcs_storage_path` (#63657). GCS fault tolerance no longer requires an external Redis instance. We've also added a public API for topology-aware scheduling (#63479, #63740).

# Ray Data

### 🎉 New Features
* Enable `DataSourceV2` by default via `DataContext.use_datasource_v2` (#64821)
* New task-based hash shuffle v2 (`ShuffleMapOp` → `ShuffleReduceOp`) with `join`, multi-input reduce, downstream map fusion, and reducer remote args, behind an env flag (#63598, #64538, #64687, #64438, #64302, #64532, #64481)
* Add a `Catalog` abstraction with a `UnityCatalog` implementation that can be passed to `read_*`, and Unity Catalog write support for Parquet and Iceberg (#64193, #64519)
* Add `read_zarr` for Zarr datasets (#63003) and `read_lerobot` for LeRobot v3 datasets (#63821)
* Add `PushdownCountFiles` optimization to answer `count()` from Parquet footers (#64763)
* Add common subexpression elimination to the expression optimizer (#63974)
* Add GPU support for `Aggregate` (#63708)
* Make dataset iteration metrics queryable per split (#64608)
* Add custom operator stats to capture worker-side metrics during task execution (#64221)
* Refactor usage collection into an extensible `UsageCallback` (#64500)

# Release v5.15.0

## New Model additions

### Meta Muse Glimmer

Muse Glimmer, released today, is Meta’s new multimodal model, especially designed for agentic use cases. Distilled from Muse to 30B parameters, and released under the Apache 2.0 license, it can be deployed to local setups for privacy-aware applications such as coding, document analysis, personal assistants, Claw- or Hermes-like setups.

Muse Glimmer is a dense 30B parameter model consisting of:
- 2B ViT-style encoder for vision (Perception Encoder)
- 28B parameter text decoder

We're covering it in the following blogpost: http://hf.co/blog/muse-glimmer

<img width="960" height="1787" alt="image" src="https://github.com/user-attachments/assets/3d8e548e-f84f-4269-8bd0-a12722d7ab01" />

---

### GraniteMoeSWA & GraniteSWA

Typer removed click as a dependency in favour of vendoring it, but spaCy imports from Click, so the requirement needs to be added.

<!-- Release notes generated using configuration in .github/release.yml at 1.61.1 -->



**Full Changelog**: https://github.com/streamlit/streamlit/compare/1.61.0...1.61.1

# Change log

## Features :point_up:

* https://github.com/qdrant/qdrant/milestone/50 - TurboQuant 4-bit as a datatype of primary vector storage. Only store 4-bit quantized vectors and spare disk space on original vectors. [[docs](https://qdrant.tech/documentation/manage-data/vectors/#turbo4)]
* https://github.com/qdrant/qdrant/pull/9669, https://github.com/qdrant/qdrant/pull/9684, https://github.com/qdrant/qdrant/pull/9950 - Unify definition of memory usage strategy for collection components. Use `"memory": "cold" / "cached" / "pinned"` to define memory behavior for each individual collection component. Allows for more fine-grained control over memory usage and performance. [[docs](https://qdrant.tech/documentation/ops-configuration/memory-tiers/)]
* https://github.com/qdrant/qdrant/pull/9683 - Allow `"match": {"prefix": "..."}` in `filter` to match keywords by prefix, must be enabled in keyword index. [[docs](https://qdrant.tech/documentation/search/filtering/#prefix-match)]
* https://github.com/qdrant/qdrant/pull/9661 - Per-query IDF corpus for sparse vector search [[docs](https://qdrant.tech/documentation/search/text-search/full-text-search/#per-tenant-idf-statistics)]
* https://github.com/qdrant/qdrant/pull/9899 - Slice filtering condition: sliced scroll / deterministic sampling [[docs](https://qdrant.tech/documentation/search/filtering/#slice)]
* https://github.com/qdrant/qdrant/pull/10035 - Global quota API [[docs](https://qdrant.tech/documentation/ops-configuration/quotas/)]
* https://github.com/qdrant/qdrant/pull/9338 - Add routing token for deterministic read routes [[docs](https://qdrant.tech/documentation/scaling/consistency-guarantees/#read-affinity)]

## Improvements :point_left:

* https://github.com/qdrant/qdrant/pull/9113 - Use batched reads in vector and payload storage
* https://github.com/qdrant/qdrant/pull/9409 - Utilize `io_uring` for payload storage
* https://github.com/qdrant/qdrant/pull/9448 - Lower default update queue length, reduce from 1M to 200
* https://github.com/qdrant/qdrant/pull/9500 - Use Entry API to avoid redundant map double-lookups, improving performance
* https://github.com/qdrant/qdrant/pull/9332 - Enable single file mmap vector storage by default for immutable segments
* https://github.com/qdrant/qdrant/pull/9376 - Add option to explicitly disable BM25 stemmer; deprecate "none" hack

MLflow 3.15.1 is a patch release that includes bug fixes and documentation updates.

### Bug fixes:

- [Model Registry] Skip `env_pack` on ARM client images (#24762, @qyc)
- [Scoring / Tracking] Harden version parsing against missing/non-PEP440 versions on Databricks Serverless (#24799, @PattaraS)

### Documentation updates:

- [Docs] Clarify scorer versioning documentation (#24769, @nihalmenon)

# Keras 3.12.4

Keras 3.12.4 is a security patch release that hardens dataset loading and model file handling against insecure deserialization and decompression-bomb attacks.

## Security Fixes

* **Restrict unpickling when loading IMDB and Reuters datasets** — Replaces `np.load(allow_pickle=True)` with a restricted unpickler that only permits numpy array reconstruction, preventing arbitrary code execution via crafted `.npz` files (CWE-502). ([#23047](https://github.com/keras-team/keras/pull/23047)) by @LinZiyuu
* **Verify all intermediary H5 groups when navigating H5 files** — Manually resolves nested H5 group paths to verify group types at each step, preventing potential path traversal. ([#23168](https://github.com/keras-team/keras/pull/23168)) by @hertschuh
* **Reject decompression-bomb members on the `.keras` asset extraction path** — Adds per-member decompression-ratio checks before extracting `.keras` archives to disk, preventing disk-exhaustion attacks via crafted archives. ([#23101](https://github.com/keras-team/keras/pull/23101)) by @LinZiyuu
* **Restrict unpickling when loading CIFAR datasets** — Replaces bare `cPickle.load` in CIFAR-10/100 batch loading with the numpy-only `RestrictedUnpickler`, blocking arbitrary code execution via pickle gadgets. ([#23252](https://github.com/keras-team/keras/pull/23252)) by @SABITHSAHEB

---

## Contributors

Thank you to all the contributors who made this release possible! 🎉

* @LinZiyuu — Security hardening for IMDB, Reuters, and `.keras` asset extraction ([#23047](https://github.com/keras-team/keras/pull/23047), [#23101](https://github.com/keras-team/keras/pull/23101))
* @hertschuh — H5 group verification ([#23168](https://github.com/keras-team/keras/pull/23168))
* @SABITHSAHEB — CIFAR dataset pickle restriction ([#23252](https://github.com/keras-team/keras/pull/23252))

## v3.0.0

Release date: July 29, 2026

| Milvus Version | Python SDK Version | Node.js SDK Version | Java SDK Version | Go SDK Version |
| -------------- | ------------------ | ------------------- | ---------------- | -------------- |
| 3.0.0          | 3.0.1              | 3.0.3               | 3.0.5            | 3.0.0          |

Milvus 3.0.0 is officially released! Building on the lake-native architecture introduced in [3.0-beta](https://milvus.io/docs/release_notes.md#v30-beta), this release completes what the beta started: External Collection covers more lakehouse workflows; schema supports online add / backfill / drop; the sparse index is rebuilt around SINDI; StructArray and faceted search round out the retrieval engine; FAISS passthrough, and TEXT extend index and modality choices; and Woodpecker runs as a standalone service.

If you are new to the 3.0 line, the Core 3.0 features recall section below summarizes the capabilities introduced in 3.0-beta; the [3.0-beta release notes](https://milvus.io/docs/release_notes.md#v30-beta) have the full write-ups.

### What's new in 3.0.0 (since 3.0-beta)

#### External Collection: more complete lakehouse workflows

3.0-beta introduced External Collection: reference lake files in place, build indexes, and search them without copying data into Milvus. This release extends it toward complete lakehouse retrieval workflows. External fields can now feed function output fields such as BM25 sparse vectors, MinHash signatures, and text embeddings, so text and model-derived retrieval fields are built inside Milvus without copying the source table. Refresh also supports additive schema evolution: when the external table gains new columns, Milvus patches the affected segments instead of rebuilding the collection.

This release also adds a `milvus-table` external format that treats Milvus Snapshot metadata and Storage V3 manifests as an external source, so a collection snapshot can itself be served as an external table — batch and serving systems get a shared, manifest-backed view of the same data.

# Highlights

<img width="1250" height="560" alt="peft-v0 20 0" src="https://github.com/user-attachments/assets/6194fae9-f3c3-48ca-b43b-40c774bfe6eb" />

This release adds no less than nine new PEFT methods and puts a lot of work into the surrounding infrastructure, for example adding a new image generation benchmark for the method comparison suite and greatly improving the documentation structure.

## New Methods

### HiRA

@hqsiswiliam added ["HiRA: Parameter-Efficient Hadamard High-Rank Adaptation for Large Language Models"](https://openreview.net/forum?id=TwJrTz9cRS) to PEFT (#2668). Instead of adding the low-rank product `BA` to the base weight, HiRA multiplies it elementwise (Hadamard product) with the frozen base weight. Because the base weight itself is full rank, the resulting update is no longer constrained to be low rank, while the trainable parameter count stays the same as LoRA's.

### GLoRA

@not-lain contributed GLoRA: ["One-for-All: Generalized LoRA for Parameter-Efficient Fine-Tuning"](https://arxiv.org/abs/2306.07967) in #3098. It is a flexible PEFT method that extends LoRA with configurable weight, activation, and bias adaptation, delivering richer fine-tuning with no extra inference cost. Use it when you need per-layer flexibility or stronger adaptation than vanilla LoRA. Skip it for non-Linear layers (e.g. Conv/Embedding) or when standard LoRA is already sufficient and simplicity matters.

### BEFT

@whubaichuan added ["BEFT: Bias-Efficient Fine-Tuning of Language Models"](https://arxiv.org/abs/2509.15974v2) in #3195. BEFT builds on the observation that fine-tuning bias terms alone can be competitive in low-data regimes, but goes further: rather than training *all* biases, it targets the value projection by default, as the authors found this to be most efficient. This brings the trainable parameter count down to roughly 0.01% of the total parameters.

# Version 3.11.0 (July 27, 2026)

## Engine

1. [4787c809](https://github.com/google-deepmind/mujoco/commit/4787c809) Added [geom/surfacevel](https://mujoco.readthedocs.io/en/stable/XMLreference.html#body-geom-surfacevel): the velocity of a geom's surface as seen by contacts, given as a velocity field with a constant component and a rotational component about the geom frame origin. This allows conveyor belts, treadmills and turntables to be modeled with static geoms and no degrees of freedom: friction drives touching bodies along the motion of the surface, with the field projected onto each contact's tangent plane. Surface velocities compose correctly with each other and with body motion. Note that the contact rows of `mjData.efc_vel`, and the constraint-state sensors that read them, report the velocity relative to the moving surface rather than to the geom, since that is the quantity the constraint acts on; for geoms without `surfacevel` the two are identical. Contact-point visualization draws an arrow along the surface velocity at contacts with moving surfaces.

   [![Watch video](https://img.youtube.com/vi/PdSdrqhSiZA/mqdefault.jpg)](https://youtu.be/PdSdrqhSiZA)

2. [a264d0bc](https://github.com/google-deepmind/mujoco/commit/a264d0bc) Added [geom/adhesion](https://mujoco.readthedocs.io/en/stable/XMLreference.html#body-geom-adhesion) and [pair/adhesion](https://mujoco.readthedocs.io/en/stable/XMLreference.html#contact-pair-adhesion): an adhesive force associated with a contact, useful for modeling sticky materials. Contacts can pull with up to the given force before breaking, and the friction budget becomes $\mu(f_N + \text{adhesion})$. Combined with [gap](https://mujoco.readthedocs.io/en/stable/XMLreference.html#body-geom-gap), adhesive contacts apply "adhesion at a distance", useful for modeling magnets. Resting penetration is unaffected by adhesion. [mj_contactForce](https://mujoco.readthedocs.io/en/stable/APIreference/APIfunctions.html#mj-contactforce) reports the net interface force, whose normal component can now be negative.

   [![Watch video](https://img.youtube.com/vi/GioWwB36XHI/mqdefault.jpg)](https://youtu.be/GioWwB36XHI)

3. [f0fa3d82](https://github.com/google-deepmind/mujoco/commit/f0fa3d82) Replaced midpoint integration of free bodies with [gyroscopic derivatives](https://mujoco.readthedocs.io/en/stable/computation/index.html#gefreebody) in the `implicitfast` [integrator](https://mujoco.readthedocs.io/en/stable/computation/index.html#geintegrators): the bias-force derivative of every standalone free body is applied via a local unsymmetric solve of its decoupled block, making `implicitfast` identical to `implicit` for such bodies. Unlike midpoint integration, which required vacuum and no constraints, this applies in all environments (contacts, fluid, constraints), and is compatible with discrete-time inverse dynamics. Spinning free bodies no longer gain energy, but tumbling motion is now mildly damped; models requiring long-horizon energy conservation of tumbling bodies in vacuum should use `RK4`. The [invdiscrete](https://mujoco.readthedocs.io/en/stable/XMLreference.html#option-flag-invdiscrete) flag no longer has any effect on forward dynamics.

4. [5618666a](https://github.com/google-deepmind/mujoco/commit/5618666a) Added [body/simple](https://mujoco.readthedocs.io/en/stable/XMLreference.html#body-simple) attribute ("false"/"auto") to disable the *simple body* mass matrix optimization. This is useful for domain randomization, where model parameters may change post-compilation.

5. [14c0b0c9](https://github.com/google-deepmind/mujoco/commit/14c0b0c9) [mj_setConst](https://mujoco.readthedocs.io/en/stable/APIreference/APIfunctions.html#mj-setconst) now recomputes the `mjModel.{body,geom,site}_sameframe` flags, to account for changes in body/geom/site frames after compilation.

6. [2444defc](https://github.com/google-deepmind/mujoco/commit/2444defc) Added support for [multiccd](https://mujoco.readthedocs.io/en/stable/computation/index.html#comulticcd) with arbitrarily large meshes.

### Added

- 🎨 **Redesigned interface.** Open WebUI has been visually rebuilt from the ground up. All aspects of the User Interface, from the chat view to the admin panel. Now with a narrower conversation column, lighter typography, tidier spacing, consistent menus and dropdowns, clearly outlined text boxes, and settings rearranged. [Commit](https://github.com/open-webui/open-webui/commit/aedb6bef4e2eb12234c02085a545ff395d96db18), [Commit](https://github.com/open-webui/open-webui/commit/b3255a36569f295766271b8a2b0bd969b4083b9f), [Commit](https://github.com/open-webui/open-webui/commit/ba067258dea2229a9956077b3b0d7b1c68b56f66), [Commit](https://github.com/open-webui/open-webui/commit/8dd862d3383978f21111e63fb2d6029711abed9a), [Commit](https://github.com/open-webui/open-webui/commit/263bbc77d803e83b9af4b04c0cae29705af5f072), [Commit](https://github.com/open-webui/open-webui/commit/f8ea15b84a274712dca33daa970f63ed7368043e), [Commit](https://github.com/open-webui/open-webui/commit/9f17c5960a0e47a09773da4bba12997a31222fc8), [Commit](https://github.com/open-webui/open-webui/commit/6772b1cb4f4e0d3dc166956014e6e7b9bddc721a), [Commit](https://github.com/open-webui/open-webui/commit/d3fd860c131846a9458888f9c256a9a29f3767f2), [Commit](https://github.com/open-webui/open-webui/commit/f1584b5a3764f72de2de6caad507e7c39ad19c23), [Commit](https://github.com/open-webui/open-webui/commit/2e8d92c7b1a9bb8d35f4a27ba3c73368d735c480), [Commit](https://github.com/open-webui/open-webui/commit/e58a4633b15ae53d33fc3b46cb97c76d86be325f), [Commit](https://github.com/open-webui/open-webui/commit/04b146f2cec7e6a01e9d3590eb83655c128fa3c7), [Commit](https://github.com/open-webui/open-webui/commit/e5e2cd78769639b2df83776f1b991966f922f8b4), [Commit](https://github.com/open-webui/open-webui/commit/3316ba76aabe5429596ffd130fd36be4d5c3aa6c), [Commit](https://github.com/open-webui/open-webui/commit/6fcb38fe2e0aded9b85f655cd8f3279e9f4e765e), [Commit](https://github.com/open-webui/open-webui/commit/421da674468f638f72cc5266c5a3874aa3bca3b7), [Commit](https://github.com/open-webui/open-webui/commit/d0bea60581eaa07d41f92ad8f86007e83247e061), [Commit](https://github.com/open-webui/open-webui/commit/21e180182a5096481d4cbb1a8f94212c0a515a40), [Commit](https://github.com/open-webui/open-webui/commit/d027a32ed134ae104f2f142ba45ff38e56215c5f), [Commit](https://github.com/open-webui/open-webui/commit/437c06c4795a72700295d7690d5fd65d1153372c), [Commit](https://github.com/open-webui/open-webui/commit/1bf05ebc7d135d74969438824778c9f73243ba8d), [Commit](https://github.com/open-webui/open-webui/commit/fd07e3a8e3e619f3712067765b416f0925fa80d3), [Commit](https://github.com/open-webui/open-webui/commit/d3ea51fd466a8741afc4dfd4f0d0f2f77fb6467f), [Commit](https://github.com/open-webui/open-webui/commit/4da2ff2655d9abb851805da127cf60b4d9ad1aa7), [Commit](https://github.com/open-webui/open-webui/commit/2fcb36267f034f2b83f936bfacedc20b680a2710), [Commit](https://github.com/open-webui/open-webui/commit/1428a4ddce4998cb3664a5ce37e176442dd426fa), [Commit](https://github.com/open-webui/open-webui/commit/bc8d24c951e9a2c973fc2dd1f2832a2b0855bc0e), [Commit](https://github.com/open-webui/open-webui/commit/704d07e9a20a830aad7bfc5131b0d92621cf0691), [Commit](https://github.com/open-webui/open-webui/commit/e88d2e053c2a63cce3823c9b6006f4a184c4fef2), [Commit](https://github.com/open-webui/open-webui/commit/6940297486d4a5de127efd1a5148b0adcebe87e3), [Commit](https://github.com/open-webui/open-webui/commit/9ca8cf528af1c49da3f0a2bc3c6ca95c1dedbcf5), [Commit](https://github.com/open-webui/open-webui/commit/c4efa81d08c425678c810c51b4d62716e1e57117), [Commit](https://github.com/open-webui/open-webui/commit/bb12b1a18b77d80829cedb2d5bf965808222415b), [Commit](https://github.com/open-webui/open-webui/commit/49abfbdd155dc22882fdcb09989e4f4964db16ee), [#27178](https://github.com/open-webui/open-webui/pull/27178), [Commit](https://github.com/open-webui/open-webui/commit/dcc7fb1e8ef144205531829f8a56e52171c4d63d), [Commit](https://github.com/open-webui/open-webui/commit/5c505c1119fec6170c1bc092ed162f86262a887e)
- 🤖 **Sub-agents.** Administrators can now enable sub-agents, which let a model hand parts of a task to background helper agents that run their own tool-driven conversations and report results back into the chat, tuned through new "ENABLE_SUBAGENTS", concurrency, iteration, and system-prompt settings. [Commit](https://github.com/open-webui/open-webui/commit/7088d245bb45fc69c0b22748563b9f3c6f0daa73), [Commit](https://github.com/open-webui/open-webui/commit/2f37e853d1259a901f736a823bad29dcc2c3b130), [Commit](https://github.com/open-webui/open-webui/commit/959558fd82eb2a3c980231acd500b73ba4b698b3), [Commit](https://github.com/open-webui/open-webui/commit/3005b7bc71fcbd5abc6e73c3e4caa4ea781cdb76)
- 📂 **Folder pages.** Opening a folder now takes you to its own page, where its chats load a page at a time, can be sorted by title or last updated, and you can start a new chat straight from the folder. [Commit](https://github.com/open-webui/open-webui/commit/409fb39717be9ab7becd9e8c01801a08c5bae318)
- ⏲️ **Chat timers.** The assistant can now set a timer that brings a prompt back into the conversation later, after a delay or at a set time, and can drop it automatically if you read the chat or reply before it fires. [Commit](https://github.com/open-webui/open-webui/commit/b23ddeb2800098c6352203ec8fbe9fca40ba415c)
- 🔔 **Notification targets.** Notifications now have their own settings tab where you can send them to several webhook destinations, each picking which events it wants, from chats finishing or failing to channel messages and calendar alerts, with a test button and a choice between always notifying or only when you are away, and any webhook you already had is carried over for you. [Commit](https://github.com/open-webui/open-webui/commit/c55e373b994d3a14c99a97f44261422012f63266), [Commit](https://github.com/open-webui/open-webui/commit/cf235738f5a44db415012b3b0ebc1f6e752f5439), [Commit](https://github.com/open-webui/open-webui/commit/200d447f6289faca42f2a666bbabae2c7f3ebadf), [#24750](https://github.com/open-webui/open-webui/issues/24750)
- 🗯️ **Full replies in channels.** A reply from the assistant in a channel is now saved and shown in full, with its reasoning, tool calls and other structured parts, where it previously came through blank. [Commit](https://github.com/open-webui/open-webui/commit/498cdab9a548d7d2fd19c389204ee26236fc7efe), [#26720](https://github.com/open-webui/open-webui/pull/26720), [#27409](https://github.com/open-webui/open-webui/pull/27409), [#26707](https://github.com/open-webui/open-webui/issues/26707), [#26656](https://github.com/open-webui/open-webui/issues/26656)
- 📣 **Notifications from the assistant.** The assistant can now send you a notification itself when something is worth your attention, so a long task can reach you after you have moved on to something else. [Commit](https://github.com/open-webui/open-webui/commit/c55e373b994d3a14c99a97f44261422012f63266), [Commit](https://github.com/open-webui/open-webui/commit/200d447f6289faca42f2a666bbabae2c7f3ebadf)
- 🌎 **Share a chat with anyone holding the link.** A shared chat can now be set to Open so it opens without signing in, with visitors no longer bounced to the sign-in page on their way to it, which administrators must first allow through a new "Chats Open Sharing" permission that stays off by default, and such pages ask search engines not to index them. [Commit](https://github.com/open-webui/open-webui/commit/1f0dc90abe879a55f654f2333e29fb0f630831c7), [Commit](https://github.com/open-webui/open-webui/commit/0e0d08382ac0d05b1ad98c47e8a2e37df2a185bb)
- 🔖 **Chat variables.** A model's system prompt can now declare fields such as text boxes and dropdown lists that you fill in for a conversation, with the values saved alongside the chat and carried over when it is forked or cloned. [Commit](https://github.com/open-webui/open-webui/commit/bef8ae4b2f05ca49ed88a02ab7a3cdc11b62c4f1), [Commit](https://github.com/open-webui/open-webui/commit/4e869011cd5040b5d6a197fc83d5f50d2425dbc2), [Commit](https://github.com/open-webui/open-webui/commit/1e88367cc837b39c0e9958fefbe053803336dce2), [Commit](https://github.com/open-webui/open-webui/commit/8cbb7f765cfc9c9b3237a6c5593cd93f849033f0), [Commit](https://github.com/open-webui/open-webui/commit/b35e2d265a4e4a2f2a075b31917d48be1dd9ef19), [Commit](https://github.com/open-webui/open-webui/commit/239cb740077a14e452ad002a1e671a09ab558e40), [#26915](https://github.com/open-webui/open-webui/discussions/26915)
- 🗄️ **LDAP group synchronization.** Administrators can now map LDAP groups to Open WebUI groups from the authentication settings, with optional automatic creation of missing groups, so a user's group memberships are kept in step with the directory each time they sign in. [#27263](https://github.com/open-webui/open-webui/pull/27263), [#18015](https://github.com/open-webui/open-webui/issues/18015)
- 👥 **Restrict sharing with groups.** Admins can now stop resources from being shared with entire groups through a new "USER_PERMISSIONS_ACCESS_GRANTS_ALLOW_GROUPS" permission, which stays enabled by default so existing group sharing keeps working untouched. [Commit](https://github.com/open-webui/open-webui/commit/4ed19d504bd30c0fc801e9228d9816669ec1c09c), [Commit](https://github.com/open-webui/open-webui/commit/f84dabe3d97ff701c097055023f28f3f2f7ebd07), [Commit](https://github.com/open-webui/open-webui/commit/77da3d8c81b9a6fda4354619de94f5d433328d8e), [Commit](https://github.com/open-webui/open-webui/commit/84e4d6ef8277f4b4f3ac4d355b3219e9b5a37268), [#27124](https://github.com/open-webui/open-webui/pull/27124)
- 🤝 **Shared folder collaboration.** People with access to a shared folder can now use its files and system prompt as knowledge in chat and, with write access, rename and manage the folder, all according to their read or write permission. [Commit](https://github.com/open-webui/open-webui/commit/797293c74957bd79e42262d1dc0fd637a45d0357), [Commit](https://github.com/open-webui/open-webui/commit/caa2457c17e592587b804f21054cc000944af75c), [Commit](https://github.com/open-webui/open-webui/commit/009715cd63d1c8b5320aba68e9a70afcde519016), [Commit](https://github.com/open-webui/open-webui/commit/53ccd718a53de25bb6d61476a6617bfb3f130a44)
- 👁️ **Chat previews in the sidebar.** Hovering a chat in the sidebar now shows a compact preview of its recent messages, so you can find the conversation you want without opening it. [Commit](https://github.com/open-webui/open-webui/commit/d0f7da4f45b8831b09b2ab3ec8f91aa354d90ba3), [Commit](https://github.com/open-webui/open-webui/commit/aaf2834db758bfec69408ab4cabcf324965c221c), [Commit](https://github.com/open-webui/open-webui/commit/1513ddaf58fe18029086880461d7cad0649a699c), [Commit](https://github.com/open-webui/open-webui/commit/93bd05271c07c249978d69abf3297fd2841900f9)
- 🕗 **Local message timestamps.** Message timestamps now appear on hover in your device's local date and time format, with the full weekday and date shown in a tooltip. [Commit](https://github.com/open-webui/open-webui/commit/797293c74957bd79e42262d1dc0fd637a45d0357), [Commit](https://github.com/open-webui/open-webui/commit/f84dabe3d97ff701c097055023f28f3f2f7ebd07)
- 📇 **User variables.** You can now store your own values in account settings, such as your role or how you like answers written, and a model's system prompt can insert them wherever they are needed. [Commit](https://github.com/open-webui/open-webui/commit/bd5d7b2e879511882429075d804d9222956f4a1a), [Commit](https://github.com/open-webui/open-webui/commit/212eec408ca320edfa2271e604d10e14b6a9bc1a), [Commit](https://github.com/open-webui/open-webui/commit/793a43d9c48225925929eb312fe2b70d5914d1da)
- 🧺 **Automations that file their chats away.** An automation can now be pointed at one of your folders, from the dialog, the editor or by asking the assistant, so each run lands there instead of loose in your chat list, and the folder is cleared automatically if it is later deleted. [Commit](https://github.com/open-webui/open-webui/commit/f798d05586a140f1a6b51f1e51b2b2a63d079d45), [Commit](https://github.com/open-webui/open-webui/commit/bab71ed08b5af6f4a8ff2daa02792baae9edab03), [Commit](https://github.com/open-webui/open-webui/commit/db5c092299471444c356216d5ef39b382ba1aa1e)
- 🔵 **See what you have not read yet.** Folders in the sidebar now carry a count of chats with something new in them, a folder's own page marks unread chats with a dot, shows a spinner on any still generating, clears the dot as you open one, and keeps itself up to date as replies finish elsewhere, unread chats sort to the top of a folder, and you can mark a single chat unread again mark everything in a folder read, or mark every chat read at once from the sidebar. [Commit](https://github.com/open-webui/open-webui/commit/f798d05586a140f1a6b51f1e51b2b2a63d079d45), [Commit](https://github.com/open-webui/open-webui/commit/f867825bf3b7699bc2bd967ef46b2bb63e48b098), [Commit](https://github.com/open-webui/open-webui/commit/1de36d600f7191c28a98bf4b347b44cf8f1bef43), [Commit](https://github.com/open-webui/open-webui/commit/85c47fb467177ed811ba77dfe62461dfbe8e2548), [Commit](https://github.com/open-webui/open-webui/commit/b7489bbc6c4e376c017edffd8da3c2eb4e6c1c8e), [Commit](https://github.com/open-webui/open-webui/commit/3cd72ee6a8e93dc39a4d4c173117056e25326c8a), [Commit](https://github.com/open-webui/open-webui/commit/6f93ecd4fd77b0d51a5fbbd2fc3fd6d151036a55), [Commit](https://github.com/open-webui/open-webui/commit/e5a08d52208e8b1ed07ff94906e27d174146b1ca), [Commit](https://github.com/open-webui/open-webui/commit/8ddf119570b3c0b04b673d41cb2363370c23939b)

## Migration

**Settings and credentials now live in a backend-managed store.**

Provider settings and credentials previously persisted in the webview's `localStorage`. In 0.8.4, Jan reads and writes them from a backend-managed store instead (secrets go to the OS keyring). Your existing settings and API keys are copied over automatically on first launch - no manual action is needed, and from this version on all new changes are saved to the backend store.

Your old `localStorage` data is **not deleted** - it's kept as a snapshot so you can downgrade to a pre-0.8.4 build and still have your previous settings. Any changes you make in 0.8.4+ live only in the backend store, so a downgrade sees the older snapshot, not your latest settings.

---

## What's Changed
* chore(flatpak): add 0.8.3 release notes to appdata by @qnixsynapse in https://github.com/janhq/jan/pull/8343
* docs: update changelogs for v0.8.3 and update documentation by @qnixsynapse in https://github.com/janhq/jan/pull/8336
* Merge release/v0.8.3 to main by @qnixsynapse in https://github.com/janhq/jan/pull/8344
* i18n(ja): update Japanese translations (#8264) by @mahirhir in https://github.com/janhq/jan/pull/8348
* i18n(ja): complete common.json translations and fix terminology (#8264) by @mahirhir in https://github.com/janhq/jan/pull/8349
* i18n(ja): complete settings.json translations (#8264) by @mahirhir in https://github.com/janhq/jan/pull/8352
* fix(markdown): force code blocks LTR in RTL language context by @qnixsynapse in https://github.com/janhq/jan/pull/8350
* feat(chat): add toggle for folding interim text into reasoning trace by @qnixsynapse in https://github.com/janhq/jan/pull/8363
* fix(rag): repair embeddings, stop chat-model reloads, and parse more PDFs by @qnixsynapse in https://github.com/janhq/jan/pull/8360

# PyTorch 2.13.0 Release Notes

- [Highlights](#highlights)
- [Backwards Incompatible Changes](#backwards-incompatible-changes)
- [Deprecations](#deprecations)
- [New Features](#new-features)
- [Improvements](#improvements)
- [Bug fixes](#bug-fixes)
- [Performance](#performance)
- [Documentation](#documentation)
- [Developers](#developers)

# Highlights

<table>
  <tr><td><strong>FlexAttention</strong> lands on Apple Silicon (MPS), with up to ~12x speedup over SDPA on sparse patterns, and gains a deterministic backward path on CUDA for reproducible gradient computation.</td></tr>
  <tr><td><strong>CuTeDSL "Native DSL" backend</strong> gives Inductor a second high-performance code path (alongside Triton) for key GPU operations, with faster compilation. [Prototype]</td></tr>
  <tr><td><strong><code>nn.LinearCrossEntropyLoss</code></strong> combines the final prediction and loss computation to cut peak GPU memory by up to 4x for large-vocabulary language model training.</td></tr>
  <tr><td><strong>torchcomms</strong>, a new communications backend for PyTorch Distributed, improves fault tolerance, scalability, and debuggability for large-cluster training.</td></tr>
  <tr><td><strong>FSDP2</strong> now overlaps reduce-scatter and all-gather communications via a dedicated process group (opt-in), increasing distributed training throughput.</td></tr>

## New Pipelines

### Cosmos 3

[**Cosmos 3**](https://huggingface.co/docs/diffusers/main/api/pipelines/cosmos3) is NVIDIA's unified world foundation model (WFM) for Physical AI — a single omni-model built on a Mixture-of-Transformers (MoT) architecture that combines world generation, physical reasoning, and action generation, replacing the separate Predict, Reason, and Transfer models from earlier Cosmos releases. A single `Cosmos3OmniTransformer` runs a Qwen-style language model in parallel with a diffusion generation pathway, joined by a 3D multimodal RoPE. This release also lands video-to-video and action-conditioned generation, and a sound encoder.

- PR: [https://github.com/huggingface/diffusers/pull/13818](https://github.com/huggingface/diffusers/pull/13818)
- Docs: [https://huggingface.co/docs/diffusers/main/api/pipelines/cosmos3](https://huggingface.co/docs/diffusers/main/api/pipelines/cosmos3)

Thanks to @atharvajoshi10, @yzhautouskay, and @MaciejBalaNV for the contributions.

### Ideogram 4

[**Ideogram 4**](https://huggingface.co/docs/diffusers/main/api/pipelines/ideogram4) is a flow-matching text-to-image model that uses a multimodal text encoder and an asymmetric classifier-free guidance scheme: a dedicated `unconditional_transformer` produces the negative branch with zeroed text features, while the main `transformer` consumes the full packed text + image sequence. The pipeline ships with structured prompt upsampling and LoRA loading support.

- PR: [https://github.com/huggingface/diffusers/pull/13859](https://github.com/huggingface/diffusers/pull/13859)
- Docs: [https://huggingface.co/docs/diffusers/main/api/pipelines/ideogram4](https://huggingface.co/docs/diffusers/main/api/pipelines/ideogram4)

Thanks to @JinLiIdeogram for the contribution.

### Breaking Changes:
- Relaxed Gymnasium version range (from `"gymnasium>=0.29.1,<1.3.0"` to `"gymnasium>=0.29.1,<2.0"`)
- `pandas` and `matplotlib` are no longer core dependencies; they are now optional and only required for loading results and plotting (moved to `stable-baselines3[extra]`).
- Moved `read_json` and `read_csv` helper functions to test files
- Raised `torch` minimum version from 2.3 to 2.8 to mitigate https://github.com/advisories/GHSA-887c-mr87-cxwp

### Bug Fixes:
- Fixed deprecated error Taxi-v3 from gymnasium v1.3.0 in tests

### [SB3-Contrib]
- Optimized tests (faster to run)
- Fixed dead link for `RecurrentPPO`.

### [RL Zoo]


### [SBX] (SB3 + Jax)

- Added support for `rollout_buffer_class` and `rollout_buffer_kwargs` arguments in `PPO` and `OnPolicyAlgorithmJax` constructors, as in Stable Baselines3. (@Trenza1ore)
- Updated Jax dependency

## FSDP2 Improvements

This release brings a large batch of FSDP2 fixes and quality-of-life improvements: correct dtype handling on load, sharding of embeddings/norms, QLoRA crash prevention, and a more robust auto-wrap policy.

  - Fsdp2 fully_shard embedding and norm by @SunMarc in #4015
  - Fix fsdp2 load full state dict dtype mismatch by @SunMarc in #4021
  - Fix region compilation fsdpv2 by @SunMarc in #4022
  - [FSDP2] Cast model to uniform dtype before fully_shard to fix mixed-dtype AssertionError by @roycho96 in #3985
  - [FSDP2] Auto-exclude non-floating frozen Params4bit from fully_shard to prevent QLoRA crash by @roycho96 in #3987
  - fix(FSDP2): auto-wrap policy ignoring _no_split_modules fallback by @JohnGiorgi in #3999
  - fix: use key-based matching in fsdp2_load_full_state_dict by @roycho96 in #3982
  - fix: add missing model_has_params4bit guard to fsdp2_load_full_state_dict call by @roycho96 in #3981
  - Fix to-fsdp2: drop REMOVED / NOT_YET_IMPLEMENTED FSDP1 keys instead of leaking them by @lollinng in #4065
  - Prevent double-wrapping models in prepare_model() by @joshuaswanson in #3977

## AMD ROCm support

Accelerate now works end-to-end on AMD ROCm devices. Thanks @Abdennacer-Badaoui!

- Make accelerate work end-to-end on AMD ROCm by @Abdennacer-Badaoui in #4025

- Fix `dialog` dependency requirement to be `>= 1.1.0`.

Version: `1.5.9`
Git ref: `refs/tags/1.5.9`
Build Date: `2026-05-05T05:55`
PIP Package: `chroma-1.5.9.tar.gz`
Github Container Registry Image: `:1.5.9`
DockerHub Image: `:1.5.9`

## What's Changed
* [ENH](frontend): block functions on topology dbs by @rescrv in https://github.com/chroma-core/chroma/pull/6836
* [ENH](faults): Add Tilt fault injection CLI by @rescrv in https://github.com/chroma-core/chroma/pull/6881
* [CHORE]  Debug TimeoutError in test_add.py by @rescrv in https://github.com/chroma-core/chroma/pull/6905
* [ENH]: Enable rebuilds for sharded collections by @tanujnay112 in https://github.com/chroma-core/chroma/pull/6916
* [ENH]: Group by support with sharding by @sanketkedia in https://github.com/chroma-core/chroma/pull/6909
* [CHORE]: Denormalize tenant and database into collection_compaction_cursors table by @tanujnay112 in https://github.com/chroma-core/chroma/pull/6940
* [CHORE]  Use normalized record sets for test add by @rescrv in https://github.com/chroma-core/chroma/pull/6935
* [ENH]: Add workflow to build and publish service container images by @jasonvigil in https://github.com/chroma-core/chroma/pull/6944
* [ENH] - Updates language around Chroma Cloud to be more representative. by @tjkrusinskichroma in https://github.com/chroma-core/chroma/pull/6952
* [ENH]: Add change stream to collection compaction cursors by @tanujnay112 in https://github.com/chroma-core/chroma/pull/6955
* [BUG] Switch to storing DOCKERHUB_USERNAME as var by @jasonvigil in https://github.com/chroma-core/chroma/pull/6962
* [CHORE]: Standardize Tilt CI image build on root docker-bake.hcl by @jasonvigil in https://github.com/chroma-core/chroma/pull/6958

# Release 2.21.0

## TensorFlow

### Breaking Changes

* Support for Python 3.9 has been removed starting with TF 2.21.
* The TensorBoard (TB) dependency has been removed starting with TF 2.21.

### Major Features and Improvements

* `tf.lite`
    * Adds int8 and int16x8 support for SQRT operator.
    * Adds int16x8 support for EQUAL and NOT_EQUAL operators.
    * Adds support for int2 type.
    * Adds support for int2/int4 in tfl.cast .
    * Adds support for SRQ int2 in tfl.fully_connected.
    * Adds support for int4 in tfl.slice.
    * Adds support for uint4 type.

## What's Changed
* misc(gha): expose action cache url and runtime as secrets by @mfuntowicz in https://github.com/huggingface/text-generation-inference/pull/2964
* feat: support max_image_fetch_size to limit by @drbh in https://github.com/huggingface/text-generation-inference/pull/3339
* Maintenance mode by @LysandreJik in https://github.com/huggingface/text-generation-inference/pull/3344
* Maintenance mode by @LysandreJik in https://github.com/huggingface/text-generation-inference/pull/3345
* fix(num_devices): fix num_shard/num device auto compute when NVIDIA_VISIBLE_DEVICES == "all" or "void" by @oOraph in https://github.com/huggingface/text-generation-inference/pull/3346


**Full Changelog**: https://github.com/huggingface/text-generation-inference/compare/v3.3.6...v3.3.7

## What's Changed
* Fix docs dotnet core typo by @lach-g in https://github.com/microsoft/autogen/pull/6950
* Fix loading streaming Bedrock response with tool usage with empty argument by @pawel-dabro in https://github.com/microsoft/autogen/pull/6979
* Support linear memory in RedisMemory by @justin-cechmanek in https://github.com/microsoft/autogen/pull/6972
* Fix message ID for correlation between streaming chunks and final mes… by @smalltalkman in https://github.com/microsoft/autogen/pull/6969
* fix: extra args not work to disable thinking by @liuyunrui123 in https://github.com/microsoft/autogen/pull/7006
* Add thinking mode support for anthropic client by @SrikarMannepalli in https://github.com/microsoft/autogen/pull/7002
* Fix spurious </think> tags caused by empty string reasoning_content in streaming by @Copilot in https://github.com/microsoft/autogen/pull/7025
* Fix GraphFlow cycle detection to properly clean up recursion state by @Copilot in https://github.com/microsoft/autogen/pull/7026
* Add comprehensive GitHub Copilot instructions for AutoGen development by @Copilot in https://github.com/microsoft/autogen/pull/7029
* Fix Redis caching always returning False due to unhandled string values by @Copilot in https://github.com/microsoft/autogen/pull/7022
* Fix OllamaChatCompletionClient load_component() error by adding to WELL_KNOWN_PROVIDERS by @Copilot in https://github.com/microsoft/autogen/pull/7030
* Fix finish_reason logic in Azure AI client streaming response by @litterzhang in https://github.com/microsoft/autogen/pull/6963
* Add security warnings and default to DockerCommandLineCodeExecutor by @ekzhu in https://github.com/microsoft/autogen/pull/7035
* Fix: Handle nested objects in array items for JSON schema conversion by @kkutrowski in https://github.com/microsoft/autogen/pull/6993
* Fix not supported field warnings in count_tokens_openai by @seunggil1 in https://github.com/microsoft/autogen/pull/6987
* Fix(mcp): drain pending command futures on McpSessionActor failure by @withsmilo in https://github.com/microsoft/autogen/pull/7045
* Add missing reasoning_effort parameter support for OpenAI GPT-5 models by @Copilot in https://github.com/microsoft/autogen/pull/7054
* Update version to 0.7.5 by @ekzhu in https://github.com/microsoft/autogen/pull/7058

- Added support for all GPT-5 models.
- Added support for Grok-4 via `xai/grok-4` and `openrouter/x-ai/grok-4` model names.
- Added support for `gemini/gemini-2.5-flash-lite-preview-06-17` model, by Tamir Zahavi-Brunner.
- `/clear` now prints “All chat history cleared.” so you know it worked, by Zexin Yuan.
- `/undo` output now shows only the first line of each commit message, making it easier to read.
- Added support for `openrouter/moonshotai/kimi-k2` model, by Jack Harrington.
- Display model announcements with no-arg `/model` command.
- Fixed an issue where new settings for an existing model didn't replace the old ones, by Andrew Grigorev.
- Fixed analytics to support the latest PostHog SDK event-capture API.
- Bumped dependencies to pick up latest litellm==1.75.0.

- Aider wrote 88% of the code in this release.

## 1.10.1

### Bug Fixes:
* fix image upscale on cpu ([#16275](https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/16275))

## What's Changed
* fix: resolve colab unsupported image type issue by @mashb1t in https://github.com/lllyasviel/Fooocus/pull/3506

**Full Changelog**: https://github.com/lllyasviel/Fooocus/compare/v2.5.4...v2.5.5

## Highlights
- Added SGLang worker for vision language models, lower latency and higher throughput https://github.com/lm-sys/FastChat/pull/2928
- Vision langauge WebUI https://github.com/lm-sys/FastChat/pull/2960
- OpenAI-compatible API server now supports image input https://github.com/lm-sys/FastChat/pull/2928
- Added LightLLM worker for higher throughput https://github.com/lm-sys/FastChat/blob/main/docs/lightllm_integration.md
- Added Apple MLX worker https://github.com/lm-sys/FastChat/pull/2940

## What's Changed
* fix specify local path issue use model from www.modelscope.cn by @liuyhwangyh in https://github.com/lm-sys/FastChat/pull/2934
* support openai embedding for topic clustering by @CodingWithTim in https://github.com/lm-sys/FastChat/pull/2729
* Remove duplicate API endpoint by @surak in https://github.com/lm-sys/FastChat/pull/2949
* Update Hermes Mixtral by @teknium1 in https://github.com/lm-sys/FastChat/pull/2938
* Enablement of REST API Usage within Google Colab Free Tier by @ggcr in https://github.com/lm-sys/FastChat/pull/2940
* Create a new worker implementation for Apple MLX by @aliasaria in https://github.com/lm-sys/FastChat/pull/2937
* feat: support Model Yuan2.0, a new generation Fundamental Large Language Model developed by IEIT System by @cauwulixuan in https://github.com/lm-sys/FastChat/pull/2936
* Fix the pooling method of BGE embedding model by @staoxiao in https://github.com/lm-sys/FastChat/pull/2926
* SGLang Worker by @BabyChouSr in https://github.com/lm-sys/FastChat/pull/2928
* Update mlx_worker to be async by @aliasaria in https://github.com/lm-sys/FastChat/pull/2958
* Integrate LightLLM into serve worker by @zeyugao in https://github.com/lm-sys/FastChat/pull/2888
* Copy button by @surak in https://github.com/lm-sys/FastChat/pull/2963