Inference

What's Changed

• Use rust-toolchain.toml before rustup on Dockerfile-{cuda,cuda-all} by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/842
• fix(backend): replace bare except with Exception in device check by @llukito in https://github.com/huggingface/text-embeddings-inference/pull/821
• Set version 1.9.3 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/849

New Contributors

• @llukito made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/821

Full Changelog: https://github.com/huggingface/text-embeddings-inference/compare/v1.9.2...v1.9.3

What's Changed

• Fix auto-truncate false setting by @vrdn-23 in https://github.com/huggingface/text-embeddings-inference/pull/836
• Set pad_token_id as nullable & add support for rope_parameters by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/832
• docs: add Homebrew installation to README by @Peredery in https://github.com/huggingface/text-embeddings-inference/pull/834
• feat: support pplx-embed-v1 by @mkrimmel-pplx in https://github.com/huggingface/text-embeddings-inference/pull/824

New Contributors

• @Peredery made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/834
• @mkrimmel-pplx made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/824

Full Changelog: https://github.com/huggingface/text-embeddings-inference/compare/v1.9.1...v1.9.2

What's Changed

🚨 Fix

• Fix support for containers w/ CUDA 13.0+ by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/831

When releasing ghcr.io/huggingface/text-embeddings-inference:cuda-1.9 with CUDA 12.9 and cuda-compat-12-9 there was an issue when running that same container on instances with CUDA 13.0+, as the cuda-compat-12-9 set in LD_LIBRARY_PATH was leading to a CUDA_ERROR_SYSTEM_DRIVER_MISMATCH = 803, which is now solved with a custom entrypoint that dynamically includes the cuda-compat on the LD_LIBRARY_PATH depending on the instance CUDA version.

Full Changelog: https://github.com/huggingface/text-embeddings-inference/compare/v1.9.0...v1.9.1

<img width="1800" height="972" alt="text-embeddings-inference-v1 9 0" src="https://github.com/user-attachments/assets/fe3751d1-1a3a-4b1f-8cf5-5c2326c14a62" />

What's changed?

🚨 Breaking changes

• Default HiddenAct::Gelu to GeLU + tanh in favour of GeLU erf by @vrdn-23 in https://github.com/huggingface/text-embeddings-inference/pull/753

Default GeLU implementation is now GeLU + tanh approximation instead of exact GeLU (aka. GeLU erf) to make sure that the CPU and CUDA embeddings are the same (as cuBLASlt only supports GeLU + tanh), which represents a slight misalignment from how Transformers handles it, as when hidden_act="gelu" is set in config.json, GeLU erf should be used. The numerical differences between GeLU + tanh and GeLU erf should have negligible impact on inference quality.

• Set --auto-truncate to true by default by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/829

--auto-truncate now defaults to true, meaning that the sequences will be truncated to the lower value between the --max-batch-tokens or the maximum model length, to prevent the --max-batch-tokens from being lower than the actual maximum supported length.

🎉 Additions

• Add --served-model-name for OpenAI requests via HTTP by @vrdn-23 in https://github.com/huggingface/text-embeddings-inference/pull/685
• Extend download_onnx to download sharded ONNX by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/817
• Add support for llama 2 by @michaelfeil in https://github.com/huggingface/text-embeddings-inference/pull/802
• Add support for blackwell architecture (sm100, sm120) by @danielealbano in https://github.com/huggingface/text-embeddings-inference/pull/735
• Mf/add-support-for-llama-3-and-nemotron by @michaelfeil in https://github.com/huggingface/text-embeddings-inference/pull/805
• Add support for DebertaV2 by @vrdn-23 in https://github.com/huggingface/text-embeddings-inference/pull/746
• Add bidirectional attention and projection layer support for Qwen3-based models by @williambarberjr in https://github.com/huggingface/text-embeddings-inference/pull/808

🐛 Fixes

• Fix reading non-standard config for past_key_values in ONNX by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/751
• Fix TruncationDirection to deserialize from lowercase and capitalized by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/755
• Fix sagemaker-entrypoint* & remove SageMaker and Vertex from Dockerfile* by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/699
• Bug: Critical accuracy bugs for model_type=qwen2: no causal attention and wrong tokenizer by @michaelfeil in https://github.com/huggingface/text-embeddings-inference/pull/762
• Fix config.json reading w/ aliases for ORT by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/786
• Fix HTTP error code for validation by @vrdn-23 in https://github.com/huggingface/text-embeddings-inference/pull/818
• Fix to acquire the permit in a blocking way by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/726
• Read Hugging Face Hub token from cache if not provided by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/814
• Align the normalize param between the gRPC and HTTP /embed interfaces by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/810

⚡ Improvements

• Serialization in tokio thread instead of blocking thread, 50% reduction in latency for small models by @michaelfeil in https://github.com/huggingface/text-embeddings-inference/pull/767
• Remove default --model-id argument by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/679
• feat: better Tokenization # workers heuristic by @michaelfeil in https://github.com/huggingface/text-embeddings-inference/pull/766
• add faster index select kernel by @michaelfeil in https://github.com/huggingface/text-embeddings-inference/pull/773
• feat: speedup Parallel safetensors download by @michaelfeil in https://github.com/huggingface/text-embeddings-inference/pull/765
• feat: startup time: add cloned tokenzier fix, saves ~1-20s cold start time by @michaelfeil in https://github.com/huggingface/text-embeddings-inference/pull/772
• Adjust the warmup phase for CPU by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/792

📄 Other

• Skip Gemma3 tests when HF_TOKEN not set by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/812
• Bump Rust 1.92, CUDA 12.6, Ubuntu 24.04 and add Dockerfile-cuda-blackwell-all by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/823
• Update rustc version to 1.92.0 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/826
• Add use_flash_attn for better FA + FA2 feature gating by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/825
• Update CUDA to 12.9 w/ cuda-compat-12-9 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/828
• Upgrade GitHub Actions for Node 24 compatibility by @salmanmkc in https://github.com/huggingface/text-embeddings-inference/pull/782
• Lint: cargo fmt and clippy fix warnings by @michaelfeil in https://github.com/huggingface/text-embeddings-inference/pull/776
• Fix rustfmt on backend/candle/tests/*.rs files by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/800
• Upgrade GitHub Actions to latest versions by @salmanmkc in https://github.com/huggingface/text-embeddings-inference/pull/783
• Update version to 1.9.0 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/830

🆕 New Contributors

• @salmanmkc made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/782
• @danielealbano made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/735
• @williambarberjr made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/808

Full Changelog: https://github.com/huggingface/text-embeddings-inference/compare/v1.8.3...v1.9.0

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

Bug Fixes

• Fix error code for empty requests by @vrdn-23 in https://github.com/huggingface/text-embeddings-inference/pull/727
• Fix the infinite loop when max_input_length is bigger than max-batch-tokens by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/725
• Fix reading modules.json for Dense modules in local models by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/738

Tests, Documentation & Release

• Add test_gemma3.rs for EmbeddingGemma by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/718
• Fix OpenAI client usage example for embeddings by @ZahraDehghani99 in https://github.com/huggingface/text-embeddings-inference/pull/720
• Handle HF_TOKEN in ApiBuilder for candle/tests by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/724
• Fix cargo install commands for candle with CUDA by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/719
• Update version to 1.8.3 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/745

New Contributors

• @ZahraDehghani99 made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/720
• @vrdn-23 made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/727

Full Changelog: https://github.com/huggingface/text-embeddings-inference/compare/v1.8.2...v1.8.3

What's Changed

• Add missing backslash by @philsupertramp in https://github.com/huggingface/text-generation-inference/pull/3311
• Revert "feat: bump flake including transformers and huggingface_hub versions" by @drbh in https://github.com/huggingface/text-generation-inference/pull/3323
• fix: remove azure by @drbh in https://github.com/huggingface/text-generation-inference/pull/3325
• Fix mask passed to flashinfer by @danieldk in https://github.com/huggingface/text-generation-inference/pull/3324
• Update iframe sources for streaming demo by @coyotte508 in https://github.com/huggingface/text-generation-inference/pull/3327
• Revert "Revert "feat: bump flake including transformers and huggingfa… by @drbh in https://github.com/huggingface/text-generation-inference/pull/3326
• Revert "feat: bump flake including transformers and huggingface_hub versions" by @drbh in https://github.com/huggingface/text-generation-inference/pull/3330
• Patch version 3.3.6 by @tengomucho in https://github.com/huggingface/text-generation-inference/pull/3329

New Contributors

• @philsupertramp made their first contribution in https://github.com/huggingface/text-generation-inference/pull/3311
• @coyotte508 made their first contribution in https://github.com/huggingface/text-generation-inference/pull/3327

Full Changelog: https://github.com/huggingface/text-generation-inference/compare/v3.3.5...v3.3.6

🔧 Fixed Intel MKL Support

Since Text Embeddings Inference (TEI) v1.7.0, Intel MKL support had been broken due to changes in the candle dependency. Neither static-linking nor dynamic-linking worked correctly, which caused models using Intel MKL on CPU to fail with errors such as: "Intel oneMKL ERROR: Parameter 13 was incorrect on entry to SGEMM".

Starting with v1.8.2, this issue has been resolved by fixing how the intel-mkl-src dependency is defined. Both features, static-linking and dynamic-linking (the default), now work correctly, ensuring that Intel MKL libraries are properly linked.

This issue occurred in the following scenarios:

• Users installing text-embeddings-router via cargo with the --feature mkl flag. Although dynamic-linking should have been used, it was not working as intended.
• Users relying on the CPU Dockerfile when running models without ONNX weights. In these cases, Safetensors weights were used with candle as backend (with MKL optimizations), instead of ort.

The following table shows the affected versions and containers:

Version	Image
1.7.0	ghcr.io/huggingface/text-embeddings-inference:cpu-1.7.0
1.7.1	ghcr.io/huggingface/text-embeddings-inference:cpu-1.7.1
1.7.2	ghcr.io/huggingface/text-embeddings-inference:cpu-1.7.2
1.7.3	ghcr.io/huggingface/text-embeddings-inference:cpu-1.7.3
1.7.4	ghcr.io/huggingface/text-embeddings-inference:cpu-1.7.4
1.8.0	ghcr.io/huggingface/text-embeddings-inference:cpu-1.8.0
1.8.1	ghcr.io/huggingface/text-embeddings-inference:cpu-1.8.1

More details: PR #715

Full Changelog: v1.8.1...v1.8.2

<img width="1200" height="648" alt="text-embeddings-inference-v1 8 1-embedding-gemma(1)" src="https://github.com/user-attachments/assets/8ad8fb64-cee4-409f-8488-1d10f5ffe995" />

Today, Google releases EmbeddingGemma, a state-of-the-art multilingual embedding model perfect for on-device use cases. Designed for speed and efficiency, the model features a compact size of 308M parameters and a 2K context window, unlocking new possibilities for mobile RAG pipelines, agents, and more. EmbeddingGemma is trained to support over 100 languages and is the highest-ranking text-only multilingual embedding model under 500M on the Massive Text Embedding Benchmark (MTEB) at the time of writing.

• CPU:

docker run -p 8080:80 ghcr.io/huggingface/text-embeddings-inference:cpu-1.8.1 \
    --model-id google/embeddinggemma-300m --dtype float32

• CPU with ONNX Runtime:

docker run -p 8080:80 ghcr.io/huggingface/text-embeddings-inference:cpu-1.8.1 \
    --model-id onnx-community/embeddinggemma-300m-ONNX --dtype float32 --pooling mean

• NVIDIA CUDA:

docker run --gpus all --shm-size 1g -p 8080:80 ghcr.io/huggingface/text-embeddings-inference:cuda-1.8.1 \
    --model-id google/embeddinggemma-300m --dtype float32

Notable Changes

• Add support for Gemma3 (text-only) architecture
• Intel updates to Synapse 1.21.3 and IPEX 2.8
• Extend ONNX Runtime support in OrtRuntime
  • Support position_ids and past_key_values as inputs
  • Handle padding_side and pad_token_id

What's Changed

• Adjust HPU warmup: use dummy inputs with shape more close to real scenario by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/689
• Add extra_args to trufflehog to exclude unverified results by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/696
• Update GitHub templates & fix mentions to Text Embeddings Inference by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/697
• Disable Flash Attention with USE_FLASH_ATTENTION by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/692
• Add support for position_ids and past_key_values in OrtBackend by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/700
• HPU upgrade to Synapse 1.21.3 by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/703
• Upgrade to IPEX 2.8 by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/702
• Parse modules.json to identify default Dense modules by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/701
• Add padding_side and pad_token_id in OrtBackend by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/705
• Update docs/openapi.json for v1.8.0 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/708
• Add Gemma3 architecture (text-only) by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/711
• Update version to 1.8.1 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/712

Full Changelog: https://github.com/huggingface/text-embeddings-inference/compare/v1.8.0...v1.8.1

What's Changed

• [gaudi] Refine rope memory, do not need to keep sin/cos cache per layer by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3274
• Gaudi: add CI by @baptistecolle in https://github.com/huggingface/text-generation-inference/pull/3160
• [gaudi] Gemma3 sliding window support by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3280
• xpu lora support by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3232
• Optimum neuron 0.2.2 by @dacorvo in https://github.com/huggingface/text-generation-inference/pull/3281
• [gaudi] Remove unnecessary reinitialize to HeterogeneousNextTokenChooser to m… by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3284
• [gaudi] Deepseek v2 mla and add ep to unquantized moe by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3287
• [gaudi] Fix the CI test errors by @yuanwu2017 in https://github.com/huggingface/text-generation-inference/pull/3286
• Hpu gptq gidx support by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3297
• Migrate to V2 Pydantic interface by @emmanuel-ferdman in https://github.com/huggingface/text-generation-inference/pull/3262
• Xccl by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3252
• Multi modality fix by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3283
• some gptq case could not be handled by ipex. but could be handle by t… by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3298
• fix outline import issue by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3282
• HuggingFaceM4/Idefics3-8B-Llama3 crash fix by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3267
• Optimum neuron 0.3.0 by @tengomucho in https://github.com/huggingface/text-generation-inference/pull/3308
• Disable Cachix pushes by @danieldk in https://github.com/huggingface/text-generation-inference/pull/3312
• chore: prepare version 3.3.5 by @tengomucho in https://github.com/huggingface/text-generation-inference/pull/3314
• feat: bump flake including transformers and huggingface_hub versions by @drbh in https://github.com/huggingface/text-generation-inference/pull/3313

Full Changelog: https://github.com/huggingface/text-generation-inference/compare/v3.3.4...git

<img width="3600" height="1944" alt="text-embeddings-inference-v1 8 0(2)" src="https://github.com/user-attachments/assets/50df05b6-3821-4e2a-8de0-3e5c911b2a27" />

Notable Changes

• Qwen3 support for 0.6B, 4B and 8B on CPU, MPS, and FlashQwen3 on CUDA and Intel HPUs
• NomicBert MoE support
• JinaAI Re-Rankers V1 support
• Matryoshka Representation Learning (MRL)
• Dense layer module support (after pooling)

What's Changed

• [Docs] Update quick tour by @NielsRogge in https://github.com/huggingface/text-embeddings-inference/pull/574
• Update README.md and supported_models.md by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/572
• Back with linting. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/577
• [Docs] Add cloud run example by @NielsRogge in https://github.com/huggingface/text-embeddings-inference/pull/573
• Fixup by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/578
• Fixing the tokenization routes token (offsets are in bytes, not in by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/576
• Removing requirements file. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/585
• Removing candle-extensions to live on crates.io by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/583
• Bump sccache to 0.10.0 and sccache-action to 0.0.9 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/586
• optimize the performance of FlashBert Path for HPU by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/575
• Revert "Removing requirements file. (#585)" by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/588
• Get opentelemetry trace id from request headers by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/425
• Add argument for configuring Prometheus port by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/589
• Adding missing head. prefix in the weight name in ModernBertClassificationHead by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/591
• Fixing the CI (grpc path). by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/593
• fix xpu env issue that cannot find right libur_loader.so.0 by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/595
• enable flash mistral model for HPU device by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/594
• remove optimum-habana dependency by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/599
• Support NomicBert MoE by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/596
• Remove duplicate short option '-p' to fix router executable by @cebtenzzre in https://github.com/huggingface/text-embeddings-inference/pull/602
• Update text-embeddings-router --help output by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/603
• Warmup padded models too. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/592
• Add support for JinaAI Re-Rankers V1 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/582
• Gte diffs by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/604
• Fix the weight name in GTEClassificationHead by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/606
• upgrade pytorch and ipex to 2.7 version by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/607
• upgrade HPU FW to 1.21; upgrade transformers to 4.51.3 by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/608
• Patch DistilBERT variants with different weight keys by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/614
• add offline modeling for model jinaai/jina-embeddings-v2-base-code to avoid auto_map to other repository by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/612
• Add mean pooling strategy for Modernbert classifier by @kwnath in https://github.com/huggingface/text-embeddings-inference/pull/616
• Using serde for pool validation. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/620
• Preparing the update to 1.7.1 by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/623
• Adding suggestions to fixing missing ONNX files. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/624
• Add Qwen3Model by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/627
• Add HiddenAct::Silu (remove serde alias) by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/631
• Add CPU support for Qwen3-Embedding models by @randomm in https://github.com/huggingface/text-embeddings-inference/pull/632
• refactor the code and add wrap_in_hpu_graph to corner case by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/625
• Support Qwen3 w/ fp32 on GPU by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/634
• Preparing the release. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/639
• Default to Qwen3 in README.md and docs/ examples by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/641
• Fix Qwen3 by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/646
• Add integration tests for Gaudi by @baptistecolle in https://github.com/huggingface/text-embeddings-inference/pull/598
• Fix Qwen3-Embedding batch vs single inference inconsistency by @lance-miles in https://github.com/huggingface/text-embeddings-inference/pull/648
• Fix FlashQwen3 by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/650
• Make flake work on metal by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/654
• Fixing metal backend. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/655
• Qwen3 hpu support by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/656
• change HPU warmup logic: seq length should be with exponential growth by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/659
• Update version to 1.7.3 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/666
• Add last token pooling support for ORT. by @tpendragon in https://github.com/huggingface/text-embeddings-inference/pull/664
• Fix Qwen3 Embedding Float16 DType by @tpendragon in https://github.com/huggingface/text-embeddings-inference/pull/663
• Fix fmt by re-running pre-commit by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/671
• Update version to 1.7.4 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/677
• Support MRL (Matryoshka Representation Learning) by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/676
• Add Dense layer for 2_Dense/ modules by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/660
• Update version to 1.8.0 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/686

New Contributors

• @NielsRogge made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/574
• @cebtenzzre made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/602
• @kwnath made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/616
• @randomm made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/632
• @lance-miles made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/648
• @tpendragon made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/664

Full Changelog: https://github.com/huggingface/text-embeddings-inference/compare/v1.7.0...v1.8.0

Noticeable Changes

Qwen3 was not working fine on CPU / MPS when sending batched requests on FP16 precision, due to the FP32 minimum value downcast (now manually set to FP16 minimum value instead) leading to null values, as well as a missing to_dtype call on the attention_bias when working with batches.

What's Changed

• Fix Qwen3 Embedding Float16 DType by @tpendragon in https://github.com/huggingface/text-embeddings-inference/pull/663
• Fix fmt by re-running pre-commit by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/671
• Update version to 1.7.4 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/677

Full Changelog: https://github.com/huggingface/text-embeddings-inference/compare/v1.7.3...v1.7.4

Noticeable Changes

Qwen3 support included for Intel HPU, and fixed for CPU / Metal / CUDA.

What's Changed

• Default to Qwen3 in README.md and docs/ examples by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/641
• Fix Qwen3 by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/646
• Add integration tests for Gaudi by @baptistecolle in https://github.com/huggingface/text-embeddings-inference/pull/598
• Fix Qwen3-Embedding batch vs single inference inconsistency by @lance-miles in https://github.com/huggingface/text-embeddings-inference/pull/648
• Fix FlashQwen3 by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/650
• Make flake work on metal by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/654
• Fixing metal backend. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/655
• Qwen3 hpu support by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/656
• change HPU warmup logic: seq length should be with exponential growth by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/659
• Update version to 1.7.3 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/666
• Add last token pooling support for ORT. by @tpendragon in https://github.com/huggingface/text-embeddings-inference/pull/664

New Contributors

• @lance-miles made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/648
• @tpendragon made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/664

Full Changelog: https://github.com/huggingface/text-embeddings-inference/compare/v1.7.2...v1.7.3

Fix for Neuron models exported with batch_size 1.

What's Changed

• [gaudi] gemma3 text and vlm model intial support. need to add sliding window … by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3270
• Neuron backend fix by @dacorvo in https://github.com/huggingface/text-generation-inference/pull/3273

Full Changelog: https://github.com/huggingface/text-generation-inference/compare/v3.3.3...v3.3.4

Neuron backend update.

What's Changed

• Remove useless packages by @yuanwu2017 in https://github.com/huggingface/text-generation-inference/pull/3253
• Bump neuron SDK version by @dacorvo in https://github.com/huggingface/text-generation-inference/pull/3260
• Perf opt by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3256
• [gaudi] Vlm rebase and issue fix in benchmark test by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3263
• Move the _update_cos_sin_cache into get_cos_sin by @yuanwu2017 in https://github.com/huggingface/text-generation-inference/pull/3254
• [Gaudi] Remove optimum-habana by @yuanwu2017 in https://github.com/huggingface/text-generation-inference/pull/3261
• [gaudi] HuggingFaceM4/idefics2-8b issue fix by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3264
• [Gaudi] Enable Qwen3_moe model by @yuanwu2017 in https://github.com/huggingface/text-generation-inference/pull/3244
• [Gaudi]Fix the integration-test issues by @yuanwu2017 in https://github.com/huggingface/text-generation-inference/pull/3265
• [Gaudi] use pad_token_id to pad input id by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3268
• chore: prepare release 3.3.3 by @dacorvo in https://github.com/huggingface/text-generation-inference/pull/3269
• [gaudi] Refine logging for Gaudi warmup by @regisss in https://github.com/huggingface/text-generation-inference/pull/3222
• doc: fix README by @dacorvo in https://github.com/huggingface/text-generation-inference/pull/3271

Full Changelog: https://github.com/huggingface/text-generation-inference/compare/v3.3.2...v3.3.3

Notable change

• Added support for Qwen3 embeddigns

What's Changed

• Adding suggestions to fixing missing ONNX files. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/624
• Add Qwen3Model by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/627
• Add HiddenAct::Silu (remove serde alias) by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/631
• Add CPU support for Qwen3-Embedding models by @randomm in https://github.com/huggingface/text-embeddings-inference/pull/632
• refactor the code and add wrap_in_hpu_graph to corner case by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/625
• Support Qwen3 w/ fp32 on GPU by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/634
• Preparing the release. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/639

New Contributors

• @randomm made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/632

Full Changelog: https://github.com/huggingface/text-embeddings-inference/compare/v1.7.1...v1.7.2

What's Changed

• [Docs] Update quick tour by @NielsRogge in https://github.com/huggingface/text-embeddings-inference/pull/574
• Update README.md and supported_models.md by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/572
• Back with linting. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/577
• [Docs] Add cloud run example by @NielsRogge in https://github.com/huggingface/text-embeddings-inference/pull/573
• Fixup by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/578
• Fixing the tokenization routes token (offsets are in bytes, not in by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/576
• Removing requirements file. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/585
• Removing candle-extensions to live on crates.io by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/583
• Bump sccache to 0.10.0 and sccache-action to 0.0.9 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/586
• optimize the performance of FlashBert Path for HPU by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/575
• Revert "Removing requirements file. (#585)" by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/588
• Get opentelemetry trace id from request headers by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/425
• Add argument for configuring Prometheus port by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/589
• Adding missing head. prefix in the weight name in ModernBertClassificationHead by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/591
• Fixing the CI (grpc path). by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/593
• fix xpu env issue that cannot find right libur_loader.so.0 by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/595
• enable flash mistral model for HPU device by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/594
• remove optimum-habana dependency by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/599
• Support NomicBert MoE by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/596
• Remove duplicate short option '-p' to fix router executable by @cebtenzzre in https://github.com/huggingface/text-embeddings-inference/pull/602
• Update text-embeddings-router --help output by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/603
• Warmup padded models too. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/592
• Add support for JinaAI Re-Rankers V1 by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/582
• Gte diffs by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/604
• Fix the weight name in GTEClassificationHead by @kozistr in https://github.com/huggingface/text-embeddings-inference/pull/606
• upgrade pytorch and ipex to 2.7 version by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/607
• upgrade HPU FW to 1.21; upgrade transformers to 4.51.3 by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/608
• Patch DistilBERT variants with different weight keys by @alvarobartt in https://github.com/huggingface/text-embeddings-inference/pull/614
• add offline modeling for model jinaai/jina-embeddings-v2-base-code to avoid auto_map to other repository by @kaixuanliu in https://github.com/huggingface/text-embeddings-inference/pull/612
• Add mean pooling strategy for Modernbert classifier by @kwnath in https://github.com/huggingface/text-embeddings-inference/pull/616
• Using serde for pool validation. by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/620
• Preparing the update to 1.7.1 by @Narsil in https://github.com/huggingface/text-embeddings-inference/pull/623

New Contributors

• @NielsRogge made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/574
• @cebtenzzre made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/602
• @kwnath made their first contribution in https://github.com/huggingface/text-embeddings-inference/pull/616

Full Changelog: https://github.com/huggingface/text-embeddings-inference/compare/v1.7.0...v1.7.1

Gaudi improvements.

What's Changed

• upgrade to new vllm extension ops(fix issue in exponential bucketing) by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3239
• Nix: switch to hf-nix by @danieldk in https://github.com/huggingface/text-generation-inference/pull/3240
• Add Qwen3 by @yuanwu2017 in https://github.com/huggingface/text-generation-inference/pull/3229
• fp8 compressed_tensors w8a8 support by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3242
• [Gaudi] Fix the OOM issue of Llama-4-Scout-17B-16E-Instruct by @yuanwu2017 in https://github.com/huggingface/text-generation-inference/pull/3245
• Fix the Llama-4-Maverick-17B-128E crash issue by @yuanwu2017 in https://github.com/huggingface/text-generation-inference/pull/3246
• Prepare for 3.3.2 by @danieldk in https://github.com/huggingface/text-generation-inference/pull/3249

Full Changelog: https://github.com/huggingface/text-generation-inference/compare/v3.3.1...v3.3.2

This release updates TGI to Torch 2.7 and CUDA 12.8.

What's Changed

• change HPU warmup logic: seq length should be with exponential growth by @kaixuanliu in https://github.com/huggingface/text-generation-inference/pull/3217
• adjust the round_up_seq logic to align with prefill warmup phase on… by @kaixuanliu in https://github.com/huggingface/text-generation-inference/pull/3224
• Update to Torch 2.7.0 by @danieldk in https://github.com/huggingface/text-generation-inference/pull/3221
• Enable Llama4 for gaudi backend by @yuanwu2017 in https://github.com/huggingface/text-generation-inference/pull/3223
• fix: count gpu uuids if NVIDIA_VISIBLE_DEVICES env set to all by @drbh in https://github.com/huggingface/text-generation-inference/pull/3230
• Deepseek r1 by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3211
• Refine warmup and upgrade to synapse AI 1.21.0 by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3234
• fix the crash in default ATTENTION path by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3235
• Switch to punica-sgmv kernel from the Hub by @danieldk in https://github.com/huggingface/text-generation-inference/pull/3236
• move input_ids to hpu and remove disposal of adapter_meta by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3237
• Prepare for 3.3.1 by @danieldk in https://github.com/huggingface/text-generation-inference/pull/3238

New Contributors

• @kaixuanliu made their first contribution in https://github.com/huggingface/text-generation-inference/pull/3217

Full Changelog: https://github.com/huggingface/text-generation-inference/compare/v3.3.0...v3.3.1

Notable changes

• Prefill chunking for VLMs.

What's Changed

• Fixing Qwen 2.5 VL (32B). by @Narsil in https://github.com/huggingface/text-generation-inference/pull/3157
• Fixing tokenization like https://github.com/huggingface/text-embeddin… by @Narsil in https://github.com/huggingface/text-generation-inference/pull/3156
• Gaudi: clean cuda/rocm code in hpu backend, enable flat_hpu by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3113
• L4 fixes by @mht-sharma in https://github.com/huggingface/text-generation-inference/pull/3161
• setuptools <= 70.0 is vulnerable: CVE-2024-6345 by @Narsil in https://github.com/huggingface/text-generation-inference/pull/3171
• transformers flash llm/vlm enabling in ipex by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3152
• Upgrading the dependencies in Gaudi backend. by @Narsil in https://github.com/huggingface/text-generation-inference/pull/3170
• Hotfixing gaudi deps. by @Narsil in https://github.com/huggingface/text-generation-inference/pull/3174
• Hotfix gaudi2 with newer transformers. by @Narsil in https://github.com/huggingface/text-generation-inference/pull/3176
• Support flashinfer for Gemma3 prefill by @danieldk in https://github.com/huggingface/text-generation-inference/pull/3167
• Get opentelemetry trace id from request headers instead of creating a new trace by @kozistr in https://github.com/huggingface/text-generation-inference/pull/2648
• Bump sccache to 0.10.0 by @alvarobartt in https://github.com/huggingface/text-generation-inference/pull/3179
• Fixing CI by @Narsil in https://github.com/huggingface/text-generation-inference/pull/3184
• Add option to configure prometheus port by @mht-sharma in https://github.com/huggingface/text-generation-inference/pull/3187
• Warmup gaudi backend by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3172
• Put more wiggle room. by @Narsil in https://github.com/huggingface/text-generation-inference/pull/3189
• Fixing the router + template for Qwen3. by @Narsil in https://github.com/huggingface/text-generation-inference/pull/3200
• Skip {% generation %} and {% endgeneration %} template handling by @alvarobartt in https://github.com/huggingface/text-generation-inference/pull/3204
• doc typo by @julien-c in https://github.com/huggingface/text-generation-inference/pull/3206
• Pr 2982 ci branch by @drbh in https://github.com/huggingface/text-generation-inference/pull/3046
• fix: bump snaps for mllama by @drbh in https://github.com/huggingface/text-generation-inference/pull/3202
• Update client SDK snippets by @julien-c in https://github.com/huggingface/text-generation-inference/pull/3207
• Fix HF_HUB_OFFLINE=1 for Gaudi backend by @regisss in https://github.com/huggingface/text-generation-inference/pull/3193
• IPEX support FP8 kvcache/softcap/slidingwindow by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3144
• forward and tokenize chooser use the same shape by @sywangyi in https://github.com/huggingface/text-generation-inference/pull/3196
• Chunked Prefill VLM by @mht-sharma in https://github.com/huggingface/text-generation-inference/pull/3188
• Prepare for 3.3.0 by @danieldk in https://github.com/huggingface/text-generation-inference/pull/3220

New Contributors

• @kozistr made their first contribution in https://github.com/huggingface/text-generation-inference/pull/2648
• @julien-c made their first contribution in https://github.com/huggingface/text-generation-inference/pull/3206

Full Changelog: https://github.com/huggingface/text-generation-inference/compare/v3.2.3...v3.3.0